Cultural issues in the classification and diagnosis of mood and anxiety disorders

WPAWorld PsychiatryOFFICIAL JOURNAL OF THE WORLD PSYCHIATRIC ASSOCIATION (WPA)

Volume 11, Supplement 1 June 2012

THE ICD-11 CLASSIFICATION OF MOOD AND ANXIETY DISORDERS: BACKGROUND

AND OPTIONS

Guest Editors

Mario MajGeoffrey M. Reed

ISSN 1723-8617

The World Psychiatric Association (WPA)

The WPA is an association of national psychiatric societiesaimed to increase knowledge and skills necessary for work inthe field of mental health and the care for the mentally ill. Itsmember societies are presently 135, spanning 117 differentcountries and representing more than 200,000 psychiatrists.

The WPA organi zes the World Congress of Psychiatryevery three years. It also organizes international and regionalcongresses and meetings, and thematic conferences. It has 66scientific sections, aimed to disseminate information and pro-mote collaborative work in specific domains of psychiatry. Ithas produced several educational programmes and series ofbooks. It has developed ethical guidelines for psychiatricpractice, including the Madrid Declaration (1996).

Further information on the WPA can be found on the web-site www.wpanet.org.

WPA Executive CommitteePresident – P. Ruiz (USA)President-Elect – D. Bhugra (UK)Secretary General – L. Küey (Turkey) Secretary for Finances – T. Akiyama (Japan)Secretary for Meetings – T. Okasha (Egypt)Secretary for Education – E. Belfort (Venezuela)Secretary for Publications – M. Riba (USA)Secretary for Sections – A. Javed (UK)

WPA SecretariatGeneva University Psychiatric Hospital, 2 Chemin du PetitBel-Air, 1225 Chêne-Bourg, Geneva, Switzerland. Phone:+41223055737; Fax: +41223055735; E-mail: [email protected].

World Psychiatry

World Psychiatry is the official journal of the WorldPsychiatric Association. It is published in three issues per yearand is sent free of charge to psychiatrists whose names andaddresses are provided by WPA member societies and sec-tions.

Research Reports containing unpublished data are wel-come for submission to the journal. They should be subdividedinto four sections (Introduction, Methods, Results, Discussion).References should be numbered consecutively in the text andlisted at the end according to the following style: 1. Bathe KJ, Wilson EL. Solution methods for eigenvalue

problems in structural mechanics. Int J Num Math Engng1973;6:213-26.

2. McRae TW. The impact of computers on accounting.London: Wiley, 1964.

3. Fraeijs de Veubeke B. Displacement and equilibrium modelsin the finite element method. In: Zienkiewicz OC, HollisterGS (eds). Stress analysis. London: Wiley, 1965:145-97.All submissions should be sent to the office of the Editor.

Editor – M. Maj (Italy).Associate Editor – P. Ruiz (USA).Editorial Board – D. Bhugra (UK), L. Küey (Turkey), T.Akiyama (Japan), T. Okasha (Egypt), E. Belfort (Venezuela), M.Riba (USA), A. Javed (UK).Advisory Board – H.S. Akiskal (USA), R.D. Alarcón (USA),J.A. Costa e Silva (Brazil), J. Cox (UK), H. Herrman (Austra-lia), M. Jorge (Brazil), H. Katschnig (Austria), F. Lieh-Mak(Hong Kong-China), F. Lolas (Chile), J.J. López-Ibor (Spain),J.E. Mezzich (USA), D. Moussaoui (Morocco), P. Munk-Jorgensen (Denmark), F. Njenga (Kenya), A. Okasha (Egypt),J. Parnas (Denmark), V. Patel (India), N. Sartorius (Switz-erland), C. Stefanis (Greece), M. Tansella (Italy), A. Tasman(USA), S. Tyano (Israel), J. Zohar (Israel).

Office of the Editor – Department of Psychiatry, University ofNaples SUN, Largo Madonna delle Grazie, 80138 Naples, Italy. Phone: +390815666502; Fax: +390815666523; E-mail: [email protected].

Managing Director and Legal Responsibility - Emile Blomme(Italy)Published by Elsevier S.r.l., Via P. Paleocapa 7, 20121 Milan,Italy.

World Psychiatry is indexed in PubMed, Current Contents/Clinical Medicine, Current Contents/Socialand Behavioral Sciences, Science Citation Index, and EMBASE.

All back issues of World Psychiatry can be downloaded free of charge from the PubMed system(http://www.pubmedcentral.nih.gov/tocrender.fcgi?journal=297&action=archive).

The development of the ICD-11 classification of mood and anxiety disorders 3 M. Maj, G.M. Reed

How global epidemiological evidence can inform the revision of ICD-10 classification of depression 6 and anxiety disorders L.H. andRade, Y.-P. WanG

Specifiers as aids to treatment selection and clinical management in the ICD classification of mood 11 disorders d.j. MikLoWitz, M.B. FiRst

Challenges in the implementation of diagnostic specifiers for mood disorders in ICD-11 17 M.B. FiRst

Cultural issues in the classification and diagnosis of mood and anxiety disorders 26 s. CHakRaBaRti, C. BeRLanGa, F. njenGa

Bipolar disorders in ICD-11 31 s.M. stRakoWski

Changes needed in the classification of depressive disorders: options for ICD-11 37 e. PaYkeL, L.H. andRade, F. njenGa, M.R. PHiLLiPs

Differentiating depression from ordinary sadness: contextual, qualitative and pragmatic approaches 43 M. Maj

Severity of depressive disorders: considerations for ICD-11 48 j.L. aYuso-Mateos, P. LoPez-GaRCía

Dysthymia and cyclothymia in ICD-11 53 M.R. PHiLLiPs

Psychotic and catatonic presentations in bipolar and depressive disorders 59 s. CHakRaBaRti

Mixed states and rapid cycling: conceptual issues and options for ICD-11 65 M. Maj

How should melancholia be incorporated in ICD-11? 69 d. Moussaoui, M. aGouB, a. kHouBiLa

Postpartum depression and premenstrual dysphoric disorder: options for ICD-11 73 M.L. FiGueiRa, V. VideiRa dias

Disruptive mood dysregulation with dysphoria disorder: a proposal for ICD-11 77 e. LeiBenLuFt, R. uHeR, M. RutteR

Generalized anxiety disorder in ICD-11 82 M.k. sHeaR

Agoraphobia and panic disorder: options for ICD-11 89 d.j. stein

Specific and social phobias in ICD-11 94 P.M.G. eMMeLkaMP

Hypochondriasis in ICD-11 100 d.j. stein

Contents

SOMMARIO_3.indd 1 06/07/12 14:02

SOMMARIO_3.indd 2 06/07/12 14:02

3

The World Health Organization (WHO) is currently revis-ing the ICD-10 (1), which was approved in 1990, making the current period the longest in the history of the ICD without a major revision. WHO’s responsibility for international defini-tion and nomenclature of diseases and the standardization of diagnostic procedures is one of the organization’s core con-stitutional functions (2). WHO Member States have agreed by international treaty to use the ICD as a basis for reporting health information that is usable and comparable across countries. The ICD is also used by Member States for pur-poses including treatment selection and eligibility, reimburse-ment, outcome and health service evaluation, health policy and priorities, and resource allocation.

Within the context of the overall WHO policies, priorities, and procedures for the development of ICD-11, currently slated for approval by the World Health Assembly in 2015, the WHO Department of Mental Health and Substance Abuse has been assigned responsibility for managing the technical work of developing the chapter on mental and be-havioural disorders. The inclusion of mental and behavioural disorders alongside all other diagnostic entities in health care is an important feature of the ICD. A classification that uses a common framework across all disease areas is more likely to be used by all specialties and general health care workers in a similar way and to yield comparable statistics on ser-vices. A common classification framework facilitates the search for related mechanisms of etiology, pathophysiology, and comorbidity, and provides for conceptual parity of psy-chopathology with the rest of the medical system for clinical, administrative, and financial functions in health care (3).

In developing the ICD-11 classification of mental and be-havioural disorders, the Department of Mental Health and Substance Abuse has specified that substantial changes to existing mental disorders categories and definitions should be made through a transparent, international, multidisci-plinary, and multilingual process that involves the direct par-ticipation of a broad range of stakeholders and is as free as possible from conflicts of interests. To assist the Department in all phases of the mental and behavioural disorders revision process, the WHO has appointed an International Advisory Group (3), on which the WPA and other international asso-ciations representing relevant health professions are official-ly represented.

The Department and the Advisory Group have clearly in-dicated that the most important goal of the ICD-11 classifica-tion of mental and behavioural disorders will be to help WHO

The development of the ICD-11 classification of moodand anxiety disordersMario Maj1,2, Geoffrey M. reed3

1World Psychiatric Association; 2Department of Psychiatry, University of Naples SUN, Naples, Italy; 3Department of Mental Health and Substance Abuse, World Health Organization, Geneva, Switzerland

Member States to reduce the disease burden associated with mental disorders. In order for the ICD-11 classification of mental and behavioural disorders to be a more effective tool for meeting international public health goals, the new system will need to be usable for implementation throughout the world at the point where people with mental health needs are most likely to come into contact with the health system. Peo-ple are only likely to have access to the most appropriate mental health services when the conditions that define eligi-bility and treatment selection are supported by a precise, valid, and clinically useful classification system (3).

In May 2012, the WHO’s governing body, comprised of the Ministers of Health of all 194 WHO Member States, passed a resolution on the “Global burden of mental disor-ders and the need for a comprehensive, coordinated response from health and social sectors at the country level”. Accord-ing to the resolution, mental health problems are of major importance to all societies and are significant contributors to the burden of disease and the loss of quality of life, and have huge economic and social costs. The resolution urges a series of specific actions by WHO Member States and by the WHO as an organization, including: development by countries of comprehensive policies and strategies for early identifica-tion, support, treatment and recovery of persons with mental disorders; prioritizing mental health treatment in programs addressing health and development and allocating appropri-ate resources to this effort; and addressing policies related to equitable access to affordable, quality and comprehensive health services that integrate mental health into all levels of the health care system, including psychosocial interventions and medication and addressing physical health care needs. The ICD-11 classification of mental and behavioural disor-ders will be an integral part of the framework for accomplish-ing these objectives.

In order to assist the WHO in developing a classification of mental and behavioural disorders that is scientifically val-id, clinically useful, and globally applicable, the WHO has appointed a series of Working Groups focused on particular areas of the classification, which are charged with providing recommendations for the structure of the classification with-in their areas and developing the content that will be required for various versions of the classification. The Working Group on the Classification of Mood and Anxiety Disorders in-cludes M. Maj, Chair (Italy), L.H. Andrade (Brazil), J. Angst (Switzerland), J.L. Ayuso-Mateos (Spain), C. Berlanga (Mex-ico), S. Chakrabarti (India), P.M.G. Emmelkamp (Nether-

003_005.indd 3 06/07/12 14:03

4 World Psychiatry 11:S1 - June 2012

lands), M.L. Figueira (Portugal), E. Leibenluft (USA), D.J. Miklowitz (USA), D. Moussaoui (Morocco), F. Njenga (Ke-nya), E. Paykel (UK), M.R. Phillips (P.R. China), M.K. Shear (USA), and D.J. Stein (South Africa), assisted by consultants M.B. First and S.M. Strakowski and WHO Secretariat Mem-bers S. Saxena (Director, WHO Department of Mental Health and Substance Abuse) and G.M. Reed.

Accurate, reliable and useful diagnosis of mood and anxi-ety disorders represents an important challenge in the devel-opment of ICD-11, particularly given the high prevalence of these disorders and the large proportion of the disease bur-den of mental disorders for which they account (4). In order to facilitate an examination of diagnostic boundaries be-tween and within mood and anxiety disorders and to avoid artificial reification of categories (5), the WHO chose to ap-point a single Working Group charged with examining both sets of disorders.

The high rate of co-occurrence among mood and anxiety disorders – partly an artifact of overlapping symptomatology in diagnostic descriptions – has led to questions regarding whether they in fact represent distinct disorders. However, while available evidence suggests that mood and anxiety dis-orders are closely related, the existence of “pure” types sup-ports the conceptualization of them as distinct groups of dis-orders. In translating the varied phenomenology of mood and anxiety disorders into a standard classification system, some have argued that the assignment of multiple diagnostic cate-gories is the best way to convey the complexity of a given clinical presentation. Others hold that artificial “comorbidi-ty” should be eliminated from the classification to the greatest extent possible. The Working Group has been asked to con-sider how best to improve clinical utility in relation to presen-tations that fulfill current diagnostic guidelines for multiple disorders, as well as in situations where anxiety or depressive symptoms are an important clinical feature of presentations assigned to the other diagnostic group.

Other issues that the Working Group was asked to con-sider include:

• The clinical utility of mood and anxiety diagnoses in day-to-day practice around the world;

• The extent of the influence of culture on mood and anxiety symptomatology and diagnosis, and how this may be bet-ter captured in disorder descriptions;

• Diagnostic descriptions across the life span: older adults, adolescents and children;

• Diagnostic thresholds and their relationship to functional status;

• Whether mood and anxiety disorder diagnoses should be assigned in the context of life events in response to which symptoms of depression or anxiety would be considered normative (e.g., bereavement);

• The potential utility of severity or other qualifiers for use with specific mood and anxiety disorder categories;

• The overall groupings of mood and anxiety disorders with-in the classification.

As a part of its work, the Working Group was asked: a) to review available scientific evidence and clinical information on use, clinical utility, and experience with ICD-10 mood and anxiety diagnostic categories in various countries around the world and within various health care settings; and b) to review proposals for mood and anxiety diagnostic categories in the DSM-5, and consider how these may or may not be suited for global applications.

A complicating factor in developing a satisfactory classifi-cation of disorders in the ICD-11 is the enormous range of settings in which it must be useful. At one end of the spec-trum are technologically sophisticated specialist facilities in developed countries with highly educated, specialized, mul-tidisciplinary staff. At the other end are community-based primary care programs in rural settings in developing coun-tries with very limited facilities and infrastructure. In the lat-ter type of setting, identification and treatment of mental disorders must be carried out by health care workers who are often not physicians, are highly unlikely to be mental health professionals, and may have limited formal professional training of any kind.

To help countries to reduce the disease burden associated with mental disorders, the WHO and the Advisory Group have indicated that the classification system must be usable and useful for health care workers who are not highly trained mental health professionals to identify people with mental disorders in need of treatment (3). At the same time, in order for the system to be widely used and influential, it must also meet the needs of highly trained specialists in developed countries. This suggests that there will be a need for more than one version of the classification system, the complexity and characteristics of which can be matched to the needs, charac-teristics, and resources of the setting. With the ICD-11, there appears to be a unique opportunity to produce such a multi-layered system, since the possibility of “telescoping” views of ICD is explicitly envisioned as a core aspect of the system, made possible by the electronic nature of its development.

The WHO Department of Mental Health and Substance Abuse envisions developing at least three versions of the ICD-11 classification of mental and behavioural disorders: a) a version for use by mental health professionals in daily clin-ical practice, equivalent to the Clinical Descriptions and Di-agnostic Guidelines for ICD-10 (6); b) a much simpler ver-sion for primary care professionals, focusing both on the types of problems that are typically presented in these set-tings and on those conditions that account for the highest proportions of the disease burden of mental disorders and for which detection in primary care settings must be improved; and c) a version for use in research. Additional adaptations for specific purposes are also possible, based on the source material to be developed by the Working Groups. The Work-ing Groups have been asked to focus on the version for men-tal health specialists (equivalent to the Clinical Descriptions and Diagnostic Guidelines) in the initial phase of their work, though keeping in mind the requirements of primary care and research usage.

003_005.indd 4 06/07/12 14:03

5World Psychiatry 11:S1 - June 2012

The issue of the “metastructure” or overarching architec-ture of the ICD-11 classification of mental and behavioural disorders has primarily been managed by the Advisory Group, with input from the Working Groups on specific areas. In the ICD-10, the numbers of large groupings, or “blocks”, of dis-orders was artificially constrained by the decimal coding sys-tem used in the classification, such that it was possible to have a maximum of only ten major groupings of disorders within the mental and behavioural disorder chapter (corre-sponding to the digits 0 to 9). This meant that some group-ings were created that were not based on clinical utility or scientific evidence. In the ICD-10, one block (F30-F39) is devoted to mood (affective) disorders, which includes manic episode, bipolar disorder, depressive episode, recurrent de-pressive disorder, and persistent mood (affective) disorders (cyclothymia and dysthymia). Anxiety disorders, on the oth-er hand, represent only a portion of a different block (F40-F49) called “Neurotic, stress-related, and somatoform disor-ders”.

The ICD-11 will use a different coding structure that is not based on a decimal numbering system, such that there can be a larger number of blocks or groupings within the chapter. The Advisory Group, in consultation with the Working Groups, has proposed a metastructure for mental and behav-ioural disorders consisting of about twenty-two groupings of disorders (the exact number will depend on pending deci-sions about how to handle specific categories). According to the proposed metastructure, there will be separate, adjacent blocks for bipolar and related disorders, depressive disor-ders, and anxiety and fear-related disorders. According to this structure, cyclothymia (if retained) would be grouped under bipolar and related disorders, and dysthymia (if re-tained) would be grouped under depressive disorders. There would be separate blocks for disorders specifically associated with stress and for obsessive-compulsive and related disor-ders rather than these categories being grouped with anxiety disorders. Responsibility for developing initial proposals for stress-related and obsessive-compulsive spectrum disorders has been assigned to other Working Groups.

The Working Group on the Classification of Mood and Anxiety Disorders is still engaged in its work, and has not yet produced its final recommendations. The papers in this supplement reflect the initial evaluations of Working Group members with respect to the series of specific topics within the group’s scope of responsibility, based on the group’s ini-tial review of the evidence and consideration of various alter-natives. Where appropriate, the papers follow the general template of: a) considering historical issues in the classifica-tion of particular categories and the approaches to these cat-egories taken in the ICD-10 and the DSM-IV (7), b) describ-ing the problems that have arisen in the implementation of the ICD-10 and the DSM-IV; c) summarizing the currently available proposals for DSM-5 related to these categories; and d) presenting the options for ICD-11, taking into ac-

count the issues of clinical utility, global applicability, and use outside specialty mental health settings. None of the options presented in the supplement represent decisions taken by the Working Group or the official positions of the Working Group, the Advisory Group, or the WHO.

The Working Group is publishing the results of its initial evaluation now in order to stimulate scientific discussion and exchange regarding how the ICD-11 classification of mood and anxiety disorders can better fulfill the purposes and aims described above, particularly in terms of clinical utility and global applicability. We believe that it is important that such discussion and debate start now, before the Work-ing Group’s proposals are fully formulated, and that such exchange should be an ongoing feature of the ICD revision. A vigorous and transparent process will best serve the inter-ests of global psychiatry as well as global public health in developing a classification of mood and anxiety disorders that can be a more effective tool for reducing the burden of mental and behavioural disorders throughout the world.

Note

M. Maj is a member of the WHO International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders and the Chairman of the ICD Revision Working Group on the Classification of Mood and Anxiety Disorders. G.M. Reed is Senior Project Officer, Revision of ICD-10 Men-tal and Behavioural Disorders, Department of Mental Health and Substance Abuse, WHO. The views expressed in this ar-ticle are those of the authors and, except as specifically noted, do not represent the official policies or positions of the Inter-national Advisory Group, the Working Group on the Classi-fication of Mood and Anxiety Disorders, or the WHO.

References

1. World Health Organization. International classification of diseases and related health problems, 10th revision. Geneva: World Health Organization, 1992.

2. World Health Organization. Basic documents, 46th edition. Ge-neva: World Health Organization, 2007.

3. International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders. A conceptual framework for the revi-sion of the ICD-10 classification of mental and behavioural disor-ders. World Psychiatry 2011;10:86-92.

4. World Health Organization. The global burden of disease: 2004 up-date. Geneva: World Health Organization, 2008.

5. Hyman SE. The diagnosis of mental disorders: the problem of reifi-cation. Annu Rev Clin Psychol 2010;6:155-79.

6. World Health Organization. The ICD-10 classification of mental and behavioural disorders: clinical descriptions and diagnostic guidelines. Geneva: World Health Organization, 1992.

7. American Psychiatric Association. Diagnostic and statistical manu-al of mental disorders, 4th ed., text revision. Washington: American Psychiatric Association, 2000.

003_005.indd 5 06/07/12 14:03


How global epidemiological evidence can informthe revision of ICD-10 classification of depressionand anxiety disorders

Anxiety and mood disorders are the most prevalent classes of mental disorder, affecting respectively around 7% and 5% of respondents in the year prior to interview in community-based samples (1-3). In primary care settings (4), 24% of at-tendees receive a current ICD-10 diagnosis of an anxiety or mood disorder, and 9.5% receive multiple diagnoses. Depres-sion is the most common mental disorder diagnosis in these settings (7.1% for men and 12.8% for women), followed by generalized anxiety disorder (GAD) (5.7% for men and 9.2% for women). In addition, among the ten most frequently used ICD-10 diagnostic categories in daily clinical practice by psy-chiatrists around the world, six are mood or anxiety disorders (5), with depressive episode the most commonly used catego-ry. Therefore, these two groups of disorders have great impor-tance in both primary care and specialty settings, as well as for public health.

Problems wItH Current ClassIfICatIons

There are many problems related to the diagnosis of de-pression and anxiety disorders according to current classifi-cations. First, high rates of co-occurrence among classes of disorders and within classes represent a particular challenge for classification. Second, symptomatic but subthreshold presentations have been well described in referred (6), non-referred (7), and community-dwelling samples (8). Subthresh-old presentations are more prevalent than the corresponding ICD-10 diagnoses (9), accounting for nearly half of all psy-chological problems detected in general practice settings (7), with levels of psychological distress, disability and perceived health comparable to diagnosable disorders (9). Third, het-erogeneous and culturally-linked somatic presentations of depression and anxiety are frequent among attendees in pri-

Laura H. andrade, Yuan-Pang Wang

Section of Psychiatric Epidemiology - LIM 23, Department/Institute of Psychiatry, University of São Paulo Medical School, Rua Dr. Ovídio Pires de Campos, 785, CEP 05403-010, São Paulo, Brazil

This article reviews current evidence from epidemiological surveys that may be relevant for improving the ICD classification of depression and anxiety disorders. Topics of special interest are the boundaries of these disorders, subthreshold presentations, patterns of comorbidity, and cultural variations in somatic presentations. Proposals for hierarchical structures that model the core factors underlying these disorders are presented. The need for a better description of mixed anxiety and depressive disorder in ICD-11 is highlighted. Questions regarding how to model somatic symptoms related to depression and anxiety disorders and how to take the cultural components of these disorders into account are raised. The challenge of reformulating the current classification in order to enhance clinical utility should not cast aside basic conceptu-alizations of psychopathology.

Key words: Mood disorder, anxiety disorder, epidemiology, comorbidity, classification, mental disorders, ICD

World Psychiatry 2012;11(Suppl. 1):6-10

mary care settings (10).

Comorbidity

The high co-occurrence of mood and anxiety disorders can be viewed as an artifact of the large number of polythetic categorical concepts that have characterized diagnostic sys-tems since DSM-III (11). The DSM-III was the first classifica-tion intended for clinical use that relied on highly explicit operational criteria for mental disorders, giving rise to a neo-Kraepelinian approach that continues to characterize the field (12). With the suppression of hierarchical rules in DSM-III-R and DSM-IV, the likelihood of multiple diagnoses in-creased, due to both genuine correlations among distinct syndromes and artifactual comorbidity based on shared symptoms and arbitrary boundaries (13). According to epide-miological studies, depression and anxiety disorders have very high rates of co-occurrence, although the onset of anxi-ety disorders typically precedes that of depression (14,15).

To explain the comorbidity of anxiety and depression, an empirical model postulating a three-component structure for anxiety and depressive syndromes was proposed by Clark and Watson (16,17), consisting of a non-specific general af-fective distress component, “anhedonia” or the lack of posi-tive affect as a specific component of depression, and “phys-iological hyperarousal” or somatic tension as a specific com-ponent of anxiety. The general affective distress component – called negative affectivity or neuroticism – is conceptualized as a stable, heritable, and highly general trait dimension char-acterized by temperamental sensitivity to negative stimuli, high predisposition to experience negative moods like fear, anxiety, sadness, guilt, and hostility, as well as somatic com-plaints, negativistic appraisal of the self and others, pessimism

006_010.indd 6 06/07/12 14:04


and low self-esteem (16). An overlap of genetic contributions to depression and GAD was reported by Hettema et al (18), arguing against the current grouping of these disorders into distinct polythetic categories with clear-cut boundaries.

The phenomenon of comorbidity is important as an indi-cator of overall clinical severity and has significant treatment implications. Comorbid cases are usually more severe than single disorder cases, and have an earlier age of onset (19) and higher level of symptomatology (20). Comorbidity of GAD and depression has been found to be associated with an in-creased likelihood of hypertension, and the two disorders seem to interact synergistically, increasing the risk of both all-cause and cardiovascular mortality (21). In the World Mental Health Surveys, comorbid depression-anxiety disor-der was more strongly associated with several physical condi-tions than were single disorders (22).

In the National Comorbidity Survey Replication, Kessler et al (23) found that among the twelve highest bivariate cor-relations (exceeding 0.60), six occurred among classes of anxiety and mood disorders: bipolar disorder (depressive episode with mania-hypomania), double depression (depres-sive episode with dysthymia), anxious depression (depressive episode with GAD), comorbid mania-hypomania and atten-tion-deficit/hyperactivity disorder, panic disorder with agora-phobia, and comorbid social phobia with agoraphobia. In the São Paulo Megacity Mental Health Survey (24), of the eight pairs of disorders showing the highest correlations, four were represented by the association of mood and anxiety disorders, between and within classes: double depression (depressive episode with dysthymia), bipolar disorder (depressive epi-sode with mania-hypomania), phobic disorders (agoraphobia with specific phobia and social phobia), and panic disorder and agoraphobia.

The observation of symptom overlap and high levels of comorbidity within and among anxiety and mood disorders has stimulated theoretical and empirical work focused on the core structure underlying these disorders. Goldberg et al (25) recently proposed a general factor very similar to negative affectivity/neuroticism as the defining feature of an “emo-tional disorders” cluster as part of a proposal for a meta-structure for ICD-11 and DSM-5. Similarly, other investiga-tors (26-29) set forth a higher-order internalizing disorder factor, with two components – “anxious-misery” (depressive disorders and GAD) and “fear” (phobias and panic disor-ders) – to cover these two classes of disorders. Finally, Zin-barg and Barlow (30) suggested a hierarchical structure, with negative affectivity as a higher-order factor, and several lower order factors: social anxiety, generalized dysphoria, fear of fear, agoraphobia, simple fears, and obsessions-compulsions.

Despite the attractiveness of these models, others have questioned the clinical utility of the above clusters (31). Some authors have criticized Krueger’s (27) structural mod-el, which provided the empirical basis for Goldberg et al’s proposed ”emotional disorders” cluster, on methodological grounds (32,33), including the use of top-down diagnostic interviews to derive categories, with no attention to issues of

threshold, developmental stage and age, and the limited number of disorders considered for analysis (33).

A change of this magnitude in the meta-structure of the ICD-11 classification of mood and anxiety disorders is prob-ably not feasible (34). However, a step forward would be the inclusion of a subtype in the classification of depressive epi-sode indicating the presence of anxiety symptoms.

subthreshold presentations

Particularly in primary care settings, patients with sub-threshold disorders but who experience considerable distress are common (35), partly as a consequence of the arbitrary cut-off points of existing polythetic diagnostic systems. In population samples, comorbidity extends along a continuum of severity from co-occurring subsyndromal depressive and anxiety symptoms to the higher end of severity where “diag-nosable” depression and anxiety disorders coexist (8).

The category of mixed anxiety-depression disorder is de-fined in ICD-10 as the presence of both anxiety and depres-sive symptoms not sufficient to justify a diagnosis of either of the two disorders considered separately. In DSM-IV, this disorder is listed in Appendix B as a provisional diagnosis. In DSM-5, the disorder had been proposed for inclusion in the mood disorders section, defined by the presence for at least 2 weeks of three or four depressive symptoms, which must include depressed mood and/or loss of interest or pleasure, accompanied by two or more symptoms of anxious distress, including irrational worry, preoccupation with unpleasant worries, having trouble relaxing, motor tension, or fear that something awful may happen, but this proposal has now been withdrawn.

Mixed anxiety-depression disorder as defined in ICD-10 is common in the general population. In the UK National Sur-vey of Psychiatric Morbidity (8), it was found to be the most prevalent mental disorder (8.8%) and to be associated with a level of distress greater than non-cases and similar to indi-viduals with anxiety disorders. In a prospective cohort study of 250 primary care patients, Walters et al (7) also found high-er distress at 3-month follow-up among those with mixed anxiety-depression disorder as compared to those with no diagnosis, and lower mental health-related quality of life. However, many individuals with mixed anxiety-depression disorder did improve at 1 year, which led the authors to won-der whether the inclusion of this condition as a diagnostic category in DSM-5 could bring to “medicalization of many people who have minor, self-limiting symptoms of distress”. Barkow et al (36), using data from the World Health Organi-zation (WHO) Collaborative Study on Psychological Prob-lems in General Healthcare (PGHC), noted the heterogeneity of presentation and outcome among individuals with mixed anxiety-depression disorder, with the majority remitting at one-year follow-up.

The WPA-WHO global survey of psychiatrists’ attitudes toward mental disorder classification (5) found that mixed

006_010.indd 7 06/07/12 14:04


anxiety-depression disorder was the fourth most frequently used diagnostic category among nearly five thousand psy-chiatrists in 44 countries, even though it was among the cat-egories they rated as being most difficult to use and as having the lowest goodness of fit or accuracy in describing their pa-tients.

Schmidt et al (37) explored whether mixed anxiety-de-pression disorder is a condition distinct from other Axis I conditions as well as from nonclinical levels of anxiety (i.e., a taxon). They confirmed the existence of a taxon with a base rate which is substantially higher than the prevalence rates of mixed anxiety-depression disorder suggested in epidemio-logical reports, and which is associated with an increased risk for incidence of other anxiety and mood disorder diag-noses over time. Based on this research, the authors made a series of recommendations for changes to the criteria set for mixed anxiety-depression disorder in DSM-5.

Given the heterogeneity of mixed anxiety-depression dis-order, the ICD-11 classification should provide explicit guid-ance for diagnosing it. The presence of significant distress and impairment may be required for the diagnosis.

somatic presentations of mood and anxiety disorders

Somatic presentations of mood and anxiety disorders have long been recognized as indicating serious psychopathology but varying considerably across population groups (e.g., women, geriatric populations, certain cultural groups) and settings (e.g., primary care as compared to psychiatric set-tings). Distress associated with physical complaints is closely related to the use of health services (38), thus offering an op-portunity to identify and treat mood and anxiety disorders.

However, little systematic epidemiological effort has been dedicated to this topic. The existing literature has highlighted three key issues: a) a substantial proportion of patients with depressive and anxiety disorders somatize their distress (39); b) many patients with a depressive illness also have a physi-cal illness (40); c) the presentation of a somatic symptom by a patient with a diagnosable depressive illness can be an im-portant determinant of “hidden psychiatric morbidity” in the medical setting, in particular for some cultural groups (41). The somatization of psychological distress is associated with greater psychosocial disability, more visits to clinicians, high-er risk for suicide, poorer treatment response, and worse clinical outcome.

Exploring the psychopathological profile of primary care attendees in 14 countries in the PGHC Study, Simon et al (42) found that the most common somatic presentations of depression and anxiety were musculoskeletal pain and fa-tigue. Of patients who met criteria for a depressive disorder, the proportion who reported only somatic symptoms as the reason for visiting the physician was found to be 69% overall, ranging from 45% in Paris to 95% in Ankara. In a recent reanalysis of these data, Simms et al (43) demonstrated that much of the variance associated with somatic complaints

can be subsumed under the umbrella of an internalizing fac-tor. That is, somatic symptoms without clear physical causes may be better understood as non-specific manifestations of distressing internalizing psychopathology rather than an in-dependent class of disorder (43).

Somatic symptoms are an important component of de-pression and anxiety disorders, though they can be shaped by cultural factors. Current debates on the expression of psy-chiatric syndromes have been focused on the theoretical ra-tionales of the “etic” and “emic” approaches which underline either biological universality or cultural diversity, respective-ly (44). The “etic” approach to mental disorders assumes the validity of Western nosology for the rest of the world. The bulk of epidemiological research conducted worldwide sup-ports this approach. On the other hand, the role played by culture in the experience, expression, diagnosis and manage-ment of mental illness (the “emic” approach) has gained in-creasing attention by scholars of medical anthropology in healthcare, particularly in low-income countries. According to this approach, psychiatric classifications can be viewed as a product of the Western culture. The task of building a “cul-turally and biologically correct” system of classification will require an incorporation of both methods. The trend in in-ternational epidemiology is to regard culture as a key vari-able, particularly when the research is based in non-Western societies.

Discussions of how to integrate cultural variations – par-ticularly in a diagnostic classification intended to be used worldwide – can be informed by the explanatory models of somatic presentations in relation to mood and anxiety disor-ders (45).

There are two problems in ICD-10 in respect to somatic presentation. First, the term “somatic syndrome” in ICD-10 is used as an equivalent for melancholic features, and should be replaced by another descriptive term. Second, as there is no separate coding for prominent somatic complaints as mani-festations of anxiety and depression, we suggest incorporating this pattern of symptoms into diagnostic descriptions.

ConClusIons

Although some investigators (46,47) have called for a ma-jor revision of current classifications systems, there is a reluc-tance to base such changes on inconclusive and inconsistent evidence from studies using different methodologies (31-33).

However, empirical evidence has demonstrated that is pos-sible to develop cross-cutting higher-order dimensional rat-ings relevant to mood and anxiety disorders without making dramatic changes in the current classification of mental dis-orders.

Watson and Clark (46) have proposed a reorganization of diagnostic classes by adopting a fully dimensional taxonomy, identifying the basic symptom dimensions underlying cur-rent mood and anxiety disorders and then organizing them into a fully quantitative hierarchical model. For Krueger and

006_010.indd 8 06/07/12 14:04


Bezdjian (11), both categorical and dimensional concepts are important and the most effective classification should have both. Symptoms, severity, and impairment measures could be incorporated in categorical diagnosis. More recently, King-don et al (48) also proposed a reorganization of disorders into broad categories with clinically relevant specifiers. They proposed that anxiety and mood disorders should be placed together under a broader “emotional disorders” category or as “internalizing disorders”. They suggested that, for GAD, a personality specifier might be particularly helpful (49), e.g., emotional disorder plus avoidant personality. Grouping so-matic presentations together could improve clarity through the use of a specifier to denote cultural aspects.

Based on a review of the available epidemiological data, the most important ways for the ICD-11 to improve the clas-sification of depression and anxiety disorders might be to model comorbidity, to improve the description of mixed anx-iety-depression disorder, and to identify the core symptoms of emotional disorders in order to achieve a better understand-ing of the significance of somatic presentations. As different users of nosology may have conflicting views about what con-stitutes progress in psychiatric classification based on their divergent needs, simple criteria and potential specifiers or subtypes could be included as options in order to meet the needs of both practitioners and academic researchers.

It is important to consider that the majority of epidemio-logical surveys rely on top-down diagnostic interviews with a priori assumption based on distinct Western categories, which fail to account for contextual features, qualitative characteristics of anxiety and mood disorders (50), and cul-tural variations in the presentation of these disorders. The challenge of reformulating the current classification in order to enhance clinical utility should not cast aside basic concep-tualizations of psychopathology.

note

L.H. Andrade is a member of the WHO ICD Revision Working Group on the Classification of Mood and Anxiety Disorders, reporting to the International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders. The views expressed in this article are those of the authors and, except as specifically noted, do not represent the official policies or positions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.

The DSM-5 website was accessed on December 5, 2011. DSM-5 proposals have not been finalized and are subject to change.

references

1. Andrews G, Stewart G, Morris-Yates A et al. Evidence for a general neurotic syndrome. Br J Psychiatry 1990;157:6-12.

2. Andrade L, Walters EE, Gentil V et al. Prevalence of ICD-10 mental

disorders in a catchment area in the city of São Paulo, Brazil. Soc Psychiatry Psychiatr Epidemiol 2002;37:316-25.

3. Demyttenaere K, Bruffaerts R, Posada-Villa J et al. Prevalence, sever-ity, and unmet need for treatment of mental disorders in the World Health Organization World Mental Health Surveys. JAMA 2004; 291:2581-90.

4. Gater R, Tansella M, Korten A et al. Sex differences in the prevalence and detection of depressive and anxiety disorders in general health care settings: report from the World Health Organization Collabora-tive Study on Psychological Problems in General Health Care. Arch Gen Psychiatry 1998;55:405-13.

5. Reed GM, Mendonça Correia J, Esparza P et al. The WPA-WHO Global Survey of Psychiatrists’ Attitudes towards Mental Disorders Classification. World Psychiatry 2011;10:118-31.

6. Sherbourne CD, Wells KB, Hays RD et al. Subthreshold depression and depressive disorder: clinical characteristics of general medical and mental health specialty outpatients. Am J Psychiatry 1994;151: 1777-84.

7. Walters K, Buszewicz M, Weich S et al. Mixed anxiety and depres-sive disorder outcomes: prospective cohort study in primary care. Br J Psychiatry 2011;198:472-8.

8. Das-Munshi J, Goldberg D, Bebbington PE et al. Public health sig-nificance of mixed anxiety and depression: beyond current classifi-cation. Br J Psychiatry 2008;192:171-7.

9. Rucci P, Gherardi S, Tansella M et al. Subthreshold psychiatric dis-orders in primary care: prevalence and associated characteristics. J Affect Disord 2003;76:171-81.

10. Kirmayer LJ, Robbins JM, Dworkind M et al. Somatization and the recognition of depression and anxiety in primary care. Am J Psy-chiatry 1993;150:734-41.

11. Krueger RF, Bezdjian S. Enhancing research and treatment of men-tal disorders with dimensional concepts: toward DSM-V and ICD-11. World Psychiatry 2009;8:3-6.

12. Klerman GL. Paradigm shifts in USA psychiatric epidemiology since World War II. Soc Psychiatry Psychiatr Epidemiol 1990;25:27-32.

13. Boyd JH, Burke JD, Jr, Gruenberg E et al. Exclusion criteria of DSM-III. A study of co-occurrence of hierarchy-free syndromes. Arch Gen Psychiatry 1984;41:983-9.

14. Merikangas KR, Angst J, Eaton W et al. Comorbidity and boundaries of affective disorders with anxiety disorders and substance misuse: results of an international task force. Br J Psychiatry 1996;168(Sup-pl. 30):58-67.

15. Andrade L, Caraveo-Anduaga JJ, Berglund P et al. The epidemiology of major depressive episodes: results from the International Consor-tium of Psychiatric Epidemiology (ICPE) Surveys. Int J Methods Psychiatr Res 2003;12:3-21.

16. Watson D, Clark LA. Negative affectivity: the disposition to experi-ence aversive emotional states. Psychol Bull 1984;96:465-90.

17. Clark LA, Watson D. Tripartite model of anxiety and depression: psychometric evidence and taxonomic implications. J Abnorm Psy-chol 1991;100:316-36.

18. Hettema JM, Neale MC, Myers JM et al. A population-based twin study of the relationship between neuroticism and internalizing dis-orders. Am J Psychiatry 2006;163:857-64.

19. Andrade L, Eaton W, Chilcoat H. Lifetime co-morbidity of panic attacks and major depression in a population-based study: age of onset. Psychol Med 1996;26:991-6.

20. Andrade L, Eaton W, Chilcoat H. Lifetime comorbidity of panic at-tacks and major depression in a population-based study: symptom profiles. Br J Psychiatry 1994;165:363-9.

21. Phillips AC, Batty GD, Gale CR et al. Generalized anxiety disorder, major depressive disorder, and their comorbidity as predictors of all-cause and cardiovascular mortality: the Vietnam experience study. Psychosom Med 2009;71:395-403.

22. Scott KM, Bruffaerts R, Tsang A et al. Depression-anxiety relation-ships with chronic physical conditions: results from the World Men-tal Health Surveys. J Affect Disord 2007;103:113-20.

006_010.indd 9 06/07/12 14:04


23. Kessler RC, Chiu WT, Demler O et al. Prevalence, severity, and co-morbidity of 12-month DSM-IV disorders in the National Comor-bidity Survey Replication. Arch Gen Psychiatry 2005;62:617-27.

24. Viana MC, Teixeira MG, Beraldi F et al. São Paulo Megacity Mental Health Survey – a population-based epidemiological study of psy-chiatric morbidity in the São Paulo metropolitan area: aims, design and field implementation. Rev Bras Psiquiatr 2009;31:375-86.

25. Goldberg DP, Krueger RF, Andrews G et al. Emotional disorders: cluster 4 of the proposed meta-structure for DSM-V and ICD-11. Psychol Med 2009;39:2043-59.

26. Krueger RF, Caspi A, Moffitt TE et al. The structure and stability of common mental disorders (DSM-III-R): a longitudinal-epidemio-logical study. J Abnorm Psychol 1998;107:216-27.

27. Krueger RF. The structure of common mental disorders. Arch Gen Psychiatry 1999;56:921-6.

28. Vollebergh WA, Iedema J, Bijl RV et al. The structure and stability of common mental disorders: the NEMESIS study. Arch Gen Psychia-try 2001;58:597-603.

29. Slade T, Watson D. The structure of common DSM-IV and ICD-10 mental disorders in the Australian general population. Psychol Med 2006;36:1593-600.

30. Zinbarg RE, Barlow DH. Structure of anxiety and the anxiety disor-ders: a hierarchical model. J Abnorm Psychol 1996;105:181-93.

31. First MB. Reorganizing the diagnostic groupings in DSM-V and ICD-11: a cost/benefit analysis. Psychol Med 2009;39:2091-7.

32. Kessler RC, Cox BJ, Green JG et al. The effects of latent variables in the development of comorbidity among common mental disorders. Depress Anxiety 2011;28:29-39.

33. Wittchen HU, Beesdo K, Gloster AT. The position of anxiety disor-ders in structural models of mental disorders. Psychiatr Clin North Am 2009;32:465-81.


35. Kroenke K, Mangelsdorff AD. Common symptoms in ambulatory care: incidence, evaluation, therapy, and outcome. Am J Med 1989; 86:262-6.

36. Barkow K, Heun R, Wittchen HU et al. Mixed anxiety-depression in

a 1 year follow-up study: shift to other diagnoses or remission? J Affect Disord 2004;79:235-9.

37. Schmidt NB, Kotov R, Bernstein A et al. Mixed anxiety depression: taxometric exploration of the validity of a diagnostic category in youth. J Affect Disord 2007;98:83-9.

38. Meltzer H, Gill H, Pattierew M et al. Physical complaints, service use and treatment of adults with psychiatric disorders. London: HMSO, 1995.

39. Haug TT, Mykletun A, Dahl AA. The association between anxiety, depression, and somatic symptoms in a large population: the HUNT-II study. Psychosom Med 2004;66:845-51.

40. Stordal E, Bjelland I, Dahl AA et al. Anxiety and depression in indi-viduals with somatic health problems. The Nord-Trøndelag Health Study (HUNT). Scand J Prim Health Care 2003;21:136-41.

41. Kirmayer LJ. Cultural variations in the clinical presentation of de-pression and anxiety: implications for diagnosis and treatment. J Clin Psychiatry 2001;62(Suppl. 13):22-8.

42. Simon G, Gater R, Kisely S et al. Somatic symptoms of distress: an international primary care study. Psychosom Med 1996;58:481-8.

43. Simms LJ, Prisciandaro JJ, Krueger RF et al. The structure of depres-sion, anxiety and somatic symptoms in primary care. Psychol Med 2011;20:1-14.

44. Patel V. Cultural factors and international epidemiology. Br Med Bull 2001;57:33-45.

45. Patel V, Gwanzura F, Simunyu E et al. The phenomenology and explanatory models of common mental disorder: a study in primary care in Harare, Zimbabwe. Psychol Med 1995;25:1191-9.

46. Watson D. Rethinking the mood and anxiety disorders: a quantita-tive hierarchical model for DSM-V. J Abnorm Psychol 2005;114: 522-36.

47. Widiger TA. A dimensional model of psychopathology. Psychopa-thology 2005;38:211-4.

48. Kingdon D, Afghan S, Arnold R et al. A diagnostic system using broad categories with clinically relevant specifiers: lessons for ICD-11. Int J Soc Psychiatry 2010;56:326-35.

49. Tyrer P, Baldwin D. Generalised anxiety disorder. Lancet 2006;368: 2156-66.

50. Maj M. When does depression become a mental disorder? Br J Psy-chiatry 2011;199:85-6.

006_010.indd 10 06/07/12 14:04

11

Specifiers as aids to treatment selection and clinical management in the ICD classification of mood disorders

A major purpose of diagnostic manuals such as the ICD-10 and the DSM-IV is to guide clinicians in accurately iden-tifying patients’ conditions so that more effective treatment decisions can be made (1). Diagnostic definitions that are easy for clinicians to apply in multiple settings, and that have the backing of sound empirical research, have the potential to bring us closer to the “personalized medicine” approach now favored in psychiatry: the tailoring of treatments vali-dated in randomized trials to the individual needs of patients (2). Much needs to be done in the current revision of the ICD-10 to maximize the clinical utility of existing diagnoses to better predict treatment response.

The purpose of this paper is to review a selected set of clinical utility problems relevant to the mood disorder sec-tions of ICD-10. We focus on three representative but non-exhaustive issues: the importance of an illness severity di-mension in depressive disorders and bipolar disorder, the use of a mixed episode specifier (depressive episode with simul-taneous hypomanic or manic features, or manic/hypomanic episodes with simultaneous depressive features) and the in-clusion of a childhood-onset specifier, along with greater discussion of developmental presentations, for bipolar disor-der and depressive disorders.

There are arguments for including more specifiers in the ICD-11 chapter on mood disorders, or reorganizing the top-level categories (see examples below) to allow for distinc-tions among illness or course subtypes. We refer to the paper by First (3) in this supplement for a more detailed discussion of the challenges posed by the ICD-11 structure on the use of specifiers and subcategories.

DaviD J. Miklowitz1, Michael B. First2

1Division of Child and Adolescent Psychiatry, University of California, Los Angeles (UCLA) Semel Institute for Neuroscience and Behavior, David Geffen School of Medicine at UCLA, Los Angeles, CA; 2Department of Psychiatry, Columbia University and Department of Clinical Phenomenology, New York State Psychiatric Institute, New York, NY, USA

The World Health Organization (WHO) has suggested that changes to the ICD-10 should be based in large part on clinical utility: the degree to which they help clinicians communicate with patients, family members and providers, make differential diagnoses, formulate current treat-ment plans, and predict future treatment needs. The clinical utility of the existing mood disorder diagnoses, however, is compromised by their heterogeneity. The DSM classification has addressed this heterogeneity by the provision of different types of “specifiers” that can be used to describe various aspects of the disorder, from symptom pattern of the current episode (e.g., melancholic features) to longitudinal course (e.g., seasonal pattern). Integration of certain illness qualifiers or subcategories into ICD-11 may increase the clinical utility of the mood disorder classifications. In this paper, we focus on three particular specifiers that may increase the clinical utility of the depressive disorder and bipolar disorder categories: a well-operationalized severity dimension, a broad mixed episode phenotype, and the specification of childhood illness onset. Clinical vignettes are used to illustrate the types of patients whose presentations would be better classified with the addition of diag-nostic specifiers.

Key words: Depressive disorder, depression, bipolar disorder, course specifiers, childhood-onset, mixed episodes, age at onset


What IS ClInICal utIlIty?

A diagnostic definition, subtype, or course specifier has clinical utility when it aids the clinician or researcher in the following: a) conceptualizing the diagnostic entity; b) com-municating information to patients, caregivers, and providers; c) using the definition in practice to aid in differential diagno-sis; d) choosing effective interventions, and e) predicting the patient’s course of illness and future treatment needs (3-5).

Clinical utility is a concept that is distinct from diagnostic reliability (the degree to which two clinicians arrive at the same diagnosis based on the same information) or validity (the relationship of the category to criterion variables such as specific neurobiological markers or responses to treatments). Nonetheless, significant limitations in diagnostic reliability and/or validity will compromise clinical utility. For example, if a diagnostic definition is so grossly unreliable that two cli-nicians are not able to agree on the same diagnosis when evaluating the same patient, then the goal of effectively com-municating clinical information is compromised. Similarly, if a diagnosis is lacking in predictive validity (a feature that overlaps to some degree with clinical utility (6)), then a clini-cian’s ability to communicate information to patients and families about the patient’s prognosis and likely treatment response is greatly hindered.

The categories in ICD-10 and DSM-IV vary in their clini-cal utility. For example, the diagnosis of bipolar I disorder has high clinical utility in that its application usually leads to specific treatment recommendations (mood stabilizers and/or atypical antipsychotic medications). In contrast, certain diagnostic definitions, such as anorexia nervosa and bulimia nervosa, have lower clinical utility by virtue of their leaving a large percentage of patients without a specific diagnosis (7).

011_016.indd 11 06/07/12 14:06


Moreover, many of the DSM criteria sets are overly complex and virtually impossible to remember, compromising their usability in clinical settings, and consequently the validity of clinical data that are used to make health care decisions.

Among the diagnostic areas with problematic clinical util-ity in ICD-10 are the severity rating criteria for depressive disorder (mild, moderate, severe); the lack of inclusion of subthreshold mixed episodes, which can occur in depressive disorders or bipolar disorder; and the lack of consideration of developmentally-specific presentations of mood disorder. In this paper, we focus on the addition of diagnostic specifi-ers that may increase the clinical utility of the mood disorder diagnoses.

the uSe of SpeCIfIerS

Many DSM-IV and ICD-10 categories are heterogeneous: patients assigned the same diagnostic label can have signifi-cant variability in their clinical presentations. A majority of the criteria sets in the DSM are polythetic, meaning that in-stead of requiring a particular set of clinical features, they re-quire a minimum number of items drawn from a much larger set. While providing a certain degree of real-life flexibility, these criteria sets also result in potentially thousands of differ-ent symptom combinations, any of which would justify mak-ing the diagnosis. As an example, the diagnosis of borderline personality disorder can be made from any one of 93 combi-nations of the 9 symptoms in its criterion set. Moreover, it is now understood that some DSM and ICD categories describe patients with a widely diverse set of underlying pathophysi-ological mechanisms. This diversity in mechanisms may help explain the wide array of treatments that, in aggregate, can be shown to be effective in treating a person with a particular diagnosis.

To reduce this diagnostic heterogeneity, a strategy adopted by the DSM system is to provide specifiers that identify more homogeneous subsets of patients. Given the broad defini-tions of mood disorders and the fact there are essentially only four primary mood disorders (i.e., depressive disorder, bipolar disorder, dysthymic disorder, and cyclothymic disor-der), such specifiers, often with elaborate criteria sets of their own, have been included since DSM-III.

Most of the specifiers appearing in the mood disorders sec-tion of DSM-IV (which are being carried over to DSM-5) indicate subgroups that are more likely to respond to particu-lar interventions. For example, the “melancholic features” specifier is used to select patients who are thought to be more likely to respond to antidepressant medications or electro-convulsive therapy than those without melancholia (8). The “seasonal pattern” specifier may predict response to light therapy. The “rapid cycling” specifier for bipolar disorder sug-gests that the patient will not respond well to selective sero-tonin reuptake inhibitors or tricyclic antidepressants (9).

In ICD-10, only the melancholia subtype of mild and moderate depressive episode (called “somatic syndrome” in

ICD-10) is included. One complication is that, unlike DSM, the ICD requires that all specifiers be assigned unique codes, which creates added complexity and limits the number of specifiers that practically can be included (2). Another op-tion would be to reorganize the top-level categories – such as seasonal affective disorder – to have a depressive and a bipo-lar subtype, as opposed to having a seasonal pattern specifier that modifies both recurrent depressive disorders and bipolar disorder.

example 1:DImenSIonal SeverIty ratIngS for DepreSSIon

Both ICD-10 and DSM-IV recognize the importance of including dimensional ratings of severity for depressive epi-sodes in depressive disorder and for manic and depressive episodes in bipolar disorder. Severity ratings may help to pre-dict treatment response in both disorders. In a meta-analysis of six randomized pharmacotherapy trials, Fournier et al (10) found that, among patients with Hamilton Rating Scale for Depression (HRSD, 11) scores of 25 or higher prior to treat-ment, there was a greater separation at follow-up between antidepressant treatment and placebo than among those scoring below 25. Dimensional severity ratings may also be useful for predicting response to treatments in bipolar de-pression. In a meta-analysis of lamotrigine trials, Geddes et al (12) found that effect sizes relative to placebo or other treatments were consistently higher when the baseline HRSD score was 24 or above.

The current ICD-10 classification of depressed patients as having mild, moderate or severe depression takes a step to-ward identifying subgroups of patients who are more and less likely to respond to antidepressant treatment. It is not clear, however, whether thresholds of severity that partition de-pressed patients into groups are supported by epidemiologi-cal studies (13). When quality of life is the criterion variable, clear distinctions emerge between patients with ICD-10 de-pressive disorder and healthy controls, but few differences emerge between patients with ICD-10 mild, moderate, and severe depression (14).

Second, the Clinical Descriptions and Diagnostic Guide-lines contain specific descriptions of patients with mild, moderate, or severe depression, but it is not clear that physi-cians use these descriptions in practice. Further, physicians may differ in their interpretation of these descriptions de-pending on their comparison populations. For example, phy-sicians working in inpatient settings with extremely ill pa-tients may classify a patient as having moderate depression if he or she is able to live independently after discharge. In contrast, a clinician in a private practice serving uncompli-cated, “worried well” patients may view severe depression as being anything that does not respond quickly to antidepres-sants. In that sense, this tripartite division may not satisfy the communication criterion of the clinical utility definition.

Consider the following vignette:

011_016.indd 12 06/07/12 14:06


Two patients, Angus and Sunil, were admitted to separate inpatient units with depressive disorders. Angus was ad-mitted to a private hospital with a small inpatient unit. Sunil was admitted to an inner-city public hospital with a large number of patients and a broad range of illness pre-sentations. Both were classified as having a severe depres-sive episode by ICD-10, and antidepressant treatment was recommended for both. During their inpatient stays, both patients were evaluated for a research project concerning treatments for depression, which included a HRSD score as well as a categorical diagnosis. At hospital admission, Angus’s score was 23 whereas Sunil’s was 44. Over the 4 weeks after hospital discharge, Angus showed a 20% im-provement in his HRSD score, whereas Sunil showed a 50% improvement. Despite their equivalent severity clas-sification, dimensional severity ratings proved to be impor-tant in predicting the degree of treatment response shown by these two patients after hospital discharge.

Most of the depression rating scales provide cutoffs for des-ignating depression as mild, moderate or severe. The HRSD, which is familiar to many clinicians, defines high severity as a score of 25 or above. The Quick Inventory of Depressive Symptoms, Clinician- or Patient-Rated Versions (QIDS, 15) can be used to distinguish between subsyndromal depression (scores of 13-27) and syndromal depression (scores of 28 or higher). The Beck Depression Inventory–II uses scores of 14–19 to indicate mild depression, 20–28 for moderate de-pression, and 29 or above for severe depression (16).

The DSM-5 Mood Disorders Work Group has proposed use of the Patient Health Questionnaire-9 (PHQ-9, 17) as a severity specifier (see www.dsm5.org). The PHQ-9 has nine items corresponding to the DSM-IV individual criteria for major depressive disorder. However, recent research has questioned its utility as a severity measure as opposed to a screening tool (18,19).

Although it may not be practicable to include specific de-pression rating scales in the ICD-11, descriptions of patients who score in the moderate or high range on these scales may inform the operational definitions of depression severity in the Clinical Descriptions and Diagnostic Guidelines, which may have the effect of increasing the reliability of the severity designations across practice settings. Although questions about the cross-setting and cross-cultural use of such scales will be raised, their supplemental use may nonetheless im-prove clinical utility.

Estimating the number and severity of comorbid disorders or symptoms may also be useful to assess severity. A 1-year follow-up of 1000 outpatients with bipolar disorder found that patients with comorbid anxiety disorders (32%) had fewer days well, less likelihood of recovery from depression, earlier relapse, lower quality of life and more impairment of role function compared to those without anxiety disorders. Furthermore, impairment at follow-up increased with the number of comorbid anxiety disorders diagnosed at baseline (20). Patients with complex and comorbid presentations in

depressive disorder or bipolar disorder may be candidates for more aggressive pharmacological or psychosocial treatments.

example 2:BroaDer DefInItIonS of mIxeD affeCtIve StateS

ICD-10 and DSM-IV recognize that syndromal manic and depressive episodes can coexist in the same person during the same week. However, their definitions appear to be too restrictive and result in poor diagnostic coverage in clinical settings. Clinicians often ignore the requirement that epi-sodes of both poles must be syndromal and instead include patients with a variety of subthreshold depressive or manic symptoms (21).

The DSM-5 Mood Disorders Work Group is considering a broad mixed features specifier that would include patients with major depressive, hypomanic or manic episodes with subthreshold symptoms of the opposite polarity (see www.dsm5.org). In this proposal, patients with mania or hypoma-nia would have a mixed features specifier if they have had three depressive symptoms nearly every day during the (hy-po)manic episode. Likewise, predominantly depressed pa-tients, even if they carry the diagnosis of major depressive disorder, would have a mixed features specifier if they have had three manic symptoms during their depressive episode. These criteria would exclude the use of overlapping symp-toms (i.e., distractibility, irritability, insomnia or hypersom-nia, or indecisiveness) in tabulating evidence for the opposite mood pole.

Evidence supporting this recommendation comes from studies indicating that patients with major depression plus one or two concomitant manic symptoms have more epi-sodes of depression and hypomania, increased treatment uti-lization, more lifetime suicide attempts, increased substance abuse, and an earlier age at onset of mood disorder (22-25). Patients with bipolar disorder who have manic symptoms during depressive episodes often show a poor response to lithium (26) and may become destabilized on antidepressants (27). The mixed hypomania state is relatively common: of 1044 psychiatric visits among bipolar patients with hypoman-ic symptoms in the Stanley Foundation Network study, 57% met criteria for mixed hypomania; these patients were more frequently women than men (28). Finally, the National Insti-tute of Mental Health Collaborative Study found that patients with subthreshold manic symptoms accompanying a major depressive episode were more likely to have a family history of bipolar disorder and were more likely to develop bipolar disorder over an 18-year follow-up than major depressive dis-order patients without manic symptoms (29).


Reese, a 43 year-old man with major depression, typically woke late and slept for most of the day, complaining that he had little energy to do anything. However, his energy level increased at night when he attempted to go back to

011_016.indd 13 06/07/12 14:06


sleep. His thoughts, which were often ruminative, sped up at those times as he fantasized about elaborate and dra-matic ways of committing suicide. At these times he felt angry and irritable, and particularly resentful of his wife who seemed to have no trouble sleeping. He eventually fell asleep after 3 am, and the cycle began again the next day.

This patient would not fulfill the DSM-IV diagnostic crite-ria for a mixed episode or for bipolar II disorder because: a) he had never had a syndromal manic or mixed episode, and b) he did not have clear episodes of hypomania between his depressive episodes. Nonetheless, he was at high risk for sui-cide and did not respond well to antidepressants. He eventu-ally responded to a combination of divalproex and quetiap-ine, an atypical antipsychotic medication, even though he retained the major depressive disorder diagnosis.

Thus, including a specifier or subtype that alerts clinicians to symptoms of the opposite pole may help specify the condi-tions under which depressed patients develop bipolar disor-der or respond poorly to antidepressants. It would suggest closer monitoring of suicide risk, increased risk for substance abuse or other impulse-driven behaviors, and a more refrac-tory course of illness. It may eventually inform specific treat-ment decisions (for example, the choice between lithium, anticonvulsants or specific atypical antipsychotics). Finally, a mixed specifier or subtype would allow practitioners to ac-knowledge mixed presentations of depressive disorder with-out having to apply the label of bipolar illness.

example 3:age at IllneSS onSet

Depressive disorders have long been known to occur in childhood or early adolescence, affecting as many as 9% of all adolescents (30). Furthermore, there is evidence from multiple countries that childhood-onset bipolar disorder oc-curs in about 1.9% of children in clinical samples, even when strict criteria are used (31).

Significant controversies exist about the definition, ascer-tainment, and boundaries of early-onset bipolar disorder (32,33). Early-onset bipolar disorder may be characterized by shorter episodes, more polarity switches, more mixed presen-tations, and a greater incidence of psychosis than the adult-onset disorder (34,35). Among offspring of parents with bipo-lar disorder I or II, subthreshold forms of the disorder can be detected as much as 10 years prior to onset of full bipolar disorder (36,37). Agreement is substantial that early-onset forms of both major depressive disorder and bipolar disorder have a significant impact on functionality (e.g., school failure) and quality of life (30,32). Early diagnosis and intervention may lead to a reduction in suicide rates, a decrease in sub-stance abuse disorders, and a decrease in school dropout, even among youth who do not go on to develop bipolar dis-order or major depressive disorder (38,39).

Developmental presentations of mood disorders are not

well covered by ICD-10 or DSM-IV, and there is currently no way to specify that an adult or an adolescent had a childhood onset. At minimum, the “associated features” discussion in ICD-11 should take greater account of the different develop-mental presentations of childhood-onset mood disorders, even if no new categories or specifiers are added. The Clini-cal Descriptions sections should include more developmen-tally-oriented discussion to accompany the diagnostic crite-ria and case examples that illustrate the differences between teens and adults in symptom presentation (for example, de-pressed mood may present as boredom or irritability in ado-lescents).

A specifier or subtype code for age at onset (i.e., child-hood, adolescence, adulthood) of first mood symptoms may help distinguish a group of patients with distinctive course features and treatment responses and thus may improve the clinical utility of the ICD-11 mood categories. For example, an adult with depressive disorder who had episodic periods of intense irritability with grandiosity and decreased need for sleep as a youngster may be at risk for manic onset (40). Thus, such patients may not be good candidates for antidepressant monotherapy. Adolescents or adults with attention-deficit/hyperactivity disorder (ADHD) are often indiscriminately treated with psychostimulants, which may be a poor choice among patients who have a history of depression in child-hood or adolescence. Adult patients with generalized anxiety disorder who have a history of manic or mixed symptoms in adolescence may become activated by antidepressants. An adult with bipolar disorder who has had a lengthy history of mood episodes dating back to childhood may be less likely to respond to cognitive-behavioral therapy (41).


Rosie, a 19 year-old female, was diagnosed with ADHD and “borderline personality features”. The latter was based on her history of impulsive sexual encounters, rageful re-sponses to her mother and father, rapidly shifting interper-sonal attachments, poor peer functioning, and self-destruc-tive behavior. A thorough diagnostic assessment revealed that Rosie had developed a major depressive episode at age 14 following the breakup of her first romantic relationship. During that time she began self-cutting with glass and ra-zors, smoking tobacco and marijuana, and being truant from school. Her ADHD medication was increased to little effect, with the exception that her sleep became more dis-turbed. Her mother reported numerous short periods of hy-pomania in her mid- to late-teens, each lasting 1-2 days, in which she had elevated mood, increased energy, increased sexual drive, and decreased need for sleep.

Does Rosie have a borderline personality disorder? Is her impulsive behavior consistent with ADHD? Or is her clinical presentation more consistent with bipolar disorder, or per-haps major depressive disorder with mixed features? The clinician reasoned that her early onset of mood symptoms might affect her treatment response and decided against an

011_016.indd 14 06/07/12 14:06


antidepressant. Rosie was eventually stabilized on a combi-nation of an atypical antipsychotic, divalproex, and a low-dose psychostimulant.

Thus, age at onset (childhood, adolescence, or adulthood) may define a subgroup of patients with a more refractory and complex course of illness, and thus should be considered as an additional specifier or subtype in ICD-11. As discussed earlier, adding a mixed features specifier may further help clinicians to classify a large group of early-onset youth who have depressive symptoms with one or more symptoms of mania or hypomania.

ConCluSIonS

The clinical utility of the mood categories in ICD-11 may be improved by considering illness specifiers that describe subgroups of patients with depressive disorder or bipolar dis-order who may be more or less responsive to certain treat-ments. In this paper, we have focused on three specifiers. First, the addition of a well-operationalized depression se-verity dimension (informed by dimensional rating scales such as the HRSD or QIDS) may increase the reliability with which clinicians identify patients with depression or bipolar disorder who are most likely to respond to antidepressants or other agents (e.g., lamotrigine). Second, including a mixed features specifier may inform the prognosis of patients with depressive disorder – notably, the increased likelihood of manic or hypomanic switches and eventual conversion to bipolar disorder I or II – and assist in selecting patients for mood stabilizer or atypical antipsychotic treatments.

Third, ICD-11 should consider including an age at onset specifier that would identify patients with a childhood or adolescent onset of syndromal or subsyndromal mood symp-toms. These patients are at risk for a variety of negative out-comes in adulthood. Among patients with first onsets of de-pression or bipolar disorder, knowing that symptoms have had a childhood onset may signal the need for more intensive early intervention before the illness becomes highly recurrent. At minimum, ICD-11 should take greater account of the dif-ferent developmental presentations of childhood-onset mood disorders, even if no new categories or specifiers are added.

In addition to increasing clinical utility, these changes may generate hypotheses for the next generation of research on treatment selection for patients with mood disorders.

acknowledgements

The preparation of this article was supported by National Institute of Mental Health grants MH077856 and MH073871, and a research award from the Brain and Behavior Research Foundation (Dr. Miklowitz).

note

D.J. Miklowitz is a member of the WHO ICD Revision Working Group on the Classification of Mood and Anxiety Disorders, reporting to the International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders. M.B. First is a consultant to the WHO Department of Mental Health and Substance Abuse and to the International Advi-sory Group for the Revision of ICD-10 Mental and Behav-ioural Disorders. The views expressed in this article are those of the authors and, except as specifically noted, do not rep-resent the official policies or positions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.

The DSM-5 website was accessed on November 10, 2011. DSM-5 proposals have not been finalized and are subject to change.

references


2. Insel TR. Translating scientific opportunity into public health im-pact: a strategic plan for research on mental illness. Arch Gen Psy-chiatry 2009;66:128-33.

3. First MB. Challenges in the implementation of diagnostic specifiers for mood disorders in ICD-11. World Psychiatry 2012;11(Suppl. 1): 17-25.

4. First MB, Pincus HA, Levine JB et al. Clinical utility as a criterion for revising psychiatric diagnoses. Am J Psychiatry 2004;161:946-54.

5. First MB. Clinical utility in the revision of the Diagnostic and Statis-tical Manual of Mental Disorders (DSM). Prof Psychol Res Pr 2010; 41:465-73.

6. Mullins-Sweatt SN, Widiger TA. Clinical utility and DSM-V. Psychol Assess 2009;21:302-12.

7. Fairburn C, Cooper Z, Bahn K et al. The severity and status of eating disorder NOS: implications for DSM-V. Behav Res Ther 2007;45: 1705-15.

8. Thase ME, Sachs GS. Bipolar depression: pharmacotherapy and related therapeutic strategies. Biol Psychiatry 2000;48:558-72.

9. Schneck CD, Miklowitz DJ, Miyahara S et al. The prospective course of rapid cycling bipolar disorder. Am J Psychiatry 2008;165: 370-7.

10. Fournier JC, DeRubeis RJ, Hollon SD et al. Antidepressant drug effects and depression severity: a patient-level meta-analysis. JAMA 2010;303:47-53.

11. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psy-chiatry 1960;12:56-62.

12. Geddes JR, Calabrese JR, Goodwin GM. Lamotrigine for treatment of bipolar depression: independent meta-analysis and meta-regres-sion of individual patient data from five randomised trials. Br J Psy-chiatry 2009;194:4-9.

13. Ayuso-Mateos JL, Nuevo R, Verdes E et al. From depressive symp-toms to depressive disorders: the relevance of thresholds. Br J Psy-chiatry 2010;196:365-71.

14. Nuevo R, Leighton C, Dunn G et al. Impact of severity and type of depression on quality of life in cases identified in the community. Psychol Med 2010;40:2069-77.

15. Rush AJ. Handbook of psychiatric measures. Washington: Ameri-can Psychiatric Publishing, 2000.

011_016.indd 15 06/07/12 14:06


16. Beck AT, Steer RA, Brown GK. Beck Depression Inventory - II. San Antonio: The Psychological Corporation, 1996.

17. Kroenke K, Spitzer RL, Williams JB et al. The Patient Health Ques-tionnaire somatic, anxiety, and depressive symptom scales: a system-atic review. Gen Hosp Psychiatry 2010;32:345-59.

18. Cameron IM, Cardy A, Crawford JR et al. Measuring depression severity in general practice: discriminatory performance of the PHQ-9, HADS-D, and BDI-II. Br J Gen Practice 2011;61:e419-26.

19. Reynolds WM. The PHQ-9 works well as a screening but not diag-nostic instrument for depressive disorder. Evid Based Ment Health 2010;13:96.

20. Otto MW, Simon NM, Wisniewski SR et al. Prospective 12-month course of bipolar disorder in outpatients with and without comorbid anxiety disorders. Br J Psychiatry 2006;189:20-5.

21. Frank E. Proposed revisions to the concept of mixed episodes in DSM-5: the path traveled. Bipol Disord 2011;13:14.

22. Agosti V, Stewart JW. Hypomania with and without dysphoria: com-parison of comorbidity and clinical characteristics of respondents from a national community sample. J Affect Disord 2008;108:177-82.

23. Goldberg JF, Perlis RH, Bowden CL et al. Manic symptoms during depressive episodes in 1,380 patients with bipolar disorder: findings from the STEP-BD. Am J Psychiatry 2009;166:173-81.

24. Sato T, Bottlender R, Schröter A et al. Frequency of manic symptoms during a depressive episode and unipolar ‘depressive mixed state’ as bipolar spectrum. Acta Psychiatr Scand 2003;107:268-74.

25. Swann AC, Moeller FG, Steinberg JL et al. Manic symptoms and impulsivity during bipolar depressive episodes. Bipolar Disord 2007; 93:206-12.

26. Swann AC, Bowden CL, Morris D et al. Depression during mania: treatment response to lithium or divalproex. Arch Gen Psychiatry 1997;54:37-42.

27. Goldberg JF, Perlis RH, Ghaemi SN et al. Adjunctive antidepressant use and symptomatic recovery among bipolar depressed patients with concomitant manic symptoms: findings from the STEP-BD. Am J Psychiatry 2007;164:1348-55.

28. Suppes T, Mintz J, McElroy SL et al. Mixed hypomania in 908 pa-tients with bipolar disorder evaluated prospectively in the Stanley Foundation Bipolar Treatment Network: a sex-specific phenome-

non. Arch Gen Psychiatry 2005;62:1089-96.29. Fiedorowicz JG, Endicott J, Leon AC et al. Subthreshold hypoman-

ic symptoms in progression from unipolar major depression to bipo-lar disorder. Am J Psychiatry 2011;168:40-8.

30. Cicchetti D, Toth SL. The development of depression in children and adolescents. Am Psychol 1998;53:221-41.

31. Van Meter AR, Moreira AL, Youngstrom EA. Meta-analysis of epi-demiologic studies of pediatric bipolar disorder. J Clin Psychiatry 2011;72:1250-6.

32. Carlson GA. Broadening bipolar disorder – by design or by acci-dent? World Psychiatry 2011;10:195-6.

33. Youngstrom E, Van Meter A, Algorta GP. The bipolar spectrum: myth or reality? Curr Psychiatr Rep 2010;12:479-89.

34. Birmaher B, Axelson D, Goldstein B et al. Four-year longitudinal course of children and adolescents with bipolar spectrum disorders: the Course and Outcome of Bipolar Youth (COBY) study. Am J Psychiatry 2009;166:795-804.

35. Luby JL, Navsaria N. Pediatric bipolar disorder: evidence for pro-dromal states and early markers. J Child Psychol Psychiatry 2010; 51:459-71.

36. Shaw JA, Egeland JA, Endicott J et al. A 10-year prospective study of prodromal patterns for bipolar disorder among Amish youth. J Am Acad Child Adolesc Psychiatry 2005;44:1104-11.

37. Youngstrom E, Youngstrom JK, Starr M. Bipolar diagnoses in com-munity mental health: Achenbach Child Behavior Checklist profiles and patterns of comorbidity. Biol Psychiatry 2005;58:569-75.

38. Miklowitz DJ, Chang KD, Taylor DO et al. Early psychosocial inter-vention for youth at risk for bipolar disorder: a 1-year treatment development trial. Bipolar Disord 2011;13:67-75.

39. Wilens TE, Biederman J, Kwon A et al. Risk of substance use disor-ders in adolescents with bipolar disorder. J Am Acad Child Adolesc Psychiatry 2004;43:1380-6.

40. Leibenluft E, Cohen P, Gorrindo T et al. Chronic vs. episodic irritabil-ity in youth: a community-based, longitudinal study of clinical and diagnostic associations. J Child Adolesc Psychopharmacol 2006;16: 456-66.

41. Scott J, Paykel E, Morriss R et al. Cognitive behaviour therapy for severe and recurrent bipolar disorders: a randomised controlled trial. Br J Psychiatry 2006;188:313-20.

011_016.indd 16 06/07/12 14:06

17

Challenges in the implementation of diagnostic specifiers for mood disorders in ICD-11

Two basic approaches to classification in developing a di-agnostic system for mental disorders have been character-ized: “lumping” and “splitting”. A “lumping” strategy favors fewer diagnostic entities that are defined relatively broadly and heterogeneously, with the assumption that differences among cases are not as important as their similarities. In con-trast, a “splitting” strategy leads to a greater number of more narrowly defined entities, based on the assumption that the differences among cases are more important than their simi-larities in terms of defining etiologically and therapeutically homogeneous entities.

Although it is widely assumed that the DSM-IV and ICD-10 invariably favor a splitting strategy (1), the extent to which current classifications appear to reflect a lumping or splitting approach varies substantially across diagnostic domains. The DSM-IV and ICD-10 mood disorders sections can be seen as employing a “lumping” strategy in which there are only a few distinct disorders that are defined relatively broadly. For ex-ample, depression is conceptualized as a single syndromal entity that varies along a continuum of severity and accord-ing to its chronicity. Presentations of depression range from mild persistent cases (i.e., dysthymic disorder) to episodic forms of varying severity based on the number of symptoms and their impact on functioning (i.e., mild, moderate, and severe depressive episodes). In contrast, the DSM-IV and ICD-10 anxiety disorders are split into a much larger number of distinct forms, differentiated by the course of the anxiety (e.g., panic disorder, generalized anxiety disorder), focus of anxiety (e.g., social phobia, specific phobia, agoraphobia, obsessive-compulsive disorder), and causal factors (e.g., post-traumatic stress disorder).

The adoption of a lumping strategy for mood disorders, particularly for classifying depressive disorders, reflects the view of depression as a unitary construct representing a final common pathway derived from a variety of etiological and

Michael B. First

Department of Psychiatry, Columbia University and Department of Clinical Phenomenology, New York State Psychiatric Institute, 1051 Riverside Drive – Unit 60, New York, NY, USA

The “lumping” strategy employed in the DSM and ICD mood disorders sections requires that clinically relevant diagnostic specificity be indi-cated via the provision of various subtypes and specifiers. The ICD requirement that subtypes and specifiers have unique diagnostic codes presents significant challenges both in terms of providing needed diagnostic specificity and facilitating harmonization with the forthcoming DSM-5, which makes extensive use of uncodable specifiers. This paper begins with a historical review to explicate why the DSM and ICD have adopted different approaches to diagnostic coding and discusses their implications for administrative utility. It then reviews the various specifiers and subtypes for mood disorders contained in the various editions of the DSM and ICD, the number of which has steadily increased over each edition to a total of 16 in DSM-5. Finally, given the ICD requirement that all subtypes and specifiers be coded, the paper concludes with a discussion of various strategies to maximize the inclusion of the most clinically important specifiers in ICD-11 without compromising the usability of the system.

Key words: DSM-5, ICD-11, mood disorder classification, clinical utility, diagnostic codes


pathophysiological sources (2) that “accounts for the shared clinical features seen in the heterogeneous groups of depres-sive disorders”. Over the past 30 years, all major psychiatric classification systems – starting with the Feighner criteria (3), and continuing with the Research Diagnostic Criteria (RDC, 4), DSM-III and subsequent DSM revisions, and ICD-10 – have used a single set of descriptive criteria for all episodes of clinical depression, despite evidence of heterogeneity in terms of pathophysiological mechanisms and treatment response.

One of the most important aspects of clinical utility in a diagnostic system is facilitating treatment selection (5). If it were the case that simply meeting the diagnostic criteria for a depressive disorder was sufficient to determine the optimal treatment, then having a single unitary diagnosis would suf-fice. However, the long-recognized inconsistent response to various depression treatments suggests the value of identify-ing subgroups of cases that are more likely to respond to spe-cific treatments based on severity, phenomenology, course, and other factors. The DSM and ICD mood and anxiety dis-orders sections illustrate the two contrasting approaches for providing such diagnostic specificity. For the “lumped” mood disorders, diagnostic specificity is indicated via the provision of various subtypes and specifiers, whereas for anxiety disor-ders, diagnostic specificity is achieved by virtue of the group-ing having been split into many distinct disorders.

From an administrative perspective, however, the primary goal of the DSM-IV and ICD-10 is to facilitate the reporting of diagnostic information by health care professionals and facilities to administrative authorities for the purposes of data collection and payment for services. To this end, the diagnostic codes in the system constitute the primary data elements for record keeping, data analysis, and reimburse-ment. Consequently, any diagnostic distinction that is impor-tant for these purposes needs to be accommodated by the provision of a unique code for each diagnostic entity. In this

017_025.indd 17 06/07/12 14:07


regard, the DSM-IV and ICD-10 classifications differ signifi-cantly in their approach. The DSM-IV (and DSM-5) classifi-cations make frequent use of uncodable specifiers and sub-types for indicating diagnostic specificity. In contrast, struc-tural requirements of the ICD-10 (and ICD-11) classifica-tions mandate that each disorder, subtype, and specifier that is important enough to be included in the system have its own diagnostic code.

As the ICD-11 Mood and Anxiety Disorders Working Group considers its options regarding how to revise the ICD-10 to make it more clinically useful (6), the ICD requirement that subtypes and specifiers be codable presents significant challenges both in terms of providing needed diagnostic specificity and facilitating harmonization with the forthcom-ing DSM-5 (7), particularly the sections on bipolar and de-pressive disorders. This paper begins with a historical review discussing how diagnostic coding has evolved over the vari-ous editions of the DSM, in order to explicate why the DSM and ICD have adopted such different approaches to diagnos-tic coding, and then discusses the implications of these ap-proaches for the administrative utility of the systems. Next the paper addresses the usage of subtypes and specifiers in the mood disorders sections of ICD and DSM, beginning with a general discussion of the usage of the terms “type”, “subtype” and “specifier”, and then addresses the specifics of how the adoption of a lumping strategy for mood disorders necessitates the extensive use of subtypes and specifiers as indicators of clinically relevant diagnostic features. Finally, based on the ICD requirement that all subtypes and specifi-ers be coded, the paper concludes with a discussion of vari-ous strategies to maximize the inclusion of the most clini-cally important specifiers without compromising the usabil-ity of the system.

HIstorICal baCKgrounD

The first edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-I, 8) employed two different cod-ing systems. The DSM-I was explicitly developed to serve as the mental disorders section of the American Medical As-sociation’s Standard Nomenclature of Diseases and Opera-tions (9), a medical nomenclature “set up for use by physi-cians, specialists, and hospitals to secure standard and uni-form terminology in the diagnosis of the diseases of individ-ual patients” (8). In order to facilitate the storage of diagnos-tic information in medical record systems, each DSM-I dis-order and subtype was assigned a unique diagnostic code (which could extend up to 8 digits) that conformed to the coding conventions of the Standard Nomenclature. A medi-cal nomenclature, which contains approved terms and defi-nitions, is to be contrasted with a statistical classification, which is developed for the collection of statistical informa-tion about groups of patients in a facility, state, country, or the world. In order to allow the international comparison of health statistics, all countries that are members of the World

Health Organization (WHO) have agreed to collect health statistics using the WHO’s International Classification of Diseases (ICD). In order to facilitate the collection of inter-national health statistics, each DSM-I category also listed the corresponding ICD-6 code (10) in parentheses.

The introduction of the DSM-II (11) begins with the state-ment that it “reflects the growth of the concept that the people of all nations live in one world”. As such, the DSM-II was essentially thought of as an adaptation of the ICD-8 (12) for use by psychiatrists in the United States. The DSM-I coding system from the Standard Nomenclature was entirely dropped in favor of adopting only the ICD-8 diagnostic codes. To ac-commodate diagnostic categories and subtypes that were not part of ICD-8, unused ICD-8 digits at the 4th digit level and occasionally 5th digit codes were assigned. Like the DSM-I, each DSM-II diagnosis and subtype had its own unique diag-nostic code.

For the DSM-III (13), the diagnostic codes were for the most part taken from the ICD-9-CM (14), a “clinical modifi-cation” of the ICD-9 originally developed in 1977 by the Na-tional Center for Health Statistics in the United States. While largely maintaining compatibility with the ICD-9 at the 3- and 4-digit level, additional diagnostic specificity was pro-vided largely through the provision of 5th digit codes. Al-though the majority of DSM-III categories were assigned the ICD-9-CM diagnostic code that was conceptually the closest (for example, the DSM-III category of dysthymia was as-signed the ICD-9-CM code for neurotic depression), some DSM-III categories were assigned unique diagnostic codes not contained in the ICD-9-CM, in order to insure that every DSM-III disorder and subtype had a unique code. For ex-ample, the presence of mood-incongruent psychotic features in a depressive or manic episode could be coded in the 5th digit using the unofficial ICD-9-CM code “7”. Diagnostic specifiers indicating illness course were also available for some disorders (e.g., substance use disorders), each with its own unique code (e.g., “1” for continuous, “2” for episodic, “3” for in remission, and “4” for unspecified).

Starting with the DSM-III-R (15), however, the DSM con-vention for diagnostic coding assignments changed. Al-though the DSM-III approach insured that each diagnosis and subtype could be uniquely entered into record keeping systems, the creation of codes that were not legitimate ICD-9-CM codes caused problems when used in data systems designed to accept only valid ICD-9-CM codes, i.e., those used for reporting diagnoses for U.S. government purposes, such as Medicare. Thus, invalid ICD-9-CM-type codes were eliminated in the DSM-III-R, which had two implications. First, because of the more limited number of available ICD-9-CM diagnostic codes, a number of DSM-III-R categories had to share the same diagnostic code. Second, all of the subtypes and specifiers that had been assigned unofficial 5th digit codes could no longer be assigned any 5th digit code at all. Moreover, new subtypes and specifiers proposed by the DSM-III-R work groups could not be assigned their own di-agnostic codes. Consequently, 31 specific DSM-III-R disor-

017_025.indd 18 06/07/12 14:07


ders had uncoded subtypes or specifiers. While this solution was clearly suboptimal in terms of limiting the ability of the DSM-III-R diagnostic distinctions to be expressed in the di-agnostic codes and thus be available to record keeping and reporting systems, the lack of available ICD-9-CM diagnostic codes made this a necessity.

The practice of including uncoded subtypes and specifiers exploded in DSM-IV, with an additional 94 disorders having uncoded specifiers and subtypes (although 73 of these in-volved specifiers that applied to the various types of sub-stance-related disorders, e.g., with and without physiological dependence applying to nine of the specific substance de-pendence categories). Depending on the inclinations of the various DSM-IV work groups, virtually any diagnostic fea-ture that potentially had important research or clinical impli-cations was added to DSM-IV. Moreover, given that their uncodable nature made them clearly “optional” for use by clinicians or researchers, the threshold for accepting work group proposals for new specifiers was relatively low, much lower than for including new disorders (16).

Although the diagnostic codes in DSM-5 will be adapted from the ICD-10-CM, which will become official in the Unit-ed States in October 2014, a majority of the DSM-5 subtypes and specifiers will continue to be uncoded, because of a lack of corresponding ICD-10-CM designations.

In contrast to the DSM-III-R, DSM-IV, and DSM-5 sce-narios, wherein a majority of the specifiers and subtypes are uncodable, a structural requirement of the ICD-10 classifica-tion (as well as the ICD-9-CM and ICD-10-CM classifica-tions) is that all specifiers and subtypes have unique diagnos-tic codes and are thus captured as distinct diagnostic entities in record keeping systems. In the ICD-10 Clinical Descrip-tions and Diagnostic Guidelines, some subtypes and speci-fiers are coded at the 4th character level (e.g., the phenome-nological subtypes of F20 Schizophrenia) and others are coded at the 5th character level (e.g., with and without pan-ic disorder for F40.0 Agoraphobia). However, the diagnostic codes in the “official” version of ICD-10 (i.e., the version voted on by the World Health Assembly) extend only to the 4th character level and are thus the only diagnostic distinc-tions that are formally required in reporting of health statis-tics by WHO Member States. Subtypes and specifiers that are coded at the 4th character level are thus presumably more diagnostically important than those at the 5th character lev-el, as the latter are often truncated by local record keeping systems.

ImplICatIons of utIlIzIng unCoDable subtypesanD speCIfIers on aDmInIstratIve utIlIty

The fact that a majority of the subtypes and specifiers in-cluded in DSM-IV are not codable has serious implications for their administrative utility, i.e., in facilitating diagnostic communication between clinicians and health care systems. Virtually all ICD-oriented health care record keeping sys-

tems have been designed to process and store ICD-9-CM diagnostic codes, each of which is unique to a particular di-agnostic entity, with the accompanying diagnostic terminol-ogy captured only in text fields. Since there is a one-to-one relationship between each diagnostic code and its accompa-nying diagnostic term, the record keeping systems generally ignore the diagnostic terms when processing the diagnosis fields in a medical record, claims form, or data base, thus al-lowing the system to store only the diagnostic code. The di-agnostic information contained in uncodable specifiers and subtypes – which, if it is included at all, is indicated solely in diagnostic text fields – is lost to the system, completely com-promising their administrative, research, and public health utility. Uncodable specifiers and subtypes cannot subse-quently be used in health reporting, outcomes evaluation, treatment planning, health services and policy research or any other application based on data from health encounters.

Consider, for example, two patients with recurrent severe depression. One patient has had a postpartum onset of her current melancholic episode, whereas the second patient has a current atypical episode, which has been occurring in a seasonal pattern for the past several years. The treatment plan for these two patients would be quite different. How-ever, because the clinically salient features (i.e., postpartum onset, melancholic features vs. seasonal onset atypical fea-tures) are not codable, both of these patients would have the same diagnostic code (recurrent depressive disorder, current episode severe) and would not be distinguishable in an ICD-based record keeping system. Consequently, it would be im-possible to conduct analyses on how many patients have postpartum onset to their depression or a seasonal pattern or to develop a differential reimbursement rate reflecting the varied treatment costs associated with these two different clinical pictures.

subtypes anD speCIfIers In DsmanD ICD mooD DIsorDers

The terms “subtype” and “specifier” (and its synonyms “modifier” and “qualifier”) are not well defined in the do-main of medical classification. The dictionary definition of subtype is “a type that is subordinate to or included in an-other type” (17), with type being defined as “a lower taxo-nomic category selected as a standard of reference for a higher category” (17). Presumably, rules that are analogous to those that determine whether a category should be classi-fied as a single disorder or split into two or more separate disorders (i.e., whether they represent distinct nosological entities) should apply as well to splitting an entity into vari-ous “types”. A search of the ICD medical classification for diagnostic entities that include the word “type” indicates that this rule does generally apply to medical conditions with a common presentation but separable underlying pathophysi-ological processes. For example, while individuals with the diagnosis “diabetes mellitus” present with high blood sugar

017_025.indd 19 06/07/12 14:07


accompanied by excessive production of glucose-laden urine, the two types of diabetes mellitus involve different un-derlying mechanisms: in type 1 diabetes, the elevated blood sugar is due to the absence of insulin production, whereas in type 2 diabetes, the elevated blood glucose is caused by insu-lin resistance at the cellular level.

Owing to the lack of formal definitions for the terms “sub-type” and “specifier” in DSM-III, DSM-III-R, DSM-IV and ICD-10, their conventions must be inferred from their usage. Because of the limited information about underlying patho-physiology, using “type” to indicate different pathophysio-logical processes, as is done in some other areas of medicine, inevitably indicates presumed rather than established under-lying differences in etiology. For example, the various sub-types of schizophrenia can be interpreted to reflect an as-sumption that these phenomenological differences represent different underlying processes and thus have different prog-nostic and treatment implications. Specifiers (or qualifiers, as they are referred to in DSM-III), on the other hand, are gen-erally used to indicate notable clinical features of a disorder that may have prognostic, treatment, or etiological implica-tions but are not presumed to represent a different underlying etiology or pathophysiology. For example, DSM-IV provides a “with injurious behavior” specifier for stereotyped move-ment disorder because of its management implications.

The introductory sections of DSM-IV do offer definitions of a subtype vs. specifier, which primarily revolve around whether the defined groupings are mutually exclusive. Ac-cording to DSM-IV, subtypes “define mutually exclusive and jointly exhaustive phenomenological subgroupings within a diagnosis”. In contrast, “specifiers are not intended to be mu-tually exclusive or jointly exhaustive…. [They are intended to] provide an opportunity to define a more homogeneous subgrouping of individuals within the disorder who share certain features (e.g., Major Depressive Disorder with psy-

chotic features)”. Given that the ICD hierarchical structure requires that all categories at the same level of the hierarchy be mutually exclusive and jointly exhaustive, the DSM-IV-defined distinction between subtypes and specifiers is not meaningful in the context of the ICD classification. There-fore, for the purpose of the ICD, the terms “subtype” and “specifier” should be considered as functionally synony-mous.

The mood disorders sections of DSM and ICD essentially include two disorders, each defined on a severity continuum: bipolar disorder/cyclothymia and major depressive disorder/dysthymia. Beyond this fundamental diagnostic distinction, the ability to communicate additional specification in terms of course, phenomenology, prognosis and treatment respon-sivity for the most part requires the use of subtypes or speci-fiers. Complicating the situation regarding the use of sub-types and specifiers for DSM and ICD mood disorders is the fact that episodic mood disorders are defined in terms of depressive, manic, hypomanic, and mixed episodes, each of which has its own syndromal definition. Specifiers, there-fore, can be used either to indicate noteworthy aspects at the episode level (e.g., melancholic type to indicate the presence during the depressive episode of non-reactive anhedonia ac-companied by a certain pattern of symptoms such as anorex-ia and early morning awakening) or features at the disorder level (e.g., seasonal pattern, which describes a regular tem-poral relationship between the onset of episodes and a par-ticular time of the year).

Tables 1 and 2 present the various mood disorders speci-fiers that have been used in the DSM and ICD systems, start-ing with DSM-III and through to the proposed changes in DSM-5. Table 1 contains those specifiers that apply at the disorder level, and Table 2 contains those specifiers that are defined at the episode level. The DSM-III and ICD-10, both constrained by the need to have a unique diagnostic code for

Table 1 Specifiers/subtypes of mood disorders that apply at the disorder level

Applies to: Potential clinical utility DSM-III DSM-III-R ICD-10 DSM-IV DSM-5

Type I/Type II Bipolar Encourages recognition of abnormal elevated/irritable

mood states in recurrent depression

X X

Single episode/recurrent

MDD Indicates need for maintenance treatment

X X X X X

Current episode manic, depressive, mixed, hypomanic

Bipolar Indicates specific current treatment needs

X X X X X

Rapid cycling Bipolar Indicates specific treatment needs (e.g., less likely to respond

to lithium)

X X

Seasonal pattern Bipolar/MDD Indicates utility of light therapy X X X

With interepisode recovery/without interepisode recovery

Bipolar/MDD Indicates need for interepisode treatment

X

MDD – Major depressive disorder

017_025.indd 20 06/07/12 14:07


each specifier, had 6 and 7 different types of specifiers respec-tively. The DSM-III-R, the first edition to employ uncoded specifiers, had 9. The DSM-IV expanded this to 15 different types of specifiers. The DSM-5 Work Group has proposed three additional specifiers (“with mixed features”, “with low/medium/high suicide risk”, and “with mild/moderate/severe anxiety”), but also proposed removing two specifiers (“chron-ic” and “with interepisode recovery/without interepisode recovery”) for a grand total of 16 different types of specifiers being proposed for mood disorders in DSM-5 (see www.dsm5.org).

Given that a primary goal of the development of the men-tal and behavioural disorders chapter of ICD-11 is to im-prove its clinical utility (6), and given that the main reason for utilizing specifiers is to identify clinically useful group-ings, the ICD-11 Mood and Anxiety Working Group will need to consider which specifiers should be formally incor-porated into the ICD-11 mood disorders sections. Besides the 16 subtypes/specifiers that are proposed for the DSM-5 (along with the three types that were in DSM-IV and dropped from DSM-5), additional subtypes/specifiers have been pro-posed, both in the literature (e.g., “depression after child-hood trauma”, “depressive reaction to separation stress”,

“late-life depression” (18)) and in another article in this sup-plement (“childhood onset” (19)).

As was the case with the ICD-10, however, the coding structure of the ICD-11 places a limitation on the number of specifiers that can comfortably be accommodated without making the classification impossibly unwieldy. While the precise coding format for ICD-11 has yet to be finalized, one significant change will be the removal of the ICD-10 restric-tion of having only 10 available digits that applied to the 2nd through the 5th character. In contrast, the number of avail-able choices at each character level in ICD-11 will be much greater. Some character placeholders may use Roman letters, in which case there will be 24 places available (“I” and “O” will not be used because of potential confusion with “1” and “0”). However, some character levels may use two-digit num-bers, increasing the number of potential values to 100 (00-99). While this increase in range of possible values will pro-vide much greater flexibility in terms of expanding the num-ber of groupings at a particular level of the hierarchy, it has limited positive impact for providing unique diagnostic codes for the various subtypes.

In the mental disorders section of the ICD-11, the main benefit of the increase in the number of available values is

Table 2 Specifiers/subtypes of mood disorders at the episode level

Applies to: Potential clinical utility DSM-III DSM-III-R ICD-10 DSM-IV DSM-5

Mild, moderate, severe Depressive, manic, mixed

Treatment selection for depressive: psychotherapy for mild-moderate, medication for

moderate-severe

X X X X

In remission (partial/full)

Depressive, manic, mixed,

hypomanic

X X X X X

Psychotic Depressive, manic, mixed

Treatment (need for antipsychotic medication)

X X X X X

Mood-congruent/mood-incongruent psychotic features

Depressive, manic, mixed

Shared genetic vulnerability with schizophrenia (?)

X X X X X

Mixed features Depressive, manic

More difficult to treat, higher suicidality rates

X

Chronic Depressive More difficult to treat X X

Melancholic features Depressive Treatment (need for somatic treatment)

X X X X X

Atypical features Depressive Treatment (avoid use of tricyclic antidepressants)

X X

Catatonic features Depressive, manic, mixed

Treatment (ECT) X X

Postpartum onset Depressive, manic, mixed

Future prognosis X X

With mild/moderate/severe anxiety

Depressive More difficult to treat, risk of suicide

X

With low/medium/high suicide risk

Depressive, manic, mixed

Indicates higher safety risk X

ECT – electroconvulsive therapy

017_025.indd 21 06/07/12 14:07


that it allows for a much larger number of major diagnostic groupings. In ICD-10, the first level of the classification rep-resents the major medical domains within the ICD (e.g., in-fectious diseases, diseases of the eye, etc.). The fact that the first character in ICD-10 was alphabetic allowed for 26 dif-ferent major divisions of the ICD-10 diagnostic pie. The men-tal disorders were put into the 5th division and assigned the letter “F”. However, within the ICD-10 mental disorders sec-tion, given that the value of the 2nd character ranged from 0 to 9, the entire mental disorders pie had to be split into no more than 10 sections, forcing the ICD-10 developers to cre-ate some groupings with little conceptual or clinical coher-ence (e.g., F6, Disorders of Adult Personality and Develop-ment, included personality disorders, habit and impulse dis-orders, gender identity disorder, and disorders of sexual preference).

Because the second character in ICD-11 will not be lim-ited to 10 divisions, the DSM-5 and ICD-11 Advisory Groups were able to divide the mental disorders into 19 or 20 group-ings, based on validity, utility, and common sense consider-ations. This change has implications for the ICD-11 mood and anxiety disorders classification as well. It has been pro-posed that what used to be contained within a single second level section in ICD-10 (i.e., F3 Mood Disorders block) will be split into two second level sections in ICD-11 and DSM-5: Bipolar Disorders and Depressive Disorders. Similarly, it has been proposed that the ICD-10 F4 Neurotic, Stress-Related, and Somatoform Disorders block be split into at least three sections in ICD-11 and DSM-5: Anxiety Disorders, Obses-sive-Compulsive and Related Disorders, and Trauma and Stress-Related Disorders.

However, the coding limitation on specifiers imposed by the ICD-11 structure has to do with the way that mood spec-ifiers are used in the DSM and ICD: multiple different speci-fiers can apply to the same diagnostic entity. For example, consider the case of a woman who develops a full depressive episode of moderate symptom severity four weeks after giv-ing birth to her child and who symptomatically is extremely anxious, experiencing religious delusions about her child, expressing suicidal ideation (but without a plan), and is an-hedonic, psychomotorically retarded and not eating, with her depression being worse in the morning. Moreover, dur-ing the initial evaluation, the clinician also discovers a his-tory of mania. Her diagnosis, using the nine applicable spec-ifiers proposed for DSM-5, would be: bipolar I disorder, cur-rently depressive, severe, with mood-congruent psychotic features, with melancholic features, with severe anxiety, with medium suicide risk, with postpartum onset. Current ICD-10 and DSM-IV coding rules only allow for the numeric coding of the fact that she is bipolar, with a current severe depressive episode with psychotic features (F31.4 in ICD-10, 296.54 in DSM-IV-TR). Even though the DSM theoretically allows the clinician to indicate the other clinical features as specifiers by appending them to the text field of the diagnosis, the fact that these specifiers cannot be captured by record keeping sys-tems is very likely to discourage their use by clinicians.

One solution to this problem would be for ICD-11 to as-sign unique diagnostic codes to each of the clinically relevant mood specifiers. Unfortunately, the structural characteristics of the ICD classification present significant challenges to implementing such a plan. The crux of the problem is that, in a hierarchical diagnostic system such as the ICD, each inde-pendent diagnostic distinction, even one that is binary in nature (e.g., a specifier indicating the presence or absence of melancholic features) requires its own unique character placeholder to express its variability. Given that most of the specifiers that apply to the mood disorders can be present or absent independent of the presence or absence of the other specifiers, each one requires its own separate character. In the case outlined above, character placeholders would be required not just for the specifiers that are present but also to indicate those that are absent. Using the coding rules (see Table 3) indicating unique codes for both the presence and absence of each proposed DSM-5 specifier, a 15-character diagnostic code would be required for properly coding bipo-lar I disorder, currently depressive, severe, with mood-incon-gruent psychotic features, with melancholic features, with severe anxiety, with medium suicide risk, with postpartum onset (see Table 4). Unfortunately, a 15-character code is hardly a satisfactory solution. For the sheer sake of clinical usability, 6 or 7 characters is generally as lengthy as any code gets in the ICD-10 system. Moreover, from a practical per-spective, given that many WHO Member States truncate codes at after four or five character level, the number of meaningful code positions is effectively even less.

As noted above, the bipolar and depressive disorder sec-tions, with their “lumped” organizational strategy, suffer uniquely from this coding problem, because of the combina-tion of having only a few broadly defined categories with mul-tiple independent specifiers. For ICD diagnostic groupings that use a splitting strategy, such as the anxiety disorders, the 3rd character in the code represents the major diagnostic divi-sions among the anxiety disorders (i.e., panic disorder, agora-phobia, social phobia, specific phobia, separation anxiety disorder, generalized anxiety disorder), leaving the 4th, 5th, and if needed 6th characters available for subtypes and speci-fiers.

strategIes for CoDIng speCIfIers

Given that the limiting factor in developing a usable ICD-11 diagnostic coding scheme is the number of required char-acters, any strategy that reduces that number is potentially advantageous and should be considered carefully. One strat-egy to reduce the number of character placeholders is to com-bine specifiers that do not in fact occur independently. For example, the psychotic features specifier and mood-congru-ence/mood-incongruence specifiers are not independent and could be combined (0=without psychotic features, 1=with mood-congruent psychotic features, 2=with mood-incongru-ent psychotic features). It is important, however, to insure that

017_025.indd 22 06/07/12 14:07


the specifiers to be combined are truly non-independent; combining diagnostic constructs under the assumption that they are interdependent when in fact they can occur indepen-dently will limit their proper use. Such is the case with the DSM-IV/ICD-10 strategy of linking the presence of psychot-ic features with symptom severity. In this instance, “psychotic features” are assumed to be present only in severe cases of depression, as manifested by the psychotic features specifier existing only as an extension of “severe”. Thus, clinicians can only indicate the presence of psychotic features in severe cases of depression (i.e., “severe with psychotic features”); no option is provided for indicating “moderate severity with psy-chotic features” despite evidence that psychotic symptoms can occur in non-severe depression (20). Given the clinical importance of noting the presence of psychotic symptoms regardless of severity, the ICD-11 should either keep psychot-ic features as a specifier independent of symptom severity or else come up with a more complex combined specifier that allows for more combinations of severity states and psychotic symptoms.

For specifiers that occur independently, a second strategy is to merge two specifier categories into one combined cate-gory, which works as long as the combined code values re-flect all possible combinations of values in the two single specifier categories. For example, the two categories “melan-

cholic features” and “atypical features” could be combined into a single category called melancholic/atypical features as follows: 0=neither melancholic or atypical features, 1=mel-ancholic features only, 2=atypical features only, 3=both mel-ancholic and atypical features. One drawback of this ap-proach is that it tends to work best for categories with only binary (0/1) values; the higher the number of possible values for each category, the greater the number of possible com-bined values overall, which can lead to an inordinately com-plex system. For example, combining the anxiety specifier with its mild, moderate, and severe values with the suicide risk specifier with its low, medium, and high values to pro-duce a combined anxiety/suicide risk specifier would result in the need for one fewer character placeholder but would result in 16 possible combinations for the specifier, a scheme that most clinicians would likely find difficult to use.

A third strategy is to move the mood disorders section away from its strictly lumping approach by splitting the mood disorders into more diagnostic distinctions at the top level. This strategy effectively reduces the number of placeholders needed at lower levels in the hierarchy by moving the diag-nostic distinction to a higher level. Take, for example, the di-agnostic specifier “with seasonal pattern”, that applies to re-current depressive disorder and bipolar disorder. A first op-tion could be to organize the depressive disorders section

Table 3 Illustrative example of hypothetical coding rules for bipolar disorder incorporating 15 specifiers (in the classification, the first character is 05, indicating mental disorder, and the second is C, indicating bipolar disorder)

Character Specifier 0 1 2 3 4

3 Type Type I Type II

4 Current episode Manic Depressive Hypomanic

5 Psychotic features Without psychotic features

With mood- congruent psychotic

features

With mood-incongruent

psychotic features

6 Severity/Remission Mild Moderate Severe In partial remission In full remission

7 Mixed features Without With

8 Melancholic features Without With

9 Atypical features Without With

10 Catatonic features Without With

11 Anxiety None Mild Moderate Severe

12 Suicide risk None Low Medium High

13 Postpartum onset Without With

14 Rapid cycling Without With

15 Seasonal pattern Without With

Table 4 Illustrative example of the 15 character diagnostic code which would be required to represent for bipolar I disorder, currently de-pressive, severe, with mood-incongruent psychotic features, with melancholic features, with severe anxiety, with medium suicide risk, with postpartum onset, using hypothetical coding rules from Table 3

Character # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Code 05 C 0 1 1 2 0 1 0 0 3 2 1 0 0

017_025.indd 23 06/07/12 14:07


with specifiers for indicating single vs. recurrent episode at the 3rd character level, with psychotic features vs. without psychotic features at the 4th character level, severity/remis-sion at the 5th character level, and with seasonal pattern vs. without seasonal pattern at the 6th character level. Alterna-tively, the seasonal designation could be moved up to the dis-order level (at the 3rd character) by splitting “recurrent de-pressive disorder” into “recurrent seasonal depressive disor-der” and “recurrent non-seasonal depressive disorder”. This would both reduce the overall complexity of the depressive disorders classification and free up the 6th character for an-other clinically important specifier.

Finally, the strategy with arguably the greatest clinical util-ity would be to limit specifiers to only those that are the most clinically important. To assist in that regard, Tables 1 and 2 contain a column describing the presumed clinical utility of each of the possible mood specifiers. Thus, the ICD-11 Mood Disorders Working Group will likely need to review the exist-ing and proposed specifiers, evaluate each in terms of clinical utility, and select the 7 or 8 specifiers that are most clinically important (for example, see papers by Chakrabarti (21), Maj (22), and Moussaoui et al (23) in this supplement for a con-sideration of the clinical utility of several existing and pro-posed specifiers).

ImplICatIons for ICD-11/Dsm-5 HarmonIzatIon

The DSM-5 practice of extensively using uncodable sub-types and specifiers to indicate diagnostic heterogeneity pres-ents significant obstacles to any attempt to harmonize the mood disorders sections in DSM-5 and ICD-11. As discussed above, it is simply not an option for the official ICD-11 clas-sification to utilize uncodable specifiers. The ICD-11 require-ment that every diagnostic entity have a unique code com-bined with the need to ensure that the system is not so com-plex as to render it unusable in typical practice settings sug-gests that some divergence between the two systems in the classification of mood disorders is likely.

These differences in the DSM and ICD approaches to cod-ing specifiers reflect a fundamental conceptual difference in the DSM and ICD systems that goes back to the DSM-I. As noted in the historical background section, the DSM-I was explicitly thought of primarily as a medical nomenclature; it was developed to serve as the mental disorders section of the American Medical Association’s Standard Nomenclature of Diseases. In contrast, the ICD always had been, and will continue to be, primarily a statistical classification designed for the collection of data about groups of health care system users. The use of uncodable subtypes and specifiers in the DSM is thus essentially a throwback to DSM’s roots as a medical nomenclature, i.e., providing standardized defini-tions for diagnostically useful terms.

An unanswered question is whether DSM users actually utilize uncodable specifiers in their clinical practice. Al-though uncodable specifiers first appeared in the DSM-III-R,

its “Use of This Manual” introductory section made no men-tion of how such specifiers should be used, nor were they included in any of the examples of how to record diagnoses and codes. Although the “Use of This Manual” section of DSM-IV instructs the DSM user to include uncoded sub-types or specifiers “after the name of the disorder (e.g., Social Phobia, Generalized)”, no other instructions are offered. Given that multiple specifiers and subtypes are often appli-cable to many DSM-IV-TR disorders (such as mood and substance-related disorders), the lack of standardized rules as to how they should be ordered likely discourages their being systematically recorded and processed even in record keeping systems that capture text fields.

Although anecdotally it appears that clinicians routinely ignore uncodable specifiers, there are no empirical studies that have examined how or whether such specifiers are used in clinical practice. In fact, the inability to capture these spec-ifiers in clinical databases makes it nearly impossible to study their actual use. Moreover, given that many clinicians use the DSM primarily to meet administrative requirements and for reimbursement purposes, their lack of administrative utility provides a strong disincentive against their routine use. Stud-ies focusing on the clinical utility of specifiers for mood dis-orders are needed to determine priorities for including them in the ICD-11.

note

M.B. First is a consultant to the WHO Department of Mental Health and Substance Abuse and to the Internation-al Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders. The views expressed in this article are those of the author and, except as specifically noted, do not represent the official policies or positions of the Interna-tional Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.


references

1. Van Praag H. Nosologomania; a disorder of psychiatry. World J Biol Psychiatry 2000;1:151-8.

2. Akiskal HS, McKinney WT, Jr. Overview of recent research in de-pression. Integration of ten conceptual models into a comprehen-sive clinical frame. Arch Gen Psychiatry 1975;32:285-305.

3. Feighner JP, Robins E, Guze SB et al. Diagnostic criteria for use in psychiatric research. Arch Gen Psychiatry 1972;26:57-63.

4. Spitzer R, Endicott J, Robins E. Research Diagnostic criteria: ratio-nale and reliability. Arch Gen Psychiatry 1978;35:773-82.

5. First MB. Clinical utility in the revision of the Diagnostic and Statis-tical Manual of Mental Disorders (DSM). Prof Psychol Res Pr 2010; 41:465-73.

6. International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders. A conceptual framework for the revi-sion of the ICD-10 classification of mental and behavioural disor-

017_025.indd 24 06/07/12 14:07


ders. World Psychiatry 2011;10:86-93.7. First MB. Harmonisation of ICD-11 and DSM-V: opportunities and

challenges. Br J Psychiatry 2009;195:382-90.8. American Psychiatric Association. Diagnostic and statistical manu-

al of mental disorders. Washington: American Psychiatric Associa-tion, 1952.

9. Phinkett RJ, Hayden AC (eds). Standard nomenclature of diseases and operations. Philadelphia: American Medical Association, The Blakiston Company, 1952.

10. World Health Organization. Manual of the international statistical classification of diseases, injuries and causes of death (ICD-6). Ge-neva: World Health Organization, 1948.

11. American Psychiatric Association. Diagnostic and statistical manu-al of mental disorders, 2nd ed. Washington: American Psychiatric Association, 1968.

12. World Health Organization. Eighth revision of the international clas-sification of diseases. Geneva: World Health Organization, 1968.

13. American Psychiatric Association. Diagnostic and statistical manu-al of mental disorders, 3rd ed. Washington: American Psychiatric Association, 1980.

14. Commission on Professional and Hospital Activities. The interna-tional classification of diseases, ninth revision, clinical modification. Ann Arbor: Commission on Professional and Hospital Activities, 1978.

15. American Psychiatric Association. Diagnostic and statistical manu-al of mental disorders, 3rd ed. - revised (DSM-III-R). Washington: American Psychiatric Association, 1987.

16. Pincus H, Frances A, Davis W et al. DSM-IV and new diagnostic categories: holding the line on proliferation. Am J Psychiatry 1992; 149:112-7.

17. Medline Plus Merrian Webster Dictionary. www.merriam-webster.com.

18. Lichtenburg P, Belmaker RH. Subtyping major depressive disorder. Psychother Psychosom 2010;79:131-5.

19. Miklowitz DJ, First MB. Specifiers as aids to treatment selection and clinical management in the ICD classification of mood disorders. World Psychiatry 2012;11(Suppl. 1):11-6.

20. Schatzberg A, Rothschild A. Psychotic (delusional) major depres-sion: should it be included as a distinct syndrome in DSM-IV? In: Widiger T, Frances A, Pincus H et al (eds). DSM-IV sourcebook, Vol. 2. Washington: American Psychiatric Association, 1996:127-80.

21. Chakrabarti S. Psychotic and catatonic presentations in bipolar and depressive disorders. World Psychiatry 2012;11(Suppl. 1):26-30.

22. Maj M. Mixed states and rapid cycling: conceptual issues and op-tions for ICD-11. World Psychiatry 2012;11(Suppl. 1):65-8.

23. Moussaoui D, Agoub M, Khoubila A. How should melancholia be incorporated in ICD-11? World Psychiatry 2012;11(Suppl. 1):69-72.

017_025.indd 25 06/07/12 14:07


Cultural issues in the classification and diagnosis of mood and anxiety disorders

The World Health Organization (WHO) has declared that the goals of clinical utility and cross-cultural applicability will be accorded a very high priority during the process of revising the Mental and Behavioural Disorders chapter of the International Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) (1).

Given the tremendous disparities in cultural characteris-tics and socioeconomic conditions of the WHO Member States, the task of devising a globally applicable classification system is always going to be a challenging one. Nevertheless, a guiding principle of the ICD-10 classification of mental disorders has been to take into account the needs of men-tally ill people and mental health professionals worldwide, in addition to the usual scientific requirements for such classi-fications (2).

The ICD-11 needs to continue with this process by incor-porating relevant insights from the growing body of evidence concerning cultural influences on the classification, diagno-sis and treatment of mental disorders.

Culture and its influenCe on mental disorders

Although the term culture has been defined in many dif-ferent ways, most regard it as some form of shared learned behaviour, handed down the generations for purposes of in-dividual or societal growth and adaptation (3-5). These defi-nitions also include an objective/external aspect of culture consisting of artefacts, roles and institutions, as well as a sub-jective/internal aspect represented by shared beliefs, atti-tudes, values and norms, a shared identity, and a shared view of the world. Unlike older concepts of cultures as self-con-tained systems, current views lay more emphasis on the dy-namic, ever-changing nature of cultures, and the lack of clear boundaries between different cultures, especially with the

Subho Chakrabarti1, CarloS berlanga2, Frank njenga3

1Department of Psychiatry, Postgraduate Institute of Medical Education and Research, Chandigarh 160012, India; 2National Institute of Psychiatry, Ramón de la Fuente Clinical Research Division, Colonia San Lorenzo Huipulco México, D.F. 14370, México; 3Upper Hill Medical Centre, P.O. Box 73749, Nairobi, Kenya

The World Health Organization has stated that clinical utility and cross-cultural applicability will be accorded a very high priority during the process of revising the ICD-10. This brief review highlights the impact of cultural factors on clinical presentations, gender differences, and treatment realities in the area of mood and anxiety disorders, with particular emphasis on research from non-Western settings. Even a cur-sory examination of these issues reveals significant cross-cultural variations in the clinical and treatment profile of these disorders. The process of revision would greatly benefit by incorporating relevant insights from the growing body of evidence regarding socio-cultural influences on depressive and anxiety disorders. This may eventually help lessen the widespread problems of non-recognition and under-treatment of these disorders, thereby enhancing the global utility of the ICD-11.

Key words: Culture, clinical presentations, gender, treatment, ICD-11


advent of globalization (3-5). It is also clear that culture has a pervasive influence on

almost every aspect of mental disorders: from concepts of health and disease to the experience and expression of dis-tress, from clinical presentations to coping and help-seeking, and from recognition and treatment of psychiatric conditions to eventual responses to the distress or disability caused by mental illnesses. Finally, cultural effects are not just limited to patients and their communities, but usually influence the clinician, as well as the clinical encounter itself (3,5).

Presentation of mood and anxietydisorders aCross Cultures

Cross-cultural epidemiology

Despite the ongoing debate about universality versus (cul-tural) specificity (4), it is clear that many mental disorders exist in most, if not all, cultures. At the same time, there ap-pear to be considerable differences in rates and profiles of mental disorders across cultures. Apart from ethnographi-cally informed “emic” studies, large-scale epidemiological surveys conducted across different countries form the basis of our understanding of these variations across cultures (3,6).

Although apparent differences in prevalence and presen-tation of mental disorders across cultures could be conse-quences of conceptual and methodological discrepancies (6), certain epidemiological findings are consistent enough to deserve mention. Anxiety, mood and substance-use disor-ders happen to be the commonest disorders in all countries. There are wide variations in prevalence estimates of these disorders across different regions; low to middle-income countries generally have lower lifetime prevalence rates than those with higher incomes. Anxiety and mood disorders typically have early ages of onset, are usually persistent and

026_030.indd 26 06/07/12 14:04


chronic, and are often highly comorbid. They are consistent-ly associated with substantial role impairments, as well as social disadvantage (e.g., minority status or female gender). Finally, inadequate treatment levels are widespread, particu-larly in the less developed countries (7,8).

Clinical studies

Two seminal studies by the WHO have provided valuable insights regarding the clinical profile of depressive and anxi-ety disorders in different countries and cultures.

The WHO Collaborative Study on Standardized Assess-ment of Depressive Disorders, conducted in Basel, Montreal, Nagasaki, Tehran and Tokyo, revealed a “core” group of symptoms that characterized the majority of depressed pa-tients across these centres (9). Simultaneously, variations in rates of certain symptoms such as feelings of guilt and self-reproach, suicidal ideation and somatic symptoms were also noted. These findings have been partly replicated by similar studies from India, Thailand and Ghana (10-12). However, the suggestion that guilt feelings are relatively rare among Asian and African countries was disproved by these studies, as well as by subsequent research (5,10-12), while cross-cul-tural differences in suicidal attempts and ideation continue to evoke considerable research interest (13).

Presentations of anxiety and depression in primary care were more comprehensively addressed by the WHO Collab-orative Study on Psychological Problems in General Health Care (PPGHC), which used identical methods and measures to examine the form, frequency, and course of common psy-chological disorders among patients in 14 countries and five continents around the world (14). The results of this study were consistent with much of research in this area in finding large cross-national variations in prevalence of depression and anxiety disorders. Then again, the symptom pattern or latent structure of depression was generally similar across all centres (15). The high prevalence of sub-threshold and co-morbid presentations of anxiety and depression found in the epidemiological surveys was also replicated in primary care settings by the PPGHC study. About 9% of the patients suf-fered from a sub-threshold condition, which did not meet diagnostic criteria, but was associated with clinically signifi-cant symptoms and functional impairment. Nearly half the instances of depression and anxiety appeared in the same patients at the same time (14).

somatization

Early reports from India, Africa and China had propa-gated the notion that somatic symptoms were the cultural equivalents of depression in non-Western countries, and that somatization, the process by which psychological distress was converted to somatic symptoms, was typical of these cultures (3-5).

Data from the PPGHC study were used to examine some of these notions regarding the relationship between somati-zation and depression (16,17). Three different concepts of somatization were examined. The first emphasized presenta-tion with somatic symptoms at initial contact, while the sec-ond definition emphasized the association between depres-sion and medically unexplained somatic symptoms. The third definition considered somatization as a psychological defense against the awareness or expression of psychological distress, related to traits such as alexithymia. Somatization defined on the basis of presenting symptoms had the highest overall prevalence and the greatest variations across centres. These differences were partly attributable to differences in characteristics of physicians and health care systems, as well as cultural differences in help-seeking among patients. Rates of depression with medically unexplained symptoms were similar in all centres, while denial of psychological symptoms was extremely rare (17). Finally, there was a strong and con-sistent correlation between somatic symptoms and anxiety and depression across all centres (16,17).

The finding that somatization of depression and anxiety is ubiquitous and not characteristic of any particular cultural group has been replicated in a number of other community surveys, which have shown a similar balance between psy-chological and somatic symptoms in non-Western and West-ern countries (3-5). Research has also demonstrated that so-matic symptoms are a core component of the depressive syn-drome in all cultures (17). Moreover, somatic presentations are neither due to denial of psychological distress, nor due to abnormal personality traits such as alexithymia (3,17). Rather, somatic complaints are an appropriate and less stigmatizing reason to seek help, especially in certain cultures and health care systems, which endorse this process of “facultative soma-tization” (3-5,17). The high prevalence of somatic complaints at initial presentation, however, contributes a great deal to the widespread non-recognition of depressive and anxiety disor-ders. Therefore, psychological symptoms assume significance for diagnosis; fortunately, they can be elicited in a majority of “somatizing” patients on careful enquiry (3,4,17).

Culture-bound syndromes

The classification of culture-specific or culture-bound syn-dromes is a complex and unresolved issue.

First, it is unclear whether these conditions are valid clin-ical entities meriting independent categorization, or are sim-ply local variants of depressive or anxiety disorders, in which case they can be included as part of the existing groups of these disorders (18). While forcing the culture-specific syn-dromes to fit in with the existing categories runs the risk of losing the unique cultural perspective on these syndromes, creating independent categories is equally problematic, be-cause many of these conditions have multiple and heteroge-neous presentations that are difficult to fit under a single di-agnostic entity (18).

026_030.indd 27 06/07/12 14:04


Moreover, several of the classical culture-bound syn-dromes may not be syndromes at all, but represent causal explanations, illness attributions or somatic metaphors (3). Others, such as the ataque de nervios, a form of anxiety disorder frequently observed in many Latin American coun-tries, are examples of culturally-sanctioned responses to acute stress (19).

Given these uncertainties, the best possible option for the ICD-11 would be to strike a compromise between including descriptions of such syndromes and emphasising their simi-larities with the more traditional categories of mood and anxiety disorders.

expression and communication of distress

The way people experience and express emotional distress is dictated by culture and differs widely throughout the world. The terms depression or anxiety are absent from the languages of many cultures; they are used rarely in others, or are construed differently (5,20). Therefore, the concepts of depression and anxiety may be very difficult to explain and describe in languages other than English. For example, if one attempts to explain the nature of an anxiety attack in the Kiswahili (Swahili), one could get involved in a complex dis-cussion with the patient about the existence of anxiety or fear, its causes, the part played by evil spirits and so on.

Moreover, each culture also has its own emotional lexicon that encodes cultural ways of expressing distress (3,5). These culturally-mediated somatic metaphors or idioms of distress may represent the physiological equivalent of the somatic symptoms specified by conventional nosology (20). The cli-nician’s ability to understand these local idioms is crucial not only for accurate diagnosis, but also for the building of a therapeutic alliance (5).

explanatory models and causal attributions

People also explain their distress in a multitude of ways, such as physical or biological disturbances, social circum-stances, relationship problems, witchcraft or sorcery. Explan-atory models offer the idiosyncratic perspective of the patient about the origin of symptoms, why they occur, and how the illness experience evolves (21). Explanatory models, thus, represent the position from which patients may express dis-tress and can govern how patients interpret a psychiatric ex-planation of their problems (5,21).

For example, in many African, Asian or Latin-American contexts, depression is often attributed to magical or super-natural forces, while European or American patients are more likely to offer biological explanations (5). It is impor-tant for the clinician to be conversant with these alternative explanations for psychiatric disorders. If, for example, the doctor seems to dismiss the idea that sadness, weakness, or insomnia are caused by angry spirits of a dead father and,

instead, seems to attribute these feelings to some intangible changes in strange chemicals called monoamines, then the consultation may come to a premature end.

Gender differences in mood andanxiety disorders across cultures

Women have been unfailingly reported to be twice as like-ly as men to suffer from depressive and anxiety disorders. This gender imbalance is one of the most robust findings of psychiatric epidemiology, and has been replicated across dif-ferent countries and cultures (22).

Gender differences in prevalence may not have a signifi-cant bearing on clinical presentations of depression and anxiety. Although some studies have reported that women tend to report physical symptoms – e.g. pain, fatigue, sleep and appetite disturbances – more often, others such as the PPGHC study have not been able to find a consistent effect of gender on reporting of somatic symptoms (23).

Risk factors such as poverty, poor physical health and do-mestic violence have a major role in determining the greater vulnerability of women to anxiety and depression, particu-larly in developing countries (4,5). Moreover, subtle differ-ences in communication of distress and depression may exist between men and women, particularly in non-Western popu-lations (20). For example, depressed African men may not express sadness, hopelessness, or even cry in public, much less so in front of a female doctor, because this would damage their image as men. Externalizing behaviours, including gen-der-based violence, alcohol use disorders, as well as mood swings and irritability, could act as reliable pointers to de-pression in men from these cultures (20).

treatment realities aCross Cultures

In all parts of the world, the vast majority of those with anxiety and depression are treated in general health care or primary care settings (4). Milder, subthreshold presentations and a high comorbidity between anxiety and depression are characteristic of such patients (14). Unfortunately, this par-ticular profile of anxiety and depression leads to non-recog-nition and ineffective treatment of these disorders, especially in the less developed regions (3-5,7,14).

Other economic and resource constraints may also con-tribute to under-treatment (4,5). For example, a typical con-sultation in primary care in Africa lasts only a few minutes, and is usually carried out by a non-specialist person without specific training in mental health, often in an environment where laboratory investigations, drugs and other essential medical supplies are not available.

Classification systems need to take cognizance of these treatment realities if they are to be of any use in such situa-tions.

026_030.indd 28 06/07/12 14:04


oPtions for iCd-11

We have highlighted some of the key areas where social and cultural processes influence the diagnosis, assessment and treatment of depressive and anxiety disorders. Several other important areas such as the cross-cultural variations in disability, diagnostic practices, or acceptability of classifica-tions have not been addressed. Nevertheless, even this cur-sory examination suggests that the process of revision of the ICD could profit greatly from incorporation of the currently existing data on socio-cultural influences on psychiatric dis-orders around the world.

Based on this brief review, a few tentative recommenda-tions can be made for the ICD-11:

• The current psychiatric diagnostic categories are largely embedded in Western psychiatric practice. The ICD-11 needs to extend beyond traditional Euro-American psy-chiatric notions of mental illness to include non-Western concepts, clinical descriptions and illness categories (20). The inclusion of acute psychosis in the ICD-10 is an excel-lent example of such an attempt (2), which needs to be perpetuated by the ICD-11.

• Though there is considerable cross-cultural variation in the prevalence and presentation of mood and anxiety dis-orders, similarities between cultures are equally impres-sive. As suggested by the authors of another article in this supplement (24), this implies that broadly similar criteria can be used for mood and anxiety disorders across cul-tures, provided they are not applied with undue rigidity. At the same time, the ICD-11 should describe culturally me-diated variations in presentation, wherever such informa-tion about cultural differences is available. A specific pro-vision for including such information has already been made in the guidelines provided by the WHO Department of Mental Health and Substance Abuse for developing ICD-11 content (25). In this regard, priority needs to be given to somatic symptoms, because they are central to the presentation of depression, are largely determined by cul-tural processes, and if not properly delineated could lead to under-recognition and under-treatment of depression. Similarly, sub-threshold and comorbid presentations also need to be properly highlighted, since they are common and disabling, especially in primary care settings, where they contribute to missed diagnoses and inadequate treat-ment of depressive and anxiety disorders.

• With respect to culture-specific syndromes, a fine balance has to be struck between including separate, but brief de-scriptions of these concepts, as well as allowing for cross-walks linking these syndromes to equivalent categories of mood and anxiety disorders. Such a provision already ex-ists in the research version of the ICD-10 (2); consider-ation should be given for including similar options in the clinical version of the ICD-11.

• All changes made will require field testing for usefulness, ease of use, and applicability in different settings. It is crit-

ical that field testing of the ICD-11 be conducted in mul-tiple languages and in culturally diverse settings, in order to adequately examine its cross-cultural applicability.

It is apparent that the extent to which the ICD-11 is able to satisfy the requirements of global applicability and utility will determine its usefulness in alleviating the widespread non-recognition and under-treatment of mood and anxiety disorders all over the world. Therefore, consideration of cul-tural factors should be central to the process of revising the ICD.

note

S. Chakrabarti, C. Berlanga and F. Njenga are members of the WHO ICD Revision Working Group on the Classifica-tion of Mood and Anxiety Disorders, reporting to the Inter-national Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders. The views expressed in this ar-ticle are those of the authors and, except as specifically not-ed, do not represent the official policies or positions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.

references

1. Reed GM, Correia Mendonça J, Esparza P et al. The WPA-WHO global survey of psychiatrists’ attitudes towards mental disorders classification. World Psychiatry 2011;10:118-31.

2. Sartorius N. Understanding the ICD-10 classification of mental dis-orders. A pocket reference, 2nd ed. London: Science Press Limited, 2002.

3. Kirmayer LJ. Cultural variations in the clinical presentation of de-pression and anxiety: implications for diagnosis and treatment. J Clin Psychiatry 2001;62(Suppl. 13):22-8.

4. Patel V. Cultural factors and international epidemiology. Br Med Bull 2001;57:33-45.

5. Bhugra D, Mastrogianni A. Globalisation and mental disorders: overview with relation to depression. Br J Psychiatry 2004;184:10-20.

6. Ballenger JC, Davidson JR, Lecrubier Y et al. Consensus statement on transcultural issues in depression and anxiety from the Interna-tional Consensus Group on Depression and Anxiety. J Clin Psychia-try 2001;62(Suppl. 13):47-55.

7. WHO World Mental Health Survey Consortium. Prevalence, sever-ity, and unmet need for treatment of mental disorders in the World Health Organization World Mental Health Surveys. JAMA 2004; 291:2581-90.

8. Bromet E, Andrade LH, Hwang I et al. Cross-national epidemiology of DSM-IV major depressive episode. BMC Med 2011;9:90.

9. Jablensky A, Sartorius N, Gulbinat W et al. Characteristics of de-pressive patients contacting psychiatric services in four cultures. A report from the WHO Collaborative Study on the Assessment of Depressive Disorders. Acta Psychiatr Scand 1981;63:367-83.

10. Gupta R, Singh P, Verma S et al. Standardized assessment of depres-sive disorders: a replicated study from northern India. Acta Psychi-atr Scand 1991;84:310-2.

11. Lotrakul M, Saipanit R, Theeramoke W. Symptoms of depression in Thai patients. ASEAN J Psychiatry 1996;4:24-32.

12. Majodina MZ, Johnson FY. Standardized assessment of depressive

026_030.indd 29 06/07/12 14:04


disorders (SADD) in Ghana. Br J Psychiatry 1983;143:442-6.13. Bertolote JM, Fleischmann A, De Leo D et al. Suicide attempts,

plans, and ideation in culturally diverse sites: the WHO SUPRE-MISS community survey. Psychol Med 2005;35:1457-65.

14. Ustun T, Sartorius N. Mental illness in general health care. New York: Wiley, 1995.

15. Simon GE, Goldberg DP, Von Korff M et al. Understanding cross-national differences in depression prevalence. Psychol Med 2002; 32:585-94.

16. Simon G, Gater R, Kisely S et al. Somatic symptoms of distress: an international primary care study. Psychosom Med 1996:58:481-8.

17. Simon GE, Von Korff M, Piccinelli M et al. An international study of the relation between somatic symptoms and depression. N Engl J Med 1999;341:1329-35.

18. Tseng W-S. From peculiar psychiatric disorders through culture-bound syndromes to culture-related specific syndromes. Transcult Psychiatry 2006;43:554-76.

19. Guarnaccia PJ, Lloyd H, Rogler LH. Research on culture-bound syndromes: new directions. Am J Psychiatry 1999;56:1322-7.

20. Manson SM. Culture and major depression. Current challenges in

the diagnosis of mood disorders. Psychiatr Clin North Am 1995:18: 487-501.

21. Bhui K, Bhugra D. Explanatory models for mental distress: implica-tions for clinical practice and research. Br J Psychiatry 2002;181:6-7.

22. Gater R, Tansella M, Korten A et al. Sex differences in the prevalence and detection of depressive and anxiety disorders in general health care settings. Report from the World Health Organization Collab-orative Study on Psychological Problems in General Health Care. Arch Gen Psychiatry 1998;55:405-13.

23. Piccinelli M, Simon G. Gender and cross-cultural differences in so-matic symptoms associated with emotional distress. An internation-al study in primary care. Psychol Med 1997;27:433-44.

24. Paykel E, Andrade L, Njenga F et al. Changes needed in the classi-fication of depressive disorders: options for ICD-11. World Psychia-try 2012;11(Suppl. 1):37-42.


026_030.indd 30 06/07/12 14:04

31

Bipolar disorders in ICD-11

Mania and depression have accompanied human beings throughout our recorded history, and likely even earlier. For example, more than 3,000 years ago, Jewish writings reported extreme mood swings in King Saul, consistent with a bipolar illness (1). In the first century A.D., Aretaeus of Cappadocia described a syndrome similar to what we now consider bipo-lar disorder (2). Additional descriptions of recurrent manic and depressive disorders are replete throughout ancient and more recent medical history (2,3), but it was really not until E. Kraepelin described manic-depressive insanity that this condition was codified (4). Specifically, in the process of de-fining dementia praecox, Kraepelin distinguished it from manic-depressive insanity by the occurrence of recovery be-tween episodes in the latter.

The term “bipolar disorder” was introduced by K. Leon-hard in 1957 in order to distinguish individuals who experi-enced recurrent depression from those who experienced both mania and depression (3). This distinction was vali-dated with family history data shortly thereafter (5); it was further validated by the relative specificity of lithium treat-ment for the manic phase of bipolar disorder, as well as for preventing future affective episodes in bipolar illness (6). Subsequently, bipolar disorders have been codified in essen-tially every diagnostic system based upon the occurrence of manic or hypomanic episodes, even though affected indi-viduals typically spend more time in depression (7).

Bipolar disorders are common psychiatric conditions that occur in up to 3% of the worldwide population; even larger numbers of individuals report various sub-syndromal symp-toms of mania (8-10). Bipolar disorders cause considerable morbidity and mortality and are the 6th leading cause of dis-ability worldwide (11). Up to 15% of bipolar individuals commit suicide, making it perhaps the most lethal psychiat-ric illness (3). Indeed, individuals suffering from bipolar dis-orders exhibit higher rates of a wide range of medical ill-nesses, leading to a shorter lifespan than the general popula-tion (3). Although the pathophysiology of bipolar illness re-mains unknown, recent neuroimaging evidence suggests that it may arise from metabolic abnormalities within specific

Stephen M. StrakowSki

Department of Psychiatry, University of Cincinnati College of Medicine, 231 Albert Sabin Way, Cincinnati, OH 45267-0555, USA

Bipolar disorders are common worldwide and have been recognized throughout human history. Although the ICD-10 and DSM-IV concep-tualizations of bipolar illnesses are similar in many respects, particularly in the focus on mania and hypomania as defining syndromes, there are also differences that limit comparability across countries. In this article, aspects of the diagnosis of bipolar disorders within these two diagnostic systems are reviewed in order to inform considerations regarding the development of ICD-11. Specifically, this paper reviews the current DSM-IV and ICD-10 descriptions of bipolar and related conditions. It then contrasts the two in order to identify changes that might assist clinicians and better align the criteria, while continuing to acknowledge the very long history of features that identify these disorders. The article then reflects on proposed changes in the diagnosis of bipolar disorder in DSM-5 and the corresponding approach to ICD-11.

Key words: Bipolar disorders, mania, hypomania, ICD-10, DSM-5


ventral prefrontal-striatal-thalamic-amygdala networks (12). Overwhelming evidence identifies bipolar disorders as ge-netic conditions, likely arising from interactions among mul-tiple genes, with heritability rates as high as 85% (13). There is little doubt, then, that bipolar disorders are major medical conditions that represent a significant public health problem worldwide.

In general, bipolar disorders appear to occur across ethnic, racial and cultural boundaries at similar rates, although epi-demiological and clinical studies show some variability (6). A component of this variability likely reflects the complexity of the illness. Bipolar disorder is a dynamic condition that ex-hibits a broad range of symptoms spanning affective, cogni-tive, neurovegetative and behavioral domains. These symp-toms occur typically in a continuum within individuals with affective illness, without clear “points of rarity” of specific symptoms that would link them to unequivocally defined spe-cific affective syndromes (14). Because of these symptom con-tinua, a number of investigators have suggested that a “spec-trum” approach might be useful for identifying bipolar cases. Neither of the current diagnostic systems (DSM-IV, ICD-10) uses this approach (15,16). As reviewed previously (6), al-though there may be value in these spectrum models, cur-rently data are lacking which demonstrate that a spectrum approach will improve treatment outcomes, better define prognosis, or advance research efforts, so these models are not yet ready for codification into widely used clinical manu-als. Consequently, at this time, working toward improved clinical utility and alignment of DSM and ICD criteria sets appears to be the best approach when reconsidering diagnos-tic criteria.

With these thoughts in mind, I review the current DSM-IV and ICD-10 descriptions of bipolar and related conditions. I contrast the two and identify changes that might assist clini-cians and better align the criteria, while continuing to ac-knowledge the very long history of features that identify these disorders (namely mania and hypomania). I then reflect on proposed changes in the diagnosis of bipolar disorder in DSM-5 and the corresponding approach to ICD-11.

031_036.indd 31 06/07/12 14:04


BIpolar anD relateD ConDItIons In ICD-10

The ICD-10 includes two main categories relevant to bi-polar disorders as typically conceptualized: Manic Episode (F30) and Bipolar Affective Disorder (F31). In ICD-10, man-ic episode can be coded separately from bipolar affective dis-order. Manic episodes are subcategorized into “hypomania” and “with or without psychosis”.

Hypomania is defined by at least a four-day disturbance characterized by irritable or elevated mood that is “definitely abnormal for the individual” and at least three additional symptoms that cause interference with daily activities, which may include: increased activity or physical restlessness; in-creased talkativeness; difficulty in concentrating and dis-tractibility; decreased need for sleep; increased sexual energy; mild spending sprees or other types of reckless or irrespon-sible behavior; and increased sociability or over-familiarity.

A manic episode is similarly defined, although has addi-tional both primary and secondary symptoms, as well as a different temporal requirement. Namely, a manic episode requires at least one week (unless hospitalization is required) of prominent and sustained elevated, expansive or irritable mood, with three or four (if mood is only irritable) of the fol-lowing symptoms: increased activity or physical restlessness; increased talkativeness (“pressure of speech”); flight of ideas or the subjective experience of thoughts racing; loss of nor-mal social inhibitions resulting in behavior which is inap-propriate to the circumstances; decreased need for sleep; inflated self-esteem or grandiosity; distractibility or constant changes in activity or plans; behavior which is foolhardy or reckless and whose risks the subject does not recognize (e.g., spending sprees, foolish enterprises, reckless driving); and marked sexual energy or sexual indiscretions.

Mania can occur with (F30.2) or without (F30.1) psychot-ic symptoms. In mania with psychosis, the delusions and hallucinations permitted are “other than those listed as typi-cal schizophrenic (i.e. delusions other than those that are completely impossible or culturally inappropriate and hal-lucinations that are not in the third person or giving a run-ning commentary)”. The ICD-10 also recognizes Other Man-ic Episodes (F30.8) and Manic Episode, Unspecified (F30.9), diagnostic guidelines for which are not provided.

Bipolar affective disorder is a separate diagnosis from ma-nia, although it is ultimately dependent on the presence of either a current or past mania or hypomania and at least one other affective episode. Bipolar affective disorder is subcat-egorized based upon the current affective episode (hypoma-nia, mania with or without psychosis, mild or moderate de-pression, severe depression with or without psychosis, mixed episode and in remission). Again, subtypes of Other Bipolar Affective Disorders (F31.8) and Bipolar Affective Disorders, Unspecified (F31.9) are available, but are not operational-ized.

BIpolar anD relateD ConDItIonsIn DsM-IV anD DsM-5 (proposeD)

The DSM-IV identifies three subtypes of bipolar disorder: Bipolar I Disorder (296.0x-296.7), Bipolar II Disorder (296.89), and Bipolar Disorder not Otherwise Specified (NOS; 296.80). Cyclothymic Disorder (301.13) is also included under bipo-lar disorders, but is the subject of another paper in this issue (17), so will not be further reviewed here.

Bipolar I disorder is defined by the occurrence of one or more manic or mixed episodes. A manic episode is defined by a distinct period of an abnormal and persistent elevated, expansive or irritable mood, lasting at least one week, unless the individual is hospitalized. In addition to this change in mood, a DSM-IV manic episode requires three or four (if mood is only irritable) of the following symptoms: inflated self-esteem or grandiosity; decreased need for sleep; more talkative than usual (pressured speech); flight of ideas or feel-ing that thoughts are racing; distractibility; increased goal-directed activity or psychomotor agitation; and excessive involvement in pleasurable activities with a high potential for painful consequences.

These symptoms must be of sufficient severity to cause marked psychosocial impairment or require hospitalization, or psychotic symptoms must be present. The symptoms can-not be due to another medical cause or a mixed episode. In DSM-IV, manic or hypomanic episodes occurring while re-ceiving a somatic antidepressant treatment are not counted toward a diagnosis of bipolar disorder. A mixed episode is defined as the co-occurrence for at least one week of a full manic and a full depressive episode. Depressive episodes commonly occur in bipolar I disorder, but are not required for diagnosis.

Bipolar I disorder is subclassified according to the type of current or most recent episode (hypomanic, manic, or de-pressive episode). A single manic episode with no additional history of affective episodes is coded separately (296.0x) from a manic episode in which past affective syndromes have oc-curred (296.4x), but both are considered bipolar I disorder. Manic and depressive episodes are additionally categorized as mild, moderate, severe with or without psychotic features, in partial or full remission, or unspecified. The presence of seasonal pattern and rapid cycling are identified, but not coded.

Bipolar II disorder is defined by a course of illness charac-terized by at least one episode of hypomania and one or more episodes of depression. Hypomania shares all of the same symptoms of mania, with two primary differences. First, hy-pomania requires only four days of symptoms instead of the one week of mania. Second, hypomania requires an unequiv-ocal change in functioning that is observable by others, but that does not cause marked psychosocial impairment. How-ever, in order to meet criteria for bipolar II disorder, symptoms must be associated with significant distress or impairment in psychosocial functioning, which presumably, then, are relat-ed to the depressive phase of the illness.

031_036.indd 32 06/07/12 14:04


Bipolar disorder NOS includes disorders that exhibit bi-polar features, namely symptoms of mania or hypomania, but do not meet full criteria for either bipolar I or bipolar II disorders. This condition is not fully operationalized.

Several changes to the DSM-IV criteria are being pro-posed, although not yet finalized, for DSM-5 (see www.dsm5.org). The mixed episode will be eliminated, replaced with a mixed features qualifier that can be applied to manic, depres-sive or hypomanic episodes. Suicide risk severity, anxiety severity and seasonal pattern will be added as episode speci-fiers as well. Bipolar I disorder will continue to be defined by the occurrence of at least one manic episode, but criteria for mania will be different. Namely, both the mood criterion of DSM-IV and persistently increased activity or energy must be present. Then at least three additional symptoms (four if only irritable) from a list similar to DSM-IV will be required. Unlike DSM-IV, a manic episode occurring during antide-pressant treatment will count toward a bipolar I diagnosis, if symptoms persist “beyond the physiological effect of that treatment”. Proposed changes in hypomania are similar in that both mood and energy abnormalities must be present. Bipolar II disorder in DSM-5 otherwise remains largely as defined in DSM-IV.

CoMparIson anD ContrastBetween DsM-IV/5 anD ICD-10

The ICD-10 and DSM-IV both identify the occurrence of mania or hypomania as the defining event for a diagnosis of bipolar disorder. The description of mania is very similar be-tween these two systems, which is not surprising given that it is anchored in decades if not centuries of clinical practice and lore (3,18). In DSM-IV, hypomania is distinguished from mania largely by functional impact, whereas in ICD-10, the symptoms differ somewhat between hypomania and mania and both result in functional impairment, albeit to a lesser and marked degree respectively. In both diagnostic systems, the intent is similar in that hypomania is viewed as a less severe expression of mania, with less functional impairment, but otherwise largely the same symptom structure. There are no data to suggest that one or the other of these descriptions is superior, and given the high degree of similarity, it is un-likely such would be forthcoming.

Despite the similarities in defining bipolar disorders by the occurrence of mania or hypomania, and in the definitions of these syndromes, there are two major differences between ICD-10 and DSM-IV. First, ICD-10 codes manic episodes as conditions that are distinct from bipolar affective disorder, whereas DSM-IV conceptualizes even single manic episodes as an expression of bipolar I disorder. Second, DSM-IV dis-tinguishes between bipolar I and bipolar II disorders based upon the occurrence of mania or only hypomania, respec-tively, whereas in ICD-10, these syndromes are essentially combined. Additionally, in ICD-10, the content of psychotic symptoms impacts assignment to bipolar affective disorder or

schizophrenia, whereas in DSM-IV, schizophrenia is concep-tualized more as a diagnosis of exclusion. Finally, DSM-IV takes the somewhat idiosyncratic (and hard to support) posi-tion that mania occurring within the context of antidepres-sant treatment does not “count” toward a bipolar diagnosis; this consideration is not included in ICD-10. The proposal for DSM-5 to eliminate mixed episodes and require increased activation/energy as one of the two defining symptoms for mania represents a change from both of the previous diagnos-tic systems. In the next section we consider these differences when developing diagnostic descriptions for ICD-11.

optIons for ICD-11 CrIterIa for BIpolaranD relateD DIsorDers

Before changing from existing ICD-10 clinical descrip-tions and diagnostic guidelines, it is useful to be reminded of the function of diagnoses (6). In psychiatry, the major goal of identifying a “case” (i.e., making a diagnosis) is to assign treatment and provide a prognosis. A secondary goal is to provide consistency across different treatment sites around the world in order to ensure that an individual diagnosed with bipolar disorder in Calcutta will be similarly diagnosed if presenting with the same symptoms in London or Buenos Aires. Related to this latter goal, structured diagnoses also permit commonly defined groups across research programs to foster advances. As reviewed elsewhere, increases in our understanding of bipolar disorder have not yet led to new information suggesting that the current diagnostic schemes are errant (6). Therefore, when considering recommenda-tions regarding descriptions for bipolar disorders in ICD-11, I focus on three specific questions: a) Can these descriptions be made more useful to clinicians? b) Is it possible to better align the ICD and DSM systems in order to ensure compara-bility across countries (since both are widely used)? and c) Is the criteria set consistent with our current knowledge of bi-polar disorder?

A first step simply relates to nomenclature – to resolve the difference between DSM-IV’s “bipolar disorder” and ICD-10’s “bipolar affective disorder”. The latter term, although perhaps more descriptive, includes redundancy. There is no coded “unipolar affective disorder” (it is instead listed as a depressive disorder) and there is no “bipolar nonaffective disorder”. Consequently, the term “bipolar disorder” is pre-ferred for its precision and conciseness.

Both ICD and DSM systems rely on the occurrence of ma-nia or hypomania to distinguish bipolar disorders from other affective and psychiatric conditions. There is no evidence sug-gesting a better approach at this time. Mania is a complex syndrome that is dynamic and multi-dimensional, including affective, cognitive, neurovegetative and psychosocial com-ponents. It is characterized by extreme mood states, the most common of which are irritability, euphoria and expansiveness (3). Depression or dysphoria also frequently occurs during mania (even in the absence of mixed states) and relatively

031_036.indd 33 06/07/12 14:04


rapid switches among moods (lability) are very common (3). Additionally, previous studies suggest that excessive energy or psychomotor activation occurs in nearly all individuals expe-riencing mania, such that, coupled with the extreme mood states noted, this dyad is central to the diagnosis (3,19). A number of other commonly occurring symptoms are used to better define a reliable diagnosis of mania, including pres-sured speech, flight of ideas/racing thoughts, grandiosity, de-creased need for sleep, distractibility and impulsive behavior. These symptoms have each been found to occur in large ma-jorities of manic samples (3,19,20). When diagnosing mania, it is necessary to identify the temporal association of these symptoms; only if they occur during the same period of time do they cumulatively support this diagnosis. Finally, the symptoms are considered to be present only if they represent a significant change from the individual’s typical behavior. For example, if a person’s usual personality is one of being ebullient and talkative, then these symptoms are not consid-ered part of a manic episode; in contrast, if his/her typical personality is quiet and withdrawn, then they could be so considered.

Although the ICD and DSM systems both rely on mania and hypomania to define bipolar disorders, they differ on whether hypomania occurring alone identifies a separate con-dition than that associated with mania. This distinction codi-fied in DSM-IV as bipolar I and bipolar II disorders does ap-pear to have merit. Although the two conditions have many similarities, there may be differences in antidepressant mono-therapy response (21), neurocognitive measures (21-23), ge-netic effects (21,24), and neuroimaging findings (21,25,26) that support the value of separating individuals who experi-ence only hypomania (bipolar II disorder) from those who experience mania as well as hypomania (bipolar I disorder). Studies contrasting these conditions are relatively rare, as many more studies have combined bipolar I and II disorders. Additional work is needed to further validate this separation. Nonetheless, this additional level of diagnostic refinement has clinical utility for potentially guiding treatment.

Bipolar I disorder, then, can be conceptualized as an epi-sodic mood disorder defined by the occurrence of at least one lifetime episode of mania. An episode of mania is strongly predictive of a typical bipolar course of illness that is com-prised of recurrent manic, hypomanic, depressive or mixed episodes with intervening euthymic periods. Bipolar II disor-der is conceptualized as an episodic mood disorder defined by the occurrence of both hypomania and depression. Be-cause hypomania is often difficult to identify, and it is rare that individuals seek treatment for this phase of illness, in order to improve diagnostic validity, we recommend that two or more episodes of hypomania be required to establish this diagnosis.

It has been recognized for many years that certain indi-viduals appear to express only “unipolar” mania, somewhat inconsistent with the “bipolar” nomenclature. The specific rate of individuals with bipolar disorder who experience only mania appears to be relatively low, probably in the range of

5-16% (3). However, calculating a specific rate is difficult, since not all bipolar depressed subjects come to clinical atten-tion or report untreated depressive episodes, and the age of risk for depression is virtually the entire lifespan so that, tech-nically, unipolar manic disorder could only be diagnosed with certainty posthumously. Investigations have typically failed to find consistent differences in various clinical validating fac-tors (e.g., family history, lithium response, outcome) between bipolar subjects who did or did not experience depression (3). For example, recently, Merikangas et al (27) found relatively few differences in a variety of clinical variables between bipo-lar teens who only experienced mania compared with teens who expressed both mood states. Differences that were ob-served were largely those that would be predicted by the pres-ence of depression. Consequently, at this time, so-called “unipolar” mania seems best considered an atypical variant of bipolar I disorder rather than a separate coded disorder.

The term “bipolar” implies that individuals who experi-ence mania also experience symptoms on an “opposite” (de-pressed) pole. However, depressive symptoms are very com-mon during the course of mania (3,18,19). Moreover, defin-ing boundaries between mixed and “pure” manic or depres-sive episodes has been difficult to achieve in a manner that meaningfully impacts prognosis and treatment decisions (18,28). In DSM-IV, full criteria for both mania and depres-sion are required before diagnosing a mixed episode; this ap-proach appears to be too restrictive, however, limiting effec-tive management of mixed-state individuals (18,28). McEl-roy et al (28) critically reviewed the concept of mixed or dysphoric mania (“mania with mixed features”) and identi-fied that “mixed” mania had specific prognostic and treat-ment implications distinct from “pure” mania. From their review, they proposed that mania accompanied by at least three distinct depressive symptoms represented a mixed case. Mixed depressive episodes have also been described, in which the predominant mood state is depression accompa-nied by symptoms of mania and often psychomotor agitation (29). The specific number of manic symptoms during depres-sion necessary to impact relevant outcome and treatment variables is not known. Clearly, more research is needed looking at the utility of these various approaches to better understand mixed episodes (30).

Although bipolar disorder is defined by the occurrence of mania or hypomania, these syndromes are not always due to bipolar illness. Indeed, manic or manic-like syndromes can occur in response to a wide variety of brain injuries (e.g., stroke), medication or drug exposure (e.g., corticosteroids or cocaine), as well as other psychiatric conditions (e.g., schizoaffective disorder) (31). When reviewing the complete clinical picture for a given individual, then, the occurrence of mania or hypomania suggests a diagnosis of bipolar disorder, but other potential causes must be considered and ruled out. There is no evidence that antidepressants induce mania in people not at risk for bipolar disorder; consequently, antide-pressant-associated mania is considered to result from an underlying latent bipolar disorder.

031_036.indd 34 06/07/12 14:04


By definition, diagnoses of bipolar disorder and schizo-phrenia are mutually exclusive. Because the course, outcome and treatment options for bipolar disorder are typically more positive than for schizophrenia, it seems preferable to favor a bipolar diagnosis until sufficient evidence clearly demands a diagnosis of schizophrenia. Consequently, it is recommended that schizophrenia be considered a diagnosis of exclusion. Moreover, the validity of using the content of delusions to make this distinction across widely variable cultures may be problematic (32,33). So-called first-rank symptoms may have more validity in some populations than others, so weighting these symptoms or any content-based symptoms toward di-agnostic specificity must be considered carefully (32,33).

ConClusIons

Changes should not be made in diagnostic definitions and descriptions just for the sake of making changes. However, I believe that the recommendations made here will aid clini-cians in more easily identifying and managing the complex conditions we call bipolar disorders. Additionally, by better aligning DSM and ICD, diagnosis and management can prog-ress similarly across practice sites, cultures and countries.

It is important to recognize that medical evidence to date has limited precision for many components of psychiatric diagnosis. For example, there are insufficient data to demon-strate that a one week course of mania is different than 6 days or 10 days. In these circumstances, then, it seems reasonable to maintain historical perspectives until better information becomes available.

Future studies that examine more closely how symptom combinations impact treatment and outcome are needed to further refine our diagnostic systems. Until objective neuro-biological markers are identified (i.e., a so-called “gold stan-dard”), these advances, as they deviate from historical ap-proaches, are likely to proceed slowly.

acknowledgement

The author thanks J. Angst, D.J. Miklowitz and H. Grunze for their input in the preparation of an earlier version of this paper.

note

S.M. Strakowski is a consultant to the WHO Working Group on the Classification of Mood and Anxiety Disorders, reporting to the International Advisory Group for the Revi-sion of ICD-10 Mental and Behavioural Disorders. The views expressed in this article are those of the author and, except as specifically noted, do not represent the official policies or po-sitions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.

The DSM-5 website was accessed on May 18, 2012. DSM-5 proposals have not been finalized and are subject to change.

references

1. Ben-Noun L. What was the mental disease that afflicted King Saul? Clinical Case Studies 2003;2:270-82.

2. Angst J, Marneros A. Bipolarity from ancient to modern times: con-ception, birth, and rebirth. J Affect Disord 2001;67:3-19.

3. Goodwin FK, Jamison KR. Manic depressive illness. New York: Ox-ford University Press, 2007.

4. Kraepelin E. Manic-depressive insanity and paranoia. Edinburgh: Livingstone, 1921.

5. Angst J, Perris C. On the nosology of endogenous depression: a comparison of the findings of two studies. Archiv Psychiatr Nerven-krank 1968;210:373-86.

6. Strakowski SM, Fleck DF, Maj M. Broadening the diagnosis of bi-polar disorder: benefits vs. risks. World Psychiatry 2011;10:181-6.

7. Judd LL, Akiskal HS, Schettler PJ et al. The long-term natural his-tory of the weekly symptomatic status of bipolar I disorder. Arch Gen Psychiatry 2002;59:530-7.

8. Narrow WE, Rae DS, Robins LN et al. Revised prevalence estimates of mental disorders in the United States: using a clinical significance criterion to reconcile 2 surveys‘ estimates. Arch Gen Psychiatry 2002; 59:115-23.

9. Judd LL, Akiskal HS. The prevalence and disability of bipolar spec-trum disorders in the US population: re-analysis of the ECA data-base taking into account subthreshold cases. J Affect Disord 2003; 73:123-31.

10. Hirschfeld RM, Holzer C, Calabrese JR et al. Validity of the Mood Disorder Questionnaire: a general population study. Am J Psychiatry 2003;160:178-80.

11. Murray CJ, Lopez AD. The Global Burden of Disease: a comprehen-sive assessment of mortality and disability from diseases, injuries and risk factors in 1990 and projected to 2020. Cambridge: Harvard School of Public Health, 1996.

12. Strakowski SM. Integration and consolidation: a neurophysiologi-cal model of bipolar disorder. In: Strakowski SM (ed). The bipolar brain: integrating neuroimaging and genetics. New York: Oxford University Press, 2012:253-74.

13. Nurnberger JI, Jr. General genetics of bipolar disorder. In: Strakows-ki SM (ed). The bipolar brain: integrating neuroimaging and genet-ics. New York: Oxford University Press, 2012:187-202.

14. Phelps J, Angst J, Katzow J et al. Validity and utility of bipolar spec-trum models. Bipolar Disord 2008;10:179-93.

15. American Psychiatric Association. Diagnostic and statistical manu-al of mental disorders, 4th ed. Washington: American Psychiatric Association, 1994.

16. World Health Organization. ICD-10 classification of mental and behavioral disorders: clinical descriptions and diagnostic guidelines. Geneva: World Health Organization, 1992.

17. Phillips MR. Dysthymia and cyclothymia in ICD-11. World Psychia-try 2012;11(Suppl.1):53-8.

18. Cassidy F, Murry E, Forest K et al. Signs and symptoms of mania in pure and mixed episodes. J Affect Disord 1998;50:187-201.

19. Serretti A, Olgiati P. Profiles of “manic” symptoms in bipolar I, bi-polar II and major depressive disorders. J Affect Disord 2005;84:159-66.

20. Swann AC, Janicak PL, Calabrese JR et al. Structure of mania: de-pressive, irritable, and psychotic clusters with different retrospec-tively-assessed course patterns of illness in randomized clinical trial participants. J Affect Disord 2001;67:123-32.

21. Vieta E, Suppes T. Bipolar II disorder: arguments for and against a distinct diagnostic entity. Bipolar Disord 2008;10:163-78.

22. Simonsen C, Sundet K, Vaskinn A et al. Neurocognitive profiles in

031_036.indd 35 06/07/12 14:04


bipolar I and bipolar II disorder: differences in pattern and magni-tude of dysfunction. Bipolar Disord 2008;10:245-55.

23. Solé B, Martínez-Arán A, Torrent C et al. Are bipolar II patients cognitively impaired: a systematic review. Psychol Med 2011;41: 1791-803.

24. Wang Z, Li Z, Chen J, Huang J et al. Association of BDNF gene polymorphism with bipolar disorders in Han Chinese population. Genes Brain Behav (in press).

25. Chou Y-H, Wang S-J, Lin C-L et al. Decreased brain serotonin trans-porter binding in the euthymic state of bipolar I but not bipolar II disorder: a SPECT study. Bipolar Disord 2010;12:312-8.

26. Ha TH, Ha K, Kim JH et al. Regional brain gray matter abnormalities in patients with bipolar II disorder: a comparison study with bipolar I patients and healthy controls. Neurosci Lett 2009;456:44-8.

27. Merikangas KR, Cui L, Kattan G et al. Mania with and without de-pression in a community sample of US adolescents. Arch Gen Psy-chiatry (in press).

28. McElroy SL, Keck PE, Jr, Pope HG, Jr et al. Clinical and research implications of the diagnosis of dysphoric or mixed mania or hypo-mania. Am J Psychiatry 1992;149:1633-44.

29. Koukopoulos A, Koukopoulos A. Agitated depression as a mixed state and the problem of melancholia. Psychiatr Clin North Am 1999;22:547-64.

30. Maj M. Mixed states and rapid cycling: conceptual issues and op-tions for ICD-11. World Psychiatry 2012;11(Suppl.1):65-8.

31. Sax KW, Strakowski SM. The co-occurrence of bipolar disorder with medical illness. In: Tohen M (ed). Comorbidity in affective disor-ders. New York: Dekker, 1998:213-28.

32. Arnold LM, Keck PE, Jr. Collins J et al. Ethnicity and first-rank symptoms in patients with psychosis. Schizophr Res 2004;67:207-12.

33. Strakowski SM, McElroy SL, Keck PE, Jr et al. Racial influence on diagnosis in psychotic mania. J Affect Disord 1996;39:157-62.

031_036.indd 36 06/07/12 14:04

37

Changes needed in the classification of depressive disorders: options for ICD-11

This paper focuses on the classification and description of non-bipolar depressive disorders in ICD-11. The starting point is the existing classification in ICD-10. In particular, we consider the ICD-10 diagnoses of depressive disorder and recurrent depressive disorder.

We address four principal questions:

• What changes are needed to the definitions of depressive episode and recurrent depressive disorder in the ICD-10 Clinical Descriptions and Diagnostic Guidelines?

• Are separate categories for depressive episode and recur-rent depressive disorder needed?

• How should the overlap of depression, anxiety, and phys-ical complaints be addressed in the Clinical Descriptions and Diagnostic Guidelines?

• How should mood and anxiety disorders be grouped in the ICD-11 classification system to reflect the overlap be-tween them while preserving clinically important distinc-tions?

The relevant diagnoses in ICD-10 are as follows. The first is F32 Depressive Episode, with further digits for mild and moderate depressive episodes, with or without somatic symp-toms, and for severe depressive episode with or without psy-chotic symptoms. The second is F33 Recurrent Depressive Disorder. This has similar additional digits to Depressive Episode, primarily for coding the current episode, with an additional code for Recurrent Depressive Disorder, Currently in Remission.

BaCkgrounD of ICD-10

In earlier versions of the ICD, there were considerable problems regarding mood disorders. Prior to pioneering

EugEnE PaykEl1, laura H. andradE2, Frank njEnga3, MicHaEl r. PHilliPs4

1Department of Psychiatry, University of Cambridge, 18B Trumpington Road, Cambridge, CB2 8AH, UK; 2Department/Institute of Psychiatry, University of São Paulo Medical School, Rua Dr. Ovidio Pires de Campos, 785, CEP 05403-010, São Paulo, Brazil; 3Upper Hill Medical Centre, P.O. Box 73749, Nairobi, Kenya; 4Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine, Shanghai 201108, People’s Republic of China, and Emory University School of Medicine, Atlanta, GA, USA

This paper considers changes required to the classification of non-bipolar depressive disorders in ICD-11, using ICD-10 as a starting point. The new classification needs to recognize recent epidemiological findings, cross-cultural variations in presentation of depression and differ-ences in provision of health care among WHO Member States. Recommended changes in organization include making the main distinction between bipolar and unipolar disorders, and use of only one overall depressive episode diagnosis, with the distinction between first and recur-rent episode to become a subtype or be abandoned. Some redrafting of the criteria for depressive episode is suggested, in order to simplify. Subtypes are discussed, with possible addition of two subtypes (seasonal affective disorder and postpartum depression) and an improved way of handling mixed anxiety-depressive disorders.

key words: Depression, classification, recurrence, subtypes, anxiety, ICD-10, ICD-11


work by Angst and Perris published in the 1960s (1), the distinction between bipolar and unipolar disorder was not recognized. The ICD-8 simply regarded manic-depressive disorder and endogenous depression as synonymous.

The ICD-9 classification of mood disorders was still prob-lematic (2). Although published in 1978, it did not clearly in-corporate the newer knowledge. Affective disorders were scat-tered among five different sections, plus subcategories in four other sections. Hidden within the multiple subcategories of affective psychosis (296) was a weak bipolar-unipolar distinc-tion between 296.1 Manic-Depressive, Depressed (which was in effect unipolar) and 296.3 Manic-Depressive Circular, De-pressed (which was bipolar). Most users did not realize this, so the distinction was in practice very erratically recorded.

The ICD-10 was a considerable step forward. First, the affective disorders were mainly brought together in one sec-tion. Second, the bipolar-unipolar distinction became prom-inent. Third, the subtypes of severity, psychotic, and somatic syndrome, and the distinction between the latter two, were improvements on previous ICD versions. Fourth, dysthymia, which had first been introduced in DSM-III, was incorpo-rated. Fifth, definitions were provided in the official diagnos-tic system.

There are also some disadvantages to the ICD-10 mood disorders section (2). First, the classification is complex. Its fundamental organization is into single episode, recurrent, and persistent forms, although this is not stated explicitly. This organization obscures the whole classification, particu-larly the bipolar-unipolar distinction. The schema is hard to teach to medical students and others new to mental health. The separation into episodic and persistent disorders is use-ful, but the single episode - recurrent distinction is less so.

Secondly, the formulation of the criteria for depressive episode in the Clinical Descriptions and Diagnostic Guide-lines needs some improvement in organization and wording.

037_042.indd 37 06/07/12 14:08


The third disadvantage is the use of two different diagnos-tic sets of criteria, one clinical and the other for research. The Research Criteria omit some disorders and define some oth-ers differently and restrictively, again a problem not always explicit and requiring a very detailed reading. It has often been unclear in research publications which set of criteria is being used. In fact the Research Criteria have not been much used and many researchers in countries officially employing the ICD-10 have preferred to use the successive DSM sys-tems for research purposes.

Lastly, some subtypes of depression which clinicians fre-quently use in practice are entirely absent.

EpIDEmIologICal nEEDs In rEvIsIon

Depression is one of the most prevalent and disabling men-tal disorders. In population-based samples, its prevalence var-ies, probably due to genetic vulnerability, environment risk factors and measurement factors. In a recent report (3), with data from 18 countries, the average lifetime prevalence of DSM-IV major depression episode varied from 14.6% in high-income to 11.1% in low- to middle-income countries. The 12-month prevalence was around 5.5%. Across all coun-tries, the risk period of onset ranged from mid-late adoles-cence to early 40s, with approximately half of cases having their first episode around 25 years, which is consistent with previous findings (4). Female gender and being separated from a partner are consistent socio-demographic correlates.

Major depression is also strongly comorbid with anxiety and substance use disorders. Comorbidity with anxiety ex-tends along a continuum of severity from subsyndromal co-existent depression and anxiety symptoms to the higher end of severity where “diagnosable” depression and anxiety coex-ist. Subthreshold symptoms have been associated with fluc-tuation of symptom severity across the life course due to life events, particularly in women and old age (5).

ClInICal utIlIty

An important goal for ICD-11 is to be a useful vehicle for World Health Organization (WHO) Member States to re-duce the disease burden of mental and behavioural disor-ders. Recent global data demonstrates widespread use of ICD, and also that the most important perceived purpose of the classification is clinical utility (6).

In low-income countries, many of which are in Africa and Asia, and where the majority live on less than a dollar a day, psychiatrists are fewer than one to 1-2 million people. Con-tact with the medical profession in primary care is almost entirely with non-physician medical practitioners, who often have no more than five minutes to spend with each patient. Studies (7) have demonstrated that it is possible to train non-physician primary care workers to recognize and treat effec-tively common psychiatric disorders, as defined using the

modified WHO primary care guidelines. These simplified guidelines also make it practical to align reporting systems with the very limited pharmaceutical supplies available.

Cultural variations are also relevant to utility. In some cul-tures, a patient with depression can be the subject of simul-taneous management by a Christian priest, a traditional healer and a physician, creating little or no conflict in the mind of the patient or his family. In East Africa (7), for ex-ample, although 8-12% of people living in the community and 20-25% of those seen in primary care have a diagnosable depressive illness, most people do not seek help from mod-ern health services, but find solace within their family, tradi-tional or religious healers. Many patients present with mul-tiple physical complaints which are misdiagnosed, leading to expensive and inappropriate investigations.

The ICD-11 mood disorders section should therefore be usable in primary care in low-income countries, appropriate for a brief consultation, useful for dialogue with the specialist and to inform policy, easily linked to essential medical sup-plies and also useful for training purposes.

Cross-Cultural symptom varIatIon

There is evidence of a common core of symptoms in de-pression that extends across cultures and can be elicited by questioning. There is also evidence that psychological ex-pression of depressive symptoms is gradually replacing ear-lier somatic expression in some cultures. Studies from China report that, without changing diagnostic criteria, use of alter-native probe questions that capture different ways in which individuals experience the symptoms of depression can iden-tify depressed individuals who would not otherwise be rec-ognized (8,9). These findings suggest that criteria for depres-sion can be used across cultures, provided they are not ap-plied with undue rigidity. Cultural differences are discussed further in a separate paper in this supplement (10).

proposED ChangEs for DEprEssIvE DIsorDErs In Dsm-5

Several of the proposed changes for DSM-5 are relevant for the revision of the mood disorders section of the ICD-10. The current plan is to divide the mood disorders section of DSM-IV into two separate chapters: depressive disorders and bipolar and related disorders. Within the depressive dis-orders chapter, four new diagnoses are being considered: disruptive mood dysregulation disorder (severe recurrent temper outbursts that start before the age of 10), chronic de-pression (an updated version of dysthymia), premenstrual dysphoric disorder, and mixed anxiety/depression. More-over, the current plan is to reorganize the depressive disor-ders “not otherwise specified” (NOS) category – an overused category that provides minimal clinical information – into separate, more distinctive groups.

037_042.indd 38 06/07/12 14:08


According to current proposals (www.dsm5.org), the ba-sic criteria for major depressive disorder in the DSM-5 will remain the same as those in DSM-IV, but there may be some minor changes. The most controversial recommendation has been the suggestion to remove the bereavement exclusion criterion for major depressive disorder, on the grounds that depressive symptoms after the death of a loved one are not different from those seen after other types of major losses (11). The coding scheme is changed to code psychotic fea-tures independently, allowing for these to be present in indi-viduals who have moderate depression. The window for the postpartum onset specifier is extended to 6 months, in keep-ing with research.

Recognizing that symptoms exist on a continuum of sever-ity, the drafters of DSM-5 have also recommended that di-mensional assessments of depression, mania, anxiety and suicidality augment the dichotomous diagnostic classifica-tion of depressive disorders. Clinicians will be expected to include dimensional assessments of these in the clinical management of the patient every time they make a diagnosis.

These proposed changes are not finalized and modifica-tions being debated are shown in the “hot topics” section of the DSM-5 website. Given the different target users of the two diagnostic systems, some of the recommended changes to DSM-5 may not be relevant for ICD-11.

ChangEs nEEDED to thE ICD-10 ClInICal DEsCrIptIons anD DIagnostIC guIDElInEs for DEprEssIvE DIsorDErs

Depressive disorders are centrally characterized by lower-ing of mood, with closely associated loss of interest and plea-sure. These are accompanied by a number of additional symptoms, including such disturbances of thought content as reduced self esteem, ideas of guilt, pessimistic views of the future, suicidal feelings; somatic disturbances of sleep, ap-petite and weight, usually in the direction of loss, sometimes of increase; impaired activity and function, with decreased energy, tiredness, impaired concentration, loss of libido, im-paired role function in work and social activity.

The usual ways of defining thresholds for clinical disorder and severity have been by number of symptoms and impair-ment of role function. Low mood itself is difficult to quantify. Although research evidence is sparse, what appears to occur clinically is that, as depression becomes more severe, it be-comes more pervasive, leading to increase in number and severity of additional symptoms. There is good evidence (e.g., 12,13) that role function becomes impaired as depression becomes more severe, although the relationship is not one to one. The impairment is most clearly seen at work.

Problems with the characterization of depressive episode in the ICD-10 Clinical Descriptions and Diagnostic Guide-lines arise from the criteria being partly specified at the begin-ning of the section of F32, and later re-specified, amplified, and defined by severity in the descriptions of mild, moderate

and severe depressive episode, requiring the user to refer back to the initial definition for the specification of the ac-tual symptoms to be counted. These problems could easily be handled by some reorganization and simplification.

A further issue is whether severity criteria should be re-tained. ICD-10 defines three levels of severity: mild depres-sion, which requires at least two from the core symptoms of depressed mood, loss of interest and enjoyment, and reduced energy, with two more additional symptoms; moderate de-pression, requiring at least two symptoms from the core plus three additional symptoms; severe depression, requiring all three core symptoms plus at least three additional symptoms. There are additional qualifiers to these levels in terms of se-verity of the symptoms and impairment of role function. The criteria for mild depression also imply a minimum threshold for a depressive episode.

These severity criteria produce additional complexity and are rather rigid. Moreover, it is unlikely that they are used consistently by all workers in the field. However, they do add very usefully to the classification. The threshold for a depres-sive episode is a key feature, since depressed feelings are very common in the general population, and over-diagnosis of clinical disorder needs to be avoided. Degrees of severity are important guides to treatment, since moderate depression re-sponds best to antidepressants, but in the most severe depres-sions electroconvulsive therapy may be a better treatment.

The use of separate ICD-10 Diagnostic Criteria for Re-search is an issue extending to all disorders. The problems are only minor for depressive disorders, while they are greater in some other sections of ICD-10. If separate Research Criteria are to be retained, they should be brought to coincide more with the Clinical Guidelines, although they may be expressed more explicitly. Since this issue, although important, con-cerns the system as a whole rather than depressive disorders, it will not be discussed further in this paper.

Is it appropriate to have role function included in diagnos-tic criteria which are predominantly symptomatic? There is a strong case for its retention. A weakness of all symptom-based definitions is that they fail to define the threshold of intensity at which a symptom is present, something which in practice is difficult to do with rigour. Impaired work function has been embedded in severity measurement of depression since the earliest depression rating scale, the Hamilton Scale (14), and it is included in DSM-IV criteria. Although rating scales, which incorporate severity ratings for individual symptoms, are better than symptom counts for defining severity, their use is not practical for routine clinical practice in global settings.

Depressive episode and recurrent depressive disorder

ICD-10 separates the diagnosis of a single depressive epi-sode from a recurrence. Although this is also present in DSM-IV, the case for its retention is weak, and it would simplify the classification to abandon it as a major classifying feature, including it as a feature at subtype level.

037_042.indd 39 06/07/12 14:08


There have been a very large number of follow-up studies of outcome, relapse and recurrence in depression (e.g., 15). Over a period of years, somewhere between 50% and 80% of first depressive episodes will be followed by at least one recurrence. Prognosis is better and the need for maintenance treatment less after a first episode than after many episodes, but a major diagnostic division in which a high proportion of subjects will ultimately change diagnosis is of limited value.

Related to this issue is the question of persistent milder depression, or dysthymia. When it was introduced into psy-chiatric nosology in DSM-III, it did enable explicit identifica-tion of milder persistent depressions with short remissions, which had been evident in clinical practice. It delineates a form of mood disorder which can have an adverse impact on the life of the sufferer. It has also generated much research (16). However, it has become clear that the majority of pa-tients with dysthymia do also experience more severe depres-sive episodes. There is a spectrum of depressive disorder in some people, from minor to severe, with fluctuations over longitudinal course, variously labelled, running through sub-threshold depression, recurrent brief depression, prodromal symptoms of episodes, full major depressive episodes, in-complete remission with residual symptoms, early relapse and later recurrence. Some revision of diagnoses may be re-quired to encompass this spectrum. Dysthymia is discussed further in a separate paper in this supplement (17).

subtypes of depressive disorder

The inclusion of subtypes of depression within an official classification requires a balance between conflicting needs. Subtypes provide useful information, may guide treatment and illuminate prognosis. However, boundaries of subtypes are often difficult to define. Many depression subtypes have been proposed over the years, and then fallen away as they are found to be less useful. Busy clinicians, not finding sub-types useful for their practice, tend not to employ them, so that information collected from official diagnoses becomes unreliable.

A recent meta-review of depressive subtyping models (18) suggested five molar categories. There are symptom-based subtypes, such as melancholia, psychotic depression, atypi-cal depression and anxious depression; aetiologically-based subtypes, such as adjustment disorders, perinatal depression, drug-induced depression; time of onset-based subtypes, such as seasonal affective disorder; gender-based subtypes (e.g., female depression); and lastly treatment resistant depression.

The ICD-10 opted for relative simplicity. The primary sub-types of depressive disorder included are based on severity: mild, moderate, severe without psychotic symptoms, severe with psychotic symptoms. A fifth digit allows coding of pres-ence or absence of somatic symptoms (DSM melancholia) for mild or moderate depression.

The DSM-IV is more complex, with single episode includ-ing a variety of specifiers: severity/psychotic/in remission,

chronic, with catatonic features, with melancholic features, with atypical features, with postpartum onset. Major depres-sive disorder, recurrent has other longitudinal course speci-fiers, including the seasonal pattern. DSM-5 proposals in-clude all these, plus several others indicating the presence of mixed features (hypomanic symptoms) and the presence and severity of anxiety and suicide risk.

The ICD system is used in a wider variety of settings than the DSM, with varying resources and mental health staff, so ease of use is more important. Subtypes of severity, somatic syndrome (melancholia), psychotic, and catatonic features are discussed in other papers in this supplement (19-21). Anxiety is discussed in a section below.

Atypical depression, included in DSM-IV and DSM-5, is more questionable. At one time the subject of much study and a copious literature (22) and linked to response to mono-amine oxidase inhibitors, the atypical diagnosis is now less used, as antidepressant response appears to be less specific, and the nature of the syndrome less clear. Moreno and An-drade (23), using epidemiological data, found that atypical features tended to occur more in the bipolar than in the uni-polar depression class, which would suggest that the subtype has been misplaced in the past. In the interest of simplifica-tion, there seems to be a good case not to include atypical depression in ICD-11.

For two DSM-IV subtypes, the case for inclusion in ICD-11 is stronger. The first is seasonal affective disorder: this diagnosis is now fully established among clinicians, its epi-demiology and features have been well documented, and it has a specific treatment, bright light (24). The case for inclu-sion as a subtype in ICD-11 is strong.

The second subtype is postpartum depression. Although the specificity may be less strong, the diagnosis is important because of implications for the baby in immediate manage-ment and cognitive development, and for risk in future preg-nancies. For early onset postpartum affective disorder, there is a link with bipolar disorder, and depression may herald increased future risk of bipolar switch. The literature has also extended to a link with depression during pregnancy, and the term perinatal depression is also used (25). The case for in-clusion requires further consideration (26).

Depression and anxiety

Depression and anxiety commonly co-occur, and some anxiety symptoms are frequent in depressive disorders. In milder disorders in the community, as found in general pop-ulation surveys and seen clinically in primary care, mixed and less specific disorders are the most common pictures.

Mixed disorders of depression and anxiety were handled differently in ICD-10 and DSM-IV. ICD-10 has a diagnosis of F41.2 Mixed Anxiety and Depressive Disorder, within F41 Other Anxiety Disorders, in the larger group of Neurotic, Stress-Related and Somatoform Disorders. This is explicitly defined as a mild disorder that is subthreshold with respect

037_042.indd 40 06/07/12 14:08


to the diagnostic requirements for either depressive or anxi-ety disorder. For more severe disorder, the instructions are to code the predominant diagnosis, or both. It should be noted that anxious depression is not the same as agitated depres-sion. The term agitation, although associated with anxious mood, is classically used to describe actual overactivity of an anxious or tense kind.

The DSM-IV and its predecessors viewed mixed anxiety-depressive states as depressive, based on the idea that pres-ence of depression overruled anxiety. DSM-5 present pro-posals include a new mood disorder, mixed anxiety/depres-sion, similar in description to ICD-10. There is also a subtype proposed within major depressive disorder (with anxiety, mild to severe).

The diagnosis of mixed anxiety and depressive disorder is useful for milder disorders, and the case for its retention is strong, either within anxiety disorders or depressive disor-ders. For more severe disorders, neither the ICD-10 nor the DSM-IV solution is satisfactory. The instruction in ICD-10 to code predominant diagnosis is reasonable, but does not document the mixture. The alternative of coding both diag-noses involves specifying two disorders for what is truly a single mixed episode of illness. The DSM-IV solution of hav-ing depression overrule anxiety is arbitrary.

An alternative approach for more severe mixed pictures in ICD-11 would be to assign them to one or other disorder, depending on a judgement of which is more severe, but with a subtype in each for coding the other symptoms (i.e., depres-sive disorder with prominent anxiety symptoms, and anxiety disorder with prominent depressive symptoms). This solu-tion has much in its favour. Anxiety disorders and depressive disorders are very important in any official classification and merit adequate specification.

Depression with physical symptoms

The term “somatic syndrome” in ICD-10 refers to the equivalent of DSM melancholia. There is no separate coding for depression with prominent somatic complaints. DSM-IV has codes within Other Mood Disorders for mood disorder due to a medical condition, and for substance-induced mood disorder. Neither concerns presence of somatic complaints, and in general, the specificity of the mood disorder in either case is not sufficient to justify in a classification aiming to be simpler and with greater clinical utility.

Presentations of depression in which somatic complaints are prominent are common, especially in primary care. This is also influenced by cultural differences, as discussed in a separate paper in this supplement (10). Some languages pri-marily have somatic terms to express lowered mood. There is suggestive evidence that Western cultures have become more attuned to using psychologized terms to express de-pression over the last 50 years, and that similar processes may now be gradually happening in other cultures.

In view of the need to avoid subtype proliferation, and the

contemporary change towards psychologized terms, we do not propose separate coding for such depressions. Once the depressive diagnosis has been established and the patient can be brought to accept it, treatment and prognostic impli-cations are not different from other depressions. The Clinical Descriptions and Diagnostic Guidelines could be modified to incorporate mention of somatic symptoms.

overall organization and grouping of mood disorders

In spite of its advance on ICD-9, further revision of the overall organization of mood disorder diagnoses in ICD-10 is needed to bring it further into line with modern concepts.

The main division that needs to be made is that between bipolar disorder and unipolar depressive disorder, as in DSM-IV. This is discussed in a separate paper in this supplement (27). Bipolar disorders should form a separate group, but within the mood disorders, since there is evidence of com-monality with unipolar disorder in the risk of diagnostic switch of diagnosis when a first manic episode develops, and the symptom differences between bipolar and unipolar de-pression are comparatively weak.

The overriding organization into single episode and recur-rent disorder in ICD-10 is not useful, given the high risk that one will ultimately become the other. These might become subtypes, or might possibly, for simplification, be omitted.

The separation of the persistent disorder of dysthymia is of more value and should preferably be retained, possibly with some changes of terminology to reflect the wider longi-tudinal spectrum.

ConClusIons

A number of options and recommendations for ICD-11 depressive disorders arise:

• Consider redrafting the criteria for depressive episode in the Clinical Descriptions and Diagnostic Guidelines, in order to simplify and reorganize. This is recommended.

• The group responsible for ICD-11 as a whole should con-sider abandoning the separate Research Diagnostic Crite-ria or, if retained, make these correspond more closely to the Clinical Descriptions and Diagnostic Guidelines.

• Consider only one overall depressive episode diagnosis. This is recommended. Either make first and recurrent de-pressive episode subtypes or abandon the distinction.

• Consider adding two additional subtypes: seasonal affec-tive disorder and postpartum depression.

• Consider a subtype for depressive disorder (above mild level) with prominent anxiety symptoms, and the equiva-lent for anxiety disorders.

• Consider a subtype for physical/somatic presentation. This is not recommended, in view of the need to avoid subtype proliferation. The possibility of somatic symptoms might

037_042.indd 41 06/07/12 14:08


be added to the Clinical Descriptions and Diagnostic Gui- delines.

• Consider reorganizing the mood (affective) disorders sec-tion of ICD-10 to make the distinction between bipolar and unipolar disorders, in keeping with modern concepts. This is recommended.

note

E. Paykel, L.H. Andrade, F. Njenga and M.R. Phillips are members of the WHO ICD Revision Working Group on the Classification of Mood and Anxiety Disorders, reporting to the International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders. The views expressed in this article are those of the authors and, except as specifi-cally noted, do not represent the official policies or positions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.

The DSM-5 website was accessed on October 24, 2011. DSM-5 proposals have not been finalized and are subject to change.

references

1. Angst J, Perris C. On the nosology of endogenous depression: a comparison of the findings of two studies. Archiv Psychiatr Nerven-krank 1968;210:373-86.

2. Paykel ES. Mood disorders: review of current diagnostic systems. Psychopathology 2002;35:94-9.

3. Bromet E, Andrade LH, Hwang I et al. Cross-national epidemiology of DSM-IV major depressive episode. BMC Med 2011;9:90.

4. Andrade L, Caraveo-Anduaga JJ, Berglund P et al. The epidemiology of major depressive episodes: results from the International Consor-tium of Psychiatric Epidemiology (ICPE) Surveys. Int J Methods Psychiatr Res 2003;12:3-21.

5. Chen LS, Eaton WW, Gallo JJ et al. Empirical examination of current depression categories in a population-based study: symptoms, course, and risk factors. Am J Psychiatry 2000;157:573-80.

6. Reed GM, Mendonça Correia J, Esparza P et al. The WPA-WHO Global Survey of Psychiatrists’ Attitudes Towards Mental Disorders Classification. World Psychiatry 2011;10:118-31.

7. Jenkins R, Kiima D, Okonji M et al. Integration of mental health into primary care and community health working in Kenya: context, ra-tionale, coverage and sustainability. Ment Health Fam Med 2010; 7:37-47.

8. Phillips MR, Shen QJ, Liu XH et al. Assessing depressive symptoms in persons who die of suicide in mainland China. J Affect Disord 2007;98:73-82.

9. Phillips MR, Zhang JX, Shi QC et al. Prevalence, associated disabil-ity and treatment of mental disorders in four provinces in China, 2001-2005: an epidemiological survey. Lancet 2009;373:2041-53.

10. Chakrabarti S, Berlanga C, Njenga F. Cultural issues in the classifica-tion and diagnosis of mood and anxiety disorders. World Psychiatry 2012;11(Suppl.1):26-30.

11. Zisook S, Kendler KS. Is bereavement-related depression different than non-bereavement-related depression. Psychol Med 2007;37: 779-94.

12. Weissman MM, Paykel ES. The depressed woman: a study of social relationships. Chicago: University of Chicago Press, 1974.


14. Hamilton M. Development of a rating scale for primary depressive illness. Br J Soc Clin Psychol 1967;6:278-96.

15. Kennedy N, Abbott R, Paykel ES. Remission and recurrence of de-pression in the maintenance era: long-term outcome in a Cambridge cohort. Psychol Med 2003;33:827-38.

16. WPA Dysthymia Working Group. Dysthymia in clinical practice. Br J Psychiatry 1995;166:174-83.

17. Phillips MR. Dysthymia and cyclothymia in ICD-11. World Psychia-try 2012;11(Suppl.1):53-8.

18. Baumeister H, Parker G. Meta-review of subtyping models. J Affect Disord 2012;139:126-40.

19. Ayuso-Mateos JL, Lopez-García P. Severity of depressive disorders: considerations for ICD-11. World Psychiatry 2012;11(Suppl.1):48- 52.

20. Moussaoui D, Agoub M, Khoubila A. How should melancholia be incorporated in ICD-11? World Psychiatry 2012;11(Suppl.1):69-72.

21. Chakrabarti S. Psychotic and catatonic presentations in bipolar and depressive disorders. World Psychiatry 2012;11(Suppl.1):59-64.

22. Paykel ES. Clinical efficacy of reversible and selective inhibitors of monoamine oxidase A in major depression. Acta Psychiatr Scand 1995;91(Suppl. 386):22-7.

23. Moreno DH, Andrade LH. Latent class analysis of manic and de-pressive symptoms in a population-based sample in São Paulo, Bra-zil. J Affect Disord 2010;123:208-15.

24. Partonen T, Pandi-Perumal SR (eds). Seasonal affective disorder: practice and research. Oxford: Oxford University Press, 2010.

25. Meltzer-Brody S. New insights into perinatal depression: pathogen-esis and treatment during pregnancy and postpartum. Dialogues in Clinical Neuroscience 2011;13:89-100.

26. Figueira ML, Videira Dias V. Postpartum depression and premen-strual dysphoric disorder: options for ICD-11. World Psychiatry 2012; 11(Suppl.1):73-6.

27. Strakowski SM. Bipolar disorders in ICD-11. World Psychiatry 2012; 11(Suppl.1):31-6.

037_042.indd 42 06/07/12 14:08

43

Differentiating depression from ordinary sadness: contextual, qualitative and pragmatic approaches

According to available epidemiological data, major de-pressive disorder is the most common mental disorder, with a lifetime prevalence in the general population ranging from 10 to 25% in women and from 5 to 12% in men (1). Although these figures are frequently quoted in the psychiatric litera-ture, they are viewed by many – both outside and within the psychiatric field – with a certain degree of skepticism.

From outside the psychiatric field, it has been pointed out that “determining when relatively common experiences such as sadness should be considered evidence of some disorder requires the setting of boundaries that are largely arbitrary, not scientific, unlike setting the boundaries for what consti-tutes cancer or pneumonia” (2). From within the psychiatric field, it has been stated that “based on the high prevalence rates identified in both the ECA and the NCS, it is reasonable to hypothesize that some syndromes in the community rep-resent transient homeostatic responses to internal or external stimuli that do not represent true psychopathologic disor-ders” (3).

The latter argument has been articulated by Horwitz and Wakefield (4), according to whom “the DSM definition of major depressive disorder fails to exclude from the disorder category intense sadness that arises from the way human be-ings naturally respond to major losses”. As a consequence, normal sadness is sometimes treated as if it were depressive disorder, which “may undermine normal recovery by dis-rupting normal coping processes and use of informal support networks”.

This criticism is likely to be increasingly endorsed by the public opinion in the years to come. In fact, a recent popula-tion study (5) concluded that the public does not tend to perceive depressive symptoms as an indication of mental dis-order when they occur in the context of adverse life events. It appears, therefore, urgent to articulate a convincing response

Mario Maj

Department of Psychiatry, University of Naples SUN, Naples, Italy

How can depressive disorder be differentiated from “normal” sadness? This paper summarizes three approaches. The first, called “contextual”, assumes that, contrary to “normal” sadness, depression is either not triggered by a life event, or, if triggered by an event, is disproportionate to that event in its intensity and duration, and in the degree of functional impairment it produces. The second approach, called “qualitative”, maintains that there is always a qualitative difference between the experience of depression and that of ordinary sadness, a difference that has been lost as part of the oversimplification of psychopathology occurring during recent decades. The third approach, called “pragmatic”, as-sumes that there is a continuum of depressive states in the general population, and that the boundary between depressive disorder and ordinary sadness has to be based on pragmatic grounds (i.e., on issues of clinical utility). This third approach is adopted by the ICD-10 and DSM-IV, which regard depression as a “disorder” when it reaches a given threshold in terms of severity, duration and degree of suffering or functional impairment, thus deserving clinical attention. This paper examines the weaknesses of the above three approaches, as well as the lessons we can take from each of them, and considers their implications for the development of the ICD-11.

Key words: Depression, normal sadness, mental disorder, clinical utility, context, ICD-10, DSM-IV


to the question “When does depression become a mental dis-order?” This article conceptualizes three approaches to this issue (the “conceptual”, the “qualitative” and the “pragmat-ic”), pointing out their weaknesses, the lessons we may take from each of them, and their implications for the develop-ment of the ICD-11 classification of mental disorders.

The “conTexTual” approach

The “contextual” approach assumes that there is always a basic difference between “true” depression and “normal” sad-ness. The latter is always triggered by a life event and appears to be proportionate to that event. The former is either not triggered by a life event or, if triggered by an event, is dispro-portionate to that event in its intensity and duration, and in the degree of the functional impairment it produces (6). This is the approach endorsed by Wakefield, according to whom the diagnosis of depression should be “excluded if the sadness response is caused by a real loss that is proportional in mag-nitude to the intensity and duration of the response” (7).

This “contextual” approach is certainly appealing to lay people and to several clinicians (although nearly two-thirds of the almost 5,000 psychiatrists participating in a recent WPA-World Health Organization (WHO) survey (8) rejected it), but has some important weaknesses.

First, the presence itself of a depressive state can lead to a significant increase in reports of recent stressful events (9), since many depressed people tend to attribute a significance to events which are in themselves neutral. Not surprisingly, Wakefield himself, in a study based on a community sample, found that as many as 95% of depressive episodes were re-ported by the respondents to have been triggered by a life event (10).

043_047.indd 43 06/07/12 14:06


Second, the presence of a depressive state may expose a person to adverse life events. In fact, the relationship be-tween depression and so-called “dependent” events (i.e., events which can be interpreted as a consequence of the de-pressive state, such as being fired from a job or being left by a fiancé) is much stronger than the relationship between de-pression and other events (11).

Third, whether an adverse life event has been really deci-sive in triggering a depressive state may be difficult to estab-lish in many cases, and in any case would require a subjective judgment by the clinician, likely resulting in poor reliability. This has been well known since the 1930s, when Sir Aubrey Lewis, testing a set of criteria aimed to distinguish between “contextual” and “endogenous” depression, concluded that most depressive cases were “examples of the interaction of organism and environment”, so that “it was impossible to say which of the factors was decidedly preponderant” (12).

Fourth, the few studies comparing definitely situational with definitely non-situational major depressive disorder, de-fined according to Research Diagnostic Criteria (RDC, 13), have reported that the two conditions were not different with respect to demographic, clinical and psychosocial variables (14). Similarly, in a study comparing five groups of depressed patients differing by the level of psychosocial adversity expe-rienced prior to the depressive episode, Kendler et al (15) found that the groups did not differ significantly on several clinical, historical and demographic variables.

Finally, the clinical utility of the proposed contextual exclu-sion criterion in terms of prediction of treatment response appears very uncertain. Currently available research evidence suggests that response to antidepressant medication in major depressive disorder is not related to whether or not the de-pressive state was preceded by a major life event (16). Fur-thermore, interpersonal psychotherapy is based on the as-sumption that depression is often understandably related to a disturbing life event, and that “if the patient can solve the life problem, depressive symptoms should resolve as well” (17). This begs the question of whether we should conclude that all cases in which interpersonal psychotherapy is effec-tive are not “true” cases of depression.

Further research is clearly needed to explore the applica-bility and reliability of a contextual exclusion criterion in the diagnosis of depression, and its clinical utility for the predic-tion of treatment response and clinical outcome. However, at the present state of knowledge, it would be unwise to dis-allow the diagnosis of major depression, in a person meeting the current severity, duration and impairment criteria for that diagnosis, just because the depressive state occurs in the con-text of a significant life event (6).

The “qualiTaTive” approach

The “qualitative” approach assumes that there is always a qualitative difference between the experience of “true” de-pression and that of “normal” sadness. This difference has

been lost as part of the oversimplification of psychopathol-ogy related to the development of operational diagnostic cri-teria over the last four decades (6). According to this ap-proach, the oversimplification has operated at two levels: the level of individual phenomena, where the need for specific professional expertise to discriminate psychopathological manifestations from other expressions of impaired mental well-being has been de-emphasized, and the level of syndro-mal description, where the fact that there is a gestalt of the depressive syndrome, beyond the sum of depressive symp-toms, has been ignored (18,19).

Within this approach, it has been maintained that “normal forms of negative mood, such as despair or sadness, must not be mistaken as depressed mood, characterized by a lack of holothymia and being an emotional feeling only known to depressed persons” (19). Indeed, the notion that depressed mood, at least in endogenous depression, is perceived by the patient as being qualitatively different from the feeling he/she would have following the death of a loved one, has a long tradition (e.g., 20). However, early authors did not clarify how the perception of depressed mood is actually different from the experience of ordinary sadness.

This issue has been addressed in some more recent studies. A good example is a qualitative study by Healy (21). In the first part of the study, a sample of depressed patients were asked to describe in their own words their current mood. “The commonest primary description was of the experience of lethargy and inability to do things, whether because of tiredness, a specific inability to summon up effort, a feeling of being inhibited or an inability to envisage the future.” “The next most common description was of a sense of detachment from the environment.” “The next most common descriptor was of physical changes that were described in terms of feel-ing that the subject was coming down with a viral illness, ei-ther influenza or glandular fever, along with descriptions of aches and pains and, in particular, headaches or numbness of the head or tight bands around the head.” In the second part of the study, the patients were asked to select, from a list of adjectives, those best describing their current mood. The words from the list that were endorsed most frequently were as follows: dispirited, sluggish, wretched, empty, washed out, awful, dull, bothered, listless, tightened up, exhausted, gloomy, burdened. The author emphasizes that both the descriptions and the selected adjectives are different from those that peo-ple who are simply miserable or unhappy would be likely to offer spontaneously (21).

Along the same line, some studies carried out in patients with severe or chronic physical illness have described the differential features between true depression and “demoral-ization”. According to Clarke and Kissane (22), a depressed person has lost the ability to experience pleasure generally, whereas a demoralized person is able to experience pleasure when he/she is distracted from thoughts concerning the de-moralizing event. Moreover, the demoralized person feels inhibited in action by not knowing what to do, feeling help-less and incompetent, while the depressed person has lost

043_047.indd 44 06/07/12 14:06


motivation and drive, and is unable to act even when an ap-propriate direction of action is known.

However, taxonomic research on depression, based on Meehl’s taxometric methods and latent class analysis (e.g., 23), has failed to support the idea that a qualitative difference exists between diagnosable depression and non-pathological mood states, arguing instead in favour of a continuum of depressive states, but with the possible existence of a sub-type, roughly corresponding to melancholia, which may be qualitatively different (24). It is possible, therefore, that the qualitative differences from “understandable” sadness sug-gested by some studies apply to a subtype of depression rath-er than to major depressive disorder as a whole.

Further research is needed to explore the subjective experi-ence of depressed persons, and the differences between this experience and that of ordinary sadness. A more precise char-acterization of individual depressive symptoms is needed, as well as an exploration of the differential predictive value of individual symptoms and specific symptom clusters. Further studies on the validity and the clinical utility of the construct of melancholia are also warranted. For the time being, how-ever, it would be hard to maintain that the difference between depression and ordinary sadness is fundamentally a qualita-tive one rather than one of degree or severity (6).

The “pragmaTic” approach

The “pragmatic” approach assumes that there is a con-tinuum of depressive states in the general population, and posits that the boundary between depression and “normal” sadness should be based on pragmatic grounds, i.e., on issues of clinical utility (6). This is what both the DSM-IV and the ICD-10 currently try to achieve, regarding depression as a “disorder” when it reaches a given threshold in terms of se-verity, duration and degree of suffering or functional impair-ment, thus deserving clinical attention.

The threshold fixed by the DSM-IV for the diagnosis of major depression, however, is not only arbitrary, but also not based on reasonably solid pragmatic grounds. In particular, the number of depressive symptoms required by the DSM-IV criteria (at least five, one of which must be either depressed mood or loss of interest or pleasure) does not currently have an acceptable level of empirical support.

An increasing number of DSM-IV depressive symptoms has been found to correlate in a monotonic fashion with a greater risk for future depressive episodes, a greater func-tional impairment, a higher physical comorbidity, and a more frequent family history of mental disorders (25). When a point of rarity has been reported, it usually corresponded to a threshold higher than that fixed by the DSM-IV. For in-stance, Kendler and Gardner (26) found that the risk for fu-ture depressive episodes was substantially greater in subjects with seven or more symptoms than in those with six symp-toms, and Klein (27) reported that the risk for mood disorder was significantly higher in relatives of patients with six or

more depressive symptoms than in both those with four or five symptoms and those with non-affective disorder.

The notion that the threshold fixed by the DSM-IV may be too low is also supported by some research concerning pre-diction of response to pharmacological treatment. Paykel et al (28) found that the superiority of amitriptyline over placebo was more substantial when the initial score on the 17-item Hamilton Rating Scale for Depression (HRSD-17) was be-tween 16 and 24, less substantial when it was between 13 and 15, and non-significant when it was between 6 and 12. The authors reported that 13% of patients with RDC major de-pression were among those with HRSD-17 scores between 6 and 12, while 34% had a score between 13 and 15. So, almost one half of the patients with a diagnosis of major depression according to RDC (which are almost identical to DSM-IV criteria) were in the groups showing a non-significant or “less substantial” response to pharmacotherapy. Similarly, Elkin et al (29) found that, among patients with an RDC diagnosis of major depressive disorder, those with an initial score of less than 20 on the HRSD-17 (more than 60% of the sample) did not recover more frequently with imipramine than with pla-cebo plus clinical management, whereas patients with an initial score of 20 or more did significantly better.

On the other hand, it has been repeatedly reported that the psychosocial impairment associated with the presence of two to four depressive symptoms is comparable to that as-sociated with the presence of five or more symptoms (e.g., 30). Furthermore, Thase (31) reported a point of rarity be-tween depressed patients seeking outpatient care and age-matched healthy controls corresponding to an HRSD-17 score between 7 and 10 (which is likely to be lower than the threshold fixed by the DSM-IV).

These findings clearly require replication, but seem to sug-gest that more than one threshold may be required in the characterization of depressive states in order to maximize clinical utility. These thresholds may need to be based on the overall severity of depressive symptoms rather than, or in addition to, their number (6).

The ICD-10 approach seems more consistent with the above research evidence than that of the DSM-IV. In fact, the ICD-10 fixes a threshold for mild depressive episode requir-ing the presence of at least four depressive symptoms (includ-ing at least two of the core symptoms of depressed mood, loss of interest and enjoyment, and increased fatiguability), none of which should be present to an intense degree, and a threshold for severe depressive episode requiring the pres-ence of at least seven depressive symptoms, including all the above mentioned core symptoms, some of which should be of severe intensity. It is further specified that “an individual with a mild depressive episode is usually distressed by the symptoms and has some difficulty in continuing with ordi-nary work and social activities, but will probably not cease to function completely”, whereas “during a severe depressive episode it is very unlikely that the sufferer will be able to continue with social, work, or domestic activities, except to a very limited extent”.

043_047.indd 45 06/07/12 14:06


lessons To be TaKen fromThe Three approaches

The lesson that may be taken from the “contextual” ap-proach is that the life context in which a depressive state occurs is important, if not in the diagnosis of depression, at least in the subsequent stage of clinical characterization of the individual case, which is the one that guides the choice of treatment. In particular, the decision whether to imple-ment a psychotherapy and the choice of the type of psycho-therapy should be based on this contextual understanding of the depressive state rather than solely on the clinician’s theo-retical orientation.

The lesson that may be taken from the “qualitative” ap-proach is that we need a more in-depth exploration of the subjective experience of depressed persons and of the differ-ences between this experience and that of ordinary sadness. We may need a more precise description of individual de-pressive symptoms (currently available glossaries appear in-sufficient in this respect). The individual depressive symp-toms, and specific clusters of them, should be examined with respect to their predictive value, especially concerning clini-cal outcome and treatment response. In addition, the valid-ity and clinical utility of the construct of melancholia should be further explored.

The lesson we can take from the “pragmatic” approach is that our clinical practice and our research on depression should be informed by the currently predominant evidence of the existence of a continuum of depressive states. Depression should not be conceptualized and presented as an all-or-none disease, and biological research based on that assumption is probably bound to fail. The threshold for a depressive state requiring clinical attention may be lower than that fixed by the DSM-IV, but the threshold for a depressive state requiring pharmacological treatment is likely to be higher.

An analogy seems to emerge between depression and some common physical diseases such as hypertension and diabe-tes, which also occur on a continuum, with at least two iden-tifiable thresholds: one for a condition deserving clinical at-tention and another for a state requiring pharmacological intervention. The refinement of thresholds for the diagnosis of mild depressive episode and severe depressive episode rep-resents a priority in the development of the ICD-11 classifica-tion of mood disorders.

note

M. Maj is a member of the WHO International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders and the Chairman of the ICD Revision Working Group on the Classification of Mood and Anxiety Disorders. The views expressed in this article are those of the author and, except as specifically noted, do not represent the official policies or positions of the International Advisory Group, the Working Group on the Classification of Mood and Anx-

iety Disorders, or the WHO.

references

1. American Psychiatric Association. Diagnostic and statistical manu-al of mental disorders, 4th ed. Washington: American Psychiatric Association, 1994.

2. Kutchins H, Kirk SA. Making us crazy. DSM: the psychiatric bible and the creation of mental disorders. New York: Free Press, 1997.

3. Regier DA, Kaelber CT, Rae DS et al. Limitations of diagnostic cri-teria and assessment instruments for mental disorders: implications for research and policy. Arch Gen Psychiatry 1998;55:109-15.

4. Horwitz AV, Wakefield JC. The loss of sadness. How psychiatry transformed normal sorrow into depressive disorder. Oxford: Ox-ford University Press, 2007.

5. Holzinger A, Matschinger H, Schomerus G et al. The loss of sad-ness: the public’s view. Acta Psychiatr Scand 2011;123:307-13.

6. Maj M. When does depression become a mental disorder? Br J Psy-chiatry 2011;199:85-6.

7. Wakefield JC. Diagnosing DSM-IV. 1. DSM-IV and the concept of disorder. Behav Res Ther 1997;35:633-49.

8. Reed GM, Mendonça Correia J, Esparza P et al. The WPA-WHO Global Survey of Psychiatrists’ Attitudes Towards Mental Disorders Classifications. World Psychiatry 2011;10:118-31.

9. Cohen MR, Winokur G. The clinical classification of depressive dis-orders. In: Mann JJ (ed). Phenomenology of depressive illness. New York: Human Sciences Press, 1988:75-96.

10. Wakefield JC, Schmitz MF, First MB et al. Extending the bereave-ment exclusion for major depression to other losses: evidence from the National Comorbidity Survey. Arch Gen Psychiatry 2007;64: 433-40.

11. Williamson DE, Birmaher B, Anderson BP et al. Stressful life events in depressed adolescents: the role of dependent events during the depressive episode. J Acad Child Adolesc Psychiatry 1995;34:591-8.

12. Lewis A. Melancholia: a clinical survey of depressive states. J Ment Sci 1934;80:277-378.

13. Spitzer RL, Endicott J, Robins E: Research Diagnostic Criteria (RDC). New York: Biometrics Research, New York State Psychiatric Insti-tute, 1975.

14. Hirschfeld RMA, Klerman GL, Andreasen NC et al. Situational ma-jor depressive disorder. Arch Gen Psychiatry 1985;42:1109-14.

15. Kendler KS, Myers J, Halberstadt LJ. Should the diagnosis of major depression be made independent of or dependent upon the psycho-social context? Psychol Med 2010;40:771-80.

16. Anderson IM, Nutt DJ, Deakin JFW, on behalf of the Consensus Meeting and endorsed by the British Association for Psychophar-macology. Evidence-based guidelines for treating depressive disor-ders with antidepressants: a revision of the 1993 British Association for Psychopharmacology guidelines. J Psychopharmacol 2000;14: 3-20.

17. Markowitz JC, Weissman MM. Interpersonal psychotherapy: prin-ciples and applications. World Psychiatry 2007;6:149-56.

18. van Praag HM. The diagnosis of depression in disorder. Aust N Z J Psychiatry 1998;32:767-72.

19. Helmchen H, Linden M. Subthreshold disorders in psychiatry: clin-ical reality, methodological artifact, and the double-threshold prob-lem. Compr Psychiatry 2000;41(Suppl. 1):1-7.

20. Kiloh LG, Garside RF. The independence of neurotic depression and endogenous depression. Br J Psychiatry 1963;109:451-63.

21. Healy D. Dysphoria. In: Costello CG (ed). Symptoms of depression. New York: Wiley 1993:23-42.

22. Clarke DM, Kissane DW. Demoralization: its phenomenology and importance. Aust N Z J Psychiatry 2002;36:733-42.

23. Ruscio J, Ruscio AM. Informing the continuity controversy: a taxo-metric analysis of depression. J Abnorm Psychol 2000;109:473-87.

24. Grove WM, Andreasen NC, Young M et al. Isolation and character-

043_047.indd 46 06/07/12 14:06


ization of a nuclear depressive syndrome. Psychol Med 1987;17:471-84.

25. Kessler RC, Zhao S, Blazer DG et al. Prevalence, correlates, and course of minor depression and major depression in the National Comorbidity Survey. J Affect Disord 1997;45:19-30.

26. Kendler KS, Gardner CO, Jr. Boundaries of major depression: an evaluation of DSM-IV criteria. Am J Psychiatry 1998;155:172-7.

27. Klein DN. Symptom criteria and family history in major depression. Am J Psychiatry 1990;147:850-4.

28. Paykel ES, Hollyman JA, Freeling P et al. Predictors of therapeutic

benefit from amitriptyline in mild depression: a general practice placebo-controlled trial. J Affect Disord 1988;14:83-95.

29. Elkin I, Shea T, Watkins JT et al. National Institute of Mental Health Treatment of Depression Collaborative Research Program. General effectiveness of treatments. Arch Gen Psychiatry 1989;46:971-82.

30. Broadhead WE, Blazer DG, George LK et al. Depression, disability days, and days lost from work in a prospective epidemiologic survey. JAMA 1990;264:2524-8.

31. Thase ME. Evaluating antidepressant therapies: remission as the optimal outcome. J Clin Psychiatry 2003;64(Suppl. 13):18-25.

043_047.indd 47 06/07/12 14:06


Severity of depressive disorders: considerationsfor ICD-11

Since the publication of the DSM-IV and ICD-10, there has been an extensive debate centering on how these systems have considered the classification of depressive disorders. However, very little research has focused on the way in which the severity levels of depressive symptomatology have been formulated in the two systems, despite the fact that, for clini-cians, assessment of the severity of the clinical picture is a central aspect of evaluation and a decisive element in treat-ment management. Severity of depression has been reported to predict the probability of recovery (1), and to be related to the occurrence of suicidal ideation (2) and the length of de-pressive episodes (3,4).

Severity of depression has been found to have a close re-lationship with treatment response to antidepressants. The pioneering study of Paykel et al (5) provided evidence that amitriptyline was superior to placebo in probable or definite major depression, but not minor depression, as defined by the Research Diagnostic Criteria (6). Amitriptyline was also superior to placebo in subjects with initial scores on the Hamilton Depression Scale of 13-15, and of 16 or more, but not in those with lower scores. Later research provided fur-ther evidence of the relationship between clinical severity and response to pharmacological treatment (7), reporting that severity and duration of the index episode predicted out-come in the short-term treatment of major depression.

This paper reviews how severity has been considered in the DSM and ICD systems. We summarize ICD-10 and DSM- IV approaches and problems arising from their application, the proposals for DSM-5 and options to be considered for ICD-11. We specifically consider in which way distinctions on severity of depressive symptomatology could have an im-pact on clinical utility, global applicability and applicability outside specialty mental health settings.

During the review process, and when weighing possible

José Luis Ayuso-MAteos, PiLAr LoPez-GArcíA

Department of Psychiatry, Instituto de investigación de la Princesa (IIS-IP), Universidad Autónoma de Madrid; Centro de Investigación Biomédica en Red de Salud Mental, CIBERSAM, Madrid, Spain

The severity of depression is crucial in its clinical assessment; however, few research studies have focused on how the severity levels of depres-sive symptomatology have been formulated in the two dominant classification systems, ICD-10 and DSM-IV. In this article, we review the ICD-10 and DSM-IV approaches, and problems arising from their consideration of severity levels. The current DSM-IV severity classification of depression – into mild, moderate or severe – is determined according to symptom count, severity of the symptoms, and degree of functional impairment. The Clinical Descriptions and Diagnostic Guidelines of the ICD-10 use symptom count in their severity classification but also refer to functional status. The DSM-5 Mood Disorders Work Group is considering an overall severity dimension for mood disorders. We present five possible options to be considered for the classification of severity of depression in the ICD-11: a) keeping the classification in three levels, as in ICD-10; b) including symptom count, syndromal impairment, and symptom severity in the definition of severity; c) considering an over-all severity dimension, as proposed for DSM-5; d) reducing the categories to mild and moderate/severe; e) developing a new severity classifica-tion system that is not only based on symptom counts, but also gives different weight to some symptoms.

Key words: Depression, severity, DSM-5, ICD-11, classification


options for improving the current classification, we have tak-en into account the fact that only a very small percentage of individuals with mental disorders will ever see a psychiatrist or any other type of mental health professional. Therefore, psychiatrists, clinical psychologists and psychiatric nurses cannot be envisioned as the primary users or sole profes-sional constituency for the ICD classification system – many other professional groups will also be using the classification. This includes other medical specialties, especially primary care physicians, as well as nurses and even lay health care workers, who deliver the majority of primary and mental health care in some developing countries.

DSM-IV anD SeVerIty

In the DSM-IV criteria for major depression, severity is the first of the “episode specifiers”. Episodes can be classified as “mild”, “moderate” or “severe”. A diagnosis of “mild” de-pression should be considered when few, if any, symptoms in excess of the five required to make the diagnosis are present, and symptoms result in only minor functional impairment. “Moderate” depression should be the diagnosis if symptoms or functional impairment are between “mild” and “severe”. Finally, the term “severe” depression is used when nearly all of the possible symptoms are present, and they markedly in-terfere with functioning. The DSM-IV specifies that severe depression can occur with or without psychotic symptoms.

These severity subtypes rely, therefore, on three different measures of severity: number of criteria symptoms, severity of the symptoms, and degree of functional disability. The cri-teria for defining levels of severity were based on expert opin-ion and, as pointed out by Lux et al (8), were neither empiri-cally developed nor subsequently validated. Parker (9-11)

048_052.indd 48 06/07/12 14:07


argues that the DSM-IV describes depressive disorder as a single condition varying by severity, and that this model has not generated any replicable biological correlates, and has not been informative in identifying treatment-specific effects.

Many clinicians are uncomfortable with applying the term “major depression” to individuals who may meet the DSM-IV criteria, but have depression that is short-lived or is rather mild. It was proposed that this issue could be partially cir-cumvented by use of the modifier “mild” after “major depres-sion”. However, this solution would appear to involve a con-tradiction in terms (12).

Very few empirical studies have been conducted to evalu-ate the coherence and validity of the DSM-IV definition of severity of major depression. One of these was by Lux et al (8), in a sample of 1,015 twins from a population-based reg-istry who met criteria for major depression in the year prior to interview. The authors examined the inter-relationships of the three severity measures and their associations with a wide range of potential validators, including demographic factors, risk for future episodes, risk of major depression in the co-twin, characteristics of the depressive episode, the pattern of comorbidity, and personality traits. They found that correla-tions between the three severity measures were significant but moderate. Factor analysis indicated the existence of a general severity factor, although it was not highly coherent. The three severity measures showed differential predictive ability for most of the validators. In their conclusions, the authors stated that severity of major depression as defined by the DSM-IV is a multifaceted and heterogeneous construct. The three proposed severity measures reflect overlapping but partly independent domains with differential validity as as-sessed by a wide range of clinical characteristics. They con-cluded that clinicians should probably use a combination of severity measures as proposed in DSM-IV rather than privi-leging just one of them. This recommendation is relatively similar conceptually to the approach taken in the ICD-10 Clinical Descriptions and Diagnostic Guidelines (13), which include a discussion of the function impact of depression at different levels of severity, as well as several comments about weighting symptoms in making a determination of severity.

Another issue that has been addressed in the literature is the relationship between clinical severity and specific clinical presentations of depressive syndromes. Klerman et al (14) pointed out that there is considerable overlap between the concepts of endogenous and severe depression. The question often arises as to whether a severity judgment would suffice in place of what is usually a qualitative judgment of endoge-neity.

Furthermore, the link between psychotic depression and a severity specifier has been questioned by other authors. Maj et al (15) conducted a longitudinal follow-up study of 452 patients with DSM-III major depression, differentiating those with at least one belief that fulfilled both of the DSM-III prerequisites for a delusion (i.e., being of “delusional pro-portions” and being maintained with “delusional intensity”). These patients were compared to the others with respect to

demographic, historical, and index episode features; time spent in a depressive episode during a prospective observa-tion period; and 10-year outcome. They found that some DSM-IV assumptions concerning psychotic depression were probably unwarranted (i.e., that depression with psychotic features is always “severe”; that in an individual patient, delu-sions will be either all congruent or all incongruent with de-pressed mood; and that mood-incongruent delusions are associated with a poorer prognosis).

ICD-10 anD SeVerIty

One of the major classification principles in ICD-10 is se-verity of the disorder, which was chosen because different treatments are indicated for various levels of severity, and because different levels of health care service must be planned and provided.

The current ICD model for classification of mood disor-ders takes the depressive episode as a categorical entity; the model is unitary, with disorders differentiated largely on the basis of severity. Each depressive episode can be classified as “mild”, “moderate”, “severe”, “severe with psychotic fea-tures”, or “of unspecified severity”. The model uses a list of ten depressive symptoms, and divides major depressive epi-sode into four groups: not depressed (fewer than four symp-toms), mild depression (four symptoms), moderate depres-sion (five to six symptoms), and severe depression (seven or more symptoms, with or without psychotic symptoms).

The Clinical Descriptions and Diagnostic Guidelines of the ICD-10 (13) mention that the differentiation between mild, moderate, and severe depressive episodes rests upon a complicated clinical judgment that involves the number, type, and severity of symptoms present. They explain that the extent of ordinary social and work activities is often a useful general guide to the episode’s likely degree of severity; how-ever, individual, social, and cultural influences that disrupt a smooth relationship between severity of symptoms and so-cial performance are so common, and powerful, that it is unwise to include social performance amongst the essential criteria of severity.

The way the ICD-10 Clinical Descriptions and Diagnostic Guidelines handle severity has been praised, and recom-mended for preservation in future editions (16). There have been very few studies that validate the ICD-10 classification of severity levels of depressive episodes by variables such as biological markers, prognosis, outcomes, and treatment re-sponses. One validation was done by Kessing (17), in a clin-ical sample comprising all psychiatric inpatients in Denmark discharged between 1994 and 1999 with a diagnosis of a single depressive episode, testing the power of the classifica-tion in predicting risk of relapse and suicide. The result was that patients with increasing levels of severity had a signifi-cantly higher risk. Hamalaines et al (18), in a population study in Finland, showed that more severe depressive epi-sodes were associated with a larger use of health services.

048_052.indd 49 06/07/12 14:07


IMpaCt of SeVerIty leVelS

If severity levels really matter in a clinical sense, then we should see some relationship between severity and quality of life (QoL) measures. It is reasonable to hypothesize that the more severe the level of the depressive episode, the more likely the episode will be associated with lower QoL, even though little empirical evidence exists to support this. There are relatively few studies on the distribution of QoL in the depressed population (19). Several studies have provided evidence for a continuum in the impact on QoL as the sever-ity of depression increases (e.g., 20,21), but they have gener-ally used limited sample sizes, and above all they have not focused on the distinction of depression severity according to ICD criteria.

It has been suggested that “mild” depression may be an unnecessary category, because it is limited in time, the impact on QoL that it produces is low, and it would be diagnosed with low frequency in clinical samples, compared with popu-lation samples (22,23). It is therefore relevant to examine whether subjects with mild depression differ significantly in terms of QoL from individuals without a depressive disorder.

We conducted a study in a sample of depressive cases iden-tified in the general population within the Outcomes of De-pression International Network (ODIN) project (24,25) to explore whether the type of depression (single or recurrent episodes) and the different levels of severity (mild, moderate, severe), as defined by ICD-10 Research Criteria, correlate with the impact on an individual’s QoL (26). The main hypothesis was that recurrent episodes and a higher level of severity of depression would have a higher negative impact on QoL.

Regarding type of depression, we found no difference be-tween groups of patients classified as having “single” vs. “re-current” episodes. Concerning levels of severity, we found that classification of depression into the three levels is not a strong predictor of QoL. Participants with mild depressive episodes presented a significant difference in QoL as com-pared to participants without depression. The difference be-tween non-depressed individuals and individuals with mild depression in terms of QoL was much larger than any differ-ence in QoL observed among depressed persons at different levels of severity (Table 1). Moreover, in follow-up studies, mild depressive individuals were five times more likely than

non-depressed persons to have depressive disorder after one year (17,27).

Our findings suggest that, in terms of QoL, only the distinc-tion between the two extremes of depression severity (mild and severe) may be valid. We found large differences in QoL between non-depressed individuals and those with mild de-pression, and smaller but still significant differences in QoL between those with mild and severe depression, but almost no differences in QoL between those with moderate depres-sion and either those with mild or severe episodes of depres-sion. The discontinuity in QoL between mild depression and not having depression, and then the continuity across sever-ity of depression, provides strong support for including the form of depression currently characterized as “mild” in diag-nostic systems and clearly indicates that it merits attention from the health services.

The results of our study support the relationship between level of depression and worsening QoL, but do not support partitioning depressed persons into three severity groups.

DSM-5 propoSalS for SeVerIty of DepreSSIon

The DSM-5 Mood Disorders Work Group is considering an overall severity dimension for mood disorders. The Work Group has proposed several options for measuring the sever-ity of depression. Among them are the use of questionnaires and scales, such as the Patient Health Questionnaire-9 (PHQ-9) and Clinical Global Impression (CGI) (www.dsm5.org). The question of whether a dimension such as the one reflected in the CGI would suffice across major depression is still being addressed.

One of the changes in the DSM-5 in terms of assessing depression severity for the category of major depression is the de-linking of psychotic symptoms and depression sever-ity. In fact, Maj et al (15) have provided evidence that psy-chotic symptoms can occur in the context of a depressive episode of moderate severity.

optIonS for ICD-11

On the basis of the evidence reviewed above, there are

Table 1 Comparison of Short Form - 36 (SF-36) mean scores according to the severity of depression (from 26)

Mild Moderate Severe Control F g 2

Physical functioningRole limitations, physicalBodily painGeneral health, perceptionVitality/energySocial functioningRole limitations, emotionalMental health

73.6±25.254.4±40.759.0±27.951.8±23.632.8±16.956.4±27.044.0±39.246.7±15.0

71.0±27.755.3±43.854.5±31.349.7±24.127.2±18.547.8±29.331.1±36.437.9 ±17.8

67.4±34.051.4±42.752.0±31.147.6±26.819.4±16.938.9±27.415.4±28.031.8±16.9

92.0±15.989.4±26.583.2±21.682.4±18.272.2±17.094.9±15.692.0±23.582.8±12.8

12.323.527.860.2

151.8104.297.6

235.9

.066

.119

.141

.257

.465

.376

.358

.576

All values are expressed as mean±SD. All Fs have p<0.001

048_052.indd 50 06/07/12 14:07


several options that may be considered for how the clinical severity of depression should be addressed in ICD-11.

The first option is to keep the classification as it is now, with three different levels of severity (mild, moderate, se-vere), as defined by the Clinical Descriptions and Diagnostic Guidelines for ICD-10. Making no change in this aspect of the ICD classification would obviously facilitate acceptance among clinicians already familiar with the current approach. Treatment guidelines that base their recommendations on the current operationalization of severity levels (28) would not need to be modified. In relation to health statistics, data comparability across different versions of the ICD would be maximized for World Health Organization (WHO) Member States.

A second option is to follow the recommendations given by Lux et al (8) for the DSM-5, which support the value of a clinical specifier of severity for major depression but with a text that clearly articulates the “loose” or “fuzzy” nature of the severity construct. They suggest that clinicians should take a comprehensive approach when assessing severity, including the number of symptoms, severity of symptoms, and degree of functional impairment, since dropping any one of these three would result in a loss of information. As mentioned before, the Clinical Descriptions and Diagnostic Guidelines of the ICD-10 clearly indicate that clinical judgment of sever-ity levels should be based on the number, type, and severity of symptoms present. The text includes prototypic descrip-tions of the likely functional impairment of individuals with different levels of depression severity. These should be main-tained and refined in a revised version as a model for the clinician of the comprehensive approach recommended.

A third option is to introduce a new dimensional severity scale in the ICD-11, as is being considered for the DSM-5. As has been pointed out (29), the fact that we have competing classifications of depression, with different numbers of cate-gories in each, is itself cogent evidence that natural boundar-ies – if they exist at all – are not obvious, and that, in such a situation, dimensions have several advantages. They are more flexible, and discourage naive assumptions about disease en-tities. An empirically validated dimensional measure could benefit clinical trials addressing interventions for different levels of severity of depression. However, dimensions do not provide the same ease of description and conceptualization as a typology, and it would be somewhat illogical to use a dimensional system for depression, but to use a categorical system for all the other disorders in the classification. The ongoing revision of the ICD-10 Chapter V (F): Mental and Behavioural Disorders is occurring within the context of the revision of the entire ICD-10. The overall classification has uniform requirements for the description of all the various disorders (30). An attempt to use a different system of de-scription for mental and behavioural disorders would be against the general rules of the classification, and undermine the parity of mental health with the rest of medicine for clin-ical, administrative, and funding purposes.

A fourth option is to adopt a simpler version of the severity

levels with a reduced number of categories: one that would cover what is currently defined by mild severity, and another that would combine moderate/severe clinical presentations. There are guidelines that use only these two categories, such as the recent WHO recommendations from the Mental Health Gap Action Programme (mhGAP) to scale up services for people with mental disorders, especially in low- and middle-income countries (28). These recommendations are based on systematic reviews of the best available evidence and consid-eration of values, preferences and feasibility issues from an international perspective. The clinical utility of this simplifica-tion, however, requires further testing.

A fifth option is to develop a new severity classification with levels based not only on symptom counts but giving different weights to some symptoms – such as suicidal behav-ior and psychotic symptoms – that could be, in themselves, indicators of more severe forms. This approach should be based on studies specifically directed towards this question. In the ODIN study (26), we found that different symptoms may have a different impact on QoL of depressive subjects, and adding more symptoms to those with the highest weights may not indicate an increase in severity of depression.

Depression is one of the leading causes of global disease burden and disability, and a major contributor to the mental health treatment gap (28). It has been widely assumed and enshrined in treatment guidelines that different levels of de-pression severity correspond to different indications for treat-ment, but the assumptions underlying the partitioning of de-pression into three levels of severity and the specific methods used for making this determination have received relatively little systematic attention. We hope that the development of the mental and behavioural disorders classification for ICD-11 will make conceptual and empirical contributions in rela-tion to defining depression severity that can support more effective treatment, management strategies, and resource al-location.

acknowledgements

The authors acknowledge support from the Spanish Min-istry of Health, Instituto de Salud Carlos III (FIS:PI 09/295), and from the Spanish Ministry of Science (ACI2009/1010).

note

J.L. Ayuso-Mateos is a member of the WHO International Advisory Group for the Revision of ICD-10 Mental and Be-havioural Disorders and a member of the ICD Revision Working Group on the Classification of Mood and Anxiety Disorders. The views expressed in this article are those of the authors and, except as specifically noted, do not represent the official policies or positions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.

048_052.indd 51 06/07/12 14:07



references

1. Rubenstein LV, Rayburn NR, Keeler EB et al. Predicting outcomes of primary care patients with major depression: development of a depression prognosis index. Psychiatr Serv 2007;58:1049-56.

2. Alexopoulos GS, Bruce ML, Hull J et al. Clinical determinants of suicidal ideation and behavior in geriatric depression. Arch Gen Psychiatry 1999;56:1048-53.

3. Kennedy N, Abbott R, Paykel ES. Longitudinal syndromal and sub-syndromal symptoms after severe depression: 10-year follow-up study. Br J Psychiatry 2004;184:330-6.

4. Melartin TK, Rytsala HJ, Leskela US et al. Severity and comorbidity predict episode duration and recurrence of DSM-IV major depres-sive disorder. J Clin Psychiatry 2004;65:810-9.

5. Paykel ES, Hollyman JA, Freeling P et al. Predictors of therapeutic benefit from amitriptyline in mild depression: a general practice placebo-controlled trial. J Affect Disord 1988;14:83-95.

6. Spitzer RL, Endicott J, Robins E. Research Diagnostic Criteria: ra-tionale and reliability. Arch Gen Psychiatry 1978;35:773-82.

7. Angst J, Amrein R, Stabl M. Moclobemide and tricyclic antidepres-sants in severe depression: meta-analysis and prospective studies. J Clin Psychopharmacol 1995;15(Suppl. 2):16S-23.

8. Lux V, Aggen SH, Kendler KS. The DSM-IV definition of severity of major depression: inter-relationship and validity. Psychol Med 2010; 40:1691-701.

9. Parker G, Manicavasagar V. The current model for depressive disor-ders and its impact on clinical management. In: Parker G, Mani-cavasagar V (eds). Modelling and managing the depressive disor-ders. Cambridge: Cambridge University Press, 2005:3-13.

10. Parker G, Manicavasagar V. Depression sub-typing: independence and interdependence. In: Parker G, Manicavasagar V (eds). Model-ling and managing the depressive disorders. Cambridge: Cambridge University Press, 2005:21-8.

11. Parker G. Through a glass darkly: the disutility of the DSM nosol-ogy of depressive disorders. Can J Psychiatry 2006;51:879-86.

12. Beckham E, Leber W, Youll L. The diagnostic classification of de-pression. In: Beckham E, Leber W (eds). Handbook of depression. 2nd ed. New York: Guilford, 1995:36-60.

13. World Health Organization. The ICD-10 classification of mental and behavioural disorders. Clinical descriptions and diagnostic guidelines. Geneva: World Health Organization, 1992.

14. Klerman GL. Affective disorders and anxiety disorders. Consider-ations for developing ICD-10. In: Mezzich JE, von Cranach M (eds). International classification in psychiatry: unity and diversity. Cam-

bridge: Cambridge University Press, 1988:135-42.15. Maj M, Pirozzi R, Magliano L et al. Phenomenology and prognostic

significance of delusions in major depressive disorder: a 10-year pro-spective follow-up study. J Clin Psychiatry 2007;68:1411-7.

16. Paykel ES. Mood disorders: review of current diagnostic systems. Psychopathology 2002;35:94-9.

17. Kessing LV. Severity of depressive episodes according to ICD-10: prediction of risk of relapse and suicide. Br J Psychiatry 2004;184: 153-6.

18. Hamalainen J, Isometsa E, Laukkala T et al. Use of health services for major depressive episode in Finland. J Affect Disord 2004;79: 105-12.

19. Kruijshaar ME, Hoeymans N, Bijl RV et al. Levels of disability in major depression: findings from the Netherlands Mental Health Survey and Incidence Study (NEMESIS). J Affect Disord 2003;77: 53-64.

20. Da Silva Lima AF, De Almeida Fleck MP. Subsyndromal depression: an impact on quality of life? J Affect Disord 2007;100:163-9.

21. Ravindran AV, Matheson K, Griffiths J et al. Stress, coping, uplifts, and quality of life in subtypes of depression: a conceptual frame and emerging data. J Affect Disord 2002;71:121-30.

22. Dowrick C, Casey P, Dalgard O et al. Outcomes of Depression In-ternational Network (ODIN). Background, methods and field trials. Br J Psychiatry 1998;172:359-63.

23. Keller MB, Klein DN, Hirschfeld RM et al. Results of the DSM-IV mood disorders field trial. Am J Psychiatry 1995;152:843-9.

24. Ayuso-Mateos JL, Vazquez-Barquero JL, Dowrick C et al. Depres-sive disorders in Europe: prevalence figures from the ODIN study. Br J Psychiatry 2001;179:308-16.

25. Ayuso-Mateos JL, Nuevo R, Verdes E et al. From depressive symp-toms to depressive disorders: the relevance of thresholds. Br J Psy-chiatry 2010;196:365-71.


27. Lyness JM, Heo M, Datto CJ et al. Outcomes of minor and subsyn-dromal depression among elderly patients in primary care settings. Ann Intern Med 2006;144:496-504.

28. Dua T, Barbui C, Clark N et al. Evidence-based guidelines for men-tal, neurological, and substance use disorders in low- and middle-income countries: summary of WHO recommendations. PLoS Med 2011;8:e1001122.

29. Kendell RE. The classification of depressions. In: Burrows GD (ed). Handbook of studies on depression. Amsterdam: Elsevier/North-Holland Biomedical Press, 1977:3-19.


048_052.indd 52 06/07/12 14:07

53

Dysthymia and cyclothymia in ICD-11

This paper summarizes the development of the concepts of dysthymia and cyclothymia, discusses the similarities and differences between the DSM-IV criteria for dysthymic dis-order and cyclothymic disorder and the ICD-10 diagnostic guidelines for dysthymia and cyclothymia, identifies several problems in the DSM-IV and the ICD-10 with respect to these categories, and presents options for addressing these problems that should be considered during the development and harmonization of ICD-11 and DSM-5.

HIstorICal baCkgrounD

As described by Brieger and Marneros (1), the terms dys-thymia and cyclothymia were first used in psychiatry in the mid-19th century and subsequently elaborated by Kahlbaum in 1882 to describe mood disorders that included mild forms with a favorable course.

Given Kraepelin’s dominant influence on psychiatric no-sology at the beginning of the 20th century, dysthymia – a term he did not use – fell out of favor and did not return until the advent of DSM-III in the late 1970s. The cyclothymia label, however, was adopted by Kraepelin as a mild form of manic-depressive insanity and subsequently acquired two additional meanings: in 1921 Kretschmer used it to describe a non-pathological temperament (personality) and in 1946 Schneider used it as a synonym for manic-depressive illness. These three distinct usages of cyclothymia have continued to the present, resulting in considerable confusion. Cyclothymic disorder, as originally defined by Kahlbaum (i.e., a mild form of bipolar disorder), was also included in DSM-III. Dysthy-mia and cyclothymia were included in the ICD-10, approved in 1990. The DSM-III criteria for the disorders subsequently underwent substantial revision in DSM-III-R and DSM-IV.

Dysthymic disorder is now a well-established diagnosis, with a substantial literature (2,3). The 1985 Epidemiological Catchment Area study in the United States (4) reported a

Michael R. PhilliPs

Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine, Shanghai 201108, China; Departments of Psychiatry and Global Health, Emory University, Atlanta, GA, USA

Previously conceived as affective personality disorders, more recent formulations of dysthymia and cyclothymia consider them chronic forms of subthreshold mood disorders that may have superimposed full-criteria episodes of depression, mania or hypomania. This paper identifies several problems in the ICD-10 and DSM-IV descriptions of dysthymia and cyclothymia that need to be considered during the development and harmonization of ICD-11 and DSM-5. Proposed revisions to ICD-10 include: a) regrouping dysthymia with depressive disorders and cyclothymia with bipolar disorders (in ICD-10 they are combined under the “persistent mood disorders” rubric); b) more clearly distinguishing these mood disorders from personality disorders; c) specifying that the subthreshold mood disturbance must be persistent (that is, “present more days than not”) for a period of several years; and d) clarifying the situations in which dysthymia and cyclothymia can be comorbid with other mood disorders.

key words: Dysthymia, cyclothymia, depression, mania, hypomania


lifetime prevalence of 3.5%; the 2001 national epidemiologi-cal study in Australia (5) reported a 12-month prevalence of 1.1%, and a large, four-province epidemiological study in China in 2005 (6) reported a one-month prevalence of 2.0%.

Unlike dysthymic disorder, cyclothymic disorder is sel-dom diagnosed in clinical settings, so there are few clinical studies that use this diagnosis (7). Most recent empirical studies about cyclothymia use the “cyclothymic tempera-ment” construct developed and promoted by Akiskal (8-11), which is quite different from the DSM-IV cyclothymic disor-der and the ICD-10 cyclothymia diagnosis. Most epidemio-logical studies that report on subthreshold bipolar condi-tions (including the National Comorbidity Survey Replica-tion (12)) do not distinguish cyclothymic disorder from other subthreshold conditions, so there are few reports of the prevalence of the disorder. A re-analysis of the 2004 Nether-lands Mental Health Survey (13) estimated a lifetime preva-lence of cyclothymic disorder of 0.6%.

ICD-10 anD DsM-IV approaCHes

In DSM-IV, dysthymic disorder is one of the depressive disorders and cyclothymic disorder is one of the bipolar dis-orders. In ICD-10, dysthymia and cyclothymia are combined into a separate group of mood disorders called “persistent mood [affective] disorders” that is distinct from the episodic depressive disorders and bipolar disorders.

The general outlines of the disorders are similar in the two diagnostic systems – persistent subthreshold episodes of mood symptoms that may be punctuated by full episodes of depression, mania or hypomania. Both descriptions require the presence of subjective distress or disability. Both indicate that the conditions develop early in adult life and specify early-onset and late-onset forms. And both indicate that a positive family history of mood disorders is very common, which confirms the differentiation of these conditions from

053_058.indd 53 06/07/12 14:07


personality disorders. There are, nevertheless, three substantive differences in

the descriptions across the two diagnostic systems:

• The DSM-IV criteria specify that there be no full-criteria depressive or manic episodes during the first two years of illness (though full-criteria mood episodes can be superim-posed on ongoing dysthymia or cyclothymia after the first two years of illness), while the ICD-10 diagnostic guide-lines appear to allow full mood episodes at any time during the course of dysthymia or cyclothymia. This makes the ICD-10 guidelines more inclusive but it also makes it more difficult to distinguish dysthymia from chronic depressive episodes that fluctuate in severity.

• Hypomanic episodes are included as part of the DSM-IV definition of cyclothymic disorder, while the ICD-10 defi-nition of cyclothymia only includes subthreshold hypo-manic symptoms. This means that the DSM-IV criteria for dysthymic disorder are more inclusive and, probably, identify a clinically more severe condition.

• DSM-IV provides slightly different criteria for children and adolescents (one-year duration of symptoms versus two-year duration, and allowing irritable symptoms to re-place depressive symptoms in dysthymia), whereas ICD-10 makes no distinction by age. This makes the DSM-IV diagnoses more inclusive in children and adolescents.

Other differences in the descriptions of dysthymia and cy-clothymia in ICD-10 and DSM-IV are related to the style and intended audience of the two manuals; they are not serious impediments to the harmonization of the criteria in the two diagnostic systems:

• The ICD-10 descriptions of dysthymia and cyclothymia indicate that individuals previously diagnosed with cer-tain personality conditions (including “depressive person-ality disorder” and “cycloid personality”) should be in-cluded in the dysthymia and cyclothymia categories, thus collapsing two types of conditions – affective personality disorders and affective disorders – that some observers consider distinct. DSM-IV descriptions of the disorders do not consider the relationship between these two classes of conditions.

• DSM-IV criteria for dysthymic disorder and cyclothymic disorder specifically exclude pervasive mood symptoms that are the direct result of a psychotic disorder, a general medical condition or substance use, whereas in ICD-10 affective symptoms that are secondary to medical condi-tions or substance use are considered in a different chapter and affective symptoms secondary to a psychotic disorder are not specifically excluded.

• DSM-IV defines an inter-episode period (separating two episodes of illness) as “two months without significant symptoms”, while ICD-10 allows weeks or months of nor-mal mood during the long-term course of a single episode of illness.

• DSM-IV specifically indicates the situations in which dual diagnoses (e.g., dysthymia + major depressive disor-der; cyclothymia + bipolar I disorder, or cyclothymia + bipolar II disorder) are appropriate, whereas ICD-10 does not provide specific information about this.

probleMs arIsIng froM tHe applICatIon of ICD-10 anD DsM-IV CrIterIa

There are several major and minor problems in the diag-nostic descriptions for cyclothymia and dysthymia in ICD-10 and DSM-IV that should be addressed during the develop-ment of ICD-11:

a) ICD-10 diagnostic guidelines for cyclothymia and dysthymia give conflicting impressions about whether or not the conditions include episodes that meet the full criteria for manic, hypomanic or depressive episodes

The overall description of the “persistent mood [affective] disorders” group states that “recurrent or single episodes of manic disorder, or mild or severe depressive disorder may become superimposed on a persistent affective disorder”. But the diagnostic guidelines for cyclothymia state that “in-dividual episodes of mood swings do not fulfil the criteria for any of the categories described under manic episode or de-pressive episode” and the definition of hypomania states that “abnormalities of mood and behaviour are too persistent and marked to be included under cyclothymia”.

The intention may have been to indicate that cyclothymia and dysthymia are subthreshold conditions and that the manic, hypomanic and depressive episodes that can occur during cyclothymia or dysthymia are separate diagnoses. But this is not explicitly stated and there is no clarification about whether or not comorbid mood disorder diagnoses (e.g., cy-clothymia and bipolar disorder) are indicated, so the text is quite confusing.

b) the ICD-10 description of cyclothymia and dysthymia obscures the boundary between mood disorders and personality disorders

The overall description states that “persistent affective dis-orders are classified here rather than with the personality disorders because of evidence from family studies that they are genetically related to the mood disorders, and because they are sometimes amenable to the same treatments as mood disorders”, but the diagnostic guidelines for cyclothy-mia state that the diagnosis “includes affective personality disorder, cycloid personality, and cyclothymic personality” and those for dysthymia state that the diagnosis “includes depressive personality disorder”.

It appears that the intention of the ICD-10 guidelines was

053_058.indd 54 06/07/12 14:07


to indicate that individuals who were previously diagnosed as having one of these personality conditions should now be given an affective disorder diagnosis (i.e., either dysthymia or cyclothymia). Historically the two meanings of cyclothymia – as a personality trait (Kretschmer) or as a mild form of manic-depressive insanity (Kraepelin) – were quite distinct and more recently the “cyclothymic temperament” construct promoted by Akiskal (8) remains quite different from cyclo-thymic disorder (an affective disorder). The ICD-10 state-ment that dysthymia and cyclothymia “include” older per-sonality diagnoses confounds the boundary between these two types of conditions.

c) the symptom sets and definitions of symptoms used in DsM-IV for a major depressive episode and for dysthymic disorder are different

A major depressive episode requires depressive affect or anhedonia, while dysthymic disorder requires depressive af-fect and does not even consider anhedonia. Psychomotor agitation and retardation are identified as symptoms of a ma-jor depressive episode, but they are not considered in dysthy-mic disorder. The major depressive episode symptom of recur-rent thoughts of death is replaced by hopelessness in dysthymic disorder, and the major depressive episode symptom of worthlessness or excessive guilt is replaced by low self-esteem. These differences create unnecessary diagnostic confusion.

d) the DsM-IV diagnostic criteria for cyclothymic disorder emphasize the episodic nature of the condition and downplay its persistent nature

In the original descriptions that led to its introduction into DSM-III and ICD-10, cyclothymia was characterized as a persistent condition with short periods of “normality” in-terspersed by multiple periods of depressive or hypomanic affect. However, in the DSM-IV definition the emphasis is on multiple episodes of abnormal affect, not on the persistence of the condition. Over a two year period, “numerous periods with hypomanic symptoms and numerous periods with de-pressive symptoms” could easily mean once a month. So, according to this definition, one day of irritability and one day of poor sleep each month for two years would be enough to merit the cyclothymic disorder diagnosis.

e) subthreshold hypomania + subthreshold depressive symptoms for two years (i.e., “cyclothymic disorder with subthreshold hypomania”) followed by a major depressive episode does not meet criteria of bipolar II disorder, as stated in the DsM-IV cyclothymic disorder criterion C

As discussed in the diagnostic features section of DSM-IV, the hypomanic symptoms in cyclothymic disorder do not

need to meet the full criteria of a hypomanic episode. (“How-ever, it is not necessary that any of the periods of hypomanic symptoms meet either the duration or symptom threshold criterion for a hypomanic episode.” This clarification was added in the DSM-IV-TR version, it was not in the original DSM-IV). Thus, in cases where the hypomanic symptoms of cyclothymic disorder have never met full criteria for a hypo-manic episode, the subsequent occurrence of a major depres-sive episode after two continuous years of cyclothymic disor-der would not result in a comorbid bipolar II diagnosis, be-cause there was no prior hypomanic episode. This situation would result in the uncomfortable comorbid combination of a major depressive disorder superimposed on a cyclothymic disorder (i.e., concurrent unipolar and bipolar diagnoses) or, in the ICD system, a depressive episode superimposed on cyclothymia. The bipolar II diagnosis would only apply if the preceding hypomanic symptoms had (at any point) met the full criteria for a hypomanic episode.

There appears to be a parallel omission in the exclusion criterion (criterion C) for major depressive disorder in DSM-IV. There is no exclusion for cyclothymia in the major depres-sive disorder criteria, so an individual with “cyclothymic dis-order with subthreshold hypomania” (i.e., no prior full hypo-manic episodes) who subsequently (after 2 years) developed a major depressive episode would be given comorbid major depressive disorder and cyclothymic disorder diagnoses.

A related problem occurs in the DSM-IV criterion E for dysthymic disorder, which states that any prior episode of cyclothymic disorder precludes the possibility of diagnosing dysthymic disorder. This exclusion places some individuals who have never experienced a full manic or hypomanic epi-sode permanently in the bipolar spectrum.

f) Hypomania (and subthreshold hypomania) are often not associated with distress or dysfunction, but the diagnosis of cyclothymia in ICD-10 and cyclothymic disorder in DsM-IV requires the presence of distress or dysfunction

It is unclear whether the distress or impairment referred to in the diagnostic systems are continuous or only occur when hypomanic or depressive symptoms are present. Since there is no specific requirement regarding the persistence of the mood symptoms, if distress/impairment are only present when mood symptoms are present, they could be quite infrequent (as little as one or two days a month). Moreover, by definition the symptoms of a hypomanic episode do not cause significant impairment in social or occupational functioning and, since the hypomanic symptoms in cyclothymia and cyclothymic dis-order are usually subthreshold hypomanic symptoms, they would be less severe than those seen in a hypomanic episode. It is unclear how these symptoms could meet the “clinically significant distress or dysfunction” criterion. Subthreshold de-pressive symptoms can, however, be associated with clinically significant distress or dysfunction but that would, presumably, only be during the depressive parts of the cyclothymic cycle.

053_058.indd 55 06/07/12 14:07


In the diagnostic features section of the DSM-IV, this issue is dealt with as follows: “Although some people may function particularly well during some of the periods of hypomania, overall [italics added] there must be clinically significant dis-tress or impairment in social, occupational, or other impor-tant areas of functioning as a result of the mood disturbance”. This description is far from convincing; it remains unclear how these individuals can meet this criterion.

proposals for DsM-5

The proposed changes to DSM-IV posted on the www.dsm5.org website (as of April 1, 2012) include a reorganiza-tion of the DSM-IV mood disorders chapter into two sepa-rate chapters in DSM-5: “depressive disorders” and “bipolar and related disorders”. The DSM-IV “cyclothymic disorder” category would be unchanged and placed within the DSM-5 bipolar and related disorders chapter. Two of the exclusion criteria of the DSM-IV “dysthymic disorder” diagnosis (cri-teria D and E) would be removed, and the renamed category, “chronic depressive disorder (dysthymia)”, would be placed within the DSM-5 depressive disorders chapter.

Removal of DSM-IV dysthymic disorder criterion D (“no major depressive episode has been present during the first 2 years of the disturbance”) would have the result that the three main types of persistent depressive conditions – long-standing sub-threshold depressive symptoms (“pure” dysthy-mia), major depressive disorder in partial remission, and chronic major depressive disorder (i.e., continuously meets full criteria of a major depressive episode) – would all fall under the “chronic depressive disorder (dysthymia)” rubric. The rationale provided for this change is quite convincing: a number of studies comparing different types of persistent de-pressive conditions have failed to identify any meaningful dif-ferences in demographic variables, symptom patterns, treat-ment response or family history (14-17). These studies also show that these chronic depressive states are, in comparison to non-chronic major depression, associated with poorer treatment response, higher long-term morbidity, and greater familial loading for affective disorders; so it is reasonable to combine them into a single diagnostic cluster. This change in criteria also eliminates the need to determine whether or not full criteria for a major depressive episode were fulfilled at any time during the first two years of continuous depressive symptoms (which could be decades earlier).

Removal of DSM-IV dysthymic disorder criterion E (“there has never been a manic episode, a mixed episode, or a hypomanic episode, and criteria have never been met for cyclothymic disorder”) is harder to understand and justify. This change would mean that the DSM-5 “chronic depres-sive disorder (dysthymia)” diagnosis could be given to indi-viduals who previously met criteria for a bipolar disorder if they subsequently have a continuous history of two years of depressive symptoms without intercurrent manic symptoms. This goes against the widely accepted “once bipolar always

bipolar” principle. The proposed changes for DSM-5 do not resolve other

problems in the DSM-IV criteria. Given the differences in symptom sets for major depressive episode and “chronic de-pressive disorder (dysthymia)” (described above), it is un-clear how many of the individuals with chronic major de-pressive disorder or major depressive disorder in partial re-mission in DSM-IV would actually meet the DSM-5 criteria for chronic depressive disorder (dysthymia). There would be a subset (probably a small one) of persons with chronic ma-jor depressive disorder and major depressive disorder in par-tial remission who would not convert to chronic depressive disorder (dysthymia) after two years of continuous symp-toms because they would not meet the symptomatic criteria. Thus, the chronic major depressive disorder diagnosis would need to be retained in DSM-5, or such individuals would need to be diagnosed as having “depressive disorder not else-where classified”. The difference in diagnosis could easily hinge of the presence or absence of a single symptom that the patient would need to recall from the relatively distant past.

The proposed DSM-5 revision does not address the im-portant problem of “double depression”. As in DSM-IV, pa-tients who met criteria for chronic depressive disorder (dys-thymia) who simultaneously meet full criteria for a major depressive episode would be given a comorbid diagnosis of major depressive disorder. The minority of patients who meet DSM-IV criteria for chronic major depressive disorder would, presumably, constantly have both DSM-5 diagnoses. Pa-tients with chronic depressive disorder (dysthymia) whose depressive symptoms wax and wane during the course of ill-ness would alternatively have one or two current diagnoses. This is not an ideal solution to the problem of double depres-sion. The confusion of constantly changing diagnoses and the problem of simultaneously applying two diagnoses from the same diagnostic group might be better resolved by apply-ing the single diagnosis of chronic depressive disorder (dys-thymia) to all persistent depressive disorders and adding a current clinical specifier of “with current major depressive episode” or “without current major depressive episode”.

Another problem that remains unresolved in the proposed DSM-5 approach is the differentiation of separate episodes of illness. The ICD system does not define the criteria that distinguish separate episodes of illness, leaving the decision up to clinical judgment. The diagnostic guidelines for both dysthymia and cyclothymia in ICD-10 appear to indicate that periods of normal affect and functioning of several weeks to several months are part of a single, ongoing episode of illness, not the markers of a new episode of illness. In contrast, the DSM system defines “full remission” as “a minimum of two months without significant symptoms”, so any subsequent recurrence of symptoms after such a two-month period would be considered a new episode of illness.

This method of distinguishing separate episodes of illness is troublesome for many DSM-IV and DSM-5 diagnoses, but is particularly problematic for dysthymic disorder and cyclo-thymic disorder. Based on the DSM-IV criteria and currently

053_058.indd 56 06/07/12 14:07


proposed DSM-5 criteria, a long-term diagnosis of dysthymic disorder (“chronic depressive disorder (dysthymia)” in DSM-5) or cyclothymic disorder would be interrupted by a single two-month period when the patient was feeling somewhat better. If significant symptoms subsequently recurred, the clock would need to start again on the new episode, so two years of continuous subthreshold symptoms without any full-criteria mood episodes would be needed before the dys-thymic disorder or cyclothymic disorder diagnosis could be re-applied. During this two-year period, patients who did not meet full criteria for a major depressive or manic disorder would be in a diagnostic no-man’s land.

The two-month inter-episode period specified in DSM-IV and DSM-5 is relatively short and is not necessarily associ-ated with a return to normal mood – patients can be receiving treatment and they may continue to have some depressive or manic symptoms during the two months – so many of these individuals will be experiencing a single episode of illness with fluctuating severity rather than two distinct episodes of illness. There is no uniform method of defining separate epi-sodes that can be applied across all psychiatric diagnoses, but a more stringent requirement is needed to distinguish sepa-rate episodes of persistent disorders. The inter-episode peri-od should be longer and/or there should be no residual de-pressive, manic or hypomanic symptoms during the period.

reCoMMenDatIons for ICD-11

• Dysthymia should be included as one of the depressive disorders, and cyclothymia should be included as one of the bipolar disorders. This reorganization would also ne-cessitate elimination of the ICD-10 categories of “other persistent mood [affective] disorders” and “persistent mood [affective] disorder, unspecified”.

• To distinguish dysthymia and cyclothymia from frequent, non-persistent episodes of illness, diagnostic guidelines for both conditions need to emphasize the persistence of the mood symptoms more than the frequency of the epi-sodes. The criteria could, for example, specify that “the subthreshold mood symptoms must be present for more days than not over the course of the illness”.

• It is not recommended that ICD-11 follow proposed DSM-5 requirements for dysthymic disorder and cyclothymic disorder of two initial years of subthreshold symptoms without full-criteria mood episodes. The two-year duration is arbitrary, and determining whether or not a patient had a full-criteria episode at some point in the past is often dif-ficult. The ICD approach of allowing full-criteria episodes at any point in the course of dysthymia or cyclothymia is a better fit for the natural waxing and waning of affective disorders, but it may increase the difficulty of differentiat-ing these subthreshold conditions from other types of af-fective disorders. The main feature of dysthymia and cyclo-thymia is the persistence of subthreshold symptoms for a period of years (i.e., lasting “more days than not”), so it is

relatively easy to distinguish these disorders from chronic full-episode conditions in which subthreshold states are absent or only present intermittently. However, it is much more difficult to differentiate dysthymia and cyclothymia from recurrent depressive, manic or hypomanic episodes that have prolonged periods of partial remission following each full-criteria episode. But, in fact, this distinction is not clinically important. Once the depressive or hypomanic symptoms have lasted for years without a clear return to normal, the distinction between intermittent full-criteria episodes with prolonged subthreshold periods of partial remission and prolonged subthreshold symptoms punctu-ated by intermittent full-criteria episodes is irrelevant. Based on this reasoning, it is recommended that ICD-11 diagnostic guidelines clearly state that the dysthymia or cyclothymia diagnoses should be applied once an indi-vidual has had several years of subthreshold depressive or hypomanic symptoms without a clear return to normal, whether or not there are preceding or intervening full-cri-teria depressive, manic or hypomanic episodes.

• ICD-11 needs to be much clearer about dual diagnoses than ICD-10. In the course of dysthymia or cyclothymia, dual diagnoses should be applied whenever the individual meets full criteria for a depressive, manic or hypomanic episode. When the acute symptoms resolve and the individual no longer meets the full criteria for an episode, the diagnosis reverts to dysthymia or cyclothymia alone. An alternative approach that would avoid the necessity of frequently changing the comorbid diagnoses would be to use a clinical specifier to indicate the current state (e.g., “dysthymia with-out current depressive episode” or “dysthymia with current depressive episode”). It would also be possible to add a longitudinal specifier (lifetime or over the prior two years) to dysthymia and cyclothymia that would indicate whether or not the individual had had prior full-criteria episodes (e.g., “dysthymia with persistent subthreshold symptoms” or “dysthymia with intermittent full depressive episodes”).

• It is not recommended that ICD-11 accept the DSM-5 proposal to combine the three types of persistent depres-sive symptoms (persistent subthreshold symptoms, persis-tent full-criteria episode, and persistent alternating sub-threshold symptoms and full-criteria episodes) under a single rubric – “chronic depressive disorder (dysthymia)”. Since the available evidence suggests that these three pat-terns of chronic depressive symptoms are quite similar in course, treatment response and family history, this would be an elegant solution to the complicated problem of as-signing diagnoses to chronic depressive states. But this is a major conceptual change to our current understanding of depression, so more targeted research on this issue is needed to justify such a change.

• It is not recommended that the ICD-11 follow the proposed DSM-5 definition of cyclothymic disorder that includes both full-criteria hypomanic episodes and subthreshold hy-pomanic episodes. The core cyclothymic disorder should consist of subthreshold depressive and subthreshold hy-

053_058.indd 57 06/07/12 14:07


pomanic episodes; full-criteria hypomanic episodes can be superimposed on cyclothymia but, as with full criteria de-pressive or manic episodes, they should result in a dual di-agnosis for the time that the full-criteria episode is present.

• A clearer distinction needs to be made between dysthymia and cyclothymia and the associated personality traits and personality disorders. Part of the confusion stems from ear-lier editions of ICD that used the terms cyclothymic per-sonality and depressive personality disorder. Retention of the old labels in the ICD-10 diagnostic guidelines obscures the boundary between these separate classes of conditions. These references to old labels should be deleted.

• DSM-IV and DSM-5 provide different duration and symp-tomatic criteria for children and adolescents, who only need to have one year of persistent symptoms (versus two years for adults) to be given a diagnosis of dysthymic dis-order or cyclothymic disorder, and in whom the mood disturbance in dysthymic disorder can be irritable instead of depressed. The proposed ICD-11 diagnosis of “disrup-tive mood dysregulation with dysphoria disorder” (18) introduces the concept of irritability as a symptom of dys-phoria in children, but the ICD system has not generally provided age-specific diagnostic guidelines, so, in the in-terest of simplicity, it is not recommended that ICD-11 follow DSM-5 in this regard.

• It is not recommended that the DSM-5 definition of inter-episodes (i.e., minimum of two months without significant symptoms) be adopted. There is no scientific evidence that supports this definition or any other definition of the “point of rarity” between two separate episodes of affec-tive illness, so for the present it remains best to leave this determination up to the clinician.

• DSM-IV criteria and proposed DSM-5 criteria for dysthy-mia and cyclothymia both exclude persistent mood symp-toms that are the result of an underlying psychosis, a gen-eral medical condition, or substance use. The ICD-11 di-agnostic guidelines for dysthymia and cyclothymia could, possibly, also use these exclusion criteria.

• The description of dysthymia needs to clarify the situa-tions in which dysthymia (a pervasive depressive disorder) would be converted into cyclothymia (a pervasive bipolar disorder). There is no clear, evidence-based method of de-ciding what to do when a person with current dysthymia manifests subthreshold hypomanic symptoms. To prevent the inappropriate broadening of the bipolar spectrum (19), it is recommended that a diagnosis of dysthymia only be changed to cyclothymia if the subthreshold hypo-manic symptoms are persistent or recur frequently.

note

M.R. Phillips is a member of the WHO ICD Revision Work-ing Group on the Classification of Mood and Anxiety Disor-ders, reporting to the International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders. The

views expressed in this article are those of the author and, ex-cept as specifically noted, do not represent the official policies or positions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.

The DSM-5 website was accessed on April 1, 2012. DSM-5 proposals have not been finalized and are subject to change.

references

1. Brieger P, Marneros A. Dysthymia and cyclothymia: historical origins and contemporary development. J Affect Disord 1997;45:117-26.

2. WPA Dysthymia Working Group. Dysthymia in clinical practice. Br J Psychiatry 1995;166:174-83.

3. Lima MS, Moncrieff J. A comparison of drugs versus placebo for the treatment of dysthymia. Cochrane Database Syst Rev 2000;2: CD001130.

4. Eaton WW, Kessler LG. Epidemiological field methods in psychia-try: the Epidemiological Catchment Area program. New York: Aca-demic Press, 1985.

5. Andrews G, Henderson S, Hall W. Prevalence, comorbidity, disabil-ity and service utilization: overview of the Australian national men-tal health survey. Br J Psychiatry 2001;178:145-53.

6. Phillips MR, Zhang JX, Shi QC et al. Prevalence, associated disabil-ity and treatment of mental disorders in four provinces in China, 2001-2005: an epidemiological survey. Lancet 2009;373:2041-53.

7. Van Meter A, Youngstrom EA, Youngstrom JK et al. Examining the validity of cyclothymic disorder in a youth sample. J Affect Disord 2011;132:55-63.

8. Akiskal HS. Dysthymia and cyclothymia in psychiatric practice a century after Kraepelin. J Affect Disord 2001;62:17-31.

9. Placidi GF, Signoretta S, Liguori A et al. The Semi-Structured Affective Temperament Interview (TEMPS-I): reliability and psychometric pro- perties in 1010 14-26 year old students. J Affect Disord 1998;47:1-10.

10. Van Vlkenburg C, Kluznik JC, Speed N et al. Cyclothymia and labile personality: is all folie circulaire? J Affect Disord 2006;96:177-81.

11. Oedegaard KJ, Neckelmann D, Benazzi F et al. Dissociative experi-ences differentiate bipolar-II from unipolar depressed patients: the mediating role of cyclothymia and the Type A behaviour speed and impatience subscale. J Affect Disord 2008;108:207-16.

12. Merikangas KR, Akiskal HS, Angst J et al. Lifetime and 12-month prevalence of bipolar spectrum disorder in the National Comorbid-ity Survey Replication. Arch Gen Psychiatry 2007;64:543-53.

13. Regeer EJ, Ten Have M, Rosso ML et al. Prevalence of bipolar dis-order in the general population: a reappraisal study of the Nether-lands Mental Health Survey and Incidence Study. Acta Psychiatr Scand 2004;110:374-82.

14. Blanco C, Okuda M, Markowitz JC et al. The epidemiology of chronic major depressive disorder and dysthymic disorder: results from the National Epidemiologic Survey on Alcohol and Related Conditions. J Clin Psychiatry 2010;71:1645-56.

15. Klein DM, Shankman SA, Lewinsohn PM et al. Family study of chronic depression in a community sample of young adults. Am J Psychiatry 2004;161:646-53.

16. McCullough JP, Klein DN, Keller MB et al. Comparison of DSM-III major depression and major depression superimposed on dysthymia (double depression): validity of the distinction. J Abnorm Psychol 2000;109:419-27.

17. McCullough JP, Klein DN, Borian FE et al. Group comparisons of DSM-IV subtypes of chronic depression: validity of the distinctions. Part II. J Abnorm Psychol 2008;112:614-22.

18. Leibenluft E, Uher R, Rutter M. Disruptive mood dysregulation with dysphoria disorder: a proposal for ICD-11. World Psychiatry 2012; 11(Suppl. 1):77-81.

19. Strakowski SM, Fleck DE, Maj M. Broadening the diagnosis of bi-polar disorder: benefits vs. risks. World Psychiatry 2011;10:181-6.

053_058.indd 58 06/07/12 14:07

59

Psychotic and catatonic presentations in bipolarand depressive disorders

The presence of delusions or hallucinations defines a form of mood disorder referred to as mood disorder with psychotic symptoms. Psychotic symptoms may occur as a part of unipo-lar depression (psychotic depression), or as a part of bipolar disorder, during mania, depression, or mixed episodes.

Mood disorders with psychotic symptoms are much more common than is generally realized. They also appear to be more severe forms of these disorders, which are usually as-sociated with poorer prognosis and greater morbidity than their non-psychotic counterparts (1-8). Moreover, psychotic depression has been associated with a two-fold increase in mortality, compared to non-psychotic depression (9), while the evidence also indicates a similar mortality risk among those with psychotic bipolar disorder (10). Unfortunately, recognition and treatment of psychotic forms of mood disor-der are far from ideal, partly because psychotic symptoms are often subtle, intermittent, or underreported (1-4,6,8). The lack of separate categories for psychotic mood disorders in current classifications has also been blamed for the poor un-derstanding of these disorders (4).

There has been a long-standing debate, particularly with regard to psychotic depression, as to whether psychotic mood disorders represent distinct syndromes meriting independent disorder status, or are merely more severe variants of mood disorders (1-3,6,8). Historically, psychotic depressions were considered as syndromes with pronounced neurovegetative features, with or without delusions and hallucinations (1). This view influenced early versions of existing classifications. Although subsequent versions have introduced delusions and hallucinations as central features, psychotic variants continue to denote severity, rather than a distinct identity. Similarly, though about half of those with catatonia have a mood disor-der, the syndrome has been traditionally linked more often with schizophrenia (11).

This paper reviews research evidence concerning psychot-

Subho Chakrabarti

Department of Psychiatry, Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India

It is debatable whether psychotic and catatonic forms of mood disorders represent distinct categories meriting independent status, or are merely more severe variants of these disorders. Current research evidence indicates that not only are psychotic symptoms quite common in mood disorders, but they are often associated with poorer prognosis, though not invariably so. Catatonia also appears to have a distinct pro-file, and is more commonly associated with mood disorders than with schizophrenia. Existing classifications have accorded the status of specifiers/subtypes to psychotic and catatonic symptoms of mood disorders. They also link psychotic symptoms with severity of mood episodes, and provide only limited information about the unique characteristics of mood disorders with psychosis or catatonia. Proposals for the DSM-5 have attempted to remedy some of these deficiencies. The ICD-11 could benefit by introducing some important changes in this area, such as delinking psychosis from severity, providing greater detail about mood-congruent and mood-incongruent psychotic symptoms, and incorporat-ing descriptions of the unique features of psychotic mood disorders. Additionally, catatonic presentations could be designated separately with criteria that are uniformly applicable to mood, schizophrenic and organic disorders.

Key words: Mood disorders, psychosis, catatonia, ICD, DSM


ic and catatonic presentations of mood disorders, the ap-proaches of the ICD-10 and DSM-IV, the proposals made for the DSM-5, and options for the ICD-11.

ReseaRch evidence

Mood disorders with psychotic symptoms

Prevalence

Psychotic forms of mood disorder are more frequent than once believed. Rates of psychotic depression vary depending on the setting and the type of population studied (3,4). Com-munity studies indicate that 14-19% of those with major de-pressive disorder display delusions and hallucinations (12, 13). The point prevalence rate of psychotic depression is 0.4% (13), and the lifetime prevalence is 0.6% (12). Prevalence rates of 25% have been reported among inpatients (14). Rates of psychotic depression appear to be even higher among the elderly: community samples of those over 60 years of age have yielded prevalence rates of 0.43-3.6%, while rates as high as 53% have been reported among elderly inpatients (3,4).

Although comparatively less data is available on psychot-ic symptoms in bipolar disorder, such symptoms seem to oc-cur at a much greater rate among patients of this group. Both community-based and hospital-based studies have consis-tently found that over half the patients with bipolar disorder have had psychotic symptoms during episodes of mood dis-turbance, with rates as high as 65-78% found among those with mania (5-8). Psychotic symptoms are significantly more frequent in bipolar I than bipolar II disorder, and during manic, rather than depressive episodes (6,8,15). Neverthe-less, psychosis is not uncommon even among bipolar II pa-tients (prevalence rates: 3-45%), while patients with bipolar

059_064.indd 59 06/07/12 14:07


depression are about twice as likely to have psychotic symp-toms than those with unipolar depression (15,16).

Clinical features

Delusions and hallucinations are not necessarily the only symptoms that distinguish psychotic from non-psychotic ma-jor depression. A consistent finding in literature is that spe-cific symptoms appear to be more severe in patients with psy-chotic depression. These primarily include feelings of guilt and worthlessness, psychomotor disturbance (retardation/agitation) and cognitive abnormalities (1-4). These symptoms are characteristically associated with psychotic depression even after controlling for the effects of severity and melan-cholic (formerly referred to as endogenous) symptoms (1). A number of other symptoms such as constipation, insomnia, anorexia, weight loss, anhedonia, anxiety, hypochondriasis and paranoia have also been found more frequently in psy-chotic depression, but these are less pronounced and less con-sistently present (1-3). Despite the presence of such character-istic symptoms, the diagnosis of psychotic depression remains a challenge, because of unclear boundaries between delusions and quasi-delusional symptoms (“sustained preoccupations”) often reported by patients, fluctuations in intensity of psy-chotic symptoms, and concealment by patients (17,18).

The clinical profile of bipolar disorder with psychotic symptoms is less well characterized. While certain studies have found grandiose delusions, hallucinations, mood-incongru-ent psychotic symptoms and catatonia to be more prevalent among those with psychotic mania (6-8), most of the evi-dence suggests the lack of diagnostic specificity of all types of psychotic symptoms in this condition (7,8). Additionally, bipolar depression with psychotic symptoms has been found to be similar to psychotic unipolar depression on a number of clinical variables including age of onset, gender ratio, atyp-ical features, comorbidity, and types of delusions and hallu-cinations (19). This suggests that there is nothing particu-larly unique about psychotic symptoms associated with bipo-lar depression.

Course and outcome

Psychotic unipolar depression is generally characterized by an unfavourable course and poor outcome; these features are often helpful in reliably distinguishing it from non-psy-chotic forms. Retrospective studies and prospective follow-ups of patients with psychotic depression have usually docu-mented longer times to remission, more frequent psychotic recurrences, a greater number of lifetime episodes, greater disability and residual impairment, increased mortality, and increased use of health services. Poorer outcome of psychot-ic depression is mostly evident in the short-term (less than 2 years), whereas long-term outcome is similar to non-psy-chotic depressions (1-4).

On the other hand, psychotic symptoms are of doubtful prognostic significance in bipolar disorder, possibly due to their non-specificity and inability to predict response to lithi-um (5,20). Although certain studies have found greater mor-bidity, increased symptom severity, more frequent recurrences and poorer functioning to be associated with psychotic bipo-lar disorder, particularly mania with psychosis, several others have not confirmed such an association (7,8,15,20,21). The same appears to be true for bipolar depression with psychot-ic symptoms (19). This lends support to the notion that, even though psychotic symptoms are much more frequent in bi-polar disorder, they are poor predictors of outcome in that disorder.

Neurobiology

Apart from its distinctive clinical profile and course, psy-chotic depression differs from depression without psychotic symptoms on a number of other parameters such as neuro-psychological impairment, neurotransmitter and neurohor-monal abnormalities, genetics, and treatment response (1-4).

Though there is a dearth of similar data in bipolar disorder, emerging evidence suggests that psychotic forms of bipolar disorder might also have distinctive genetic and neurocogni-tive profiles (15,22).

Mood-congruent and mood-incongruent psychotic symptoms

Current classifications emphasize the importance of deter-mining whether psychotic symptoms of mood disorder are in keeping with the dominant mood state (23). DSM-IV further asserts that mood-incongruent delusions are infrequent in major depressive disorder and signal a poorer prognosis (23). However, these assumptions have not been borne out by cur-rent research. Firstly, mood-incongruent psychotic symptoms are not as rare in psychotic depression or bipolar disorders, as they are assumed to be (2,3,24-26). Moreover, about half of patients with psychotic mood disorders present with both types of symptoms during the same episode (2,3,24). Al-though mood-incongruent psychosis has been associated with a poorer outcome in some studies of patients with psy-chotic depression, many more have not found any difference in outcome between patients with mood-congruent or mood-incongruent symptoms (3,27). In contrast, there appears to be a stronger and more consistent evidence of poor outcome among patients with mood-incongruent bipolar psychosis, especially from some of the more recently conducted and methodologically sound community-based studies (7,25,26).

Psychosis and symptom severity

The relationship between psychotic symptoms and sever-

059_064.indd 60 06/07/12 14:07


ity of mood episodes is complex and multifaceted. Many studies have reported that the global severity of depressive symptoms among patients with psychotic depression is not significantly different from those with non-psychotic depres-sion (1). Therefore, some have argued that severity of spe-cific symptoms such as guilt or psychomotor retardation, rather than overall severity, better differentiates psychotic from non-psychotic forms (1). However, several recent stud-ies of psychotic depression have actually found greater glob-al depressive symptom severity among such patients than among those without psychosis, leading to the view that overall severity of depressive symptoms is an important dis-criminator (3). A third view is that patients with psychosis tend to display psychotic symptoms only during the more severe episodes of depression, indicating the importance of individual susceptibility in the genesis of psychotic symp-toms (27). Nevertheless, the fact that patients with milder episodes of depression can also present with psychotic symp-toms (13,17) definitely suggests that severity is not the sole determinant of psychosis, and that severity and psychosis probably represent separate dimensions (2,18).

The limited data available for bipolar disorder also indi-cate that, though symptom severity is an important determi-nant, other factors such as young age might be equally sig-nificant determinants of the emergence of psychosis (8, 20).

Distinction from primary psychotic disorders

The question whether psychotic mood disorders can be adequately distinguished from primary psychotic disorders with significant mood symptoms has been the subject of ex-tensive discussions. Although the distinction of psychotic depression from schizoaffective disorder and schizophrenia can be difficult, particularly in young patients during their initial episodes (2,4), the major area of overlap is actually between psychotic bipolar disorder and schizophrenia. Sev-eral recent reviews on the subject have suggested that, though there are substantial similarities between bipolar disorder and schizophrenia in terms of clinical features, neurobiology, treatment response and course, key differences that distin-guish the two disorders exist in the very same areas (see refer-ences 28-30 for a more detailed discussion). Accordingly, there seems to be considerable agreement that, rather than abandoning the existing clinical distinction between bipolar disorders and schizophrenia, efforts should be made to im-prove the validity of this distinction by refining current crite-ria, incorporating dimensional assessments, and including information on other validators, such as family history, as a part of the diagnostic assessment (28).

catatonia and mood disorders

A debate is also ongoing about the status of catatonic syn-dromes and their association with mood disorders. Propo-

nents of the view that catatonia is a distinct subtype of mood disorders point out that catatonia is a well-defined syndrome, distinguishable from other syndromes by a recognizable clus-ter of symptoms, with a typical course and treatment response. Moreover, catatonia is more often associated with mood dis-orders than with schizophrenia (11,31). Therefore, they have argued that catatonia should be included as a separate sub-type of mood disorder (not simply as a specifier), and not linked exclusively with schizophrenia.

Detractors of this approach (32,33) have pointed out that catatonic symptoms are found quite frequently even among patients with schizophrenia, depending on the criteria used to define the condition. Furthermore, catatonia in the context of chronic schizophrenia is phenomenologically different from the acute stuporous presentations that are characteristic of mood disorders (32). Catatonic symptoms of chronic schizophrenia respond better to antipsychotics than benzo-diazepines or electroconvulsive therapy, which are standard treatments for catatonia associated with mood disorders (32,33). Thus, while it would certainly be justified to give more prominence to catatonia in future classifications, grant-ing it independent status is probably not warranted on the basis of current evidence (33).

dsM-iv and icd-10 aPPRoaches

DSM-IV provides a number of mood specifiers to increase diagnostic specificity, create more homogeneous subgroups, assist treatment selection, and aid prognostication (21,23). The conjoint specifier “severe with psychotic features” indi-cates the presence of either delusions or hallucinations dur-ing the current/most recent episode. This denotation implies that psychotic symptoms can only occur during severe de-pressive episodes, not mild/moderate ones, or during manic, and not hypomanic episodes.

Mood-congruent psychotic features are defined as delu-sions and hallucinations with contents that are consistent with typical depressive/manic themes. In contrast, the con-tent of mood-incongruent psychotic symptoms is not direct-ly related to characteristic depressive/manic themes. They are stated to be less common and associated with poorer out-come.

A change introduced in the DSM-IV is the addition of the catatonic features specifier to the mood disorders section. For this specifier to be coded, two of the five catatonic symp-toms must be present, and substance-induced or medically-induced catatonia needs to be ruled out. The catatonic spec-ifier can be applied to mild/moderate/severe episodes of re-current depression and bipolar II disorder, or depressive, manic or mixed episodes of bipolar I disorder.

The ICD-10 Clinical Descriptions and Diagnostic Guide-lines define “psychosis” as consisting of hallucinations, delu-sions and certain severe behavioural abnormalities (e.g., gross excitement and overactivity, marked psychomotor re-tardation, catatonic behaviour). Psychotic symptoms are in-

059_064.indd 61 06/07/12 14:07


cluded as subtypes of mania (not hypomania) and severe (not mild/moderate) depressive episodes. The subtype status is somewhat different from the specifier tag of the DSM-IV. Un-like the DSM, where only certain specifiers need to be coded, all ICD subtypes are integrated into health reporting. Though psychotic symptoms of mania and severe depression are rea-sonably well described in the ICD-10, there is hardly any additional information about specific features or prognostic value of these symptoms. Specification of mood-congruence and mood-incongruence is optional. Mood-incongruent psychotic symptoms are described as affectively neutral; ex-amples are provided only for mania.

Catatonic symptoms (psychomotor retardation/stupor) are mentioned as a part of severe depression with psychosis and mania with psychotic symptoms, while severe overactiv-ity/excitement leading to physical complications is men-tioned under mania with psychosis. In contrast, more detailed descriptions of catatonic symptoms are provided for the sub-types of organic catatonia and catatonic schizophrenia.

Lacunae in dsM-iv and icd-10

Both DSM-IV and ICD-10 provide reasonably compre-hensive descriptions of psychotic and catatonic presenta-tions of mood disorders. Nevertheless, some deficiencies are evident. Both classifications portray psychotic and catatonic mood disorders as severe variants of mood disorders, rather than independent categories. Lack of consistent evidence of their validity and logistic difficulties in creating independent categories appear to justify this approach (1-3). However, current evidence does not support linking of psychotic symp-toms exclusively with severe mood episodes (2,18,21). More-over, the descriptions of psychotic symptoms appear to be inadequate, especially in ICD-10. This is particularly relevant for psychotic depression, where there is compelling evidence for a more frequent occurrence of symptoms such as guilt, psychomotor retardation/agitation and cognitive abnormali-ties (1-4). Additionally, the blurring of boundaries between delusions and “sustained preoccupations” has not got any mention (2,18). Contrary to results of several studies (1), mel-ancholia (somatic symptoms) is not clearly distinguished from psychosis in ICD-10, since it is assumed that all those with psychotic symptoms will also fulfil criteria for melan-cholia.

There is similar lack of detail in descriptions of mood-congruent and mood-incongruent psychotic symptoms. The DSM-IV assumptions that mood-incongruent symptoms are rare and associated with poor prognosis are not in keeping with existing data (1,3,18,21). Additionally, there is no way to code patients with coexisting mood-congruent and mood-incongruent symptoms, who constitute about half of those with psychotic mood disorders (3,26).

The descriptions of catatonia are quite different in the two systems. They do not always tally with the descriptions of catatonic schizophrenia and catatonia due to medical condi-

tions, particularly in the ICD-10. Furthermore, there is a great deal of overlap between catatonic and psychotic symp-toms in the ICD-10, which is somewhat confusing.

PRoPosaLs foR dsM-5

According to current proposals (see www.dsm5.org), DSM-5 will retain the specifier status for both psychotic and catatonic presentations, rather than creating separate sub-types for them. However, several other important changes have been recommended. Firstly, for both major depression and bipolar disorder, the psychotic symptoms specifier will be separated from the severity specifier. Consequently, it will be possible to apply this specifier to moderate depressive episodes. Secondly, at least for depressive episodes, a hierar-chy giving precedence to mood-incongruent features will be introduced, to allow classification of episodes where mood-congruent and mood-incongruent symptoms coexist.

Additionally, major changes have been suggested to the definition and organization of catatonia. The list of catatonic symptoms will be extended from 5 to 12 symptoms, and 3 or more of these will be required for a definite diagnosis of cata-tonia. Other changes have been also put forward, which will have implications for mood disorders. For example, catato-nia will be used as a specifier, and not a subtype of schizo-phrenia. “Catatonic behavior” will be removed from the core diagnostic criteria of schizophrenia. Additionally, it will be possible to use catatonia as a specifier for four other psy-chotic disorders. Similar diagnostic criteria for catatonia will be utilized across all these different categories.

oPtions foR icd-11

Extensive reviews on the subject and the DSM-5 recom-mendations have provided useful hints regarding the possi-ble options for the ICD-11 mood disorders section. To begin with, though there is substantial evidence for a separate des-ignation for psychotic depression, evidence for bipolar disor-der with psychosis is less convincing. Moreover, for both categories of mood disorders, there is prevailing uncertainty concerning the clinical, aetiological and prognostic implica-tions of psychotic symptoms (1-4,21). Finally, incorporation of separate categories for psychotic mood disorders could lead to significant logistic problems (4). For these reasons, it would be better to retain the current status of psychotic symptoms as subtypes of mania and depression. However, episode severity needs to be delinked from psychosis, so that a diagnosis of moderate depression with psychotic symptoms can be made. Conversely, psychotic symptoms need to be clearly distinguished from melancholic (somatic) symptoms in the context of severe depression by incorporating the with/without somatic syndrome distinction for severe de-pressive episodes.

Certain other changes could be made only to the text de-

059_064.indd 62 06/07/12 14:07


scriptions. For example, there is a need to include more de-tailed descriptions of psychotic presentations in the text, par-ticularly focusing on certain features (e.g., guilt, psychomotor disturbances and cognitive abnormalities in psychotic de-pression), the blurred boundaries between delusional and quasi-delusional presentations of psychotic depression, and the implications of psychosis for outcome. More information about characteristic features of mania/depression with psy-chosis in the text might also be helpful in differentiating psy-chotic mood disorders from primary psychoses. The distinc-tion between mood-congruent and mood-incongruent psy-chotic symptoms needs to be elucidated more clearly, and the ICD-10 option to record these symptoms could be extended to include a mixed category when mood-congruent and mood-incongruent symptoms co-occur.

Mania and depression with catatonic symptoms need to be clearly delineated from the psychotic subtypes. More de-tail about catatonic presentations could be included in the text. Alternatively, catatonic subtypes could be included un-der the rubrics of “other manic” (F 30.8) and “other depres-sive” (F 32.8) episodes. Uniform criteria may have to be de-vised for catatonia as a part of mood disorders, schizophre-nia and organic catatonia. These could be similar to DSM-5, or more stringent criteria, which include a minimum dura-tion (11), could be employed.

Finally, all changes made will require field-testing for use-fulness, ease of use and applicability in different settings, in order to ensure fulfilment of the requirements of clinical util-ity and global acceptability, which are the primary objectives of the current process of revision.

note

S. Chakrabarti is a member of the WHO ICD Revision Working Group on the Classification of Mood and Anxiety Disorders, reporting to the International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders. The views expressed in this article are those of the author and, except as specifically noted, do not represent the official policies or positions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.


References

1. Schatzberg AF, Rothschild AJ. Psychotic (delusional) major depres-sion: should it be included as a distinct syndrome in DSM-IV? In: Widiger TA, Frances AJ, Pincus HA et al (eds). DSM-IV sourcebook, Vol. 2. Washington: American Psychiatric Association, 1996: 127-79.

2. Keller J, Schatzberg A, Maj M. Current issues in the classification of psychotic major depression. Schizophr Bull 2007;33:877-85.

3. Lykouras L, Gournellis R. Psychotic (delusional) major depression: new vistas. Curr Psychiatry Rev 2009;5:1-28.

4. Rothschild AJ. Clinical manual for diagnosis and treatment of psy-chotic depression. Washington: American Psychiatric Publishing, 2009.

5. Goodwin FR, Jamison KR. Manic depressive illness. New York: Ox-ford University Press, 1990.

6. Dunayevich E, Keck PE, Jr. Prevalence and description of psychotic features in bipolar mania. Curr Psychiatry Rep 2000;2:286-90.

7. Keck PE, McElroy SL, Havens JR et al. Psychosis in bipolar disor-der: phenomenology and impact on morbidity and course of illness. Compr Psychiatry 2003;44:263-9.

8. Canuso CM, Bossie CA, Zhu Y et al. Psychotic symptoms in patients with bipolar mania. J Affect Disord 2008;111:164-9.

9. Vythilingam M, Chen J, Bremner JD et al. Psychotic depression and mortality. Am J Psychiatry 2003;160:574-6.

10. Craig TJ, Ye Q, Bromet EJ. Mortality among first-admission patients with psychosis. Compr Psychiatry 2006;47:246-51.

11. Taylor MA, Fink M. Catatonia in psychiatric classification: a home of its own. Am J Psychiatry 2003;160:1233-41.

12. Johnson J, Horwath E, Weissman MM. The validity of major depres-sion with psychotic features based on a community study. Arch Gen Psychiatry 1991;48:1075-81.

13. Ohayon MM, Schatzberg AF. Prevalence of depressive episodes with psychotic features in the general population. Am J Psychiatry 2002;159:1855-61.

14. Coryell W, Pfohl B, Zimmerman M. The clinical and neuroendo-crine features of psychotic depression. J Nerv Ment Dis 1984;172: 521-8.

15. Mazzarini L, Colom F, Pacchiarotti I et al. Psychotic versus non-psychotic bipolar II disorder. J Affect Disord 2010;126:55-60.

16. Kessing LV, Jensen HM, Christensen EM. Differences in the ICD-10 diagnostic subtype of depression in bipolar disorder compared to recurrent depressive disorder. Psychopathology 2008;41:141-6.

17. Maj M, Pirozzi R, Magliano L et al. Phenomenology and prognostic significance of delusions in major depressive disorder: a 10-year follow-up study. J Clin Psychiatry 2007;68:1411-7.

18. Maj M. Delusions in major depressive disorder: recommendations for the DSM-V. Psychopathology 2008;41:1-3.

19. Benazzi F. Bipolar versus unipolar psychotic outpatient depression. J Affect Disord 1999;55:63-6.

20. Coryell W, Leon AC, Turvey C et al. The significance of psychotic features in manic episodes: a report from the NIMH collaborative study. J Affect Disord 2001;67:79-88.

21. Colom F, Vieta E. The road to DSM-V. Bipolar disorder episode and course specifiers. Psychopathology 2009;42:209-18.

22. Goes FS, Sanders LL, Potash JB. The genetics of psychotic bipolar disorder. Curr Psychiatry Rep 2008;10:178-89.

23. American Psychiatric Association. Diagnostic and statistical manu-al of mental disorders, 4th ed., text revision (DSM-IV-TR). Arling-ton: American Psychiatric Association, 2000.

24. Fennig S, Bromet EJ, Karant MT et al. Mood-congruent versus mood-incongruent psychotic symptoms in first-admission patients with affective disorder. J Affect Disord 1996;37:23-9.

25. Goes FS, Zandi PP, Miao K et al. Mood-incongruent psychotic fea-tures in bipolar disorder: familial aggregation and suggestive linkage to 2p11-q14 and 13q21-33. Am J Psychiatry 2007;164:236-47.

26. Azorin JM, Akiskal H, Hantouche E. The mood-instability hypoth-esis in the origin of mood-congruent versus mood-incongruent psy-chotic distinction in mania: validation in a French National Study of 1090 patients. J Affect Disord 2006;96:215-23.

27. Forty L, Jones L, Jones I et al. Is depression severity the sole cause of psychotic symptoms during an episode of unipolar major depression? A study both between and within subjects. J Affect Disord 2009; 114:103-9.

28. Vieta E, Phillips ML. Deconstructing bipolar disorder: a critical re-view of its diagnostic validity and a proposal for DSM-V and ICD-

059_064.indd 63 06/07/12 14:07


11. Schizophr Bull 2007;33:886-92. 29. Carpenter WT, Bustillo JR, Thaker GK et al. The psychoses: cluster

3 of the proposed meta-structure for DSM-V and ICD-11. Psychol Med 2009;39:2025-42.

30. Goldberg DP, Andrews G, Hobbs MJ. Where should bipolar disor-der appear in the meta-structure? Psychol Med 2009;39:2071-81.

31. Fink M, Shorter E, Taylor MA. Catatonia is not schizophrenia: Krae-pelin’s error and the need to recognize catatonia as an independent

syndrome in medical nomenclature. Schizophr Bull 2010;36:314-20.

32. Ungvari GS, Caroff SN, Gerevich J. The catatonia conundrum: evi-dence of psychomotor phenomena as a symptom dimension in psy-chotic disorders. Schizophr Bull 2010;36:231-8.

33. Heckers S, Tandon R, Bustillo J. Catatonia in the DSM – shall we move or not? Schizophr Bull 2010;36:205-7.

059_064.indd 64 06/07/12 14:07

65

Mixed states and rapid cycling: conceptual issuesand options for ICD-11

This paper summarizes the development of the concepts of mixed states and rapid cycling, the way these conditions are covered in the DSM-IV and ICD-10, the proposals put forward for the DSM-5, and the options to be considered for the ICD-11.


Mixed states

Mixed states were first described by Weygandt (1) and Kraepelin (2) as conditions characterized by patterns of al-teration of mood, activity and thought that were different from those observed in both depression and mania. Included among mixed states were depressive or anxious mania (marked by anxious despair, overactivity and flight of ideas); excited or agitated depression (marked by anxious depres-sion, overactivity and inhibition of thought); and depression with flight of ideas (marked by depression, motor retardation and flight of ideas).

A revival of clinical and research interest in mixed states has occurred in the past five decades.

“Dysphoric” or “mixed” mania or hypomania (i.e., mania or hypomania accompanied by prominent depressive symp-toms) has been repeatedly described since the late 1960s. It has been found to occur in a proportion of manic patients ranging from 5 to 70%. It has been associated with an earlier onset than “classic” mania, a greater number of previous hos-pitalizations; a higher prevalence of suicidal behavior, psy-chiatric comorbidity, traumatic brain injury and substance use disorder; and a better response to valproate than to lith-ium (e.g., 3,4).

“Agitated” or “mixed” depression (i.e., major depression accompanied by symptoms of thought, motor and behav-ioral overactivation interpreted as contropolar) has been ex-

Mario Maj

Department of Psychiatry, University of Naples SUN, Naples, Italy

The proposed DSM-5 approach to mixed states is different from the DSM-IV characterization: the category “mixed episode” is going to disap-pear and a specifier “with mixed features” will be introduced in the diagnosis of both major depression and mania, requiring the presence of at least three contropolar symptoms. However, the proposed definition of major depression with mixed features is likely to be controversial, and the issue of ultradian cycling (encompassed in the DSM-IV definition of mixed episode but not in the proposed DSM-5 characterization of mixed states) will probably emerge. The DSM-IV definition of rapid cycling (i.e., at least four mood episodes during the past year demar-cated by either a remission of at least 2 months or a switch to an episode of opposite polarity) will be probably retained in the DSM-5, although the threshold of four mood episodes per year has not been validated. The ICD-11 approach to mixed states and rapid cycling will have to consider the above-mentioned problems with DSM-IV and proposed DSM-5 definitions, as well as issues of clinical utility.

key words: Mixed states, dysphoric mania, mixed depression, rapid cycling, clinical utility


tensively studied since the late 1980s. It has been found to occur in a proportion of patients with unipolar major depres-sion ranging from 7 to 60%. It has been associated with an earlier onset than “pure” major depression, a more common family history of bipolar disorder; a higher prevalence of sui-cidal behavior, psychiatric comorbidity, traumatic brain in-jury and alcohol abuse; a more frequent unipolar to bipolar switch, and a poorer response to antidepressants (e.g., 5,6).

Operational diagnostic criteria have been provided for both dysphoric mania and mixed depression. For a definite diagnosis of dysphoric mania or hypomania, McElroy et al (4) required the presence of a full DSM-III-R manic or hypo-manic syndrome and the simultaneous occurrence of at least three of the following symptoms: depressed mood, markedly diminished interest or pleasure in all or almost all activities, substantial weight gain or increase in appetite, hypersomnia, psychomotor retardation, fatigue or loss of energy, feelings of worthlessness or excessive or inappropriate guilt, feelings of helplessness or hopelessness, and recurrent suicidal ideation or a specific plan to commit suicide. For a diagnosis of mixed depression, Koukopoulos and Koukopoulos (7) required the occurrence of a major depressive episode accompanied by at least two of the following symptoms: motor agitation, psy-chic agitation or intense inner tension, and racing or crowd-ed thoughts.

rapid cycling

The expression “rapid cycling” was first used by Dunner and Fieve (8) with reference to “patients whose frequency of affective episodes was at least four per year”, who were found to be “disproportionally represented in the lithium prophy-laxis failure group”.

Further studies confirmed that rapid cyclers, defined in that manner, had a poorer response to lithium and a better

065_068.indd 65 06/07/12 14:05


response to carbamazepine (e.g., 9), but a more recent sys-tematic review concluded that rapid cycling was associated with lower effectiveness of all treatments evaluated, and that direct comparisons “provided no secure evidence of superi-ority of any treatment” (10).

The rapid cycling pattern has been found to be more com-mon in women and to be associated with a greater severity of illness and more suicide attempts, work impairment, sub-stance abuse and service utilization (e.g., 11). However, the threshold of four mood episodes per year has not been vali-dated (e.g., 12), and several prospective studies have found a unimodal distribution of episode frequencies in bipolar dis-order, and a near-linear relationship between episode fre-quency and several clinical characteristics (e.g., 13).

Several empirical studies have adopted definitions of rapid cycling including two further elements: a) the requirement of a circular course (i.e., alternation of mania or hypomania with depression without an intervening normal period) at some time during the course of the illness; b) the suspension of the duration requirements for manic, hypomanic and major de-pressive episodes, based on the observation that in rapid cy-clers mood episodes are often very brief (e.g., 14). These defi-nitions have been found to be associated with a greater stabil-ity of the rapid cycling pattern on follow-up, but with a lower interrater reliability (15).

Depending on the definitions adopted and the clinical set-tings where patients were recruited, the proportion of bipolar patients presenting the rapid cycling pattern at some time dur-ing the course of the illness has been reported to vary from 12 to 24%.

It is often difficult to distinguish ultrarapid cycling (i.e., mood states shifting in polarity over weeks to days), and es-pecially ultradian cycling (i.e., mood shifts occurring within a day), from a single mixed state (e.g., 15). In fact, neither the DSM-III nor the DSM-III-R mentioned rapid cycling, but they defined a mixed episode as presenting “the full symp-tomatic picture of both manic and major depressive episodes, intermixed or rapidly alternating every few days”. The dura-tion criterion for a major depressive episode was waived, only requiring that prominent depressive symptoms persisted at least a full day.

DsM-IV anD ICD-10 approaCHes

Mixed states

According to the DSM-IV, the diagnosis of a mixed epi-sode requires that the criteria be met for both a manic epi-sode and a major depressive episode, nearly every day during at least a 1-week period. In the narrative part of the text, it is specified that “the individual experiences rapidly alternating moods (sadness, irritability, euphoria) accompanied by symp-toms of a manic episode and a major depressive episode”. This statement clarifies that the DSM-IV definition of a mixed episode encompasses ultradian cycling.

The DSM-IV characterization of mixed states has been criticized as being inconsistent with clinical and research evi-dence. In fact, apart from states of ultradian cycling, it is very rare for a mixed state to meet the criteria for both a full man-ic and a full depressive syndrome during a one-week period. So, the vast majority of mixed states described in the litera-ture are excluded by the definition.

In the ICD-10, a mixed episode is defined as “an affective episode lasting for at least 2 weeks, characterized by either a mixture or a rapid alternation (usually within a few hours) of hypomanic, manic, and depressive symptoms”. The two sets of symptoms have to be “both prominent for the greater part of the current episode of illness”. It is further specified that “it is not uncommon for depressed mood to be accompanied for days or weeks by overactivity and pressure of speech, or for a manic mood and grandiosity to be accompanied by agitation and loss of energy and libido”.

Thus, the ICD-10 definition is less restrictive than the DSM-IV one – it requires the presence of prominent manic and depressive symptoms, but not of a full manic and a full depressive syndrome – and explicitly encompasses ultradian cycling. Furthermore, contrary to the DSM-IV criteria, the ICD-10 description acknowledges the research evidence suggesting the existence of mixed depression and mixed ma-nia, and, in line with that evidence, mentions overactivity and pressure of speech among the typical contropolar symp-toms in mixed depression (while the inclusion of agitation among the typical contropolar symptoms in mixed mania appears to be less appropriate).

Notably, in the ICD-10, the occurrence of a mixed episode implies a diagnosis of bipolar disorder, even if the patient presents “prominent manic symptoms” but not a full manic syndrome.

rapid cycling

Rapid cycling is included in the DSM-IV as a specifier for the diagnosis of bipolar I or bipolar II disorder. The defini-tion requires at least four episodes of a mood disturbance during the previous 12 months that meet criteria for a major depressive, manic, mixed or hypomanic episode. Episodes are demarcated either by partial or full remission for at least 2 months or a switch to an episode of opposite polarity (e.g., major depressive episode to manic episode). Thus, duration criteria for mood episodes (2 weeks for major depressive episodes, 1 week for manic episodes, 4 days for hypomanic episodes) are not waived, and pole switching is regarded as a possibility but not a prerequisite.

This definition has been regarded as too restrictive, be-cause it excludes a vast proportion of cases with briefer mood episodes that most clinicians would classify as rapid cyclers. In addition, rapid cycling as defined by the DSM-IV has been reported to be a transient pattern in most patients (e.g., 16), while definitions waiving duration criteria for mood epi-sodes, although having a lower interrater reliability, have

065_068.indd 66 06/07/12 14:05


been found to be associated with a greater longitudinal sta-bility of the pattern (15). Finally, as mentioned above, the threshold of four mood episodes per year has not been vali-dated.

The rapid cycling pattern is not mentioned in the ICD-10.

proposals for DsM-5

Mixed states

According to current proposals for DSM-5 (see www.dsm5.org), the category “mixed episode” is going to be re-moved, while a specifier “with mixed features” is likely to be included for the diagnoses of both mania/hypomania and major depression. Based on empirical evidence suggesting that the most significant differences between “mixed” and “pure” mania/hypomania or major depression emerge when the number of contropolar symptoms is at least three (e.g., 17), this will be the threshold adopted for the specifier. Among qualifying contropolar symptoms, those that may oc-cur in both mania and depression, such as psychomotor agi-tation, irritability, insomnia (other than decreased need for sleep), distractibility and loss of appetite, will be excluded.

Thus, the proposed DSM-5 definition of mania/hypoma-nia with mixed features requires that the full criteria are met for a manic or hypomanic episode, and at least three of the following symptoms are present almost every day during the episode: prominent dysphoria or depressed mood, dimin-ished interest or pleasure in all or almost all activities, psy-chomotor retardation, fatigue or loss of energy, feelings of worthlessness or excessive or inappropriate guilt, and recur-rent thoughts of death or a suicide attempt or a specific plan for committing suicide. The proposed DSM-5 definition of major depression with mixed features requires that the full criteria are met for a major depressive episode, and at least three of the following symptoms are present almost every day during the episode: elevated or expansive mood, inflated self-esteem or grandiosity, more talkative than usual or pressure to keep talking, flight of ideas or subjective experience that thoughts are racing, increase in energy or goal-directed activ-ity, increased or excessive involvement in activities that have a high potential for painful consequences, and decreased need for sleep.

The definition of mania/hypomania with mixed features is consistent with the available research evidence and the operational diagnostic criteria previously proposed in the literature (e.g., 4) and is therefore unlikely to be criticized. On the contrary, the definition of major depression with mixed features is likely to be controversial, as it includes typical manic symptoms (such as elevated mood and grandiosity) that have been found to be rare among patients with mixed depression, while excluding symptoms (such as irritability, psychomotor agitation and distractibility) that are frequently reported in mixed depression (e.g., 18,19). This controversy will likely revive the debate about whether “agitated depres-

sion” really represents a mixed state (20).It is important to note that a diagnosis of major depression

with mixed features would not imply that the case be re-garded as bipolar. Thus, a patient with three typical manic symptoms such as elevated mood, inflated self-esteem or grandiosity, and excessive involvement in activities having a high potential for painful consequences, would be classified as unipolar.

The proposed DSM-5 characterization of mixed states, contrary to the DSM-IV definition of mixed episode, does not encompass ultradian cycling. Therefore, the question of how this latter condition will be accommodated in the man-ual is likely to emerge.

rapid cycling

No proposal has been made to revise the description of rapid cycling in the DSM-5.

optIons for ICD-11

The main issue to be addressed when deciding if mixed states and rapid cycling should be included in the ICD-11 is whether specifying these patterns has sufficient clinical util-ity. In this discussion, it will have to be acknowledged that different levels of clinical utility exist in psychiatric nosology, especially for subtyping diagnostic categories. There are di-agnostic subtypes which have clear treatment implications (i.e., they orient towards the use of a therapeutic intervention that is not the first choice for patients receiving the main di-agnosis) and subtypes which have primarily prognostic im-plications, that may be still clinically relevant (e.g., research evidence suggests that they are associated with poorer out-come or higher suicide risk).

For dysphoric mania, the evidence of a better response to valproate than to lithium and a higher prevalence of suicidal behavior compared to “pure” mania may be regarded as in-dicating sufficient clinical utility. For mixed depression, the available evidence of a poorer response to antidepressants and an increased risk of a switch to the opposite polarity may be regarded as less cogent (21), although the higher risk for suicide may still argue in favor of the inclusion of the subtype in the system. For rapid cycling, in the absence of confirmed treatment implications, it will have to be established whether the overall poor prognosis associated with the pattern is suf-ficient to support an argument for its clinical utility.

The above-mentioned problems with the proposed DSM-5 characterization of mixed states and the DSM-IV definition of rapid cycling will have to be taken into account, and the option of retaining the ICD-10 definition of mixed episode, with some refinements, will need to be considered.

065_068.indd 67 06/07/12 14:05


note

M. Maj is a member of the WHO International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders and the Chairman of the ICD Revision Working Group on the Classification of Mood and Anxiety Disorders. The views expressed in this article are those of the author and, except as specifically noted, do not represent the official policies or positions of the International Advisory Group, the Working Group on the Classification of Mood and Anx-iety Disorders, or the WHO.


references

1. Weygandt W. Ueber die Mischzustaende des manish-depressiven Ir-reseins. Munich: Lehmann, 1899.

2. Kraepelin E. Psychiatrie: Ein Lehrbuch fur Studirende und Aerzte. 8th ed. Leipzig: Barth, 1909.

3. Post RM, Rubinow DR, Uhde TW et al. Dysphoric mania: clinical and biological correlates. Arch Gen Psychiatry 1989;46:353-8.

4. McElroy SL, Keck PE, Jr, Pope HG, Jr et al. Clinical and research implications of the diagnosis of dysphoric or mixed mania or hypo-mania. Am J Psychiatry 1992;149:1633-44.

5. Akiskal HS, Benazzi F, Perugi G et al. Agitated “unipolar” depres-sion re-conceptualized as a depressive mixed state: implications for the antidepressant-suicide controversy. J Affect Disord 2005;85:245-58.

6. Fiedorowicz JG, Endicott J, Leon AC et al. Subthreshold hypoman-ic symptoms in progression from unipolar major depression to bipo-lar disorder. Am J Psychiatry 2011;168:40-8.

7. Koukopoulos A, Koukopoulos A. Agitated depression as a mixed state and the problem of melancholia. Psychiatr Clin North Am 1999;22:547-64.

8. Dunner DL, Fieve RR. Clinical factors in lithium carbonate prophy-laxis failure. Arch Gen Psychiatry 1974;30:229-33.

9. Kishimoto A, Ogura C, Hazama H et al. Long-term prophylactic effects of carbamazepine in affective disorders. Br J Psychiatry 1983; 143:327-32.

10. Tondo L, Hennen J, Baldessarini RJ. Rapid-cycling bipolar disorder: effects of long-term treatments. Acta Psychiatr Scand 2003;108:4-14.

11. Bauer MS, Calabrese J, Dunner DL et al. Multisite data reanalysis of the validity of rapid cycling as a course modifier for bipolar disorder in DSM-IV. Am J Psychiatry 1994;151:506-15.

12. Bauer M, Beaulieu S, Dunner DL et al. Rapid cycling bipolar disor-der – diagnostic concepts. Bipolar Disord 2008;10:153-6.

13. Kupka RW, Luckenbaugh DA, Post RM et al. A comparison study of rapid and non-rapid cycling bipolar disorder using daily prospec-tive mood rating in 539 out-patients. Am J Psychiatry 2005;162:1273-80.

14. Wehr TA, Sack DA, Rosenthal NE et al. Rapid cycling affective dis-order: contributing factors and treatment responses in 51 patients. Am J Psychiatry 1988;145:179-84.

15. Maj M, Pirozzi R, Formicola AM. Reliability and validity of four alternative definitions of rapid-cycling bipolar disorder. Am J Psy-chiatry 1999;156:1421-4.

16. Coryell W, Endicott J, Keller M. Rapidly cycling affective disorder. Demographics, diagnosis, family history, and course. Arch Gen Psy-chiatry 1992;49:126-31.

17. Swann AC, Steinberg JL, Lijffijt M et al. Continuum of depressive and manic mixed states in patients with bipolar disorder: quantita-tive measurement and clinical features. World Psychiatry 2009;8: 166-72.

18. Maj M, Pirozzi R, Magliano L et al. Agitated “unipolar” major de-pression: prevalence, phenomenology, and outcome. J Clin Psychia-try 2006;67:712-9.

19. Goldberg JF, Perlis RH, Bowden CL et al. Manic symptoms during depressive episodes in 1,380 patients with bipolar disorder: findings from the STEP-BD. Am J Psychiatry 2009;166:173-81.

20. Maj M, Pirozzi R, Magliano L et al. Agitated depression in bipolar I disorder: prevalence, phenomenology, and outcome. Am J Psychia-try 2003;160:2134-40.

21. First MB. DSM-5 proposals for mood disorders: a cost-benefit anal-ysis. Curr Opin Psychiatry 2011;24:1-9.

065_068.indd 68 06/07/12 14:05

69

How should melancholia be incorporated in ICD-11?

This paper reviews the historical origins of the concept of melancholia and the way this entity has been covered in di-agnostic systems up to ICD-10 and DSM-IV. It then consid-ers some recent proposals and summarizes some options for the ICD-11.


Melancholia was the original concept of depressive illness, extensively mentioned in the medical literature for over 3,500 years as a “madness” with “bodily causes” (1). The syndrome of melancholia was described during the Pharaonic times (approximately 1500 B.C.) in the Ebers Papyrus. During the 5th century B.C., Hippocrates described it in more detail as a persistent sadness with morbid thoughts caused by a disor-der of the brain. He believed that a balance of four body humours was necessary for health and that their imbalance caused illness. He explained that “black bile” concentrated under the liver and led to “melancholia” (literally black bile), because its fumes ascended to the brain and caused the dark-ening of thoughts and emotions. The relationship between melancholia and various somatic symptoms and ailments also generated the word “hypochondriasis”, coming from the black bile that gathers in the right hypochondrium. After Hippocrates, Aretaeus of Cappadocia and Galen, who both lived in the 1st century A.D., also considered melancholia as a brain disorder due to the black bile.

This view was maintained by physicians in Arabic speak-ing countries, where Hippocratic teachings were followed by Avicenna (Ibn Sina), Razes (Arrazi), Avenzoar (Ibn Zohr) and others. In Tunis, during the 10th century, Ishak Ibn Um-ran produced the first book ever written focusing exclusively on psychiatry, called Al Maqala fi Al malikhulia (“The epistle on melancholy”). Malikhulia in Arabic was taken from the Greek word melancholia. The other word used for

Driss Moussaoui, MohaMeD agoub, aDil Khoubila

Ibn Rushd University Psychiatric Centre, Casablanca, Morocco

Melancholia is a concept that has been in use for over 3,500 years. At the beginning of the 20th century, A. Meyer suggested to replace it with the concept of depression, with a strong “reactive” connotation. This influenced the ICD-6 (1948) and ICD-7 (1955) approaches, on which the DSM-I (1952) and DSM-II (1968) classifications were based. The DSM-III and DSM-IV did not include a separate category of melancholia, but rather included “with melancholic features” as a sub-category of major depression. Melancholic features corresponded to symptoms that had previously been interpreted to indicate “endogenous depression”. The ICD-10 followed the same path, and went even further by deleting the word melancholia and replacing it with the expression “somatic syndrome”. This article presents a rationale for reinstating a sub-catego-ry of melancholia in ICD-11. This may pave the way for additional research and possibly the conceptualization of melancholia as a major category in future revisions. To this end, we recommend increased focus on melancholia as a topic of research.

key words: ICD-11, DSM-5, classification, melancholia, depression


melancholia was Sawdawia, literally the black illness, refer-ring again to the black colour of the bile but also to the black perception of life by the patient (2).

In Europe, it was only with Timothy Bright’s “Treatise of melancholy” (1586) and Robert Burton’s “The anatomy of melancholy” (1621) that melancholia was regarded again as a medical condition. Other forms of depression were how-ever also mentioned in literature, such as “illness of love” in France (“la maladie d’amour ou mélancholie érotique” by Jacques Ferrand in 1623). One of the proposed treatments of melancholia was Helleborus Niger, that induces digestive bleeding and melena. Black feces were then a proof that black bile was expelled from the melancholic patient.

At the beginning of the 19th century, melancholia was re-garded in Europe as one of the five classes of insanity (along with mania with and without delirium, dementia and idiocy). “Folie circulaire” (circular insanity) was described by the French psychiatrists Falret (1854) and Baillarger (1854), and was the precursor to manic-depressive illness introduced later by Kraepelin, who insisted particularly on the periodic-ity of recurrences.

The term “depression”, first used in the 19th century by the English physician W.W. Gull, was made popular at the beginning of the 20th century by A. Meyer and progressively replaced that of melancholia. For much of the last century, melancholia signified “endogenous” and/or severe forms of depression, and also applied to a specific “involutional” con-dition in later life. Meyer maintained that depressive epi-sodes were in fact “reactions”. This concept influenced the ICD-6 (1948) and ICD-7 (1955) approaches, on which the DSM-I (1952) and DSM-II (1968) classifications were based.

MelanCHolIa In DIagnostIC systeMs

The ICD-8 and ICD-9 approach to the classification of

069_072.indd 69 06/07/12 14:06


depression was binary. On the one hand there was “manic-depressive psychosis, depressed type”, with “melancholia not otherwise specified” as one of the inclusion terms; on the other there was “depressive neurosis”. Beginning in the 1920s, however, a number of academicians had advocated a unitary perspective, according to which there was just one type of depression, with a continuum of severity from neu-rotic to psychotic cases (3). Several studies using cluster anal-ysis addressed this binary/unitary controversy, and most of them confirmed the existence of a group of patients corre-sponding to the prototype of endogenous depression, while non-endogenous cases were generally grouped into two or three clusters. Some studies confirmed that endogenous de-pression was more responsive than neurotic depression to tricyclic antidepressants (e.g., 4), while some others did not (e.g., 5). The stability over time of the diagnosis of neurotic depression was reported to be very poor in a study by Akiskal et al (6), and the reliability of both the above diagnoses was found to be modest (7).

In 1980, the DSM-III provided operational diagnostic cri-teria for “major depressive episode”, and melancholia be-came one of the subtypes of this entity, roughly correspond-ing to endogenous depression, being characterized by a loss of pleasure in all or almost all activities, a lack of reactivity to usually pleasurable stimuli, and at least three of the following symptoms: distinct quality of depressed mood (depressed mood is perceived as distinctly different from the kind of feel-ing experienced following the death of a loved one), depres-sion regularly worse in the morning, early morning awaken-ing (at least two hours before usual time), marked psychomo-tor retardation or agitation, significant anorexia or weight loss, excessive or inappropriate guilt.

In 1987, the DSM-III-R incorporated in the definition of the melancholic subtype some further elements of the con-cept of endogenous depression, such as the absence of a sig-nificant personality disturbance before the first major depres-sive episode, the occurrence of one or more previous major depressive episodes followed by complete or nearly complete recovery, and a previous good response to specific and ade-quate somatic antidepressant therapy (tricyclics, electrocon-vulsive therapy, monoamine oxidase inhibitors, lithium).

In the DSM-IV, “with melancholic features” remains one of the specifiers for the diagnosis of major depression, de-fined by the presence of either loss of pleasure in all or almost all activities or the loss of reactivity to usually pleasurable stimuli, plus at least three of the following: distinct quality of depressed mood, depression regularly worse in the morning, early morning awakening, marked psychomotor retardation or agitation, significant anorexia or weight loss, and exces-sive or inappropriate guilt.

In the ICD-10, the expression “somatic syndrome” was introduced, though clearly corresponding to what was la-belled as “melancholic features” in DSM-IV. “Somatic syn-drome” is characterized by the presence of at least four of the following symptoms: loss of interest or pleasure in activities that are normally enjoyable, lack of emotional reactivity to

normally pleasurable surroundings and events, waking in the morning 2 hours or more before the usual time, depression worse in the morning, objective evidence of definite psycho-motor retardation or agitation, marked loss of appetite, weight loss, marked loss of libido. The presence of this “so-matic syndrome” can be recorded at the fifth character level as a specifier for the diagnosis of mild or moderate depressive episode, while it is presumed that the somatic syndrome “will almost always be present in a severe depressive episode”.

reCent proposals

It has been argued that the category of major depression is too heterogeneous to represent a meaningful target for re-search, and that the unitary approach of the DSM-III and its successors may have discouraged the validation of more ho-mogeneous, separate types of depression (8). Several authors have maintained that melancholia should be regarded as a distinct disease entity. Parker et al (9) advocated this position on the basis of clinical, biological and therapeutic features. Clinical features included disturbances in affect (non-reac-tive mood, blunted affect, and pervasive anhedonia), psycho-motor retardation or spontaneous agitation, cognitive im-pairment, and vegetative dysfunction. Biological data includ-ed primarily an increased activity of the hypothalamus–pitu-itary–adrenal (HPA) axis (hypercortisolemia) and disturbances in sleep architecture. Therapeutic evidence cited in support of this position was the higher efficacy of tricyclic antidepres-sants and electroconvulsive therapy in melancholic as com-pared to non-melancholic depression (10,11).

Taylor and Fink (12) proposed a set of operational diag-nostic criteria for melancholia as a distinct disease entity, requiring: a) an episode of illness with reduced functioning characterized by an unremitting mood of apprehension and gloom that compromises normal daily activities and persists for at least two weeks; b) psychomotor disturbance as agita-tion, retardation (including stupor and catatonia), or both; c) vegetative signs of poor sleep, appetite, libido, cognition (at least two); d) at least one of the following: abnormal dexa-methasone suppression test (DST) and corticotropin releas-ing hormone (CRH) test or high night-time cortisol levels; and decreased rapid eye movement (REM) sleep latency or other sleep abnormalities.

The view that melancholia should be regarded as a distinct disease entity has been criticized by other authors. Kocsis (13), for example, noted that the same arguments made for melancholia could be put forward for other subtypes of de-pression, such as psychotic depression.

optIons for ICD-11

In considering potential revisions for the ICD-11, there are good reasons to consider using the term melancholia rather than the current “somatic syndrome”. The latter expression

069_072.indd 70 06/07/12 14:06


is unclear and potentially confusing, as there are other so-matic symptoms in depression that have nothing to do with melancholia, and somatic symptoms are obviously present in several other mental disorders. Furthermore, some symp-toms of melancholia are not actually somatic.

There are some arguments that support the option to in-corporate melancholia in the ICD-11 as an independent di-agnostic category. Melancholia has some clinical particulari-ties compared to other subtypes of depression. Diurnal mood variation with aggravation in the morning is a common fea-ture (14). In the cognitive dimension, the melancholic group appears to require longer periods for cognitive recovery and this has implications for social functioning after clinical dis-charge (15). Some brain imaging studies showed structural alterations in areas that could be related to the pathophysiol-ogy of melancholia (16). Impaired functioning of the HPA axis is common among melancholic patients and tends to persist even after recovery (17). In the therapeutic domain, there is some evidence that melancholic patients are more responsive to electroconvulsive therapy and to tricyclic anti-depressants and less responsive to psychotherapy alone (18,19).

There are also arguments against a separate category for melancholic depression. The diagnosis of depression sub-types is unstable over time. Therefore, a person with recurrent depressive disorder could have a melancholic episode on one occasion and a non-melancholic episode during the next re-currence. It would be problematic in this situation to regard the person as switching from one disorder to another. For example, Melartin et al (20) conducted a prospective, natural-istic cohort study in order to assess the relationship of melan-cholic features to psychiatric comorbidity and the course of the illness. They did not find any difference in rates of comor-bid axis I or II disorders, nor in the course of the illness, be-tween melancholic and non-melancholic cases. The non-melancholic subtype switched to melancholic in 25% of cases. In addition, among those who had melancholic fea-tures at the index episode and subsequent episodes during the 18-month follow-up, only 22% presented melancholic fea-tures during the latter. Furthermore, some melancholic pa-tients are responsive to interpersonal psychotherapy or to cognitive behaviour therapy (21). Newer antidepressants are as potent as tricyclics and different subtypes of depression do not predict response across classes of antidepressants (13). Finally, neuroendocrine features are not specific to melan-cholia: other subtypes of depression share the same biologi-cal correlates.

In view of the mixed evidence and continuing controver-sies described above, we recommend that ICD-11 reinstate the term “melancholia” or “melancholic features” that has been a familiar concept to physicians for thousands of years and eliminate the term “somatic syndrome”. However, we do not recommend conceptualizing melancholia as a distinct diagnostic category, but rather to continue to include it as a qualifier for depressive episodes. We believe that further re-search should be pursued in this area, which we hope will

provide a better understanding of pathophysiological mech-anisms and help to manage depressed patients – particularly those that are melancholic – more effectively.

note

D. Moussaoui is a member of the WHO ICD Revision Working Group on the Classification of Mood and Anxiety Disorders, reporting to the International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders. The views expressed in this article are those of the authors and, except as specifically noted, do not represent the official policies or positions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.


references

1. Taylor MA, Fink M. Melancholia. The diagnosis, pathophysiology, and treatment of depressive illness. Cambridge: Cambridge Univer-sity Press, 2006.

2. Ammar S. En souvenir de la médecine arabe. Tunis: Société Tunisi-enne de Diffusion, 1965.

3. Maj M. Development and validation of the current concept of major depression. Psychopathology 2012;45:135-46.

4. Overall JE, Hollister LE, Johnson M et al. Nosology of depression and differential response to drugs. JAMA 1966;195:946-50.

5. Raskin A, Crook TH. The endogenous-neurotic distinction as pre-dictor of response to antidepressant drugs. Psychol Med 1976;6:59-70.

6. Akiskal HS, Bitar AH, Puzantian VR et al. The nosological status of neurotic depression. Arch Gen Psychiatry 1978;35:756-66.

7. Spitzer RL, Fleiss JL. A reanalysis of the reliability of psychiatric diagnosis. Br J Psychiatry 1974;125:341-7.

8. Parker G. Classifying depression: should paradigms lost be regai- ned? Am J Psychiatry 2000;157:1195-203.

9. Parker G, Fink M, Shorter E et al. Issues for DSM-5: whither mel-ancholia? The case for its classification as a distinct mood disorder. Am J Psychiatry 2010;167:745-7.

10. Perry PJ. Pharmacotherapy for major depression with melancholic features: relative efficacy of tricyclic versus selective serotonin reup-take inhibitor antidepressants. J Affect Disord 1996;20:1-6.

11. Anderson IM. Selective serotonin reuptake inhibitors versus tricy-clic antidepressants: a meta-analysis of efficacy and tolerability. J Affect Disord 2000;58:19-36.

12. Taylor MA, Fink M. Restoring melancholia in the classification of mood disorders. J Affect Disord 2008;105:1-14.

13. Kocsis JH. Melancholia as a distinct mood disorder? Recommenda-tions for DSM-5. Am J Psychiatry 2010;167:1534.

14. Moffot APR, O’Carroll RE, Bennie J et al. Diurnal variation of mood and neuropsychological function in major depression with melan-cholia. J Affect Disord 1994;32:257-69.

15. Withall A, Harris LM, Cumming SR. A longitudinal study of cogni-tive function in melancholic and non-melancholic subtypes of major depressive disorder. J Affect Disord 2010;123:150-7.

16. Soriano-Mas C, Hernández-Ribas R, Pujol J et al. Cross-sectional and longitudinal assessment of structural brain alterations in melan-cholic depression. Biol Psychiatry 2011;69:318-25.

069_072.indd 71 06/07/12 14:06


17. Pintor L, Torres X, Navarro V et al. Corticotropin-releasing factor test in melancholic patients in depressed state versus recovery: a comparative study. Prog Neuropsychopharmacol Biol Psychiatry 2007;31:1027-33.

18. Prusoff BA, Weissman MM, Klerman G et al. Research Diagnostic Criteria subtypes of depression. Their role as predictors of differen-tial response to psychotherapy and drug treatment. Arch Gen Psy-chiatry 1980;37:796-801.

19. Strober M, Rao U, De Antonio M. Effects of electroconvulsive ther-

apy in adolescents with severe endogenous depression resistant to pharmacotherapy. Biol Psychiatry 1998;43:335-8.

20. Melartin T, Leskelä U, Rytsälä H et al. Co-morbidity and stability of melancholic features in DSM-IV major depressive disorder. Psychol Med 2004;34:1443-52.

21. Thase ME, Friedman ES. Is psychotherapy an effective treatment for melancholia and other severe depressive states? J Affect Disord 1999; 54:1-19.

069_072.indd 72 06/07/12 14:06

73

Postpartum depression and premenstrual dysphoric disorder: options for ICD-11

This paper summarizes currently available evidence on postpartum depression and premenstrual dysphoric disor-der, the ICD-10 and DSM-IV approaches to these diagnostic entities, and the options to be considered for ICD-11.

ReseaRCh evIDenCe

Postpartum depression

Postpartum depression, also called postnatal depression, is a major depressive episode occurring during puerperium, marked by a functional impairment that can be severe, in-cluding social avoidance, self-neglect and inability to ade-quately care for the newborn. Suicidal thoughts may be a prominent symptom (1). Mood lability, anxiety and excessive worry about the newborn are common. Psychotic symptoms may be present.

Studies indicate that the prevalence of postpartum depres-sion is 10 to 16% during 6-12 weeks after delivery (2-4). Some authors have reported a significantly increased inci-dence of depression during the first 3 months after delivery as compared to the pre-pregnancy period (5-7). Other studies have found that the prevalence of depression during the post-partum period is not higher than that observed in same-age non-puerperal women in the general population, and that there are no significant differences in depression prevalence between the last trimester of pregnancy and the postpartum period (8-11).

The recognition of major depressive episodes during preg-nancy and postpartum can be complex, due to the overlap of somatic symptoms of depression with physical changes that

Maria Luisa Figueira, Vasco Videira dias

Department of Psychiatry, Santa Maria Hospital, Faculty of Medicine, University of Lisbon, Avenida Professor Egas Moniz, Piso 3, 1649-035 Lisbon, Portugal

Postpartum depression is a depressive episode occurring during puerperium, which may be associated with severe functional impairment, including social avoidance, self-neglect and inability to adequately care for the newborn. In ICD-10, postpartum depression is simply classified as a depressive episode (or recurrent depressive disorder if there is a history of prior episodes) if the symptoms are consistent with the general diagnostic description for depressive episode. Postpartum depression that is not consistent with the ICD-10 diagnostic description for depres-sive episode, due to the number or nature of presenting symptoms, is classified under “Mental and behavioural disorders associated with the puerperium, not elsewhere classified”. In DSM-IV, postpartum depression is acknowledged through a postpartum onset specifier in the diag-nosis of major depressive episode. Premenstrual dysphoric disorder is a severe form of premenstrual syndrome, which interferes with work, social activities, and relationships. In ICD-10, it is stated that, if depressive episodes occur only in relation to the menstrual cycle, the diag-nosis should be “Other specified mood [affective] disorder”, with a second code for the underlying cause (“Other specified conditions associ-ated with female genital organs and menstrual cycle”). In DSM-IV, premenstrual dysphoric disorder appears in Appendix B (“Criteria sets and axes provided for further study”). In the last two decades, several studies have been conducted on both postpartum depression and premen-strual dysphoric disorder. This paper briefly reviews the available evidence and the possible options for ICD-11.

Key words: Postpartum depression, postnatal depression, premenstrual dysphoric disorder, depression, pregnancy, reproductive cycle, puerperium


are common during those phases, such as fatigue, dimin-ished concentration, sleep difficulties, and changes in appe-tite and weight (12). After a first episode of postpartum de-pression, there is an increased risk of subsequent episodes linked to childbirth. There is evidence of family factors in the susceptibility to depression in the period of 6-8 weeks after delivery (13).

The role of hormonal factors in postpartum depression remains unclear. No consistent differences in gonadal steroid levels have been demonstrated in women with vs. without postpartum depression, and differences involving the activity of the hypothalamic-pituitary-adrenal axis might indicate an adaptive response to stress in women who have experienced or are more vulnerable to postpartum depression (14).

Premenstrual dysphoric disorder

Premenstrual dysphoric disorder is a severe form of pre-menstrual syndrome in which symptoms interfere with work, social activities, and relationships. The most common symp-toms include mood swings, depressed mood or feelings of hopelessness, marked anger, increased interpersonal con-flicts, tension and anxiety, irritability, decreased interest in usual activities, difficulty concentrating, fatigue, change in appetite, feeling out of control or overwhelmed, sleep prob-lems or physical problems, such as “bloating” (15,16).

A total of 75 to 95% of women with regular menstrual cycles have some type of premenstrual symptoms with minor intensity, and without the need for medical intervention. The prevalence of premenstrual dysphoric disorder has been es-timated between 2 and 8% of premenopausal women (17,18).

073_076.indd 73 06/07/12 14:08


The disorder displays a distinctive clinical picture that, in the absence of treatment, is remarkably stable from cycle to cycle and over time (19).

The etiopathogenesis of premenstrual dysphoric disorder is at present unclear. Sexual hormones are likely to have a crucial role, but research has not confirmed the correlation between the disorder and high levels of estrogens, progester-one deficit or changes in estrogen/progesterone ratio.

hIstoRICal baCKgRounD

In classic descriptions of postpartum illness, depression was not distinguished from psychosis. In the 18th and 19th century, German psychiatrists described several cases of postpartum mania and melancholia (20). In his Traité de la folie des femmes enceintes, des nouvelles accouchées et des nourrices, L.-V. Marcé (21) described postpartum mel-ancholia as a condition with a good prognosis and a duration of 1 to 6 months.

The modern nosological point of view was expressed by Kraepelin (22), who argued that puerperium is a period of high risk for the onset of a first depressive episode and a strong trigger for recurrence of mood episodes in women with manic-depressive illness. He thought that postpartum depres-sion is not different from non-puerperal depression and should not be considered as a separate nosological entity.

The first scientific account of premenstrual dysphoric dis-order was provided by R.T. Frank (23), who described a pre-menstrual “nervous tension” related to the hormonal system, with relief after the start of menstruation. The condition was first recognized as a diagnostic entity (“late luteal phase dys-phoric disorder”) in the DSM-III-R, and proposed for further research.

ICD-10 anD DsM-Iv aPPRoaChes


In ICD-10, postpartum depression is simply classified as a depressive episode (or recurrent depressive disorder if there is a history of prior episodes) if the symptoms are consistent with the general diagnostic description for depressive epi-sode. Postpartum depression that is not consistent with the ICD-10 diagnostic description for depressive episode due to the number or nature of presenting symptoms would be clas-sified under category F53, “Mental and behavioural disor-ders associated with the puerperium, not elsewhere classi-fied”. So, this latter category should be used only for mental disorders associated with the puerperium (beginning within 6 weeks of delivery) that are not consistent with the descrip-tion for depressive episode or recurrent depressive disorder, either because insufficient information is available, or be-cause special additional features are present that make clas-sification as a depressive episode inappropriate.

The DSM-IV provides a postpartum onset specifier for major depressive episode, requiring “onset of episode within 4 weeks postpartum”. In the text of the manual, it is stated that the symptomatology of postpartum major depressive episode does not differ from that observed in non-postpar-tum episodes.


The ICD-10 briefly comments on depressive episodes re-lated to the menstrual cycle in the Clinical Descriptions and Diagnostic Guidelines, stating that, if depressive episodes occur only in relation to the menstrual cycle, the diagnosis should be “Other specified mood [affective] disorder”, with a second code for the underlying cause (“Other specified conditions associated with female genital organs and men-strual cycle”). “Premenstrual tension syndrome” is listed as a physical disorder under “Pain and other conditions associ-ated with female genital organs and the menstrual cycle”.

The DSM-IV includes premenstrual dysphoric disorder in Appendix B (“Criteria sets and axes provided for further study”). The diagnosis requires that the syndrome has oc-curred in most menstrual cycles during the past year, for most of the time during the last week of the luteal phase, has begun to remit within a few days after the onset of the follicular phase, and has been absent in the week post menses. At least five of the following symptoms are required: markedly de-pressed mood, feelings of hopelessness or self-deprecating thoughts; marked anxiety and tension; marked affective la-bility; persistent and marked anger or irritability; decreased interest in usual activities; subjective sense of difficulty in concentrating; lethargy, easy fatigability or marked lack of energy; marked change in appetite, overeating, or specific food cravings; hypersomnia or insomnia; a subjective sense of being overwhelmed or out of control; other physical symp-toms (such as breast tenderness or swelling, headaches, joint or muscle pain, a sensation of “bloating”, weight gain). At least one of the first four symptoms must be present. The disturbance markedly interferes with usual activities and re-lationships with others. These aspects must be confirmed by prospective daily ratings during at least two consecutive symptomatic cycles. The disturbance is not merely an exac-erbation of the symptoms of another disorder, such as a ma-jor depressive disorder.

PRoPosals foR DsM-5


The DSM-5 Work Group on Mood Disorders recognized that currently available evidence does not support postpar-tum depression as a separate nosological entity (see www.dsm5.org). The current DSM-5 recommendation is to retain postpartum depression as an onset specifier for major de-

073_076.indd 74 06/07/12 14:08


pression, which is consistent with the DSM-IV approach. However, the DSM-5 Work Group recommended that the duration covered by the postpartum onset specifier be ex-tended to 6 months, on the basis of the evidence suggesting a longer period for postpartum depressive risk. In fact, a large Danish registry study of hospital postpartum admissions identified an increased risk of admission with unipolar de-pressive disorders persisting during the first 5 months post-partum (24).


The Work Group has proposed the placement of premen-strual dysphoric disorder in the Mood Disorders section of DSM-5 rather than in Appendix B. The group felt that infor-mation on diagnosis, treatment and validators of the disorder has matured to a point that justifies the qualification as a category in DSM-5 (see www.dsm5.org).

oPtIons foR ICD-11


The nosological status of postpartum depression remains controversial, with insufficient data to recommend a distinct diagnostic entity. However, due to several clinical advantages of highlighting this particular syndrome as a sub-category of depressive disorders, it would be useful to consider the addi-tion of a postpartum onset specifier for depressive episodes. A postpartum onset specifier is clinically important because it has specific implications for care, particularly involving the relationship between the mother and the infant. Early recog-nition and intervention may prevent long-term effects on child development. Furthermore, emphasizing the relation to childbirth can raise the attention to the probability of fur-ther neonatal episodes.


The category of premenstrual dysphoric disorder describes patients with a specific clinical syndrome not elsewhere rep-resented in the classification system. An adequate recogni-tion of this disorder and its impact should increase the num-ber of women that can be correctly treated. However, the risk of false positives remains significant. Common negative at-titudes towards menstruation might be reinforced with this diagnosis, but ignoring it would leave a significant propor-tion of women unprotected and suffering from a condition that can be disabling.

note

M.L. Figueira is a member of the WHO ICD Revision Working Group on the Classification of Mood and Anxiety Disorders, reporting to the International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders. The views expressed in this article are those of the authors and, except as specifically noted, do not represent the official policies or positions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.


References

1. Kendell RE, Chalmers JC, Platz C. Epidemiology of puerperal psy-choses. Br J Psychiatry 1987;150:662-73.

2. Cutrona CE. Casual attributions and perinatal depression. J Abnorm Psychol 1983;92:161-72.

3. O’Hara MW. Postpartum mental disorders. In Sciarra JJ (ed). Gyne-cology and obstetrics, Vol. 6. Philadelphia: Harper & Row, 1991:1-17.

4. O’Hara MW, Swain AM. Rates and risk of postpartum depression – a meta-analysis. Int Rev Psychiatry 1996;8:37-54.

5. Brockington IF, Winokur G, Dean C. Puerperal psychosis. In: Brockington IF, Kumar R (eds). Motherhood and mental illness. London: Academic Press, 1982:37-69.

6. Kumar R, Robson KM. A prospective study of emotional disorders in childbearing women. Br J Psychiatry 1984;144:35-47.

7. Cox JL, Murray D, Chapman G. A controlled study of the onset, duration and prevalence of postnatal depression. Br J Psychiatry 1993;163:27-31.

8. Josefsson A, Berg G, Nordin C et al. Prevalence of depressive symp-toms in late pregnancy and postpartum. Acta Obstet Gynecol Scand 2001;80:251-5.

9. Evans J, Heron J, Francomb H et al. Cohort study of depressed mood during pregnancy and after childbirth. BMJ 2001;323:257-60.

10. Wisner KL, Parry BL, Piontek CM. Clinical practice. Postpartum depression. N Engl J Med 2002;18:194-9.

11. Halbreich U. Prevalence of mood symptoms and depressions during pregnancy: implications for clinical practice and research. CNS Spectr 2004;9:177-84.

12. Matthey S, Ross-Hamid C. The validity of DSM symptoms for de-pression and anxiety disorders during pregnancy. J Affect Disord 2011;133:546-52.

13. Forty L, Jones L, Macgregor S et al. Familiarity of postpartum de-pression in unipolar disorder: results of a family study. Am J Psy-chiatry 2006;163:1549-53.

14. Rubinow DR, Schmidt PJ, Meltzer-Brody S et al. Gonadal hormones in women. Concentration versus context. In: Pfaff D, Arnold A, Et-gen A et al (eds). Hormones, brain and behavior, 2nd ed. Amster-dam: Elsevier 2009:2395-428.

15. Gianetto-Berruti A, Feyles V. Premenstrual syndrome. Min Ginecol 2002;54:85-95.

16. Dickerson LM, Mazick PJ, Hunter MH. Premenstrual syndrome. American Family Physician 2003;15:1743-52.

17. Halbreich U, Borenstein J, Pearlstein T et al. The prevalence, impair-ment, impact, and burden of premenstrual dysphoric disorder (PMS/PMDD). Psychoneuroendocrinology 2003;28:1-23.

18. Epperson CN, Steiner M, Hartlage SA et al. Premenstrual dysphor-ic disorder: evidence for a new category for DSM-5. Am J Psychiatry

073_076.indd 75 06/07/12 14:08


2012;169:465-75.19. Wittchen HU, Becker E, Lieb R et al. Prevalence, incidence and

stability of premenstrual dysphoric disorder in the community. Psy-chol Med 2002;32:119-32.

20. Lanczik M, Bergant A, Klier C. Are severe psychiatric disorders in childbed of endogenous or organic nature? German contributions to the biological, nosological and psychopathological research in postpartum psychosis in the 18th and 19th century. Arch Womens Ment Health 2006;9:293-9.

21. Marcé L-V. Traité de la folie des femmes enceintes, des nouvelles accouchées et des nourrices. Paris: J.B. Baillière, 1858.

22. Kraepelin E. Psychiatrie: ein Lehrbuck für Studierende und Aerzte, Vol. 2. Leipzig: Barth, 1899.

23. Frank TT. The hormonal causes of premenstrual tension. Arch Neu-rol Psychiatry 1931;26:53.

24. Munk-Olsen T, Laursen T, Pedersen C et al. New parents and men-tal disorders: a population-based register study. JAMA 2006;296: 2582-9.

073_076.indd 76 06/07/12 14:08

77

Disruptive mood dysregulation with dysphoria disorder: a proposal for ICD-11

The proposed inclusion of disruptive mood dysregulation with dysphoria disorder (DMDDD) in ICD-11 stems from perceived shortcomings in classifying severely irritable chil-dren. Irritability is extremely common in pediatric psychopa-thology, and is also frequent in adults (1,2). It is characterized by high levels of trait anger, with the frequency, intensity, and duration of anger exceeding what is developmentally norma-tive (3). Reflecting its clinical pervasiveness, irritability or anger features in many ICD-10 categories (oppositional defi-ant disorder, hypomania, mania, depressive episode, schizo-phrenia, nonorganic insomnia, acute stress reaction, emo-tionally unstable personality disorder, dissocial personality disorder) and in non-specific and residual categories (adjust-ment disorder with predominant disturbance of other emo-tions, neurasthenia). Often, the most salient clinical manifes-tation of irritability is frequent and severe temper outbursts.

The role of irritability in psychiatric nosology has received increased attention because of the suggestion that pediatric bipolar disorder does not present with distinct manic epi-sodes as in adults, but with non-episodic severe irritability and symptoms of attention-deficit/hyperactivity disorder (ADHD). For bipolar disorder, both ICD-10 and DSM-IV require one or more manic or hypomanic episodes lasting at least several days and differing clearly from the individual’s usual presentation (4,5). Some manic and ADHD symptoms overlap (i.e., distractibility, pressured speech, psychomotor agitation, racing thoughts, increased goal-directed activity), although these are persistent in ADHD vs. episodic in mania. Similarly, while the abnormal mood in mania can be irritabil-ity instead of (or in addition to) elation, irritability qualifies as a manic symptom only if it is episodic, not characteristic of that individual.

This episodic presentation is instantiated in both the ICD-10 and DSM-IV descriptions of mania. For example, in

EllEn lEibEnluft1, Rudolf uhER2,3, MichaEl RuttER2

1Section on Bipolar Spectrum Disorders, Emotion and Development Branch, Intramural Research Program, National Institute of Mental Health, National Institutes of Health, Department of Health and Human Services, Bethesda, MD, USA; 2Medical Research Council Social, Genetic, and Developmental Psychiatry Centre, Institute of Psychiatry, Kings College, London, UK; 3Department of Psychiatry, Dalhousie University, Halifax, NS, Canada

Recently, there has been an upsurge of interest in the diagnostic classification of youth with severe, non-episodic irritability. The relevant Work Groups for both ICD-11 and DSM-5 have recommended that, to facilitate clinical care and treatment research on these severely affected children, they should be classified separately from both bipolar disorder and oppositional defiant disorder. Data indicate that severe, chronic irritability is not a pediatric presentation of bipolar disorder. Oppositional defiant disorder encompasses a broad range of severity, includes both irritable and headstrong components, and is classified as a behavioral, rather than a mood, disorder. Therefore, the draft proposals for both ICD-11 and DSM-5 include a new diagnosis, which in ICD-11 would be called disruptive mood dysregulation with dysphoria disorder. Here, we review the historical background, rationale, and criteria for such a diagnosis, as well as the differences between the ICD-11 and DSM-5 proposals.

Key words: Disruptive mood dysregulation with dysphoria disorder, children, chronic irritability, oppositional defiant disorder, bipolar disorder


DSM-IV, a manic episode consists of: a) change in mood from baseline (elated, expansive, or irritable); and b) concur-rently, onset or worsening of specific cognitive and behav-ioral symptoms. Nonetheless, researchers began suggesting that youth with non-episodic, severe irritability and ADHD symptoms should be diagnosed with mania and, hence, bipo-lar disorder (6-9). The advent of this view coincided with a marked increase in the rate at which American children re-ceived the bipolar disorder diagnosis. Pediatric outpatient visits with a bipolar disorder diagnosis climbed from 25 to 1003 per 100,000 population from 1994 to 2002, with males with comorbid ADHD being disproportionately common in pediatric vs. adult bipolar disorder samples (10). In American hospitals, the discharge diagnosis of bipolar disorder in youth increased four- to six-fold between 1996 and 2004 (11).

To address these nosological questions, Leibenluft and colleagues began to test the hypothesis that bipolar disorder presents in youth as severe, chronic irritability rather than episodic mood and behavioral changes. To operationalize the controversial phenotype reliably, the researchers defined the severe mood dysregulation syndrome, characterized by severe irritability, chronicity, and the hyperarousal symptoms common to mania and ADHD (12). Leibenluft et al tested whether severe mood dysregulation differed from clearly episodic bipolar disorder on longitudinal course, family his-tory, and brain function. Longitudinal course is a particu-larly important validator; if severe chronic irritability in youth is a presentation of bipolar disorder, then such youth should develop manic episodes as they age. A preliminary test of this hypothesis was conducted in youth with severe mood dys-regulation (N=84, mean age 11.6±2.3 years) followed for a median of 28 months. Only one developed a hypomanic or manic episode (13). Consistent with these clinical data, two epidemiologic studies that followed youth from adolescence

077_081.indd 77 06/07/12 14:05


until age 18 or 33 found that adolescent chronic irritability was associated with adult unipolar depressive and anxiety disorders, but not mania (14,15).

These studies are consistent with research on opposition-al defiant disorder. While severe mood dysregulation and op-positional defiant disorder differ (see below), both select for youth with clinically impairing irritability. Emerging data indicate that oppositional defiant disorder is comprised of headstrong and irritable dimensions, with headstrong symp-toms predicting antisocial outcomes, and irritability predict-ing unipolar depressive and anxiety disorders (16-18). Thus, studies arising from the controversy regarding pediatric bipo-lar disorder and those of oppositional defiant disorder both find that youth irritability is associated with risk for unipolar depressive disorders and anxiety disorders, but not bipolar disorder, in adulthood.

If severe, non-episodic irritability is a form of bipolar dis-order, one would expect youth with severe mood dysregula-tion to have familial loading for bipolar disorder. In a pilot study, 3% of youth with severe mood dysregulation had a bipolar parent, compared to 33% of bipolar youth (19). Fi-nally, while brain imaging measures cannot currently be used as diagnostic tools, “proof of principle” studies suggest that brain function differs between bipolar subjects and those with severe mood dysregulation when performing face pro-cessing, frustration, or response reversal tasks (20-22).

Thus, available data do not support the contention that severe, chronic irritability should be diagnosed as mania in youth. These data, coupled with the markedly increased use of the bipolar disorder diagnosis in American youth, suggest there should be increased clarity in DSM and ICD regarding the differentiation of bipolar disorder from other childhood disorders characterized by irritability.

OverarChIng ratIOnale fOr ICD-11 anD DSM-5 prOpOSalS

While research demonstrates that chronic irritability is not a pediatric presentation of bipolar disorder, the increased re-ferrals to “child bipolar” clinics highlights that many youth who need clinical attention are not covered satisfactorily by current classifications. To facilitate clinical care and treatment research, these patients should be separated from both those with bipolar disorder and those with oppositional defiant dis-order. Their condition is characterized by a chronically pre-vailing negative mood punctuated by anger outbursts, and is termed disruptive mood dysregulation with dysphoria disor-der (DMDDD) in the ICD-11 proposal and disruptive mood dysregulation disorder (DMDD) in the DSM-5 proposal.

The distinction from bipolar disorder is clear. Episodic mood elevation or severe irritability, clearly distinct from pre-vailing inter-episode mood, defines bipolar disorder and, in the ICD-11 and DSM-5 proposals, excludes the DMDDD diagnosis. However, the relative standing of DMDDD and oppositional defiant disorder is less clear. Since most subjects

with severe mood dysregulation also fulfill criteria for oppo-sitional defiant disorder, specifying that the more severe DMDDD diagnosis trumps oppositional defiant disorder avoids artificial comorbidity. However, within the DSM Work Group, there was a desire to maintain continuity with previ-ous research and practice and hence to retain separate coding for oppositional defiant disorder and DMDD. Thus, as de-tailed below, the two proposals agree on the need for a new diagnosis and on its main features, but treat the boundary between DMD(D)D and oppositional defiant disorder differ-ently.

OptIOnS fOr ICD-11

While the literature on severe mood dysregulation ema-nates largely from one American research group, European and South American publications on this condition (23-26) and related phenotypes (27-29) demonstrate the internation-al relevance of the phenotype. Therefore, the ICD-11 Work-ing Group on the Classification of Mental and Behavioural Disorders in Children and Adolescents has suggested that DMDDD be added to ICD-11, but identified as a category that requires further testing.

The Working Group felt that DMDDD met the criteria for addition because it addresses a failure in ICD-10 to classify a clinically salient problem, but that additional information needs to be provided regarding the validity of the diagnosis as a separate and definite diagnostic category. DMDDD would be placed in the mood disorders grouping. Since irri-tability relates strongly to unipolar depression rather than disruptive behavioral disorders (16), the mood element is emphasized in the name (“with dysphoria”) and the diagnos-tic guidelines.

Consistent with ICD’s more prototypic approach, salient features of a typical presentation of DMDDD are described in accessible language, thus facilitating use by medical and nonmedical professionals across setting and cultures (Table 1). The guidelines are descriptive (rather than prescriptive), leaving room for clinical judgment and avoiding the use of overly and artificially precise frequency and duration criteria that might restrict the use of the diagnosis.

For DMDDD, the proposed guidelines specify two essen-tial features: anger outbursts and prevailing negative mood. The anger outbursts must: a) involve an immoderate response to minor provocations; b) involve poor control over negative emotions such as anger; c) be frequent enough to be prob-lematic (i.e., typically, several times a week); d) be present in at least two settings (home, school, peers), and severe in one of these; and e) be persistent. The prevailing negative mood must be irritable, angry and/or sad. Consistent with ICD-11 convention, the DMDDD diagnosis is open to all age groups. However, the problems will usually develop in childhood and be established by age 12. Hopefully, the new category will stimulate research leading to more accurate description of manifestations across age groups.

077_081.indd 78 06/07/12 14:05


The current proposal for ICD-11 suggests a clear hierarchy where, if a child fits the guidelines for the diagnosis of bipolar disorder, then the diagnosis of DMDDD would not also be assigned. In turn, if a diagnosis of DMDDD is assigned, then a diagnosis of oppositional defiant disorder would not also be assigned to the same child. Thus, more specific disorders are given precedence over less specific ones, avoiding unneces-sary artificial comorbidity due to overlapping symptoms.

DSM-5 prOpOSal: DISruptIve MOOD DySregulatIOn DISOrDer

Given the high prevalence of the pediatric bipolar disorder diagnosis in American clinical settings, and data indicating that youth with severe, chronic irritability are not suffering from a developmental phenotype of bipolar disorder, the DSM-5 Work Group wanted to ensure the availability of ap-propriate diagnostic options for youth with severe, chronic irritability. To accomplish the latter, a new diagnosis of DMDD has been proposed (see www.dsm5.org). The proposed crite-ria for the new diagnosis follow closely on those for severe mood dysregulation (12). Specifically, the criteria require re-current, developmentally inappropriate temper outbursts at least three times a week and irritable, angry mood between outbursts. These symptoms are required to be present for at least a year in at least two of three settings (home, school, peers) and to have had their onset before age 10. The diagno-sis cannot be assigned in patients younger than six years old, and hypomanic episodes of one day’s duration or longer are exclusionary. Of note, the hyperarousal symptoms required by severe mood dysregulation are not part of DMDD. In the Work Group’s judgment, such symptoms could be denoted by a comorbid ADHD diagnosis.

In recommending this new diagnosis, the DSM-5 Work Group considered the data indicating that the most common lifetime DSM-IV diagnoses of youth meeting criteria for se-

vere mood dysregulation are ADHD (86.3%), oppositional defiant disorder (84.9%), anxiety disorders (58.4%), and ma-jor depressive disorder (16.4%) (1). Importantly, ADHD and oppositional defiant disorder are disruptive behavior disor-ders, not mood disorders, so severe mood dysregulation youth often fail to receive mood disorder diagnoses even though their major clinical problem is severe irritability. Also, while oppositional defiant disorder is designed in part to cap-ture pathologic irritability, a non-irritable child can meet op-positional defiant disorder criteria due to headstrong behav-iors. Finally, oppositional defiant disorder captures a wide range of impairment, from relatively mild to severe; for ex-ample, a child who experiences irritability only at home can meet oppositional defiant disorder criteria.

In the US, where diagnosis impacts on access to services, an oppositional defiant disorder diagnosis may not allow a child access to intensive services, even though severe mood dysregulation is as impairing as bipolar disorder (1). There-fore, clinicians may assign the bipolar disorder diagnosis to severe mood dysregulation youth, even though this is con-trary to the available data. Also, it is unclear whether mood stabilizers and atypical antipsychotic medications, medica-tions of choice for bipolar disorder, should be first-line treat-ment in severe mood dysregulation or related phenotypes. Most clinicians view a bipolar disorder diagnosis to be a rela-tive contraindication to treatment with serotonergic reuptake inhibitors or stimulant medication, but evidence suggests these might help some irritable patients (30,31). This high-lights the limitations of oppositional defiant disorder and of bipolar disorder as diagnoses for severely irritable youth.

The counter-proposal to the new diagnosis was a specifier to oppositional defiant disorder, and possibly other diagno-ses, denoting severe irritability (32). One argument is that severe mood dysregulation can be seen as being on a dimen-sion with oppositional defiant disorder, rather than as a dis-tinct entity. A second argument against the new diagnosis is the non-specificity of irritability and the concern that clini-

Table 1 Disruptive mood dysregulation with dysphoria disorder (DMDDD) proposal for ICD-11

Definition

DMDDD is characterized by regularly occurring immoderate responses to minor provocations associated with poor control over expression of negative emotions such as anger, plus a prevailing mood that is persistently negative (i.e., irritable, angry and/or sad). The outbursts and/or negative mood must be present across a number of settings.

Diagnostic guidelines

There are two essential features of the disorder: outbursts of anger and prevailing negative mood.The anger outbursts must involve five components. First, they must involve an immoderate response to minor provocations; second, there must be poor control over expressions of negative emotions such as anger; third, the outbursts must be sufficiently frequent to constitute a problem (such as a frequency of several times a week); fourth, in order to differentiate it from a purely relational problem, manifestations must be evident in at least two settings (such as at home, at school or with peers) and must be severe in at least one of these settings; and fifth, the manifestations must have been persistent over time. In line with the general ICD policy, a minimum duration of one month is required, but the concept is of a persistence over a much greater period of time than that. Thus, the clinician might well expect that the phenomena have gone on for at least a year.The second requirement is that there must be a prevailing mood that is persistently negative in the sense of the individual being irritable, angry and/or sad.The concept of this disorder involves an onset in the childhood years (say, before the age of 10 years). However, because of the high frequency of temper outbursts in younger children, the diagnosis should be made only if the chronological age is at least 6 years (or an equivalent developmental level).The diagnosis cannot be made if there is a history of bipolar disorder, or if the symptoms are due to a direct physiological consequence of either the use of substances or a general medical or neurological condition.

077_081.indd 79 06/07/12 14:05


cians will apply the diagnosis inappropriately, e.g., when ir-ritability occurs only during a depressive episode or in anxi-ety-provoking situations, or is secondary to the rigidity of pervasive developmental disorder. Thirdly, all of the work on the proposed phenotype emanates from one research group, and clearly needs replication. Despite the merits of these ar-guments, the DSM-5 Work Group recommended that the DMDD diagnosis be created and placed in the mood disor-ders section of DSM-5. This was based on the limited use of specifiers in clinical settings and the desire to have a diagno-sis for severely irritable children in the mood, rather than behavior, disorders section to focus clinicians’ attention on mood symptoms.

DIfferenCeS between ICD-11 anD DSM-5 prOpOSalS

The two notable differences between the proposals relate to symptom duration (consistent with different ICD-11 and DSM-5 approaches) and to the relationship between DMDDD and oppositional defiant disorder. The DSM-5 proposal re-quires symptoms to be present for at least a year. The ICD-11 proposal requires a minimum duration of one month, a uni-versal proposal for almost all disorders, while stating that clinicians might expect the syndrome to have persisted for at least a year. The rationale for the uniform duration criterion proposed for ICD-11 is to maintain consistency across diag-noses, and to avoid specifying varying duration requirements across disorders that are not based on data. This follows the ICD principle of not using duration as a major factor in diag-nosis since, for severe disorders, it is clear relatively early that a disorder is present and it is counterproductive to delay treatment. Milder forms may require longer follow-up to es-tablish if the condition merits diagnosis and clinical atten-tion. Finally ICD-11, but not DSM-5, suggests a hierarchy in which the diagnosis of DMDDD would take precedence over that of oppositional defiant disorder, such that both di-agnoses would not be assigned to the same child.

COnCluSIOnS

Recent developments in both practice and research dem-onstrate an increased awareness of the importance of severe irritability in childhood psychopathology. In the wake of this awareness, questions have arisen as to how well our current diagnostic systems meet the clinical needs of these impaired children. In response, the drafters of both ICD-11 and DSM-5 are considering the addition of DMDDD to psychiatric nosology, with the goal of improving recognition and treat-ment for this important population.

acknowledgements

The research described herein was supported (in part) by the Intramural Research Program of the National Institute of Mental Health.

note

E. Leibenluft is a member of the WHO ICD Revision Working Group on the Classification of Mood and Anxiety Disorders, reporting to the International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders. M. Rutter is a member of the WHO International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders and the Chairman of the ICD Revision Working Group on the Classification of Mental and Behavioural Dis-orders in Children and Adolescents. R. Uher is a consultant to the ICD Revision Working Group on the Classification of Mental and Behavioural Disorders in Children and Adoles-cents. The views expressed in this article are those of the authors and, except as specifically noted, do not represent the official policies or positions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.


references

1. Leibenluft E. Severe mood dysregulation, irritability, and the diag-nostic boundaries of bipolar disorder in youths, Am J Psychiatry 2011;168:129-42.

2. Fava M, Hwang I, Rush AJ et al. The importance of irritability as a symptom of major depressive disorder: results from the National Comorbidity Survey Replication. Mol Psychiatry 2009;15:856-67.

3. Spielberger CD, Reheiser EC, Sydeman SJ. Measuring the experi-ence, expression and control of anger. Issues Compr Pediatr Nurs 1995;18:207-32.

4. American Psychiatric Association. Diagnostic and statistical manu-al of mental disorders, 4th ed., text revision. Washington: American Psychiatric Association, 2000.

5. World Health Organization. The ICD-10 classification of mental and behavioural disorders. Clinical descriptions and diagnostic guide-lines. Geneva: World Health Organization, 1992.

6. Biederman J, Klein RG, Pine DS et al. Resolved: mania is mistaken for ADHD in prepubertal children. J Am Acad Child Adolesc Psy-chiatry 1998;37:1091-9.

7. Papolos D, Mattis S, Golshan S et al. Fear of harm, a possible phe-notype of pediatric bipolar disorder: a dimensional approach to di-agnosis for genotyping psychiatric syndromes. J Affect Disord 2009; 118:28-38.

8. Mick E, Spencer T, Wozniak J et al. Heterogeneity of irritability in attention-deficit/hyperactivity disorder subjects with and without mood disorders. Biol Psychiatry 2005;58:576-82.

9. Emiroglu FN, Diler RS. Pediatric bipolar disorders: from the per-spective of Turkey. J Can Acad Child Adolesc Psychiatry 2009;18: 206-14.

10. Moreno C, Laje G, Blanco C et al. National trends in the outpatient

077_081.indd 80 06/07/12 14:05


diagnosis and treatment of bipolar disorder in youth. Arch Gen Psy-chiatry 2007;64:1032-9.

11. Blader JC, Carlson GA. Increased rates of bipolar disorder diagnoses among US child, adolescent, and adult inpatients, 1996-2004. Biol Psychiatry 2007;62:107-14.

12. Leibenluft E, Charney DS, Towbin KE et al. Defining clinical phe-notypes of juvenile mania. Am J Psychiatry 2003;160:430-7.

13. Stringaris A, Baroni A, Haimm C et al. Pediatric bipolar disorder versus severe mood dysregulation: risk for manic episodes on fol-low-up. J Am Acad Child Adolesc Psychiatry 2010;49:397-405.

14. Brotman MA, Schmajuk M, Rich BA et al. Prevalence, clinical cor-relates, and longitudinal course of severe mood dysregulation in children. Biol Psychiatry 2006;60:991-7.

15. Stringaris A, Cohen P, Pine DS et al. Adult outcomes of youth irri-tability: a 20-year prospective community-based study. Am J Psy-chiatry 2009;166:1048-54.

16. Stringaris A, Goodman R. Three dimensions of oppositionality in youth. J Child Psychol Psychiatry 2009;50:216-23.

17. Loeber R, Burke J, Pardini DA. Perspectives on oppositional defiant disorder, conduct disorder, and psychopathic features. J Child Psy-chol Psychiatry 2009;50:133-42.

18. Rowe R, Costello EJ, Angold A et al. Developmental pathways in oppositional defiant disorder and conduct disorder. J Abnorm Psy-chol 2010;119:726-38.

19. Brotman MA, Kassem K, Reising MM et al. Parental diagnoses in youth with narrow phenotype bipolar disorder or severe mood dys-regulation. Am J Psychiatry 2007;164:1238-41.

20. Brotman MA, Rich BA, Guyer AE et al. Amygdala activation during emotion processing of neutral faces in children with severe mood dysregulation versus ADHD or bipolar disorder. Am J Psychiatry 2010; 167:61-9.

21. Adleman NE, Kayser R, Dickstein D et al. Neural correlates of re-versal learning in severe mood dysregulation and pediatric bipolar disorder. J Am Acad Child Adolesc Psychiatry 2011;50:1173-85.

22. Rich BA, Carver F, Holroyd T et al. Magnetoencephalography iden-

tifies different neural pathways to negative affect in youth with pe-diatric bipolar disorder and severe mood dysregulation. J Psychiatr Res 2011;45:1283-94.

23. Serra Giacobo R, Jané MC, Bonillo A et al. Somatic symptoms, se-vere mood dysregulation, and aggressiveness in preschool children. Eur J Pediatr (in press).

24. Schapir L, Zalsman M. Bipolar disorder and severe mood dysregula-tion in children and adolescents. Harefuah 2008;147:526-31.

25. Heiler S, Legenbauer T, Bogen T et al. Severe mood dysregulation: in the “light” of circadian functioning. Med Hypotheses 2011;77: 692-5.

26. Krieger FV, Pheula GF, Coelho R et al. An open-label trial of ris-peridone in children and adolescents with severe mood dysregula-tion. J Child Adolesc Psychopharmacol 2011;21:237-43.

27. Holtmann M, Buchmann AF, Esser G et al. The Child Behavior Checklist-Dysregulation profile predicts substance use, suicidality, and functional impairment: a longitudinal analysis. J Child Psychol Psychiatry 2011;52:139-47.

28. Althoff RR, Verhulst FC, Rettew DC et al. Adult outcomes of child-hood dysregulation: a 14-year follow-up study. J Am Acad Child Adolesc Psychiatry 2010;49:1105-16.

29. Pickles A, Aglan A, Collishaw S et al. Predictors of suicidality across the life span: the Isle of Wight study. Psychol Med 2010;40:1453-66.

30. Connor DF, Glatt SJ, Lopez ID et al. Psychopharmacology and ag-gression, I: A meta-analysis of stimulant effects on overt/covert ag-gression-related behaviors in ADHD. J Am Acad Child Adolesc Psy- chiatry 2002;41:253-61.

31. Coccaro EF, Lee RJ, Kavoussi RJ. A double-blind, randomized, placebo-controlled trial of fluoxetine in patients with intermittent explosive disorder. J Clin Psychiatry 2009;70:653-62.

32. Axelson DA, Birmaher B, Findling RL et al. Concerns regarding the inclusion of temper dysregulation disorder with dysphoria in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edi-tion. J Clin Psychiatry 2011;72:1257-62.

077_081.indd 81 06/07/12 14:05


Generalized anxiety disorder in ICD-11

The term generalized anxiety disorder (GAD) was first used in ICD-9, but it was not until five years later, with the publication of DSM-III, that GAD got the attention of clini-cal researchers.

Currently extensive data support the idea that GAD is a valid independent clinical entity that can be reliably diag-nosed. Yet uncertainty has affected the enterprise of estab-lishing its diagnostic criteria in each new edition of the DSM, resulting in major revisions in DSM-III-R and DSM-IV, and another major change being proposed for DSM-5.

ICD-10 criteria for GAD differ importantly from each of the recent DSM versions, but there are only a small number of studies that have used ICD criteria. In this context, a deci-sion must be made about whether and how to revise GAD criteria in ICD-11 in order to provide the most valid and usable guidance for clinicians, applicable across cultures and useful outside of mental health settings.

The purpose of this paper is to review the history of GAD criteria in DSM and ICD, including the revision currently being proposed for DSM-5, and to outline the issues and the options for revising ICD-11 criteria.

HIstorICal baCkGrounD

The origin of descriptive typology of anxiety disorders is often attributed to S. Freud’s papers from the late 1800s in which he differentiated anxiety neurosis from neurasthenia (1). The term “anxiety neurosis” was used in both DSM and ICD nomenclature during the first half of the 20th century, and is retained as an inclusion term in ICD-10, in deference to clinicians who might still use it.

The ICD-9 category for anxiety disorders, 300.0 Anxiety States, was described as follows: “Various combinations of physical and mental manifestations of anxiety, not attribut-able to real danger and occurring either in attacks or as a persisting state. The anxiety is usually diffuse and may extend to panic. Other neurotic features such as obsessional or hys-

M. Katherine Shear

Columbia University School of Social Work, Columbia University College of Physicians and Surgeons, Department of Psychiatry, New York, NY, USA

Generalized anxiety disorder (GAD) is clinically significant, disabling, and prevalent, especially among primary care patients. Clinical and community survey data consistently indicate that GAD is an independent clinical entity that can be reliably identified. Effective treatments for GAD have been developed and tested. However, there has been ongoing uncertainty about its defining characteristics and the best way to formulate valid and reliable diagnostic descriptions. DSM diagnostic criteria have continued to evolve from DSM-III to DSM-III-R and DSM-IV. The DSM-5 Work Group is proposing another major change in the criteria set for GAD. In this context, a decision must be made about how to provide useful and reliable diagnostic guidelines for GAD in ICD-11, applicable across cultures and useful outside of mental health settings. A series of options outlined in this paper incorporate accumulated research data and would ensure a higher degree of harmonization between ICD and DSM.

key words: Generalized anxiety disorder, worry, nervous apprehension, ICD, DSM


terical symptoms may be present but do not dominate the clinical picture”. The term “generalized anxiety disorder” ap-pears in ICD-9 in this context, defined as “an anxiety disor-der characterized by free-floating, persistent, and excessive worry for at least six months”. However, the origin of this term is usually inaccurately attributed to DSM-III.

The DSM-III subdivided anxiety neurosis into panic dis-order and GAD, primarily based upon seminal studies by Klein and colleagues that documented the occurrence of a syndrome whose defining feature is sudden unexpected pan-ic attacks, different from other kinds of anxiety (2). The cat-egory of GAD was created as a way to diagnose individuals with persistent anxiety symptoms who did not meet criteria for panic or phobic disorders. However, while panic disorder had specific diagnostic criteria, GAD was included as a re-sidual category, conceptualized by what it was not, rather than what it was.

This idea of GAD as a syndrome of leftover anxiety symp-toms created confusion, some of which persists until today. The DSM-III GAD criterion symptoms of apprehensive ex-pectation, motor tension, autonomic hyperactivation and vigilance/scanning were believed to be non-specific, present in all anxiety disorders, so the defining feature was the ab-sence of any other anxiety disorder. In addition, the diagnosis of GAD required an unspecified number of symptoms from at least three of four categories and a continuous anxious mood for at least one month. Impairment in social and oc-cupational functioning was expected to be “rarely more than mild”. Given these imprecisions in both concept and criteria, the diagnostic reliability of GAD was low (3).

Anxiety disorder research was flourishing in the United States and Europe in the early 1980s, and clinical investiga-tors grew interested in better understanding GAD. Clinical observations led them to challenge the idea of GAD as a re-sidual diagnosis, re-conceptualizing it as a disorder with a principal defining symptom of excessive unrealistic anxiety and worry about multiple life circumstances (4).

Borkovec and colleagues began a seminal series of studies

082_088.indd 82 06/07/12 14:06


of worry (5-10). This group developed the Penn State Worry Questionnaire (PSWQ, 11), which has been used success-fully to discriminate individuals with GAD from those with other anxiety disorders (12) as well as other psychiatric dis-orders and non-psychiatric controls (13). Borkovec also de-veloped and tested a theory of worry as a verbal cognitive avoidance strategy central to GAD, arguing that this avoid-ance helped maintain the disorder (7,9,10,14). Influenced by this thinking, worry became the key defining feature of GAD in DSM-III-R, and is retained as such in DSM-IV and in the proposal for DSM-5.

Moving the symptoms of anxious expectation to a central position was the most important change in DSM-III-R, which further stipulated that anxiety and worry were re-quired to be unrealistic or excessive and to be focused on at least two everyday life circumstances. Symptoms had to be persistent, more days than not, over a period of at least 6 months. The somatic symptoms criterion was also modified in DSM-III-R such that at least 6 of 18 symptoms from three categories, motor tension, autonomic hyperactivity and vigi-lance/scanning, be “often present when anxious”. Addition-ally, since GAD was now considered a full-fledged disorder, comorbidity was permitted if the comorbid disorder was not the focus of the apprehensive expectation, with the excep-tion that the diagnosis was not made if symptoms occurred only during the course of a mood disorder. Interestingly, im-pairment in functioning was still said to be “rarely more than mild”, which, if true, would seem more consistent with a re-sidual state than a full-fledged disorder. Later studies un-equivocally refuted the notion that impairment from this disorder is mild (15,16).

Unfortunately, the DSM-III-R revision did not solve prob-lems of reliability (17,18), and there was continuing concern about comorbidity and discriminant validity of the DSM-III-R criteria (19). Interestingly, subsequent studies in commu-nity and primary care settings have failed to confirm that GAD has higher rates of comorbidity than other anxiety dis-orders. Rather, it appears this observation was an artifact of studies conducted in anxiety disorders clinics. More repre-sentative surveys have shown that, without comorbidity, in-dividuals with GAD are less likely to seek mental health treatment than is the case with other anxiety disorders (15). However, these patients do present to primary care physi-cians for treatment of minor physical problems, and their anxiety disorder is usually not recognized (20,21). In fact, GAD has repeatedly been shown to be one of the most com-mon mental disorders in primary care settings. However, these findings were not known to the authors of DSM-IV, and the new criteria they devised were informed primarily by clinical studies.

The most important change made by the DSM-IV Work Group was removal of autonomic hyperactivity symptoms from the somatic symptoms criterion. The requirement for this criterion was simplified such that at least 3 of 6 symp-toms be endorsed, including motor tension (restlessness or feeling keyed up on edge, easily fatigued, muscle tension) or

vigilance/scanning (difficulty concentrating, sleep distur-bance, irritability). Some of the symptoms should be present more days that not for 6 months. Removal of autonomic symptoms was based upon a relatively small multisite clinical study from the United States (22) that documented lower frequency of endorsement of autonomic symptoms com-pared to those in the motor tension and vigilance/scanning areas. Importantly, though, findings of lower endorsement of autonomic symptoms were subsequently replicated in a number of other settings (e.g., 23-26). Moreover, there is evi-dence that motor tension symptoms differentiate GAD from normal worries and other anxiety disorders, using both self report (27) and physiological measures (28,29).

However, some authors (e.g., 30) have argued that the de-cision to jettison autonomic symptoms is unfortunate, as pa-tients in primary care settings may be focused on a range of bodily complaints, including those in the autonomic hyper-activity grouping. Somewhat in support of this idea, rates of endorsement of autonomic symptoms, while statistically sig-nificantly lower than the other symptom groups, are not neg-ligible (e.g., motor tension endorsed by 93% and autonomic symptoms by 88% of subjects according to Turvey et al (25)). Additionally, cross-cultural studies suggest that some non-Western populations may endorse autonomic symptoms more frequently than observed in studies conducted in the United States (31).

The central role of worry was retained in DSM-IV, though the requirement that it be unrealistic was dropped. Instead, a separate criterion was added to require that the person find it difficult to control the worry. Also, in line with other disor-ders in DSM-IV, a requirement for clinically significant dis-tress or impairment was included. The stipulation that GAD not be diagnosed when it occurred only in the presence of a mood disorder was retained.

Notably, much of the data supporting the revisions to DSM criteria were obtained from help-seeking individuals who presented to specialty anxiety disorder programs in the United States. At the time of preparation of the DSM-IV, few data were available from low- or middle-income countries and from epidemiological or primary care studies. Cross-cultural and epidemiological research is critical in consider-ing revisions to ICD-11, as are considerations of issues rele-vant to primary care and non-mental health settings.

ICD-10 anD DsM-IV

Andrews and colleagues have provided thoughtful discus-sion of the relationship between ICD-10 and DSM-IV crite-ria for GAD (32-35). The ICD, unlike DSM, provides a dif-ferent format for clinical compared to research use. This pa-per focuses primarily on the clinical criteria. However, it seems relevant to note that the ICD-10 diagnostic research criteria (DSR) are fully specified and are different from DSM-IV in several respects. Similar to DSM-IV, they require a pe-riod of at least 6 months with prominent tension, worry and

082_088.indd 83 06/07/12 14:06


feelings of apprehension. However, there is no mention of excessiveness, uncontrollability or frequency of worry. Ad-ditionally, the DSR require the occurrence of at least one symptom of autonomic arousal (palpitations, sweating, trem-bling, shaking or dry mouth), whereas the DSM-IV omits these symptoms from the criteria set. The first 18 DSR so-matic symptoms are almost identical to those listed in DSM-IV for panic disorder, and only 4 of the 6 DSM-IV somatic symptoms are included. Muscle tension and “easily fatigued” are omitted.

ICD-10 clinical guidelines for GAD indicate that “a vari-ety of worries and forebodings” are among the essential fea-tures of the disorder. It is also noted that the dominant symp-toms of the disorder are highly variable and that complaints of continuous feelings of a variety of somatic symptoms, in-cluding those of autonomic arousal, are common. Fears, worries and forebodings are said to be frequently expressed and often related to chronic environmental stress. Clinicians are advised that “the sufferer must have primary symptoms of anxiety most days for at least several weeks at a time, and usually for several months” and that symptoms should usu-ally involve elements of apprehension, motor tension and autonomic reactivity.

The ICD-10 guidelines appropriately allow for a broader range of clinical features, as there is evidence for cultural dif-ferences in anxiety symptomatology. In fact, it is possible that DSM-IV criteria are over-specified and the central focus on worry overly narrow. Several DSM-IV criteria relate to con-cern about differentiating GAD from normal worries in re-sponse to stress. The requirement that apprehensiveness be excessive and uncontrollable was importantly related to this concern, as was the stipulation that symptom duration should be at least 6 months. However, neither removal of the excessiveness criteria nor shortening the duration require-ment seem to change the demographics or associated clinical features of the disorder. On the other hand, it is not clear that the stipulation by ICD-10 that presence of panic or phobic anxiety disorder should preclude the diagnosis of GAD is warranted, if the worry is not solely focused on symptoms of the co-occurring disorder.

There is a general consensus in the field, supported by a substantial amount of clinical and theoretical research, that an independent syndrome of generalized anxiety exists. Co-morbidity with other conditions is common, but rates are not substantially different from other mood and anxiety disor-ders. What is in question is exactly how to best help clinicians and researchers to identify these patients. A recent study comparing DSM-IV and ICD-10 (research) criteria high-lights some of the issues. This study showed low concor-dance (kappa=0.39) of GAD diagnosis made by the two sys-tems. Among a sample of 10,641 people, the weighted 12-month prevalence rates for DSM-IV and ICD-10 GAD were similar (2.6% and 3.0%, respectively). However, there were 123 concordant diagnoses of GAD and 352 discrepant diagnoses. Among the latter, there were 151 cases in which DSM-IV was positive and ICD-10 negative and 201 cases

where the reverse was found. Forty-eight percent of those meeting DSM-IV criteria failed to endorse at least one symp-tom of autonomic arousal and 52% had co-occurring ex-cluded anxiety disorders. Seventy-two percent of individuals meeting ICD-10 criteria failed to endorse excessive worry (though they did report that anxiety was prominent, lasted at least 6 months and was focused on everyday events) and 35% failed to indicate that worry caused clinically significant distress. Eight percent endorsed fewer than 3 of the 6 re-quired somatic symptoms. DSM-IV diagnosis was associated with significantly greater impairment than ICD-10 diagnosis.

These results support the idea that both DSM-IV and ICD-10 are in need of updating. DSM-5 discussions have led to a proposal for revision, but the Work Group has taken the criteria in yet another direction. Their proposed changes are described in the next section. However, the issues raised in the comparative study summarized above might suggest dif-ferent DSM revisions. Correspondingly, the ICD-10 descrip-tion of GAD may also be in need of updating, and options for doing so are discussed below.

It would seem that the requirement for excessiveness of worry should be revisited in DSM-5, since almost three-fourths of the ICD-positive patients did not endorse exces-siveness. Moreover, the term “excessive” has not been opera-tionalized, and ambiguities in rating excessiveness account for the majority of continuing reliability problems (36). Re-moving the excessiveness criterion does not seem to change the demographics or clinical course, though people identi-fied may be less severe.

Another issue that might be discussed is the removal of autonomic symptoms from the DSM criteria set. Although studies show lower rates of endorsement of autonomic symptoms, the vast majority of people who meet GAD crite-ria do endorse these. At least some data fail to confirm that DSM-IV criteria show better discriminant validity compared to DSM-III-R (26) and discriminating GAD was the primary reason for removing these symptoms. Additionally, there are data suggesting cultural differences in the focus on somatic vs. cognitive symptoms (31) and these kinds of differences are of considerable importance for an international diagnos-tic system. However, neither of these changes are proposed for DSM-5.

ProPoseD CHanGes for DsM-5

The DSM-5 Work Group on Anxiety Disorders is again suggesting major revision of the GAD criteria (www.dsm5.org). Four major changes are entailed in this recommenda-tion. First, the required symptom duration has been reduced from six to three months, because the 6-month requirement excludes a sizable number of people who have clinically sig-nificant distress and impairment. Studies have consistently documented that a large number of people endorse syndro-mal level GAD symptoms of shorter duration and that dura-tion is largely unrelated to impairment, comorbidity and so-

082_088.indd 84 06/07/12 14:06


cio-demographic profile (20,37-40) or to rates of parental GAD (40) or GAD in co-twins (41). An interesting issue raised by Kessler et al (40) is that both the 6-month require-ment and the excessiveness criterion in DSM-IV are intend-ed to exclude people with high levels of worry related to a severe life stressor. This concern about GAD, with no em-pirical data, is contrasted with a very different conceptualiza-tion of major depression, in which mood symptoms need not be considered excessive and for which duration is set at just two weeks. Adjustment disorder is considered a differential diagnosis for GAD whereas it is a residual diagnosis for de-pression. As reviewed by Maj (42), differentiation of major depression from normal, transient sad mood can be challeng-ing, but data suggest that, even in response to life stress, meet-ing criteria for major depression has important clinical impli-cations.

A second proposed change in DSM-5 is removal of the criterion requiring that worry be difficult to control. The ra-tionale is that controllability is a difficult concept to opera-tionalize and appears to be redundant with the required char-acterization of the worry as excessive (43). Of note, retention of the excessiveness criteria is not without controversy. There is no such requirement for other anxiety disorders or major depression. Studies indicate that removal of the excessiveness criteria, like changing the symptom duration requirement, increases the prevalence of GAD but does not change associ-ated socio-demographic or risk factors (15,44-49).

The third major change proposed for DSM-5 is reduction in the list of somatic symptoms from six to just two (restless-ness/keyed up/on edge and muscle tension), only one of which need be endorsed. The rationale is that such a reduc-tion does not affect prevalence rates, as few people who meet other criteria for GAD endorse just one symptom (48). How-ever, studies of somatic symptoms endorsed by patients with GAD have not addressed this question directly and there is little data regarding how this change might affect the rates of diagnosis of GAD. The reduction to just two symptoms has the advantage of simplifying the criteria and the potential disadvantage of failing to provide clinicians and others with a reasonably complete picture of the types of symptoms com-monly endorsed by patients with GAD. Moreover, the poten-tial cross-cultural implications of this change are important.

The fourth change is also potentially impactful: a require-ment for a marked behavioral change in association with the anxiety is being proposed (43). If adopted, GAD would be diagnosed only if the patient reports a marked behavioral change, manifest as avoidance of situations in which a nega-tive outcome could occur, marked time and effort preparing for situations in which a negative outcome could occur, marked procrastination in behavior or decision-making due to worries, or repeatedly seeking reassurance due to worries. There is little data supporting this change. Only two small studies examined compulsive checking behaviors in GAD and found that they were endorsed with a frequency greater than previously recognized (50,51). There is no international data indicating that these new criteria are warranted. How-

ever, there is expert consensus that addition of a behavioral measure is consistent with the general structure of anxiety disorders and could improve reliability and validity of the diagnosis. It is possible that alerting clinicians to be aware of the possible occurrence of these symptoms would be helpful, though this would be based on opinion rather than on data.

oPtIons for ICD-11

There is now extensive evidence that GAD is a prevalent and impairing condition. A great deal of mechanism and treatment research has also been conducted, mostly using the DSM-IV criteria. However, controversy continues to swirl around the DSM-IV criteria. It makes sense to revisit the ICD-10 description and clinical guidelines in order to determine whether some modifications might be made to best fit our current understanding of the disorder. That said, especially given the proposal for another major revision in DSM criteria, it is not obvious which modifications will be most important and best help clinicians identify people with this disorder. Several options are considered here, with their pros and cons. A priority is placed on retaining ICD-10 con-cepts and language whenever possible.

option 1: emphasize worry about multiple everyday lifeevents as the defining feature of GaD

There is considerable evidence that worry, defined as ap-prehensive expectation focused on multiple life circumstances, is strongly endorsed by people who meet all of the other DSM-IV criteria for GAD. Authors of a recent meta-analysis point out that “there is general consensus that excessive, chronic worry primarily defines GAD” (13). This meta-anal-ysis indicates that anxiety disorder patients generally worry significantly more than non-anxious controls, but patients with GAD worry significantly more than individuals with other anxiety disorders.

Notwithstanding these findings, the general consensus that worry is the defining feature of GAD is not without its challengers. Rickels and Rynn (30) suggest that the term “ap-prehensive expectation” is preferable to “worry” and not nec-essarily synonymous. They further suggest that the overall severity of anxious symptomatology is a better way to make a diagnosis than to focus on a single defining symptom that may not be present in all severely anxious individuals. In-deed, there are patients who present with somatic anxiety and a feeling of non-specific dread. These individuals are sometimes unable to identify anything specific that worries them. In particular, several authors have pointed out that people from Eastern cultures use various body metaphors to describe distress that seems most likely to represent GAD (31,52). Nevertheless, some cross-cultural studies do support the endorsement of worry in other cultures (38,47,53).

082_088.indd 85 06/07/12 14:06


option 2a: stipulate that worry must be excessive

Whether worry needs to be excessive is a subject of debate. Individuals who describe their anxiety as excessive generally have a more severe disorder, but those who do not endorse excessiveness have similar socio-demographic profiles, simi-lar degree of chronicity and impairment, similar comorbidity and treatment seeking (48,49).

The term “excessive” is difficult to operationalize. At least one study indicates that GAD patients are more likely to re-port that others think their worry is excessive than report that they believe this themselves. This may be explained by the observation that people with GAD view worry as a positive way of coping with problems. Some authors consider that it is best for the clinician, rather than the patient, to make the judgment of excessiveness.

option 2b: require that worry must occur frequently, i.e., more hours than not, nearly every day

A recent paper tested the usefulness of two simple sever-ity measures for GAD (54). These authors found that per-cent of the day spent worrying and number of domains of worry were both good measures of severity. Both are easier to operationalize than excessiveness. Both were significant-ly higher in patients with GAD compared to those with panic disorder or social phobia. Patients with GAD worried 71.6% of the day compared to 45.1% of the day for panic disorder and 42.7% for social anxiety disorder. An alterna-tive to option 2a would be to suggest that these patients usually worry or experience anxious apprehension most of the day (i.e., more than half) rather than using the term “ex-cessive” worry. This is likely to increase reliability and pos-sibly include some of the sufferers who are excluded by the excessiveness criteria.

option 3: Modify the narrative description of somatic symptoms to focus on the ones found to be endorsed most commonly in GaD

Numerous studies document higher endorsement of mo-

tor tension symptoms than symptoms of autonomic hyper-arousal. Physiological studies indicate that motor tension is characteristic of GAD. Autonomic symptoms are omitted from the DSM-IV and DSM-5 criteria, but they are endorsed by the majority of people with GAD. In particular, gastroin-testinal symptoms have been found to be as prevalent as mo-tor tension. A recent study of differential patterns of symp-toms in GAD and depression identified muscle pains and gastrointestinal symptoms, along with emotion intensity and intolerance of uncertainty, as specific to GAD (55). Focusing on the symptoms most commonly endorsed by GAD patients should help clinicians better identify these individuals and better differentiate them from those with other anxiety disor-

ders. However, it is also possible that this will increase over-lap with somatic distress disorders.

option 4: Modify the somatic criteria used in the diagnostic guideline to focus primarily on restlessness and muscle tension and to emphasize gastrointestinal symptoms in the list of autonomic symptoms

This option retains the autonomic symptoms and still fo-cuses primarily on the key symptoms of DSM-IV and the two somatic symptoms proposed for inclusion in DSM-5 criteria. The advantage of leaving autonomic symptoms in the criteria is that one or another of these are clearly endorsed by a sig-nificant minority of individuals who meet all the other crite-ria for GAD. Autonomic symptoms might be especially im-portant in diagnosing GAD in non-Western cultures. In some populations, a majority of patients diagnosed with GAD endorse at least one of these symptoms. The disadvantage of including autonomic symptoms is that, as above, there may be overlap with somatic distress disorders. Additionally, there may be lack of concordance with DSM-IV or DSM-5 diagnosis, and individuals with these symptoms might differ in some systematic way from the DSM-diagnosed group. However, to date there is no indication that this would be the case.

option 5: Include a description of common behavioral changes

There are two small studies that suggest compulsive check-ing behaviors occur in a subgroup of patients with GAD. Otherwise, there is little data to support this change. How-ever, the DSM-5 Work Group advisors were unanimous in believing that inclusion of these symptoms will improve the reliability and validity of the diagnosis. Inclusion of these symptoms might help clinicians identify people with GAD.

option 6: Permit the diagnosis of GaD even in the presence of other disorders, as long as the anxious apprehension is not a symptom of another disorder

The origin of the hierarchical rule for GAD has roots in the initial conceptualization of this disorder as a residual di-agnosis, without unique defining features. Extensive research over the past two decades has successfully refuted this idea and shown, instead, that GAD is a serious, independent dis-order that causes considerable distress and impairment. Fail-ing to assign this diagnosis in the presence of other condi-tions will lead to under-treatment of GAD symptoms. That said, similar to all other mood and anxiety disorders, comor-bidity is expected to be high and this complicates the diag-nostic process, especially for non-mental health clinicians.

082_088.indd 86 06/07/12 14:06


ConClusIons

GAD is an important psychiatric disorder and one of the most prevalent mental disorders seen in primary care. Work over the past decades has refuted early conceptualizations of this syndrome as a residual diagnosis. However, confusion about its defining characteristics and debate about the best way to formulate valid and reliable diagnostic criteria have plagued the field.

The ICD-10 definition has important differences from DSM-III, DSM-III-R and DSM-IV criteria, and each of the latter differs in major ways from its predecessor. Currently, the DSM-5 Work Group is proposing another major change in the criteria set, that will take the DSM even farther afield from ICD. The time is right to modify the definition and di-agnostic guidelines for GAD in ICD-11. A series of options are outlined in this paper that incorporate accumulated re-search data and would foster harmonization between ICD and DSM. At the same time, effects of major changes in GAD criteria are unknown.

Field trials will inform the decision making process for DSM-5, but more research is needed to understand the ef-fects of new diagnostic criteria. Additional work is needed to confirm the clinical usefulness of new ICD-11 definition and diagnostic guidelines, should they be adopted.

note

M.K. Shear is a member of the WHO ICD Revision Work-ing Group on the Classification of Mood and Anxiety Disor-ders, reporting to the International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders. The views expressed in this article are those of the author and, except as specifically noted, do not represent the official policies or positions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.

The DSM-5 website was accessed on January 21, 2012. DSM-5 proposals have not been finalized and are subject to change.

references

1. Strachey J. Standard edition of the complete psychological works of Sigmund Freud, Vol. III. London: Hogarth Press, 1962.

2. Klein DF. Delineation of two drug-responsive anxiety syndromes. Psychopharmacologia 1964;5:397-408.

3. Di Nardo PA, O’Brien GT, Barlow DH et al. Reliability of DSM-III anxiety disorder categories using a new structured interview. Arch Gen Psychiatry 1983;40:1070-4.

4. Barlow DH, Blanchard EB, Vermilyea JA et al. Generalized anxiety and generalized anxiety disorder: description and reconceptualiza-tion. Am J Psychiatry 1986;143:40-4.

5. Borkovec TD, Robinson E, Pruzinsky T et al. Preliminary explora-tion of worry: some characteristics and processes. Behav Res Ther 1983;21:9-16.

6. Borkovec TD. Worry: a potentially valuable concept. Behav Res

Ther 1985;23:481-3.7. York D, Borkovec TD, Vasey M et al. Effects of worry and somatic

anxiety induction on thoughts, emotion and physiological activity. Behav Res Ther 1987;25:523-6.

8. Andrews V, Borkovec TD. The differential effects of inductions of worry, somatic anxiety, and depression on emotional experience. J Behav Ther Exp Psychiatry 1988;19:21-6.

9. Borkovec TD, Hu S. The effect of worry on cardiovascular response to phobic imagery. Behav Res Ther 1990;28:69-73.

10. Borkovec TD, Inz J. The nature of worry in generalized anxiety dis-order: a predominance of thought activity. Behav Res Ther 1990; 28:153-8.

11. Meyer TJ, Miller ML, Metzger RL et al. Development and validation of the Penn State Worry Questionnaire. Behav Res Ther 1990;28: 487-95.

12. Fresco DM, Mennin DS, Heimberg RG et al. Using the Penn State Worry Questionnaire to identify individuals with generalized anxiety disorder: a receiver operating characteristic analysis. J Behav Ther Exp Psychiatry 2003;34:283-91.

13. Olatunji BO, Wolitzky-Taylor KB, Sawchuk CN et al. Worry and the anxiety disorders: a meta-analytic synthesis of specificity to GAD. Applied and Preventive Psychology (in press).

14. Metzger RL, Miller ML, Cohen M et al. Worry changes decision making: the effect of negative thoughts on cognitive processing. J Clin Psychol 1990;46:78-88.

15. Kessler RC, Keller MB, Wittchen HU. The epidemiology of general-ized anxiety disorder. Psychiatr Clin North Am 2001;24:19-39.

16. Weisberg RB. Overview of generalized anxiety disorder: epidemiol-ogy, presentation, and course. J Clin Psychiatry 2009;70(Suppl. 2): 4-9.

17. Mannuzza S, Fyer AJ, Martin LY et al. Reliability of anxiety assess-ment. I. Diagnostic agreement. Arch Gen Psychiatry 1989;46:1093-101.

18. Di Nardo P, Moras K, Barlow DH et al. Reliability of DSM-III-R anxiety disorder categories. Using the Anxiety Disorders Interview Schedule-Revised (ADIS-R). Arch Gen Psychiatry 1993;50:251-6.

19. Brown TA, Barlow DH, Liebowitz MR. The empirical basis of gen-eralized anxiety disorder. Am J Psychiatry 1994;151:1272-80.

20. Maier W, Gansicke M, Freyberger HJ et al. Generalized anxiety dis-order (ICD-10) in primary care from a cross-cultural perspective: a valid diagnostic entity? Acta Psychiatr Scand 2000;101:29-36.

21. Wittchen HU, Kessler RC, Beesdo K et al. Generalized anxiety and depression in primary care: prevalence, recognition, and manage-ment. J Clin Psychiatry 2002;63(Suppl. 8):24-34.

22. Marten PA, Brown TA, Barlow DH et al. Evaluation of the ratings comprising the associated symptom criterion of DSM-III-R general-ized anxiety disorder. J Nerv Ment Dis 1993;181:676-82.

23. Starcevic V, Fallon S, Uhlenhuth EH. The frequency and severity of generalized anxiety disorder symptoms. Toward a less cumbersome conceptualization. J Nerv Ment Dis 1994;182:80-4.

24. Brown T, Marten P, Barlow D. Discriminant validity of the symp-toms constituting the DSM-III-R and DSM-IV associated symptom criteria of generalized anxiety disorder. J Anxiety Disord 1995;9: 317-28.

25. Turvey C, Stevens D, Merikangas K. Validity of the associated symp-tom criteria for DSM-IV generalized anxiety disorder. Int J Methods Psychiatr Res 1999;8:129-37.

26. Kubarych TS, Aggen SH, Hettema JM et al. Endorsement frequen-cies and factor structure of DSM-III-R and DSM-IV generalized anxiety disorder symptoms in women: implications for future re-search, classification, clinical practice and comorbidity. Int J Meth-ods Psychiatr Res 2005;14:69-81.

27. Freeston M, Dugas M, Letarte H et al. Physical symptoms associated with worry in a nonclinical population. J Anxiety Disord 1996;10: 365-77.

28. Hoehn-Saric R, McLeod DR, Zimmerli WD. Somatic manifesta-tions in women with generalized anxiety disorder. Psychophysiolog-

082_088.indd 87 06/07/12 14:06


ical responses to psychological stress. Arch Gen Psychiatry 1989; 46:1113-9.

29. Hoehn-Saric R. Psychic and somatic anxiety: worries, somatic symptoms and physiological changes. Acta Psychiatr Scand 1998; 98(Suppl. 393):32-8.

30. Rickels K, Rynn M. What is GAD? J Clin Psychiatry 2001;62(Suppl. 1):4-11.

31. Hoge EA, Tamrakar SM, Christian KM et al. Cross-cultural differ-ences in somatic presentation in patients with generalized anxiety disorder. J Nerv Ment Dis 2006;194:962-6.

32. Andrews G, Slade T. The classification of anxiety disorders in ICD-10 and DSM-IV: a concordance analysis. Psychopathology 2002; 35:100-6.

33. Andrews G, Slade T, Peters L. Classification in psychiatry: ICD-10 versus DSM-IV. Br J Psychiatry 1999;174:3-5.

34. Andrews G, Slade T, Peters L et al. Generalized anxiety disorder, obsessive-compulsive disorder and social phobia: sources of disso-nance between ICD-10 and DSM-IV. Int J Methods Psychiatr Res 1998;7:110-5.

35. Slade T, Andrews G. DSM-IV and ICD-10 generalized anxiety dis-order: discrepant diagnoses and associated disability. Soc Psychiatry Psychiatr Epidemiol 2001;36:45-51.

36. Wittchen HU, Kessler RC, Zhao S et al. Reliability and clinical valid-ity of UM-CIDI DSM-III-R generalized anxiety disorder. J Psychiatr Res 1995;29:95-110.

37. Joseph Bienvenu O, Nestadt G, Eton W. Characterizing generalized anxiety: temporal and symptomatic thresholds. J Nerv Ment Dis 1998; 186:51-6.

38. Lee S, Tsang A, Chui H et al. A community epidemiological survey of generalized anxiety disorder in Hong Kong. Commun Ment Health J 2007;43:305-19.

39. Angst J, Gamma A, Joseph Bienvenu O et al. Varying temporal cri-teria for generalized anxiety disorder: prevalence and clinical char-acteristics in a young age cohort. Psychol Med 2006;36:1283-92.

40. Kessler RC, Brandenburg N, Lane M et al. Rethinking the duration requirement for generalized anxiety disorder: evidence from the Na-tional Comorbidity Survey Replication. Psychol Med 2005;35:1073-82.

41. Kendler KS, Neale MC, Kessler RC et al. Clinical characteristics of familial generalized anxiety disorder. Anxiety 1994;1:186-91.

42. Maj M. Differentiating depression from ordinary sadness:contextual,

qualitative and pragmatic approaches. World Psychiatry 2012;11 (Suppl. 1):43-7.

43. Andrews G, Hobbs MJ, Borkovec TD et al. Generalized worry dis-order: a review of DSM-IV generalized anxiety disorder and options for DSM-V. Depress Anxiety 2010;27:134-47.

44. Beesdo-Baum K, Winkel S, Pine DS et al. The diagnostic threshold of generalized anxiety disorder in the community: a developmental perspective. J Psychiatr Res 2011;45:962-72.

45. Gordon D, Heimberg RG. Reliability and validity of DSM-IV gener-alized anxiety disorder features. J Anxiety Disord 2011;25:813-21.

46. Kessler RC, Wittchen HU. Patterns and correlates of generalized anxiety disorder in community samples. J Clin Psychiatry 2002; 63(Suppl. 8):4-10.

47. Lee S, Ma YL, Tsang A et al. Generalized anxiety disorder with and without excessive worry in Hong Kong. Depress Anxiety 2009;26: 956-61.

48. Ruscio AM, Chiu WT, Roy-Byrne P et al. Broadening the definition of generalized anxiety disorder: effects on prevalence and associa-tions with other disorders in the National Comorbidity Survey Rep-lication. J Anxiety Disord 2007;21:662-76.

49. Ruscio AM, Lane M, Roy-Byrne P et al. Should excessive worry be required for a diagnosis of generalized anxiety disorder? Results from the US National Comorbidity Survey Replication. Psychol Med 2005;35:1761-72.

50. Townsend MH, Weissbecker KA, Barbee JG et al. Compulsive be-havior in generalized anxiety disorder and obsessive-compulsive disorder. J Nerv Ment Dis 1999;187:697-9.

51. Schut AJ, Castonguay LG, Borkovec TD. Compulsive checking be-haviors in generalized anxiety disorder. J Clin Psychol 2001;57:705-15.

52. Hinton DE, Park L, Hsia C et al. Anxiety disorder presentations in Asian populations: a review. CNS Neurosci Ther 2009;15:295-303.

53. Lee JK, Orsillo SM, Roemer L et al. Distress and avoidance in gen-eralized anxiety disorder: exploring the relationships with intoler-ance of uncertainty and worry. Cogn Behav Ther 2010;39:126-36.

54. Niles AN, Lebeau R, Liao B et al. Dimensional indicators of general-ized anxiety disorder severity for DSM-V. J Anxiety Disord (in press).

55. Aldao A, Mennin DS, Linardatos E et al. Differential patterns of physical symptoms and subjective processes in generalized anxiety disorder and unipolar depression. J Anxiety Disord 2010;24:250-9.

082_088.indd 88 06/07/12 14:06

89

Agoraphobia and panic disorder: options for ICD-11

This paper addresses the question of how agoraphobia and panic disorder should be characterized in the ICD-11. It briefly reviews the historical background to the nosology of these disorders, the approaches taken by the ICD-10 and DSM-IV, the problems arising from their application, the proposals for DSM-5 and the options for ICD-11.

HIstorICAl bACkgrounD

Agoraphobia and panic attacks have long been described in the clinical literature (1). One of Kraepelin’s lectures, Ir-repressible ideas and irresistible fears, for example, clearly describes spontaneous panic attacks and subsequent avoid-ance behavior (2). Early on, Westphal also described a series of patients with unexpected and situational panic attacks and anticipatory anxiety, and coined the term “agoraphobia” (3). Not long afterwards, Freud noted that in each of the anxiety neuroses there was anxiety, anxious expectation, and phobic avoidance (4).

A key issue over time has been the question of how opti-mally to “lump” or “split” anxiety symptoms and syndromes, and more recently, the primacy of agoraphobia versus panic disorder (4). Freud, for example, argued that “anxiety neuro-sis” was a general syndrome characterized by specific uncon-scious mechanisms, a position that strongly influenced DSM-I and DSM-II (5). On the other hand, many authors, including Freud, noted that onset of panic attacks may be followed by agoraphobia (2), a position that was emphasized in DSM-III and particularly in DSM-III-R and DSM-IV (4). Differing views on these issues have contributed to impor-tant differences in the major psychiatric nosologies.

The DSM-III and DSM-IV approach to anxiety disorders reflected in part an emphasis on descriptive phenomenology (dropping the idea of “neurosis” as an organizing principle), and in part an emphasis on new findings from psychophar-macology. Klein had noted that panic disorder responded to the tricyclic antidepressant imipramine, and had suggested that this differentiated it from other anxiety disorders, which

Dan J. Stein

Department of Psychiatry, University of Cape Town, Groote Schuur Hospital, Cape Town 7925, South Africa

This paper addresses the question of how agoraphobia and panic disorder should be characterized in the ICD-11. The approaches of ICD-10 and DSM-IV to this issue are quite different. ICD-10 provides separate diagnostic categories for both agoraphobia and panic disorder, but regards agoraphobia as primary when they co-occur. In contrast, DSM-IV gives priority to panic disorder, allowing the diagnosis of agorapho-bia only in the absence of a history of panic disorder. DSM-5 proposals are bringing the DSM and ICD nosologies into closer alignment in a number of key ways. ICD-11 options are considered, with an emphasis on considerations of clinical utility, global applicability, and applica-bility outside mental health settings.

key words: Agoraphobia, panic attack, panic disorder, DSM, ICD


responded to benzodiazepines (6). Drawing on his work, DSM-III split anxiety neurosis into panic disorder and gen-eralized anxiety disorder, and split phobic neurosis into ago-raphobia, social phobia, and simple phobia (4). Whereas many had viewed panic as a severe form of anxiety, Klein gave emphasis to spontaneous panic attacks, with subse-quent anticipatory anxiety and phobic avoidance (2). The revision of DSM-III further emphasized the priority of panic disorder over agoraphobia, as what was classified as agora-phobia with panic attacks in DSM-III became panic disorder with agoraphobia in DSM-III-R (4).

Whereas DSM-III had a separate anxiety disorder chapter, ICD-9 had separate headings for anxiety states (which in-cluded panic attacks, panic disorder, and panic state) and phobic states (which included agoraphobia, animal phobia, anxiety hysteria, claustrophobia, and phobia not otherwise specified). Furthermore, whereas DSM-III-R gave increased priority to panic disorder, ICD-10 regarded phobic disorders as primary, with panic attacks as a marker of severity. Thus, Frances et al (4) concluded that “of all the sections in the classification, DSM-III-R and ICD-10 are probably most in-consistent in their treatment of the relationship and priority of panic and agoraphobia”.

ICD-10 AnD DsM-IV ApproACHes

Given the different pathways taken by the psychiatric no-sologies over time, and the different perspectives and empha-ses of the ICD and DSM systems, there continued to be im-portant differences between the approaches of ICD-10 and DSM-IV to agoraphobia and panic disorder. Indeed, in DSM-IV, the residual status of agoraphobia within the panic disorder construct became even more pronounced (7). At the level of the individual case, however, the differences between the two systems may not be as apparent: for an individual with both agoraphobia and panic disorder, ICD-10 would diagnose agoraphobia with panic disorder, while DSM-IV would diagnose panic disorder with agoraphobia.

089_093.indd 89 06/07/12 14:05


Section F40 of ICD-10 includes the diagnoses of F40.0 Agoraphobia, and F41.0 Other Anxiety Disorders, including F41.0 Panic Disorder (Episodic Paroxysmal Anxiety). In the Clinical Descriptions and Diagnostic Guidelines for ICD-10 Mental and Behavioural Disorders, the version of the clas-sification directed towards mental health clinicians, F40.0 Agoraphobia also includes specifiers of .00 (without panic disorder) and .01 (with panic disorder).

This specifier is not found in the main version of the ICD-10 and, because most countries do not use ICD-10 5th-digit specifiers in their national classification systems, it is possible that both diagnoses are applied independently in clinical set-tings. The Guidelines emphasize that, based on an interna-tional and cross-cultural perspective, there are insufficient data to reject the notion that the phobic disorder is best re-garded as the primary disorder. Thus, a panic attack occur-ring in an established phobic situation is regarded as an ex-pression of the severity of the phobia, which should be given diagnostic precedence, and panic disorder becomes a main diagnosis only in the absence of any of the phobias listed in F40.

The anxiety disorders chapter of DSM-IV includes de-scriptions of both agoraphobia and panic attacks, which are described as symptom sets rather than as diagnostic entities. Furthermore, as noted above, the DSM-IV perspective was strongly influenced by Klein’s work emphasizing that there is often progression from spontaneous panic attacks to antici-patory anxiety and to agoraphobic avoidance (6). Thus, DSM-IV provides the diagnoses of panic disorder without agoraphobia, and of panic disorder with agoraphobia. The diagnosis of agoraphobia is only given in the absence of a history of panic disorder (i.e., agoraphobia without panic disorder).

ICD-10 describes agoraphobia as an interrelated and overlapping cluster of phobias embracing fears of leaving home. Three diagnostic criteria are listed: a) the psychologi-cal or autonomic symptoms must be primarily manifesta-tions of anxiety and not secondary to other symptoms, such as delusions or obsessional thoughts; b) the anxiety must be restricted to (or occur mainly in) at least two of the following situations: crowds, public places, travelling away from home, and travelling alone; c) avoidance of the phobic situation must be, or have been, a prominent feature.

In ICD-10, the essential features of panic disorder are re-current attacks of severe anxiety (panic) that are not restrict-ed to any particular situation or set of circumstances, and are therefore unpredictable. As in other anxiety disorders, the dominant symptoms vary from person to person, but sudden onset of palpitations, chest pain, choking sensations, dizzi-ness, and feelings of unreality (depersonalization or dereal-ization) are common. There is also, almost invariably, a sec-ondary fear of dying, losing control, or going mad. Individu-al attacks usually last for minutes only, though they are some-times longer; their frequency and the course of the disorder are both rather variable. An individual in a panic attack often experiences a crescendo of fear and autonomic symptoms

which results in an exit, usually hurried, from wherever he or she may be. If this occurs in a specific situation, such as on a bus or in a crowd, the patient may subsequently avoid that situation. Similarly, frequent and unpredictable panic attacks produce fear of being alone or going into public places. A panic attack is often followed by a persistent fear of having another attack.

For a definite diagnosis, several severe attacks of auto-nomic anxiety should have occurred, within a period of about 1 month: a) in circumstances where there is no objec-tive danger; b) without being confined to known or predict-able situations; and c) with comparative freedom from anxi-ety symptoms between attacks (although anticipatory anxi-ety is common).

The DSM-IV goes on to provide a great level of detail about the diagnostic criteria for both agoraphobia and panic disorder.

In DSM-IV agoraphobia, there is anxiety about being in places or situations from which escape might be difficult (or embarrassing) or in which help may not be available in the event of having an unexpected or situationally predisposed panic attack or panic-like symptoms. Agoraphobic fears typi-cally involve characteristic clusters of situations that include being outside the home alone; being in a crowd or standing in a line; being on a bridge; and traveling in a bus, train, or automobile. In addition, the situations are avoided (e.g., travel is restricted) or else are endured with marked distress or with anxiety about having a panic attack or panic-like symptoms, or require the presence of a companion. Finally, the anxiety or phobic avoidance is not better accounted for by another mental disorder.

In DSM-IV panic disorder with or without agoraphobia, there are recurrent unexpected panic attacks, and at least one of the attacks has been followed by 1 month (or more) of one (or more) of the following: a) persistent concern about hav-ing additional attacks; b) worry about the implications of the attack or its consequences (e.g., losing control, having a heart attack, “going crazy”); c) a significant change in behavior related to the attacks. Furthermore, the panic attacks are not due to the direct physiological effects of a substance (e.g., a drug of abuse, a medication) or a general medical condition (e.g., hyperthyroidism), and they are not better accounted for by another mental disorder.

probleMs ArIsIng froM ICD-10 AnD DsM-IV

ICD-10 has been widely used in clinical settings through-out the world, but relatively little systematic data are avail-able about its clinical utility or problems in its implementa-tion. There has been considerably greater investigation of the DSM-IV diagnostic criteria given their more frequent use in research (7,8).

Wittchen et al (7) have reviewed the position of agorapho-bia in DSM-IV, and noted a number of problems. The authors argue that any conceptualization of agoraphobia as a subor-

089_093.indd 90 06/07/12 14:05


dinate, residual form of panic disorder is untenable; there is now significant evidence that agoraphobia is a diagnosis in-dependent of panic disorder. Furthermore, they argue that the DSM-IV diagnostic criteria for agoraphobia are problem-atic in a number of ways. For example, the criteria lack spec-ification for what constitutes agoraphobic situations, and do not provide cues beyond the occurrence or fear of panic-like symptoms, so that it is unclear how to diagnose patients without such symptoms.

Wittchen et al’s review draws on a broad range of data from the last two decades, including epidemiology, psycho-pathology, neurobiology, vulnerability and risk factors, clini-cal course and outcome, and correlates and consequences of agoraphobia (7). Their conclusions about the value of in-cluding agoraphobia as an independent diagnosis in the no-sology rest on a number of findings from the literature. First, epidemiological data from community studies consistently find sizeable rates (about 50%) of agoraphobia without pan-ic attacks, which do not meet DSM-IV criteria for agorapho-bia without panic disorder. Second, epidemiological data consistently, although clinical data inconsistently, indicate that agoraphobia does not always occur secondary to panic attacks or panic-like symptoms. Third, data consistently dem-onstrate that agoraphobia without panic attacks is associated with significant impairment and disability, as well as with a persistent course and low rates of spontaneous remission. Fourth, there are often differences in patterns of incidence, gender ratios, temporal progression, and response to treat-ment across panic attacks, panic disorder, and agoraphobia. Fifth, agoraphobia often emerges as a major dimension in psychometric and taxometric studies, and agoraphobia avoid-ance is an independent contributor to severity, course, and outcome in panic disorder. Sixth, specific and explicit criteria for agoraphobia, similar or identical to those used in other phobias, appear to have high interrater and test-retest reli-ability.

Craske et al (8) have reviewed the DSM-IV diagnostic cri-teria for panic attacks and panic disorder. They considered evidence regarding the degree to which the list of panic at-tack symptoms, the cutoff of four or more symptoms, and the time to peak intensity should be revised. They also raised a number of other issues, for example, the question of whether these criteria reflected knowledge about the developmental course of panic disorder, and about the cross-cultural presen-tation of this disorder. The current criteria have important strengths, and they therefore suggested a relatively limited number of changes (see next section).

proposAls for DsM-5

Perhaps the most important proposal for DSM-5 is to have agoraphobia and panic disorder as independent diagnostic entities (see www.dsm5.org). A number of changes have also been proposed for the DSM-5 diagnostic criteria for agora-phobia (7). A particularly important element of the proposal

is to structure the diagnostic criteria in a way that is analogous to that used for social anxiety disorder and specific phobia, using a similar sequence and wording. Thus, it is proposed that criterion A refers to marked fear and anxiety about two or more clusters of agoraphobia situations; criterion B refers to the fear that escape might be difficult or help might be unavailable in the event of the development of panic-like symptoms or other incapacitating symptoms; criterion C re-fers to the agoraphobic situations almost always provoking fear or anxiety; criterion D refers to the agoraphobic situa-tions being actively avoided, requiring the presence of a com-panion, or being endured with intense fear or anxiety; crite-rion E refers to the fear or anxiety being out of proportion to the actual danger posed by the agoraphobic situations; crite-rion F refers to a duration of at least 6 months; and criteria G, H, and I are standard clinical significance and diagnostic hierarchy criteria.

A number of changes have also been proposed for the DSM-5 diagnostic criteria for panic disorder (8). Again, it is suggested that there be some delineation of anxiety symp-toms from cognitive aspects, with criterion A being broken into two separate criteria, one referring to recurrent unex-pected panic attacks, and the other referring to concern/worry/behavioral change as a result of the attacks. Addition-al clarifications include deletion of the criterion referring to agoraphobia.

Finally, a number of changes have also been proposed for the DSM-5 diagnostic criteria for panic attacks (8). These include rephrasing “hot flushes” to be “heat sensations”, and reordering the list of symptoms. In addition, slight changes have been proposed in the wording of the diagnostic criteria for panic attacks to help differentiate between panic and sur-rounding anxiety, and some simplification and clarification of the distinction between expected (cued) versus unexpect-ed (uncued) panic attacks.

optIons for ICD-11

The International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders has recommend-ed that development of ICD-11 should be particularly atten-tive to issues of clinical utility, use in a broad range of coun-tries and cultures, especially low- and middle-income coun-tries, and use in primary care and other non-specialized set-tings where the vast majority of people with mental disorders receive care, if they receive it at all (9-11).

A first issue to consider for ICD-11 is the relationship of agoraphobia to panic disorder, and the key question of how many major anxiety diagnoses are required in the diagnostic system. To a considerable extent, the DSM-5 proposal to have agoraphobia and panic disorder as independent entities moves the DSM nosology closer to the ICD-10 approach. An important part of the rationale for this proposal is the argu-ment that it will improve clinical utility by providing a sim-pler diagnostic schema (very much along the lines of the

089_093.indd 91 06/07/12 14:05


ICD-10 classification without the 5th digit specifier).There is indeed growing evidence that, while agoraphobia

and panic disorder are related, they are also distinct. Panic attacks are not exclusive to agoraphobia; they occur in most anxiety disorders and many non-anxiety disorders. Indeed, panic attacks appear to predict the onset of various forms of psychopathology and, in the context of co-occurring psychi-atric disorders (e.g., anxiety, mood, eating, personality, psy-chotic, and substance use disorders), have been associated with increased symptom severity, higher rates of comorbidity and suicidality, and poorer outcomes. Conversely, panic at-tacks are not the only reason for the fear and avoidance of agoraphobia situations. Instead, a substantial number of cases of agoraphobia may not manifest panic attacks or pan-ic-like features (7,8).

Options for ICD-11 would be: a) to employ the approach in the ICD-10 Guidelines, retaining the specifier for agora-phobia (i.e., with or without panic disorder), or b) to employ the approach in the ICD-10 main volume and that proposed for DSM-5 (i.e., having agoraphobia and panic disorder as independent disorders, without specifiers). The latter ap-proach may be a more parsimonious one, with greater clini-cal utility. Certainly, ICD-11 should remove reference to the (now outdated) concept that phobia should be given diag-nostic precedence over panic disorder. One potential nega-tive consequence of this approach is that patients with ago-raphobia and panic disorder would be diagnosed with two disorders (arguably, a comorbidity that is somewhat artifac-tual) (12). However, in other areas of medicine, clinical re-alities are also often complex, with significant comorbidity.

A second issue is the overall structure of the diagnostic criteria for agoraphobia and for panic disorder. To a consid-erable extent, the DSM-5 proposal to use a similar structure for the diagnostic criteria for agoraphobia, social anxiety dis-order, and specific phobia is again consistent with the ap-proach in ICD-10, which, unlike DSM-IV, has a good deal of parsimonious analogy across these conditions.

A third issue is that of altering diagnostic criteria for ago-raphobia and panic disorder to be as clinically useful as pos-sible in the settings where ICD-11 will be used. Once again, it can be noted that the DSM-5 proposals have been consis-tent with the intention of ICD-10 and ICD-11, by paying con-siderable attention to cross-cultural issues. For example, the proposed DSM-5 agoraphobia criteria specifically note that the phrase “out of proportion” refers to the socio-cultural context. Thus, agoraphobia fears can be differentiated from rational fear of situations that appear typical of agoraphobia (e.g., leaving the house) but which are rightly perceived as dangerous (e.g., due to living in an area with a high crime rate) (7). Similarly, rephrasing of panic attack criteria has been pro-posed in order to increase cross-cultural validity (8).

On this score, the options for ICD-11 are straightforward. Where changes to criteria increase cross-cultural validity, these should be accepted. Nevertheless, additional scrutiny may be needed to ensure high standards of cross-cultural applicability. For example, the DSM-5 proposed groupings of

situations for agoraphobia are “public transportation (for ex-ample, traveling in automobiles, buses, trains, ships, planes); open spaces (for example, parking lots, market place, being on a bridge); being in shops, theaters or cinemas; standing in a line or being in a crowd, or being outside of the home alone” (see www.dsm5.org). Further attention may need to be paid as to whether such situations are representative of those faced by patients with agoraphobia in low- and middle-income countries.

ConClusIons

In summary, the DSM-5 proposals are bringing the DSM and ICD nosologies into closer alignment in a number of key ways. Furthermore, the DSM-5 proposals are based on a good deal of data collected since the publication of DSM-IV and ICD-10. In weighing options for ICD-11, it is relevant to consider also the potential benefits of adopting a similar ap-proach in DSM-5 and ICD-11. At the same time, ICD-11 needs to focus on clinical utility, global applicability, and use outside specialized mental health settings. Indeed, addition-al work is needed to assess the clinical utility of the proposed criteria for both DSM-5 and ICD-11.

note

D.J. Stein is a member of the WHO ICD Revision Working Group on the Classification of Mood and Anxiety Disorders, reporting to the International Advisory Group for the Revi-sion of ICD-10 Mental and Behavioural Disorders. The views expressed in this article are those of the author and, except as specifically noted, do not represent the official policies or positions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.


references

1. Berrios GE, Link C. Anxiety disorders. In: Berrios GE, Porter R (eds). A history of clinical psychiatry. New York: New York Univer-sity Press, 1995:545-62.

2. Nutt D, Ballenger J. Anxiety disorders: panic disorder and social anxiety disorder. Oxford: Blackwell, 2005.

3. Kuch K, Swinson RP. Agoraphobia: what Westphal really said. Can J Psychiatry 1992;37:133-6.

4. Frances A, Miele GM, Widiger TA et al. The classification of panic disorders: from Freud to DSM-IV. J Psychiatr Res 1993;27(Suppl. 1): 3-10.

5. Frances A. Classification of anxiety disorders: from Freud to DSM-IV. J Psychiatr Res 1990;24(Suppl. 1):9-10.

6. Klein DF. Delineation of two drug-responsive anxiety syndromes. Psychopharmacologia 1964;5:397-408.

7. Wittchen HU, Gloster AT, Beesdo-Baum K et al. Agoraphobia: a

089_093.indd 92 06/07/12 14:05


review of the diagnostic classificatory position and criteria. Depress Anxiety 2010;27:113-33.

8. Craske MG, Kircanski K, Epstein A et al. Panic disorder: a review of DSM-IV panic disorder and proposals for DSM-V. Depress Anx-iety 2010;27:93-112.

9. Reed GM. Toward ICD-11: improving the clinical utility of WHO’s International Classification of Mental Disorders. Prof Psychol Res Pr 2010;41:457-64.

10. Reed GM, Mendonça Correia J, Esparza P et al. The WPA-WHO

Global Survey of Psychiatrists’ Attitudes Towards Mental Disorders Classification. World Psychiatry 2011;10:118-31.


12. Maj M. Psychiatric comorbidity: an artefact of current diagnostic systems? Br J Psychiatry 2005;186:182-4.

089_093.indd 93 06/07/12 14:05


Specific and social phobias in ICD-11

The purpose of this article is to discuss the current status of research with respect to specific phobias and social phobia as defined in the ICD-10 and DSM-IV, and to provide op-tions for their classification in the ICD-11.

Both ICD-10 and DSM-IV identify three main classes of phobias: specific phobias, social phobia, and agoraphobia. The latter two are often referred to as complex phobias. In specific phobias, the fear is limited to one or more well-de-fined classes of objects or situations: e.g., certain animals, aspects of the natural environment (heights, storms), specific situations (lifts, airplanes) or blood/injuries. Social phobia is characterized by fear of one or more social situations, while agoraphobia is defined by fear of leaving home or of places or situations from which escape may be difficult, including shops, crowds, and public places, or fear of travelling (alone) in trains, buses, or planes. The emphasis in this paper will be on specific phobias and social phobia, since agoraphobia is discussed in another article in this supplement (1).

ClInICal featureS anD prevalenCe

Specific phobias

Specific phobias are focused upon, and restricted to, fear of specific objects and situations, such as animals, heights, storms, darkness, and closed spaces.

The lifetime prevalence of specific phobias in the com-munity is reported to be around 10% (2). Prevalence varies considerably across the lifespan (2,3), with the highest prev-alence among children, adolescents and young adults (12-month: 6-8%) and the lowest among people older than 65 year of age (12-month: 2%). At least two-thirds of spe-cific phobias have their first onset in childhood or adoles-cence. In children, specific fears are common, developmen-tally appropriate, and typically disappear within a few months. Specific phobias can be differentiated from normal developmental fears in that the phobic reaction is excessive and persists over time. Specific phobias that continue from

Paul M.G. EMMElkaMP

Department of Clinical Psychology, University of Amsterdam, Weesperplein 4, 1018 XA Amsterdam, The Netherlands

This paper aims to discuss problems with the current ICD-10 and DSM-IV approaches to the diagnosis and classification of specific phobias and social phobia, and to make suggestions for changes in the ICD-11. There is a clear need to include impairment and duration components in the description of specific and social phobias. The differentiation of specific phobias and social phobia into subtypes does not seem to have sufficient implications for treatment to justify its inclusion in ICD-11. The content of specific phobias and social phobia appear to vary across cultures, which suggests that phobias should be defined in relation to the sociocultural background of the individual. It is further proposed to broaden the criteria for social phobia to include the fear of offending others.

Key words: ICD, classification, specific phobias, social phobia, cultural aspects


childhood into adolescence rarely recover spontaneously. Research has consistently found that certain specific pho-

bias (animals, lightning, enclosed spaces, and darkness) are more common in women, whereas smaller gender differ-ences have been observed for phobias related to heights, fly-ing, injections, and dentists (2). In general, gender differ-ences are less pronounced before the age of 10 and increase with age.

Social phobia

The central characteristic of social phobia is a fear of scru-tiny by other people in small groups, leading to avoidance of social situations. The fear may be discrete (e.g., restricted to eating in public, public speaking, or encounters with the op-posite sex) or diffuse, involving almost all social situations. Social phobia may be experienced in terms of physical symp-toms such as blushing, sweating or trembling, or as fears of negative evaluation.

The lifetime prevalence of social phobia in Western cul-tures is reported to be between 7.2 and 12.1% (4-6). Gener-ally, the prevalence rates are higher in women, but this gen-der difference is less pronounced than in specific phobias. In community studies using DSM-IV criteria, the rate of social phobia has been found to be much lower in East Asia (7).

ICD-10 anD DSM-Iv approaCheS

Specific phobias

The DSM-IV characterizes specific phobia as a marked and persistent, excessive or irrational fear that occurs in re-sponse to real or anticipated confrontation with a circum-scribed stimulus (object or situation). The confrontation al-most invariably provokes an immediate anxiety reaction, which might reach the severity threshold of situationally bound panic attack. The person recognizes that the anxiety

094_099.indd 94 06/07/12 14:07


is excessive or unreasonable. The situations or objects are avoided or endured with intense anxiety. The fear, the associ-ated anticipatory anxiety or the avoidance behavior must interfere significantly with the person’s normal life or be as-sociated with clinically significant suffering. The fear, panic or avoidance should not be better explained by other mental disorders. In individuals under 18 years of age, symptoms must have persisted for at least 6 months.

The ICD-10 description is very similar, but does not re-quire that the phobia be recognized as unreasonable. The following features should be present for a definite diagnosis of specific phobia: a) the psychological or autonomic symp-toms must be primary manifestations of anxiety, and not sec-ondary to other symptoms such as a delusion or obsessive thought; b) the anxiety must be restricted to the presence of the particular phobic object or situation; c) the phobic situa-tion is avoided whenever possible.

The DSM-IV distinguishes five types of specific phobias: animal type (e.g., spiders, dogs, cats, snakes, and birds), natu-ral environment type (e.g., storms, heights or water), blood-injection-injury type (e.g., dental phobia), situational type (e.g., tunnels, bridges, elevators, flying, driving, or enclosed places), and a residual category (choking, vomiting, loud sounds, costumed characters). The ICD-10 lists the following specific phobias as examples: acrophobia, animal phobia, claustrophobia, examination phobia and simple phobia. These are not coded individually, and it is clear from the de-scription that specific phobias are not intended to be limited to these types. In both DSM-IV and ICD-10, fears of specific diseases such as cancer, heart disease, venereal disease and dysmorphophobia are not classified as specific phobias.

The main differences between ICD-10 and DSM-IV are the following: a) in ICD-10, examination fear (test anxiety) is included; b) ICD-10 does not require that the phobia be rec-ognized as unreasonable; c) the ICD-10 does not identify formal subtypes; d) illness phobia is not identified as a spe-cific phobia in ICD-10, while in DSM-IV it is listed as a spe-cific phobia, other type; e) in ICD-10, if criteria for depressive disorder are met before the phobic symptoms first appeared, depression should be given diagnostic precedence.

Social phobia

According to DSM-IV, social phobia or social anxiety dis-order should be diagnosed in the case of a marked and per-sistent fear of one or more social or performance situations in which the person is exposed to unfamiliar people or to pos-sible scrutiny by others. Exposure to the feared social situa-tion almost invariably provokes anxiety that might reach the severity threshold of situational bound panic attack. The per-son recognizes that the anxiety is excessive or unreasonable. The feared social or performance situations are avoided or else are endured with intense anxiety or distress. The avoid-ance, anxious anticipation, or distress in the feared social or performance situation(s) interferes significantly with the per-

son’s normal routine, occupational (academic) functioning, or social activities or relationships, or there is marked distress about having the phobia. In individuals under age 18 years, the duration is at least 6 months. The fear or avoidance is not due to the direct physiological effects of a substance (e.g., a drug of abuse, a medication) or a general medical condition, and is not better accounted for by another mental disorder.

In DSM-IV, two subtypes are distinguished: nongeneral-ized and generalized social anxiety disorder. The nongeneral-ized subtype refers to individuals who fear a single perfor-mance situation or a limited number of social situations. In the generalized subtype, individuals fear “most” social situa-tions, usually including both performance and interactional situations (e.g., initiating or maintaining conversations, par-ticipating in small groups, dating, speaking to authority fig-ures, attending parties).

The ICD-10 description is very similar, but does not re-quire that the phobia be recognized as unreasonable. For a definite diagnosis of social phobia, it is required that: a) the psychological, behavioural, or autonomic symptoms must be primarily manifestations of anxiety, and not secondary to other symptoms such as a delusion or obsessive thought; b) the anxiety must be restricted to or predominate in particular social situations; c) the phobic situation is avoided whenever possible.

The main differences between ICD-10 and DSM-IV are the following: a) the ICD-10 does not require that the phobia be recognized as unreasonable; b) the ICD-10 does not iden-tify subtypes; c) in ICD-10, if criteria for depressive disorder are met before the phobic symptoms first appeared, the for-mer should be given diagnostic precedence; d) in ICD-10, if the distinction between social phobia and agoraphobia is very difficult, precedence should be given to agoraphobia.

probleMS arISIng froM the applICatIon of ICD-10 anD DSM-Iv

Specific phobias

The diagnosis of specific phobias in DSM-IV requires that the person recognize that his or her fear of the phobic situa-tion is excessive or unreasonable. The DSM-5 Anxiety Dis-orders Work Group has proposed replacing this criterion with a clinician’s judgment that the fear is out of proportion to actual danger posed by the situation (www.dsm5.org). It has been argued that clinical judgment is preferable to self-judgment, because a number of patients lack insight into the excessiveness or unreasonableness of their fears. For exam-ple, the relatively low incidence of phobias in the elderly (3) may be due to misattribution of fears to age-related con-straints (8).

There are several indications from cross-sectional and lon-gitudinal studies that the application of the same criteria to children and adults is problematic. Childhood phobias are different from those of late adolescents and adults with re-

094_099.indd 95 06/07/12 14:07


spect to reliability and stability of diagnosis, predictive value, severity and prognosis (2).

There is no firm evidence for differentiating the various subtypes of specific phobias as done in DSM-IV. Blood-injec-tion-injury and situational phobias are clearly different from the others, but there is not enough systematic evidence for justifying a more coherent and clinically useful subtyping (2).

Illness phobia appears to be highly prevalent in the gen-eral population and is associated with significant distress and impairment. In DSM-IV, it is considered to be a specific pho-bia, other type, while in ICD-10 it is classified under hypo-chondriacal disorder. This issue is currently controversial (9).

A critical issue related to specific phobias is that of impair-ment. Zimmerman et al (10) found that adding a criterion of clinically significant impairment reduced the number of spe-cific phobia diagnoses substantially and more so than in ma-jor depression, generalized anxiety disorder and post-trau-matic stress disorder (PTSD). In the European Study of the Epidemiology of Mental Disorders (ESEMeD), specific pho-bias had lower independent impact on work loss days than panic disorder, agoraphobia, PTSD, major depressive epi-sode, and dysthymia (11). In a study using the Work and Social Adjustment Scale, patients with specific phobias were clearly less impaired than those with agoraphobia and social phobia, especially in the areas of work and relationships (12). Cramer et al (13) found that, in contrast to other anxiety disorders, specific phobias had only a small effect on quality of life and were not associated with an impairment of epi-sodic memory and executive functioning.

Social phobia

According to DSM-IV, an individual with a medical con-dition should be diagnosed with social anxiety disorder only if the social anxiety is unrelated to that condition. However, in patients with stuttering, which is considered by some as a medical condition (14), the severity of social anxiety was found to be similar to those with DSM-IV social anxiety dis-order without a medical condition. A high prevalence of so-cial anxiety disorder has also been reported in obese indi-viduals in clinical samples (15) and in patients with Parkin-son’s disease (16). Since social anxiety in the context of stut-tering, obesity and Parkinson’s disease may present with symptoms similar in nature and severity to the social anxiety disorder as defined in DSM-IV, symptoms may need to be managed the same way as in cases of “primary” social anxiety disorder. Excluding such individuals from a formal social anxiety disorder diagnosis could hinder access to potentially beneficial psychological treatment or pharmacotherapy.

The definition of the generalized subtype as anxiety expe-rienced in “most social situations” is ambiguous. Some have interpreted this in terms of quantitative difference (i.e., gen-eralized type being defined in terms of reporting anxiety in a higher number of social situations), whereas others have in-terpreted this as a qualitative difference (17). The operation-

al definitions for the generalized and nongeneralized sub-types vary across studies, which makes comparisons of find-ings rather difficult. For example, the definition of general-ized social anxiety disorder in community studies has varied from six to seven (e.g., 17,18) to thirteen to fourteen (e.g., 19) feared social situations.

A number of recent community studies show a linear dose-response relation between number of fears and psychi-atric comorbidity, role impairment, and service utilization (20), which does not support the notion of a dichotomy of generalized versus nongeneralized subtypes, but rather sug-gests the need of a dimensional measure of severity.

One way of differentiating social anxiety disorder from shyness is the impairment associated with the former condi-tion. Persons with social anxiety disorder show impairment in work functioning, tend to have fewer social supports than non-socially anxious individuals, and are more likely to at-tempt suicide (21-23). Kessler et al (24) did not find increased impairment among individuals with specific social anxiety disorder as compared to people without social anxiety, but these researchers defined the nongeneralized subtype of so-cial anxiety based solely on a fear of public speaking, which makes their results difficult to interpret.

Social fears may be dependent on particular cultural con-texts, which is reflected in a variable prevalence of social anxiety disorder across cultures. The highest rate has been found in Russia, and the lowest rates in East Asian cultures (i.e., China, Japan, Korea and Taiwan) and Africa (Nigeria and South Africa). The rates of social anxiety disorder in North America and South America (i.e., Chile, Brazil) are in between (7,25).

Taijin Kyofusho can be considered as a possible culture-specific expression of social anxiety in Japan (26). It is char-acterized by a strong belief and fear that others will be of-fended by one’s own inadequacy (e.g., emitting bodily odors, odd gaze or facial expression, blushing, loud bowel sounds). Many cases of Taijin Kyofusho would not be diagnosed as social phobia according to the criteria of DSM-IV or ICD-10.

In DSM-IV, social phobia and avoidant personality disor-der are not mutually exclusive: ‘‘avoidant personality disor-der may be a more severe variant of social phobia, general-ized, that is not qualitatively distinct’’. In clinical practice, there is a considerable overlap between DSM-IV social anx-iety disorder and avoidant personality disorder.

propoSalS for DSM-5

Specific phobias

The DSM-5 Anxiety, OC Spectrum, Posttraumatic, and Dissociative Disorder Work Group has considered changes in the DSM classification with respect to the current subtyp-ing scheme for specific phobias. The DSM-5 group has also considered whether test anxiety should be considered as a type of specific phobia, and the boundary between specific

094_099.indd 96 06/07/12 14:07


phobia and agoraphobia (8). Currently, few changes are pro-posed, apart from slightly rephrased formulations to improve clarity and consistency with the diagnostic criteria of the other anxiety disorders.

The most important proposed changes are: a) the duration of specific phobia must be at least 6 months (in DSM-IV this criterion was limited to individuals under age 18 years); b) an explanation is provided that, in the criterion “fear or anxiety is out of proportion to the actual danger posed by the spe-cific object or situation”, out of proportion refers to the so-ciocultural context; c) the impairment criterion (“the fear, anxiety or avoidance cause clinically significant distress or impairment in social, occupational, or other important areas of functioning”) may be deleted, but no definitive recom-mendation has been made in this regard; d) reference to fears of contracting an illness are removed, due to the high overlap of these symptoms with obsessive-compulsive disorder and hypochondriasis.

The following types of specific phobias are proposed for inclusion in the DSM-5: animal type (e.g., spiders, insects, dogs); natural environment type (e.g., heights, storms, wa-ter); blood-injection-injury type (e.g., needles, invasive med-ical procedures); situational type (e.g., flying, driving, bridges, tunnels, enclosed places); and other type (e.g., situations that may lead to choking or vomiting; in children, loud sounds or costumed characters). Test anxiety is not proposed for inclu-sion (www.dsm5.org).

Social phobia

While in DSM-IV the main category was social phobia, followed by social anxiety disorder in parentheses, in DSM-5 the main category will be social anxiety disorder, followed by social phobia in parentheses (www.dsm5.org).

The main issues which have been discussed in preparation of the proposal for DSM-5 are the generalized specifier, alter-native subtypes, the overlap with avoidant personality disor-der, modifying the medical exclusion criterion, and the rela-tionship with test anxiety and with selective mutism in child-hood.

The content of the fear will be defined as follows: “The individual fears that he or she will act in a way, or show anxiety symptoms, that will be negatively evaluated (i.e., be humiliating, embarrassing, lead to rejection, or offend oth-ers)”. Offending others is added to increase cultural sensitiv-ity, which might be relevant for Asian cultures (e.g., Taijin Kyofusho).

The duration of social phobia must be at least 6 months. In DSM-IV this criterion was limited to individuals under age 18 years. As in specific phobia, the explanation will be provided that, in the criterion “fear or anxiety is out of pro-portion to the actual danger posed by the social situation”, out of proportion refers to the sociocultural context.

Although evidence does not support the current DSM-IV designation of a generalized subtype, but rather supports a

view of social anxiety disorder as existing on a continuum, from lesser to greater severity, as a function of the number of feared and avoided social situations (20), this subtype is re-tained in the proposal for DSM-5. Two additional subtypes have been added. The proposed subtypes are: a) performance only (if the fear is restricted to speaking or performing in public); b) generalized (if the fear is of most social situations and is not restricted to performance situations); c) selective mutism (consistent failure to speak in specific social situa-tions in which there is an expectation for speaking, e.g., at school, despite speaking in other situations).

Another change in the current proposal for DSM-5 is a modification of the medical exclusion criterion, such that social anxiety disorder could be diagnosed in individuals with medical conditions (e.g., stuttering, Parkinson’s disease, obesity, disfigurement from burns or injury), provided that the social anxiety is either unrelated to the medical condition or is excessive.

Test anxiety is no longer considered as a specific form of social anxiety disorder. The impairment criterion (“the fear, anxiety or avoidance cause clinically significant distress or impairment in social, occupational, or other important areas of functioning”) may be deleted.

optIonS for ICD-11

Specific phobias

Specific phobias need a more extensive description in ICD-11. They should be limited to intense fear regarding one or more objects or situations, out of proportion to the actual danger, which are actively avoided or endured with intense fear or anxiety. It is important to differentiate transient epi-sodes from long-term fears, by including a duration require-ment (e.g., more than 6 months).

Results of studies of the differentiation of specific phobia into subtypes do not justify the inclusion of formal subtypes in ICD-11 (2,8). Apart from blood-injection-injury phobia, subtypes of specific phobia are not differentially responsive to specific treatments; rather, the main forms of treatment are effective across specific phobias. In order to maximize clini-cal utility, it is important to be cautious about including bur-densome additional specifications in the classification when these have no treatment implication. Nevertheless, greater harmonization of DSM-5 and ICD-11 could be achieved by including the same categories of specific phobias as descrip-tors in ICD-11, but not as coded subtypes. Given the world-wide use of ICD, it may be important to add a description of typically culture-linked specific phobias.

Examination fear is a prevalent and often disabling condi-tion. Although it is not included in the proposal for DSM-5, it seems wise to retain it as a specific phobia in ICD-11, since this may facilitate its identification and treatment (8).

There may be reasons to consider illness phobia as a spe-cific phobia, provided that the content of the fear is related

094_099.indd 97 06/07/12 14:07


to one specific disease and it does not occur in the context of obsessive-compulsive disorder or bodily distress disorder. The fear should be intense and last more than 6 months. The potential overlap with a category related to health or illness anxiety must also be considered.

An impairment requirement is important. In children and adolescents, some fears may be part of normal development, and a number of adults may have fears which have limited impact on functioning. On the other hand, the effectiveness of cognitive behaviour therapy for disabling specific phobias is now well established, but only a small fraction of individu-als with these phobias ever receive this treatment (2). When people with specific phobias do receive treatment, it is usu-ally many years if not decades after onset.

Social phobia

Social phobia needs a more extensive description in ICD-11. It should be limited to intense fear of one or more social situations, which are actively avoided or endured with in-tense fear or anxiety. It is important to differentiate transient episodes from long-term fears, by including a duration re-quirement (e.g., more than 6 months).

In the proposal for DSM-5, three specifiers are listed (per-formance only, generalized, and selective mutism). However, research to support this distinction is weak and there is no clear operational definition for the categories. The ICD clas-sification system should have global applicability and not be overly burdensome to apply.

The phenomenology of social phobia varies across cul-tures. Therefore, it would be useful to include in the ICD definition the requirement that symptoms be evaluated in the context of the person’s sociocultural background. It might be also wise to broaden the description to include anxiety about offending others, as now proposed for DSM-5. This could be accomplished by adding the following requirement to the ICD-11 diagnostic guidelines: “The individual fears that he or she will act in a way that will be humiliating or embarrass-ing, or that may result in another person feeling offended”.

Further, as proposed for specific phobias, it seems impor-tant to add an impairment criterion, in order to differentiate true clinical disorders from less debilitating forms of social anxiety (e.g., shyness).

note

P.M.G. Emmelkamp is a member of the WHO ICD Revi-sion Working Group on the Classification of Mood and Anxiety Disorders, reporting to the International Advisory Group for the Revision of ICD-10 Mental and Behavioural Disorders. The views expressed in this article are those of the author and, except as specifically noted, do not represent the official policies or positions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders,

or the WHO.The DSM-5 website was accessed on November 21, 2011.

DSM-5 proposals have not been finalized and are subject to change.

references

1. Stein D. Agoraphobia and panic disorder: options for ICD-11. World Psychiatry 2012;11(Suppl. 1):89-93.

2. Emmelkamp P, Wittchen HU. Specific phobias. In: Andrews G, Charney DS, Sirovatka PJ et al (eds). Stress-induced and fear cir-cuitry disorders: refining the research agenda for DSM-V. Arlington: American Psychiatric Association, 2009.

3. Grenier S, Schuurmans J, Goldfarb M et al. The epidemiology of specific phobia and subthreshold fear subtypes in a community-based sample of older adults. Depress Anxiety 2011;28:456-67.

4. Kessler RC, Berglund P, Demler O et al. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Co-morbidity Survey Replication. Arch Gen Psychiatry 2005;62:593-602.

5. Ruscio AM, Brown TA, Chiu WT et al. Social fears and social phobia in the USA: results from the National Comorbidity Survey Replica-tion. Psychol Med 2008;38:15-28.

6. McEvoy PM, Grove R, Slade T. Epidemiology of anxiety disorders in the Australian general population: findings of the 2007 Australian National Survey of Mental Health and Wellbeing. Aust N Z J Psy-chiatry 2011;45:957-67.

7. Hofmann SG, Asnaani A, Hinton D. Cultural aspects in social anx-iety and social anxiety disorder. Depress Anxiety 2010;27:1117-27.

8. LeBeau RT, Glenn D, Liao B et al. Specific phobia: a review of DSM-IV specific phobia and preliminary recommendations for DSM-V. Depress Anxiety 2010;27:148-67.

9. Fava GA, Freyberger HJ, Bech P et al. Diagnostic criteria for use in psychosomatic research. Psychother Psychosom 1995;63:1-8.

10. Zimmerman M, Chelminski I, Young D. On the threshold of disor-der: a study of the impact of the DSM-IV clinical significance crite-rion on diagnosing depressive and anxiety disorders in clinical prac-tice. J Clin Psychiatry 2004;65:1400-5.

11. Alonso J, Angermeyer M, Bernert S et al. Disability and quality of life impact of mental disorders in Europe: results from the European Study of the Epidemiology of Mental Disorders (ESEMeD) project. Acta Psychiatr Scand 2004;109:38-46.

12. Mataix-Cols D, Cowley AJ, Hankins M et al. Reliability and validity of the Work and Social Adjustment Scale in phobic disorders. Com-pr Psychiatry 2005;46:223-8.

13. Cramer V, Torgersen S, Kringlen E. Quality of life and anxiety disor-ders: a population study. J Nerv Ment Dis 2005;193:196-202.

14. Schneier FR, Wexler KB, Liebowitz MR. Social phobia and stutter-ing. Am J Psychiatry 1997;154:131-4.

15. Dalrymple K L, Galione J, Hrabosky J et al. Diagnosing social anxi-ety disorder in the presence of obesity: implications for a proposed change in DSM-5. Depress Anxiety 2011;28:377-82.

16. Bolluk B, Özel-Kizil ET, Akbostanci MC et al. Social anxiety in patients with Parkinson’s disease. J Neuropsychiat Clin Neurosci 2010;22:390-4.

17. Chartrand H, Cox B, El-Gabalawy R et al. Social anxiety disorder subtypes and their mental health correlates in a nationally represen-tative Canadian sample. Can J Behav Sci 2011;43:89-98.

18. Magee WJ, Eaton WW, Wittchen HU et al. Agoraphobia, simple phobia, and social phobia in the National Comorbidity Survey. Arch Gen Psychiatry 1996;5:159-68.

19. El-Gabalawy R, Cox B, Clara I et al. Assessing the validity of social anxiety disorder subtypes using a nationally representative sample. J Anxiety Disord 2010;24:244-9.

20. Bogels SM, Alden L, Beidel DC et al. Social anxiety disorder: ques-

094_099.indd 98 06/07/12 14:07


tions and answers for the DSM-V. Depress Anxiety 2010;27:168-89.21. Vriends N, Becker ES, Meyer A et al. Subtypes of social phobia: are

they of any use? J Anxiety Disord 2007;2:59-75.22. Stein MB, Torgrud LJ, Walker JR. Social phobia symptoms, sub-

types, and severity: findings from a community survey. Arch Gen Psychiatry 2000;57:1046-52.

23. Swinson RP. Social anxiety disorder. Can J Psychiatry 2005;50:305-7.

24. Kessler RC, Stein MB, Berglund P. Social phobia subtypes in the

National Comorbidity Survey. Am J Psychiatry 1998;155:613-9.25. Lewis-Fernandez R, Hinton DE, Laria AJ. Culture and the anxiety

disorders: recommendations for DSM-V. Depress Anxiety 2010; 27:212-29.

26. Kinoshita Y, Chen J, Rapee R et al. Cross-cultural study of convic-tion subtype Taijin Kyofu: proposal and reliability of Nagoya-Osaka diagnostic criteria for social anxiety disorder. J Nerv Ment Dis 2008; 196:307-13.

094_099.indd 99 06/07/12 14:07


Hypochondriasis in ICD-11

This paper addresses the question of how hypochondria-sis should be characterized in the ICD-11. The historical background to the nosology of this disorder is reviewed. The approaches of the ICD-10 and DSM-IV to the issue, as well as the problems arising from their application, are summa-rized. Proposals for the DSM-5 and options for the ICD-11 are presented.


Hypochondriasis has traditionally been characterized as excessive preoccupation with having a serious illness. The term “hypochondria” was used by Hippocrates, and derives from the Greek, literally referring to “below the cartilage”, or the anatomical area beneath the ribs, where particular organs were thought to produce the various “humours” governing mental states (1). It was, however, only sometime in the late 19th century that the term “hypochondriasis” began to refer specifically to preoccupation with having an illness, and it was in this context that Freud developed his notion of “hy-pochondriacal neurosis”, a term subsequently adopted by DSM-II (2).

Despite criticism, the term “hypochondriasis” has per-sisted in subsequent revisions of both ICD and DSM (3). Apart from the issue of terminology, a number of other de-bates surrounding this clinical entity have continued for some time. First, questions about whether hypochondriasis is best conceptualized as a primary or secondary entity, and relatedly, whether hypochondriasis comprises a distinct tax-on or whether hypochondriacal symptoms are continuous with normality (4-8). Second, the question of the extent to which hypochondriasis and somatization are related condi-tions (9,10). Third, the question of the extent to which hypo-chondriasis, anxiety disorders, and obsessive-compulsive disorder (OCD) are related entities (11,12).

The ICD-10 conceptualizes the essential feature of hypo-chondriacal disorder as consisting of one of two variants: a) persistent preoccupation with the possibility of having one or more serious and progressive physical diseases, with pa-

Dan J. Stein

Department of Psychiatry, University of Cape Town, Groote Schuur Hospital, Cape Town 7925, South Africa

This paper addresses the question of how hypochondriasis should be characterized in the ICD-11. It briefly reviews the relevant historical background to the nosology of this disorder, the approaches taken by the ICD-10 and the DSM-IV, problems arising from the application of these approaches, proposals for the DSM-5, and options for the ICD-11. There is an important opportunity to bring the DSM and ICD ap-proaches to health anxieties and preoccupations into closer alignment. ICD-11 options are considered, with an emphasis on considerations of clinical utility, global applicability, and applicability outside mental health settings.

key words: Hypochondriasis, health anxiety, illness preoccupation, DSM, ICD


tients presenting persistent somatic complaints, or b) persis-tent preoccupation with physical appearance. In spite of these two variants being conceptualized as a single disorder, it is notable that the European literature has a long tradition of recognizing and distinguishing patients who suffer from a preoccupation with an imagined flaw in physical appearance from those with concerns about illness (13).

The DSM-IV differentiates between hypochondriasis (characterized by a preoccupation with fears of having, or the idea that one has, a serious disease based on the person’s misinterpretation of bodily symptoms) and body dysmorphic disorder (characterized by preoccupations with a slight or imagined defect in physical appearance).

ICD-10 anD DsM-IV approaCHes

The ICD recognizes that hypochondriacal symptoms are found both in hypochondriacal disorder (a primary condi-tion) and in a range of other mental disorders (as a secondary symptom). Thus, in ICD-10, the term “hypochondriacal” is mentioned in the discussion of several mental disorders, in-cluding post-concussional syndrome, hebephrenic schizo-phrenia, delusional disorder, depressive episode, vomiting associated with psychological disturbances, enduring per-sonality change after psychiatric illness, conduct disorder, and tic disorders.

As a primary condition, F45.2 Hypochondriacal Disorder is classified in the ICD-10 among F45 Somatoform Disorders, which in turn fall under the umbrella of F40-48 Neurotic, Stress-Related and Somatoform Disorders. The ICD-10 states that, for a definite diagnosis, there should be: “a) persistent belief in the presence of at least one serious physical illness underlying the presenting symptom or symptoms, even though repeated investigations and examinations have iden-tified no adequate physical explanation, or a persistent pre-occupation with a presumed deformity or disfigurement; b) persistent refusal to accept the advice and reassurance of several different doctors that there is no physical illness or abnormality underlying the symptoms”. The ICD-10 notes

100_104.indd 100 06/07/12 14:05


that individuals who suffer from hypochondriasis often mis-interpret normal or commonplace sensations and appear-ances, and that the degree of conviction and the emphasis on one presumed illness over another usually varies over time.

In the DSM-IV, the diagnostic criteria for hypochondriasis include having a preoccupation with fears of having, or the idea that one has, a serious disease based on the person’s misinterpretation of bodily symptoms. The preoccupation persists despite appropriate medical evaluation and reassur-ance. The belief is not of delusional intensity (as in delu-sional disorder, somatic type) and is not restricted to a cir-cumscribed concern about appearance (as in body dysmor-phic disorder). The preoccupation causes clinically signifi-cant distress or impairment in social, occupational, or other important areas of functioning. The duration of the distur-bance is at least 6 months. Finally, the preoccupation is not better accounted for by generalized anxiety disorder (GAD), OCD, panic disorder, a major depressive episode, separation anxiety, or another somatoform disorder. In the DSM-IV, a “poor insight” specifier is used if, most of the time during the current episode, the person does not recognize that the con-cern about having a serious illness is excessive or unreason-able.

probleMs arIsIng froM ICD-10 anD DsM-IV

The name “hypochondriasis” comprises an immediate problem. Nearly 20 years ago, Barsky and Klerman (3) sug-gested that the term was confusing and pejorative. More re-cent publications have used terms such as “illness phobia”, “health anxiety”, “illness anxiety”, and “heightened illness preoccupation” (14). These different terms to some extent reflect a long-standing and ongoing debate about how indi-viduals with hypochondriacal preoccupations are best con-ceptualized (3,15,16).

A second issue is the conflation in ICD-10 between hypo-chondriasis and body dysmorphic disorder. A growing litera-ture indicates that the latter is a distinct entity, with clinical utility and diagnostic validity (17). Since the publication of ICD-10, a range of symptom measures and therapeutic inter-ventions have been developed for body dysmorphic disorder (17,18). Thus, conceptualizing it as part of hypochondriasis likely leads to individuals with this disorder not receiving optimal diagnosis, evaluation, and treatment. Further, the ICD-10 diagnostic requirement of persistent refusal to accept the advice and reassurance of several different doctors that there is no abnormality may lead to underdiagnosis of body dysmorphic disorder, as individuals with body dysmorphic symptoms may be less willing to reveal these to clinicians than other individuals are to share their concerns about med-ical illness.

The inclusion of hypochondriasis as a somatoform disor-der in both ICD-10 and DSM-IV is also potentially problem-atic. On the one hand, this may have the advantage of alert-ing clinicians that the differential diagnosis of patients pre-

senting with somatic symptoms (ranging from somatic dis-tress to hypochondriacal preoccupations) includes hypo-chondriasis. On the other hand, hypochondriasis is a disor-der characterized by alterations in cognition and emotion, many patients with mood and anxiety disorders present with somatic symptoms, and the interventions used for hypo-chondriasis are very similar to those that are effective in anxiety and OCD and related disorders (14,19).

A final concern is about the clinical utility of ICD-10 and DSM-IV diagnostic criteria for hypochondriasis in primary care settings around the globe. In their multinational study, for example, Gureje et al (20) found that very few patients in primary care fulfilled the ICD-10 diagnostic criteria for hypo-chondriasis, in large part because it was unusual for patients not to respond to reassurance given in the context of appro-priate medical evaluation. DSM-IV criteria may also yield a low prevalence of hypochondriasis in the community. Gure-je et al (20) emphasized, however, that the triad of disease conviction, associated distress and medical help-seeking is present in primary care settings in different cultures, and is associated with considerable psychiatric ill health and func-tional disability. Studies that have used measures of illness preoccupation in primary care settings have found that it is common, even in the absence of physical complaints (21,22).

proposals for DsM-5

For the DSM-5, a new diagnostic entity has been pro-posed in the somatoform disorders section, namely “com-plex somatic symptom disorder”, which is characterized by both somatization symptoms and illness anxiety symptoms (23). This rubric includes patients with DSM-IV somatiza-tion disorder, hypochondriasis, undifferentiated somatoform disorder, and pain disorder, on the basis that these disorders share the common features of somatic symptoms and cogni-tive distortions. It is argued that a single entity will increase clinical utility and diagnostic validity.

Three criteria are proposed for the diagnosis of complex somatic symptom disorder in DSM-5 (see www.dsm5.org). Criterion A is the presence of one or more somatic symptoms that are distressing and/or result in significant disruption in daily life. Criterion B is excessive thoughts, feelings, and be-haviors related to these somatic symptoms or associated health concerns (i.e., high level of health-related anxiety, dis-proportionate and persistent concerns about the medical seriousness of one’s symptoms, or excessive time and energy devoted to these symptoms or health concerns). Criterion C relates to chronicity: although any one symptom may not be continuously present, the state of being symptomatic must have persisted for at least 6 months.

The DSM-5 proposal allows for a number of optional, un-coded specifiers for complex somatic symptom disorder: pre-dominant somatic complaints (previously, somatization dis-order), predominant health anxiety (previously, hypochon-driasis), and predominant pain. The proposal notes that, if

100_104.indd 101 06/07/12 14:05


patients present solely with health-related anxiety with min-imal somatic symptoms, they may be more appropriately di-agnosed as having illness anxiety disorder.

The DSM-5 proposal indicates that several criteria must be met for a diagnosis of illness anxiety disorder. These in-clude Criterion A, that somatic symptoms are not present or, if present, are only mild in intensity; Criterion B, preoccupa-tion with having or acquiring a serious illness; Criterion C, high level of anxiety about health or having or acquiring a serious illness; and Criterion D, the person performs related excessive behaviors (e.g., checking one’s body for signs of illness), or exhibits maladaptive avoidance (e.g., avoiding doctors’ appointments and hospitals) (see www.dsm5.org).

optIons for ICD-11

The World Health Organization and its International Ad-visory Group have indicated that the revision of ICD-10 should pay particular attention to issues of clinical utility, use in a broad range of countries and cultures, and use in pri-mary care and other general medical settings (24-26).

The first issue requiring further consideration is the name of hypochondriasis, which is also related to how this entity is conceptualized. Options include health anxiety disorder, illness anxiety disorder, and illness preoccupation disorder.

“Health anxiety” is a term that has received some currency in the literature (15,27). On the other hand, anxiety about the domain of health is a cardinal symptom in GAD (for exam-ple, individuals with GAD may worry about whether they or family members will maintain their health status, will have sufficient health insurance, or will receive good health care), whereas patients with hypochondriasis are preoccupied spe-cifically with having an illness (for example, individuals with this disorder may be concerned that they have a cancer that has not yet been diagnosed).

“Illness anxiety disorder” is a term that appropriately re-flects the fact that the concern is illness-related. Furthermore, many patients with hypochondriasis manifest anxiety, there is limited evidence of partial overlap in the psychobiology of hypochondriasis and anxiety disorders such as panic disor-der, and treatments partially overlap with anxiety disorders (16,28).

“Illness preoccupation disorder” is a term that appropri-ately reflects that the characteristic symptom is preoccupa-tion with having an illness. Notably, there are both overlaps and distinctions in the phenomenology, psychobiology, and treatment of hypochondriasis and OCD and related disor-ders (12,16). Thus, for example, neuropsychological research has indicated that patients with hypochondriasis and OCD differ in performance on attentional tasks, but family re-search has found that hypochondriasis is significantly more common among first-degree relatives of OCD patients versus control probands. In one recent brain imaging study, patients with hypochondriais, OCD, but also panic disorder, shared similar alterations in frontal-striatal brain regions during a

planning task (29).Relatedly, there is the question of where to classify hypo-

chondriasis. Following the logic of the previous paragraph, options include somatoform disorder, anxiety disorders, and OCD and related disorders. There is also an extensive litera-ture on the overlap of hypochondriasis with mood disorders (1). It is relevant to note that, for the DSM-5, the proposal has been made to classify body dysmorphic disorder among OCD and related disorders (16).

Given the heterogeneity of hypochondriasis (for example, some patients display more disease phobia, while other pa-tients have greater disease conviction (11)), and given the relative sparsity of data directly comparing the phenomenol-ogy and psychobiology of hypochondriasis with that of a range of other disorders (16), there is an argument for taking a conservative approach that is not premised on premature nosological conclusions. The terms “health anxiety disorder” and “illness anxiety disorder” imply that this condition is an anxiety disorder, a conclusion that may go beyond the cur-rent evidence base. The term “illness preoccupation disor-der” appears more theoretically neutral, and so may be pref-erable at this stage.

As for the categorization of hypochondriasis, from an ICD-10 perspective, given that hypochondriacal disorder represents individuals with both hypochondriasis and body dysmorphic disorder, it would seem that the more conserva-tive option would be to place these two conditions in the same chapter of ICD-11. Should the ICD-11 OCD and re-lated disorders chapter be located between the anxiety disor-ders and the somatoform disorder chapter, then placing ill-ness preoccupation disorder and body dysmorphic disorder in that chapter would be consistent with the DSM-5 ap-proach, and may reflect the phenomenological and psycho-biological overlaps across these conditions (namely, anxiety, OCD and related, and somatoform disorders).

A third question concerns the optimal description of the disorder in ICD-11. There is a growing literature indicating that it is clinically useful to identify illness preoccupations, as there are now specific, effective treatments for individuals who have a disorder characterized by these symptoms (i.e., hypochondriasis) (14,19). Studies from several countries show that illness preoccupations are prevalent, with a signifi-cant sub-sample of these patients not having the focus of their presentation on somatic symptoms per se (20,22,30). Further-more, data show that individuals with illness preoccupations frequently present in primary care settings, and that they are underdiagnosed and undertreated (14,20). On one score, then, there seems little option: a diagnostic description that is readily applicable in primary care settings is needed.

A particularly useful study of a potential diagnostic de-scription for hypochondriasis indicates that rumination about illness plus at least one of five other symptoms form a distinct diagnostic entity that performs better than the DSM-IV criteria (30). Unfortunately, this work has not yet been widely replicated. Nevertheless, the option of a relatively simple diagnostic description, which highlights the centrality

100_104.indd 102 06/07/12 14:05


of illness preoccupations, is appealing from the perspective of ICD-11. There is ongoing debate about whether hypo-chondriasis comprises a distinct taxon or whether hypo-chondriacal symptoms are continuous with normality (8), but in clinical practice it is not always feasible to measure rigorously a broad variety of symptoms (including hypo-chondriacal symptoms), and it is crucial for primary care cli-nicians working in a broad range of medical settings to rec-ognize the diagnostic entity of hypochondriasis.

A final question regards the issue of how best to classify those patients who have limited insight into their hypochon-driacal preoccupations. In particular, there is the question of how to differentiate between hypochondriasis and delusion-al disorder, in patients with absent insight. The DSM-5 pro-posal for OCD, body dysmorphic disorder and hoarding disorder is to include an insight specifier which ranges from good insight to no insight (delusional) (17,31,32). This is consistent with a range of data indicating that, although as-sessing insight may have clinical utility, patients with OCD and related disorders who have no insight may respond to standard treatments for these conditions (17,33). Although it may well be useful to have an insight specifier for hypochon-driasis, this is a more complex issue in the context of ICD-11, because the ICD does not include uncoded specifiers (34).

ConClusIons

ICD-10 hypochondriacal disorder represents a heteroge-neous group of patients. Separating out those with hypo-chondriasis from those with body dysmorphic disorder is consistent with available evidence and would increase clini-cal utility across the multiple general medical settings in which these patients present. It would be timely to replace the term “hypochondriasis”, either with the term “illness anxiety disorder” or with the term “illness preoccupation dis-order”. The term “illness preoccupation disorder” would be the more conservative alternative.

The existing body of research on this diagnostic entity has not answered questions about the optimal placement of this new disorder, and its diagnostic description. Given that ICD-10 has previously lumped hypochondriasis and body dys-morphic disorder together, keeping them in the same group-ing would be the more conservative approach. A relatively simple description focusing on illness preoccupation or ru-mination would be clinically useful, particularly in primary care settings, and may also demonstrate diagnostic validity (30).

acknowledgements

The author wishes to thank B. Fallon and O. Gureje for their comments on a draft of this manuscript.

note

D.J. Stein is a member of the WHO ICD Revision Working Group on the Classification of Mood and Anxiety Disorders, reporting to the International Advisory Group for the Revi-sion of ICD-10 Mental and Behavioural Disorders. The views expressed in this article are those of the author and, except as specifically noted, do not represent the official policies or positions of the International Advisory Group, the Working Group on Mood and Anxiety Disorders, or the WHO.


references

1. Kenyon FE. Hypochondriacal states. Br J Psychiatry 1976;129:1-14.2. Hyler SE, Sussman N. Somatoform disorders: before and after

DSM-III. Hosp Commun Psychiatry 1984;35:469-78.3. Barsky AJ, Klerman GL. Overview: hypochondriasis, bodily com-

plaints, and somatic style. Am J Psychiatry 1983;140:273-83.4. Kellner R. Functional somatic symptoms and hypochondriasis. A

survey of empirical studies. Arch Gen Psychiatry 1985;42:821-33. 5. Hiller W, Rief W, Fichter MM. Dimensional and categorical ap-

proaches to hypochondriasis. Psychol Med 2002;32:707-18. 6. Mayou R, Levenson J, Sharpe M. Somatoform disorders in DSM-V.

Psychosomatics 2003;44:449-51.7. Hiller W, Rief W. Why DSM-III was right to introduce the concept

of somatoform disorders. Psychosomatics 2005;46:105-8.8. Longley SL, Broman-Fulks JJ, Calamari JE et al. A taxometric study

of hypochondriasis symptoms. Behav Ther 2010;41:505-14.9. Schmidt AJM. Bottlenecks in the diagnosis of hypochondriasis.

Compr Psychiatry 1994;35:306-15.10. Noyes R, Stuart S, Watson DB et al. Distinguishing between hypo-

chondriasis and somatization disorder: a review of the existing lit-erature. Psychother Psychosom 2006;75:270-81.

11. Noyes RJ. The relationship of hypochondriasis to anxiety disorders. Gen Hosp Psychiatry 1999;21:8-17.

12. Fallon BA, Qureshi AI, Laje G et al. Hypochondriasis and its rela-tionship to obsessive-compulsive disorder. Psychiatr Clin North Am 2000;23:605-16.

13. Phillips KA. Body dysmorphic disorder: the distress of imagined ugliness. Am J Psychiatry 1991;148:1138-49.

14. Harding KJ, Skritskaya N, Doherty E et al. Advances in understand-ing illness anxiety. Curr Psychiatry Rep 2008;10:311-7.

15. Noyes RJ, Stuart SP, Watson DB. A reconceptualization of the so-matoform disorders. Psychosomatics 2008;49:14-22.

16. Phillips KA, Stein DJ, Rauch SL et al. Should an obsessive-compul-sive spectrum grouping of disorders be included in DSM-V? De-press Anxiety 2010;27:528-55.

17. Phillips KA, Wilhelm S, Koran LM et al. Body dysmorphic disorder: some key issues for DSM-V. Depress Anxiety 2010;27:573-91.

18. Phillips KA. Body dsymorphic disorder: diagnostic controversies and treatment challenges. Bull Menninger Clinic 2000;64:18-35.

19. Thomson AB, Page LA. Psychotherapies for hypochondriasis. Co-chrane Database Syst Rev 2007;CD006520.

20. Gureje O, Ustün TB, Simon GE. The syndrome of hypochondriasis: a cross-national study in primary care. Psychol Med 1997;7:1001-10.

21. Looper K J, Kirmayer LJ. Hypochondriacal concerns in a commu-nity population. Psychol Med 2001;31:577-84.

22. Kirmayer LJ, Robbins JM. Three forms of somatization in primary care: prevalence, co-occurrence, and sociodemographic character-

100_104.indd 103 06/07/12 14:05


istics. J Nerv Ment Dis 1991;179:647-55. 23. Dimsdale J, Creed F, DSM-V Workgroup on Somatic Symptom Dis-

orders. The proposed diagnosis of somatic symptom disorders in DSM-V to replace somatoform disorders in DSM-IV– a preliminary report. J Psychosom Res 2009;66:473-6.

24. Reed GM. Toward ICD-11: improving the clinical utility of WHO’s international classification of mental disorders. Prof Psychol Res Pr 2010;41:457-64.

25. Reed GM, Correia Mendonça J, Esparza P et al. The WPA-WHO Global Survey of Psychiatrists’ Attitudes Towards Mental Disorders Classification. World Psychiatry 2011;10:118-31.


27. Asmundson GJ, Abramowitz JS, Richter AA et al. Health anxiety: current perspectives and future directions. Curr Psychiatry Rep 2010;12:306-12.

28. Braddock AE, Abramowitz JS. Listening to hypochondriasis and hearing health anxiety. Expert Rev Neurother 2006;6:1307-12.

29. van den Heuvel OA, Mataix-Cols D, Zwitser G et al. Common lim-bic and frontal-striatal disturbances in patients with obsessive com-pulsive disorder, panic disorder and hypochondriasis. Psychol Med 2011;41:2399-410.

30. Fink P, Ørnbøl E, Toft T et al. A new, empirically established hypo-chondriasis diagnosis. Am J Psychiatry 2004;161:1680-91.

31. Leckman JE, Denys D, Simpson HB et al. Obsessive-compulsive disorder: a review of the diagnostic criteria and possible subtypes and dimensional specifiers for DSM-V. Depress Anxiety 2010;27: 507-27.

32. Mataix-Cols D, Frost RO, Pertusa A et al. Hoarding disorder: a new diagnosis for DSM-V? Depress Anxiety 2010;27:556-72.

33. Phillips KA, Price LH, Greenberg BD et al. Should DSM’s diagnos-tic groupings be changed? In: Phillips KA, First MB, Pincus HA (eds). Advancing DSM: dilemmas in psychiatric diagnosis. Washing-ton: American Psychiatric Association, 2003:57-84.

34. First MB. Challenges in the implementation of diagnostic specifiers for mood disorders in ICD-11. World Psychiatry 2012;11(Suppl. 1):17-25.

100_104.indd 104 06/07/12 14:05

WPAVolume 11, Supplement 1 June 2012

Cultural issues in the classification and diagnosis of mood and anxiety disorders

Documents

Transcript of Cultural issues in the classification and diagnosis of mood and anxiety disorders