Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

Long-term outcome and post-treatment effects ofpsychoanalytic psychotherapy with young adults

Annika Lindgren1,2*, Andrzej Werbart1,2 and Bjorn Philips1,3

1Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden2Institute of Psychotherapy, Stockholm County Council, Sweden3Centre for Dependency Disorders, Stockholm County Council, Sweden

Objectives. The short- and long-term effects of open-ended, long-term psycho-analytic psychotherapy for young adults were investigated. Possible changes during theyear and a half follow-up, as well as predictors of change, were explored.

Design. Patients aged 18–25 years who accepted the offered psychoanalyticindividual or group psychotherapy were included. Patients filled out questionnaires andwere interviewed at intake, termination, and follow-up. Alliance data were collectedafter the second session of psychotherapy proper.

Methods. The primary outcome measures were the Symptom Checklist-90 and theInventory of Interpersonal Problems. The Helping Alliance Questionnaire-II was usedto measure alliance. Mixed model ANOVAs were used to analyse changeover time andprediction of change in relation to gender, treatment format, treatment duration, and inindividual psychotherapy, therapist- and patient-rated alliance.

Results. All outcome measures changed significantly from intake to follow-up. Nonechanged significantly during the follow-up period, but there was a tendency towardsrecurring symptoms and an improvement in one of the object relational measuresduring the follow-up. The latter was the only outcome measure that did not changesignificantly during treatment. Lower therapist-rated alliance was predictive of greaterchange in psychiatric symptoms for patients with high levels of symptoms at intake.

Conclusions. The long-term effectiveness of psychoanalytic psychotherapy for youngadults was supported. Low therapist-rated alliance implies that the therapists haveidentified problematic interactions, which might have mobilized their effort to solve theproblems. Further research on cases reporting no gain or even deterioration is needed.

Meta-analyses of psychotherapy outcome demonstrate that psychotherapy in general

affects a diverse range of disorders and conditions (Doidge, 1997; Fonagy, Roth, &

Higgitt, 2005). Earlier research indicates that different treatment modalities produce

equivalent outcome (Lambert, Bergin, & Garfeild, 2004). All psychotherapy models are

* Correspondence should be addressed to Annika Lindgren, Karolinska Institutet, Sektionen for psykoterapi, Bjorngardsgatan25, S-118 56 Stockholm, Sweden (e-mail: Annika.Lindgren.2@ki.se).

TheBritishPsychologicalSociety

27

Psychology and Psychotherapy: Theory, Research and Practice (2010), 83, 27–43

q 2010 The British Psychological Society

www.bpsjournals.co.uk

DOI:10.1348/147608309X464422

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

intended to bring long-lasting effects, and a continued progression after treatment

termination is assumed in psychoanalytic theory (Dreher, 2002, p. 25). The same

desirable outcome should also apply to psychodynamic long-term psychotherapy, as it is

based on psychoanalytic theory. However, the research support for this theoretical

assumption is weak. Published studies of long-term effects of psychotherapy, all

naturalistic and quasi-experimental in design, have reported mixed results. A Swedishstudy of subsidized psychoanalytic treatments and psychoanalytic psychotherapies

reported a positive development from pre-treatment to follow-up with decreased levels

of psychiatric suffering during the post-treatment period (Blomberg, Lazar, & Sandell,

2001; Sandell et al., 2000), especially for patients in psychoanalysis. A German

retrospective study of a similar population found that between 70 and 80% of the

patients reported that their well-being had increased and that the change achieved

during treatment was stable during the post-treatment period (Leuzinger-Bohleber,

Stuhr, Ruger, & Beutel, 2003). A study that followed 62 patients in either long-termbehaviour therapy or long-term psychoanalytic psychotherapy found that both groups

decreased their levels of suffering during treatment and the gains remained over a 7-year

follow-up (Brockmann, Schluter, & Eckert, 2006). Interpersonal problems continued to

decrease during follow-up for patients who had been in psychoanalytic psychotherapy.

A 1-year follow-up of 23 patients in psychoanalytic therapy also revealed stable change

in psychiatric symptoms and a tendency towards improvement of interpersonal

problems over the follow-up period (Leichsenring, Biskup, Kreische, & Staats, 2005).

Also, an effectiveness study of open-ended psychodynamic psychotherapy revealed asignificant change during treatment, but due to attrition rates it was not possible to

analyse data beyond the first year from inclusion (Roseborough, 2006). A meta-analysis

of psychodynamic psychotherapies with duration of at least 1 year of treatment

identified 11 studies that reported follow-up data (Leichsenring & Rabung, 2008). The

meta-analysis found that the treatments produced large effect sizes and that treatment

gains seemed to be stable over the follow-up period. However, no formal tests of post-

treatment changes were conducted.

Only two studies specifically focusing on young adults, the target population of thepresent study, have been found. Positive change in self-rated problems from intake to

one year later was reported from community-based psychodynamic psychotherapy

including both adolescents and young adults (Baruch & Fearon, 2002). However, it is not

clear whether the patients were still in treatment at the 1-year follow-up, or if and when

they had terminated their psychotherapy. A study of adolescents and young adults

participating in an in-patient multi-treatment programme reported that increase of

complexity in object relations or schemas of self and significant others was related to

improvement in expert-rated general clinical functioning during treatment, but noinformation about post-treatment changes was presented (Blatt, Stayner, Auerbach, &

Behrends, 1996).

In the present study, we have tried to overcome the most recurrent critique among

psychoanalytic scholars against the common design of outcome studies (e.g. Blatt &

Zuroff, 2005; Kazdin, 2008; Schneider et al., 2002; Westen, Novotny, & Thompson-

Brenner, 2004). To increase the external validity, we have adopted a naturalistic stance.

The psychotherapies have been conducted in accordance with standard descriptions

and procedures of psychoanalytic psychotherapy without use of an explicit treatmentmanual, and the battery of instruments included measures of symptoms as well as

concepts central to psychoanalytic theory. Instead of using psychiatric diagnoses as

inclusion criteria, a developmental approach was adopted. Following the tradition from

28 Annika Lindgren et al.

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

child and adolescent psychotherapy, the present study focuses on psychotherapy for

young adults, ages 18–25. This is understood as a transitional period, when the young

individual leaves adolescence and forms his/her position and identity within adult

society. According to psychoanalytic developmental theory, this is the period when

intellectual and emotional capacities should consolidate in order to be able to meet the

demands of life and society. It is assumed that the encounter with life tasks andexistential questions might stir heightened levels of stress, especially when the

psychological development is interrupted or disturbed for some reason (Emde, 1985;

Escoll, 1987). In addition, young adults report higher levels and prevalence of

psychiatric problems than other groups of adults (Statistics Sweden, 2006).

When experimental restrictions are kept at a minimum, it is warranted to examine

other factors that might influence outcome in addition to the treatment. Some of the

commonly explored predictor variables of outcome in other patient samples, such as

patient– and therapist-rated alliance, gender, treatment duration, and individual versusgroup therapy seem reasonable for evaluating the present patient population. Alliance,

defined as the agreement on goals and tasks, and the emotional bond between the

therapist and the patient (Bordin, 1979), has been established as a robust predictor or

mediator of outcome (Horvath & Symonds, 1991; Martin, Garske, & Davis, 2000) in

psychotherapy for adult patients. Better agreement and quality of collaboration is

predictive of a better outcome. Gender and treatment modality in general have not been

associated with outcome (Burlingame, Fuhriman, & Mosier, 2003; Clarkin & Levy, 2004;

Fuhriman & Burlingame, 1994), that is women and men tend to profit equally frompsychotherapy, and no difference in effectiveness between individual and group

psychotherapy has been found. However, we do not know if this is the case in a

population of young adults. Owing to the disparate results in the literature about the

association between treatment duration and outcome (Feaster, Newman, & Rice, 2003;

Orlinsky, Ronnestad, & Willutzki, 2004), we included duration for evaluation.

The present study had three objectives: (1) to evaluate the short- and long-term effect

of psychoanalytic psychotherapy for young adults, (2) to evaluate maintenance and

further development of therapy gains after termination, and (3) to explore whethergender, treatment duration, treatment format (individual or group psychotherapy), and

patient- and therapist-rated alliance (only in individual psychotherapy) predicts and

moderates changes in psychiatric symptoms and level of interpersonal distress during

treatment and/or during the follow-up period. Two treatment modalities, individual and

group psychoanalytic psychotherapy were included in the present study. Even though

earlier research indicates that the two produce equivalent outcome, this does not mean

that the change processes or interventions leading to change are the same in group and

individual psychotherapy. In order to stay true to the complexity of this naturalisticstudy, we still decided to treat the outcome of both psychotherapy modalities as one

outcome. Psychotherapy modality was included in the analyses of possible predictors of

outcome. In order to facilitate comparisons with other studies on either individual or

group psychotherapy, we will include descriptive data and effect sizes for both

treatments separately on our two main outcome measures.

Method

This study uses data from the Young Adult Psychotherapy Project (YAPP), an ongoing

naturalistic, prospective, and longitudinal study of young adults in psychoanalytic

Long-term outcome for young adults 29

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

psychotherapy. The research project and the participants have been fully described

elsewhere (Philips, Wennberg, Werbart, & Schubert, 2006); we will provide a brief

overview here. All psychotherapies were conducted within the regular work at the

Institute of Psychotherapy, Stockholm County Council. Patients applied for

psychotherapy through a telephone service and were offered consultation with a

psychotherapist when a vacancy was available. A few patients were referred from thecounty council’s psychiatric out-patient clinics.

ParticipantsYAPP comprised 134 young adults, aged 18–25, who applied for psychotherapy. Age,

acceptance of the offered psychoanalytic psychotherapy, and providing informed

consent to participation in the study were the inclusion criteria. No exclusion criterion

was adopted. Seventy-three per cent of the participants were women, and mean age at

inclusion was 22 years (SD ¼ 2:2). There was a great variation in both type and number

of complaints expressed by the patients. Common complaints were depressiveness,

anxiety, interpersonal problems, and low self-esteem. Self-reports of suffering andexpert assessed functioning show that the patients were moderately suffering from

different symptoms at intake, while their capacities to describe themselves and

important others in a modulated way were quite good (see Table 2). Self-reported

personality disorders, in accordance with the DSM-IV and ICD-10 Personality

Questionnaire (DIP-Q; Ottosson et al., 1995, 1998), were found in 41 patients (31%),

which is a lower prevalence than has been reported in a recent study of adult patients in

psychiatric care in Stockholm (Noren et al., 2007).

AttritionDuring the inclusion period, 170 patients applying for psychotherapy fulfilled the age

criterion. Thirty-six patients did not enter the project, as they did not wish toparticipate, failed to appear for the initial sessions, or due to administrative error.

Drop-out was defined as a treatment shorter than 3 months. A time period shorter

than 3 months was chosen as the definition of drop-out since 10 to 12 sessions once a

week is a common length of psychodynamic short-term psychotherapy, and this project

did not include short-term psychotherapy as a treatment option. Sixteen patients

dropped out of psychotherapy, all during the first month of treatment, or did not reach a

treatment contract with their psychotherapists. There are more men than expected

among the drop-outs (37.5%) compared with among completers (26.1%). Neither agenor suffering at intake differed between the groups. Further analyses of predictors of

drop-out will be reported in a separate study (Roos, Wennberg, Philips, & Werbart,

2009).

Owing to the design with open-ended psychotherapies, three patients are still in

treatment at the time of the present study and are thus excluded from the analyses. Of

the remaining 131 patients, 115 came to the research evaluation at termination and 99

patients came to the follow-up 1.5 year post-treatment. Owing to internal missing data,

that is patients not answering all questions on all questionnaires, the N for the differentquestionnaires varies slightly with each measurement point.

Following a design of intention-to-treat, all patients with data at intake (except the

three still in psychotherapy) were included in the analyses. Accordingly, all proportions

were calculated with 131 as the sample population. The analyses of prediction of

30 Annika Lindgren et al.

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

outcome from alliance were based on all available ratings at the second session

following the contract agreement in individual psychotherapy, i.e. 69 patient ratings and

73 therapist ratings.

TreatmentsAll participants were offered either individual (N ¼ 92) or group psychoanalytic

psychotherapy (N ¼ 42). The treatment option depended on what modality was

available at the moment. No formal criterion for recommending either treatment was

formulated. However, if both group and individual therapy was available when a patient

called, a clinical decision on what to offer was made by the therapist receiving the call.

The psychotherapies were not manualized but were intended to be psycho-

analytically informed psychotherapies. The treatments in general aimed at helping the

young adults to overcome developmental arrests and to better handle strains in everydaylife. All therapists met every week in clinical teams, where treatment problems and

clinical experiences were discussed. There is no process data collected that makes it

possible to evaluate the therapists interventions, but it is obvious that both therapists

and patients are describing psychodynamic psychotherapies at the interviews at

termination and follow-up ( Johansson & Werbart, 2009; Lilliengren & Werbart, 2005).

Treatment duration was adapted to the needs of the individual patient and was

formalized in a written renegotiable contract between therapist and patient. Five

psychotherapy groups were conducted within the project. One of the groups was semi-open and open-ended, while the others were closed with a time frame of 1 or 1.5 years.

The patients stayed in treatment for an average of 19 months (SD ¼ 13:8,range ¼ 1–55).

PsychotherapistsThirty-seven different therapists were involved in this study, all with a psychoanalytic

orientation (26 female and 11 male). They represented various professionalbackgrounds: physicians (6), psychologists (14), social workers (15), or other (2).

Fifteen were psychoanalysts and licensed psychotherapists, 19 were licensed

psychotherapists, and 3 had basic training in psychodynamic psychotherapy. As to

their experience in psychotherapeutic work, the mean time from license for the 34

licensed psychotherapists was 10.3 years (median 12.5 years; range 1–15 years).

Individual therapy was conducted by 34 therapists (between 1 and 7 patients each) and

group therapy by 6 therapists. Most of the therapists had been working within a clinical

project aiming at developing and providing psychoanalytic psychotherapy for youngadults.

InstrumentsThree instruments were used to measure different aspects of symptoms and health:

Symptom Checklist-90 (SCL-90; Derogatis, 1994), a self-report questionnaire was

used to assess psychiatric symptoms experienced over the last 7 days. Items are rated on

five-point Likert-scales ranging from zero (‘not at all’) to four (‘very much’). The GlobalSeverity Index (GSI) was used for the analyses in this study. The SCL-90 demonstrates

adequate reliability (Derogatis, Rickels, & Rock, 1976) and for the Swedish translation a

Cronbach’s alpha of .97 has been reported (Fridell, Zvonomir, Johansson, & Malling

Thorsen, 2002).

Long-term outcome for young adults 31

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

Self-rated health (SRH; Bjorner et al., 1996) is a single-item measure where the

respondent rates his/her present subjective mental and somatic health. The item is rated

on a seven-point Likert-scale ranging from one (‘very bad’) to seven (‘very good’). SRH

has been shown to be a good predictor of mortality and health problems (Heidrich,

Liese, Lowel, & Kiel, 2002; Larsson, Hemmingsson, Allebeck, & Lundberg, 2002; Mossey

& Shapiro, 1982; Shadbolt, Barresi, & Craft, 2002), and the item has shown good test–retest reliability in a previous study (r ¼ :92; Lorig et al., 1996).

Global Assessment of Functioning Scale (GAF; American Psychiatric Association,

1994) was used for measuring symptomatic and social functioning. The ratings were

based on interviews at intake, except for one-third of the group. Their ratings were

based on case presentations by the therapists. A group of trained raters did all

assessments, and consensus ratings were used in order to increase reliability of

assessments. An independent rater later rated 40 cases (approximately 10% of all cases

over the three time-points) and the intra-class correlation was ¼ .74.Three instruments were used to measure aspects of personality:

The Inventory of Interpersonal Problems (IIP; Horowitz, Rosenberg, Baer, Ureno, &

Villasenor, 1988) was used to measure the degree of interpersonal suffering. The 64

items circumplex version of IIP was used (Alden, Wiggins, & Pincus, 1990). Each item

was rated on five-point Likert-scales ranging from zero (‘not at all’) to four (‘very much’).

A later translation than the presently used has shown adequate psychometric properties

in a Swedish sample (Weinryb et al., 1996). The internal consistency (Cronbach’s alpha)

of the eight clusters in the present Swedish translation of IIP has been reported as .78to .91 (Rosander & Stiwne, 1997). The average of all items was calculated as a measure

of interpersonal distress.

The structural analysis of social behaviour intrex questionnaire (SASB; Benjamin,

1974, 1983) was used to assess levels of positive and negative self-concepts. The

instrument comprises 36 items about oneself rated on 11-point Likert-scales ranging

from 0 (‘not true’) to 100 (‘true’). The reported internal consistency of the Swedish

version of the questionnaire is . 0:90 for all of the subscales (Armelius, 2001).

Following Adamson and Lyxell (1996), means of clusters two through four and clusterssix through eight were used as indicators of positive and negative self-concepts.

The Differentiation-Relatedness of Self and Object Representations Scale (DRS;

Blatt & Auerbach, 2001; Diamond, Blatt, Steyner, & Kaslow, 1995) was used to evaluate

the degree of complexity and differentiation in cognitive-affective schemas of concepts

of self and others. The ratings were based on information obtained through the Object

Relation Inventory (Blatt, Wein, Chevron, & Quinlan, 1979; Diamond et al., 1995; Gruen

& Blatt, 1990). Each patient received independent ratings of representations of mother,

father, and him/herself. A group of trained raters made all assessments. A reliability studybased on part of the present material reported the inter-rater agreement as an intra-class

correlation of .71 (Hjalmdahl, Claesson, Werbart, & Levander, 2001). Consensus ratings

were used in order to increase reliability of assessments.

Finally, the Helping Alliance Questionnaire (HAq-II; Luborsky et al., 1996) was used

to measure both patient- and therapist-rated alliance in individual psychotherapy. The

instrument comprises 19 items rated on six-point Likert-scales ranging from one

(‘I strongly feel it is not true’) to six (‘I strongly feel it is true’). The HAq-II demonstrates

excellent reliability and good convergent validity with other alliance measures(Luborsky et al., 1996). The questionnaire was translated to Swedish by the principal

investigator of the present study and has been used in several studies, although no

psychometric data have been previously published. The internal consistency

32 Annika Lindgren et al.

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

(Cronbach’s alpha) of the Swedish version was .91 for the patient questionnaire and .88

for the therapist questionnaire in this sample.

ProceduresThe psychotherapists were free to accept or reject a patient based on assessment of the

patient’s motivation for treatment and a general idea about suitability for the specific

treatment modality. Only one patient was rejected. Every second patient who was

assigned to individual therapy (N ¼ 47 of 92) and all patients assigned to group therapy(N ¼ 42) underwent a research interview pre-therapy. All patients were interviewed at

termination and at follow-up 1.5 years after termination. All patients completed

background and personality questionnaires pre- and post-treatment as well as at the

follow-up. Alliance was measured every third month starting from the second

psychotherapy session. The information from patient questionnaires and research

interviews was not disclosed to the psychotherapists during the treatment.

Data analysisInspection of descriptive statistics of all measures at all time-points revealed a positively

skewed distribution pattern in GSI at termination and follow-up, and a negatively

skewed distribution in DRS at all three time-points. DRS was thus reflected and bothmeasures were then log-transformed before any analyses of outcome were performed.

To analyse outcome, a series of mixed model ANOVAs with one within fixed-factor

(time) was performed. GSI and IIP were chosen as the two primary outcome measures,

whereas the other measures were seen as secondary outcome measures. The variance

structure for analyses of all outcome measures was set to compound symmetry

heterogeneous based on Akaike’s information criterion and an inspection of estimated

variance components. Bonferroni correction of the alpha level was performed for the

repeated comparisons between time-points and across outcome variables. In order tokeep the familywise error at 5%, the alpha level was set to .003 for each separate

significance test.

Effect sizes of change were calculated for each measure and each time period (intake

to termination, termination to follow-up, and intake to follow-up) based on estimated

marginal means and standard deviation at intake obtained through the mixed model

analyses. Z-scores representing standardized values within each measure over the three

time-points were used to illustrate the amount of change in all measures.

In order to evaluate the obtained change, we compared the effect sizes with resultsfrom earlier published studies. Research on long-term psychodynamic psychotherapy is

scarce (Crits-Christoph & Barber, 2000). Only six studies with a mean treatment

duration exceeding 18 months were identified in a recently published meta-analysis of

long-term psychodynamic psychotherapy (Leichsenring & Rabung, 2008). For this

reason, the comparison was based on data from two meta-analyses on short-term

psychodynamic psychotherapy. Leichsenring, Rabung, and Leibing (2004) reported that

adult patients treated for a diverse range of psychiatric disorders achieved a within-

group effect size in the range of 0.80–0.95 for general psychiatric symptoms and socialfunctioning. The lower 95% confidence limits for these measures ranged from 0.56

to 0.65. Thus, we decided that effect sizes below 0.56 were small or modest, effect

sizes in the interval 0.56 to 0.95 were of medium size, whereas effect sizes above 0.95

where large.

Long-term outcome for young adults 33

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

To determine whether results at follow-up were equivalent to termination, we

used data from a meta-analysis of placebo or waiting list controls in randomized,

controlled studies of short-term psychodynamic psychotherapy (Leichsenring &

Rabung, 2006). An average effect size of 0.12 with a 95% confidence interval ranging

from 20.05 to 0.28 has been reported for these control groups. Effect sizes below

20.05 were interpreted as an indication of deterioration, whereas effect sizes greaterthan 0.28 were seen as indications of a tendency to improve. In addition, percentage

of improved patients was calculated based on the reliable change index (RCI;

Jacobson, Roberts, Berns, & McGlinchey, 1999; Jacobson & Truax, 1991) on GSI and

IIP. The RCI was chosen rather than the cut-off level of two standard deviations above

the mean of the norm group as a measure of clinically significant change, since

approximately 75% of the included patients in this study reported levels below this

cut-off at intake. Thus, they had no room to become clinically significantly changed.

The RCI was based on Cronbach’s alpha following recommendations from Evans,Margison, and Barkham (1998). A value above 1.96 was interpreted as a reliable

improvement, while a value below 21.96 was interpreted as a reliable deterioration.

The internal consistencies reported in the Swedish SCL-90 manual (Fridell et al.,

2002) and the Swedish IIP manual (Psykologiforlaget, 2002) were used to calculate

the RCI of GSI and IIP.

Finally, in order to further explore the outcome, analyses of the association between

change in the two primary outcome measures (GSI and IIP) and possible predictors of

change were performed. The mixed model analyses were rerun with one predictorvariable included at a time. The interaction between predictors and time was analysed

through a piecewise model that divided time into two periods, during treatment and

during follow-up (Schwartz, 1993), since the analyses of outcome indicated a difference

in rate of change during and after psychotherapy. The predictor variables were gender,

duration of treatment, treatment modality (individual or group psychotherapy), patient-

rated alliance, and therapist-rated alliance in individual psychotherapy. The analyses

were rerun with intake level of the dependent variable added as a covariate to the

analyses whenever a significant predictor was found. This was done in order to checkwhether intake level explained the significant results. No correction of alpha-level was

adapted to these exploratory analyses.

Results

Short- and long-term outcomeThe mixed model analyses showed that eight of nine outcome measures were

significantly changed at termination, and by follow-up all nine outcome variables were

significantly changed (Table 1 and Figure 1). Effect sizes of change during treatment

were modest to high, whereas the effect sizes of change from intake to follow-up were

medium to high (Tables 2 and 3).

Based on the RCI cut-off scores, 66 patients (50.4%) showed significant improvement

from intake to follow-up, whereas five patients (3.8%) showed impairment in severity on

the GSI. Concerning interpersonal problems (IIP), 46 patients (35.1%) showedimprovement and 6 patients (4.6%) deteriorated by follow-up. The corresponding

numbers for change between intake to termination were 58 patients (44.3%) improved

and 8 patients (6.1%) deteriorated on the GSI, and 41 patients (31.3%) improved and

4 patients (3.1%) deteriorated on the IIP.

34 Annika Lindgren et al.

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

Maintenance of therapeutic gainNone of the outcome variables changed significantly during the follow-up period

(Table 1). The effect sizes for the post-treatment period was low to modest (ranging

from 20.09 to 0.31). Scores on GAF, GSI, SASB-positive, and SRH were slightly reduced

Table 1. Mixed model analysis of change in outcome measures from intake to termination and follow-up

Model Intake versus TerminationaTermination versus

Follow-upa

N df F p df t p df t p

Symptom measuresGSIb 129 2/119 83.67 , .001 105 10.35 , .001* 158 0.51 .61SRH 129 2/157 78.21 , .001 157 10.84 , .001* 153 0.69 .49GAF 131 2/132 80.99 , .001 121 11.38 , .001* 118 0.59 .56Personality measuresIIP 128 2/172 40.72 , .001 174 6.83 , .001* 172 1.28 .20SASB-P 128 2/156 46.16 , .001 163 8.07 , .001* 158 0.24 .81SASB-N 128 2/150 39.34 , .001 157 6.97 , .001* 155 0.63 .52DRS-M 88 2/103 12.12 , .001 111 2.30 .023 96 2.40 .02DRS-F 88 2/104 16.48 , .001 105 3.97 , .001* 98 1.61 .11DRS-S 88 2/95 21.84 , .001 97 4.88 , .001* 99 0.53 .60

Note. SRH, Self-Rated Health; GSI, The Global Severity Index; GAF, Global Assessment of FunctioningScale; IIP, The Inventory of Interpersonal Problems; SASB, the structural analysis of social behaviorintrex questionnaire (Positive, Negative); DRS, The Differentiation-Relatedness of Self and ObjectRepresentations Scale (Mother, Father, Self). Asterick indicate significant p value after a Bonferronicorrection.a Bonferroni-correction for 18 comparisons gives a p value of .003 to keep the family wise error at analpha level of 5%.b Symptom Checklist-90-R.

Figure 1. Mean Z-scores of all outcome measures at each time-point.

Long-term outcome for young adults 35

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

during follow-up. The change was greater than an effect size of 2 .05 for GAF, GSI,

and SRH, which indicated a tendency towards deterioration. The change in positive

direction for DRS-mother was large enough to support a conclusion that the results

at termination and follow-up were not equivalent, even though the difference was

not significant.

Table 2. Descriptive statistics and effect sizes of outcome data based on estimates from the mixed

model analyses

Intake Termination Follow up

M SD M SD M SD es1 es2 es3

Symptom measuresGSIa 1.2 0.6 0.8 0.6 0.8 0.6 1.30 20.07 1.23SRH 3.0 1.4 4.7 1.5 4.6 1.5 1.27 20.09 1.18GAF 56.8 8.0 67.5 11.8 66.8 12.4 1.33 20.09 1.24Personality measuresIIP 1.3 0.5 1.0 0.5 1.0 0.5 0.57 0.11 0.69SASB-P 40.3 16.7 53.6 18.1 53.2 16.6 0.80 20.03 0.77SASB-N 34.8 16.9 23.5 17.5 22.4 15.9 0.67 0.06 0.73DRS-Mb 6.5 1.5 7.2 1.0 7.4 0.7 0.32 0.31 0.63DRS-Fb 6.7 1.4 7.5 0.8 7.6 1.0 0.54 0.22 0.76DRS-Sb 6.1 1.6 7.4 1.1 7.4 1.0 0.80 0.09 0.89

Note. SRH, Self-Rated Health; GSI, The Global Severity Index; GAF, Global Assessment of FunctioningScale; IIP, The Inventory of Interpersonal Problems; SASB, the structural analysis of social behaviorintrex questionnaire (Positive, Negative); DRS, The Differentiation-Relatedness of Self and ObjectRepresentations Scale (Mother, Father, Self). es1, Intake to termination; es2, Termination to follow-up;es3, Intake to follow-up.a Symptom Checklist-90, Descriptive statistics are based on the 86 patients with complete data for allthree time-points and are not compatible with the es.b Descriptive statistics are based on the 49 patients with complete data for all three time-points and arenot compatible with the effect size.

Table 3. Descriptive statistics and effect sizes of main outcome measures based on estimates from the

mixed model analyses in individual and group psychotherapy

Intake Termination Follow up

M SD M SD M SD es1 es2 es3

Individual psychotherapyGSIa 1.4 0.6 0.7 0.7 0.8 0.7 1.23 0.00 1.23IIP 1.3 0.6 1.0 0.6 1.1 0.6 0.63 20.18 0.45Group psychotherapyGSIa 1.1 0.5 0.7 0.6 0.6 0.5 1.29 0.26 1.55IIP 1.3 0.4 1.0 0.5 0.9 0.5 0.83 0.07 0.90

Note. GSI, Global Severity Index; IIP, The Inventory of Interpersonal Problems. es1, Intake totermination; es2, Termination to follow up; es3, Intake to follow up.a Symptom Checklist-90, Descriptive statistics are based on the patients with complete data for allthree time-points and are not compatible with the es.

36 Annika Lindgren et al.

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

Eighty-six patients completed the GSI and IIP at all three time-points. Thirty-nine

(29.8% of 131 included patients) reported a significant improvement in GSI scores at

both termination and follow-up, whereas 9 patients (6.9%) had lost their significant

improvement from termination to follow-up, and 16 patients (12.2%) had values above

the critical value of RCI 1.96 at follow-up but not at termination. The corresponding

numbers for IIP were 28 patients (21.4%) with stable improvement, 4 patients (3.1%)had lost their significant improvement, and 11 (8.4%) had reached a significant

improvement at follow-up.

Predictors and moderators of changeTherapist-rated alliance measured after session 2 was significantly related to change in

SCL-90 GSI during treatment (F½1; 122� ¼ 4:76, p ¼ :03), whereas a tendency towards asignificant association was found for patient-rated alliance in relation to the same

measure (F½1; 114� ¼ 3:67, p ¼ :06). The association between therapist-rated alliance

and change in SCL-90 GSI was negative, that is, lower levels of therapist-rated alliance

were related to greater change. The association between patient-rated alliance and

change in SCL-90 GSI showed a reversed pattern, that is, higher level of patient-rated

alliance was related to greater change in SCL-90 GSI. The tendency towards significant

association between patient-rated alliance and change in SCL-90 GSI was lost when the

intake level of SCL-90 GSI was added as a covariate to the analyses (F½1; 127� ¼ 0:968,p ¼ :33). When the same was done with therapist-rated alliance as a moderator, a three-

way interaction was found in relation to change in SCL-90 GSI during treatment

(F½1; 144� ¼ 4:274, p ¼ :04). A median split based on intake level of SCL-90 GSI showed

that the negative association between therapist-rated alliance and change in psychiatric

symptoms was significant for patients with higher ratings on SCL-90 GSI at intake

(r ¼ 2:36), but no association was found in the group of patients with lower ratings

(r ¼ :01). A significant negative main effect was found for treatment duration in relation

to IIP (F½1; 172� ¼ 4:12, p ¼ :04), meaning that patients with lower average levels ofinterpersonal distress over all three time-points stayed longer in treatment, but

treatment duration was not related to change in interpersonal distress. Neither gender,

treatment duration, nor treatment format (individual or group psychotherapy)

predicted changes on GSI and IIP up to termination or follow-up.

Discussion

The long-term effectiveness of psychoanalytic psychotherapy for young adults was

supported in the present study. There was a significant positive change in all nine

outcome variables from intake to follow-up. All outcome measures except one changed

significantly during treatment, and none of the measures changed significantly from

termination to follow-up. The rating of differentiation-relatedness in object

representation of mother changed significantly from intake to follow-up, but not

during the treatment or follow-up period separately. Effect sizes of change in alloutcome measures during treatment were moderate to high, and modest or negligible

for change during the follow-up period. An indication of recurring psychiatric symptoms

and health problems during the follow-up period was found, as well as an indication for

continued progression for one of the object relation measures. Reliable improvement of

psychiatric symptoms at follow-up was achieved by approximately 50% of the patients

Long-term outcome for young adults 37

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

while on a measure of interpersonal problems, 35% of patients showed reliable

improvement.

Lower levels of therapist-rated alliance predicted greater change in psychiatric

symptoms for patients with high levels of psychiatric symptoms at intake. A tendency

towards an association between higher patient-rated alliance and change in psychiatric

symptoms was found. None of the other possible predictors were associated withchanges in psychiatric symptoms or interpersonal distress.

The major limitation of this study is also its main asset, i.e. its naturalistic design. This

approach facilitates optimization of external validity, but the lack of a comparison group

makes it impossible to rule out the time effect. The treatment model in this study

conceptualized the patients’ problems as arrests in development, and all gains could be

true effects of maturation rather than treatment effects. However, effect sizes were

comparable to within-group effect sizes usually reported in randomized controlled

studies of short-term treatments (Leichsenring et al., 2004) and they exceed the upperlimit of the confidence interval for change occurring in psychiatric patients participating

in control groups not receiving specific treatment. This supports the conclusion that the

results presently reported are true treatment effects. In addition, the broken line of

development in almost all outcome measures illustrates a difference in amount

of change during and after treatment, which indicates that the psychotherapies have had

an effect on outcome. However, the results need to be replicated.

The theoretical assumption of continued development and improvement after

treatment termination could not be supported. In accordance with findings on short-term psychodynamic psychotherapy (Leichsenring et al., 2004), the therapeutic gains

were maintained during the follow-up. However, while lacking statistical significance,

not all variables were equivalent at termination and follow-up. Effect sizes indicated that

symptom variables displayed the most change during treatment, and a slight recurrence

of problems was in evidence. Object relational variables changed least during treatment

but continued to develop slightly, whereas measures of self-image and interpersonal

problems remained stable during the follow-up period. These results agree with

psychological theory in that symptoms are generally understood to be reactive andcontextually dependent, whereas character or personality can be understood as a

quality of functioning or ability (see e.g. Kopta, Howard, Lowery, & Beutler, 1994). Once

achieved it is either stable or continues to develop.

The present study replicated earlier results indicating that alliance moderates

outcome, whereas it has been hard to identify other consistently significant predictors

of outcome. Gender and treatment format were not associated with change in either

psychiatric symptoms or interpersonal distress, which is in accordance with earlier

findings (Clarkin & Levy, 2004). Treatment duration has sometimes been associated withoutcome and sometimes not (Orlinsky et al., 2004); however, in the current study it was

not predictive of change in either psychiatric symptoms or interpersonal distress.

A shortcoming of the present study is the lack of outcome data during treatment. The

collection of such data would have made it possible for us to evaluate the pattern of

change during the treatment, and it would have given us a possibility to evaluate if it had

been sufficient with shorter treatments than the ones given. An additional limitation of

the present study is the lack of adherence data that would have assured treatment

fidelity.Patients with lower levels of interpersonal distress stayed longer than others in

psychotherapy. Perhaps, patients with less problematic interpersonal experiences

endure the interaction with their psychotherapists better, regardless of treatment

38 Annika Lindgren et al.

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

outcome. An earlier study has reported that the presence of any personality disorder on

the one hand, and higher adaptive defensive style on the other hand, predicted longer

treatments (Perry, Bond, & Roy, 2007).

The tendency towards significant association between patient-rated alliance and

change in psychiatric symptoms during treatment went in the expected direction

(Martin et al., 2000), which supports the validity of the present translation of the HAq-II.The association was lost, however, when we controlled for the patients’ intake level of

psychiatric symptoms. Due to lack of data it was not possible to check whether early

symptom change from intake until the second session predicted the early patient-rated

alliance, which has been observed by others (Barber, Connolly, Crits-Christoph, Gladis,

& Siqueland, 2000; DeRubeis & Feeley, 1990; Feeley, DeRubeis, & Gelfand, 1999). More

studies investigating this pattern of alliance and symptom change are needed, not least

in adolescents and young adults.

Lower therapist-rated alliance predicted greater change in psychiatric suffering forpatients reporting more psychiatric symptoms at intake, which is a paradox. We do not

have other process data that could help us explain the finding, but one plausible

explanation is related to the theory of alliance rupture (eg. Safran, Muran, Samstag, &

Stevens, 2002). Therapeutic alliance has been one of the top issues within the research

community as well as within the clinical society during the last two decades. Low alliance

ratings means that the therapists have identified problematic interactions, which might

havemobilized their effort to solve theproblems. If thiswas the case, itwould also be easier

for the therapists to identify problematic interactions with patients that suffered more.To summarize, we found significant changes in all outcome measures from intake to

follow-up. Changing the perspective from variables to individuals, we found that the

majority of patients reported reduced levels of psychiatric symptoms and interpersonal

problems, and between 35 and 50% displayed statistically significant improvement at

either termination and/or follow-up. This supports the effectiveness of long-term

psychoanalytic psychotherapy for young adults both in the short and the long

perspective. However, a small group of patients got worse from intake to follow-up,

which should not be ignored. Further research on cases reporting no gain or evendeterioration is needed if we as clinicians are to be able to recommend treatment for a

specific patient.

Acknowledgements

The project was supported by grants from the Bank of Sweden, Tercentenary Foundation and the

Research and Development Secretariat at Stockholm County Council. The project has been

approved by the Regional Research Ethics Committee at the Karolinska Institutet, and all

participants have given their informed consent.

References

Adamson, L., & Lyxell, B. (1996). Self-concept and questions of life: Identity development during

late adolescence. Journal of Adolescence, 19, 569–582.

Alden, L. E., Wiggins, J. S., & Pincus, A. L. (1990). Construction of circumplex scales for the

inventory of interpersonal problems. Journal of Personality Assessment, 55, 521–536.

American Psychiatric Association (1994). Diagnostical and statistical manual of mental

disorders (4th ed.). Washington, DC: Author.

Long-term outcome for young adults 39

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

Armelius, K. (2001). Reliabilitet och validitet for den svenska versionen av SASB –

sjalvbildstestet (Reliability and validity for the Swedish version of SASB – the self image

test). Unpublished manuscript.

Barber, J. P., Connolly, M. B., Crits-Christoph, P., Gladis, L., & Siqueland, L. (2000). Alliance predicts

patients’ outcome beyond in-treatment change in symptoms. Journal of Consulting and

Clinical Psychology, 68, 1027–1032.

Baruch, G., & Fearon, P. (2002). The evaluation of mental health outcome at a community-based

psychodynamic psychotherapy service for young people: A 12-month follow-up based on self-

report data. Psychology and Psychotherapy: Theory, Research and Practice, 75, 261–278.

Benjamin, L. S. (1974). Structural analysis of social behavior. Psychological Review, 81, 392–425.

Benjamin, L. S. (1983). The Intrex user the ’s manual, part I and II. Madison, WI: Intrex

Interpersonal Institute, (Computer program manual).

Bjorner, J. B., Sondergaard Kristensen, T., Orth-Gomer, K., Tibblin, G., Sullivan, M., & Westerholm,

P. (1996). Self-rated health: A useful concept in research, prevention and clinical medicine.

No. 96:9. Stockholm: Swedish Council for Planning and Coordination of Research (FRN).

Blatt, S. J., & Auerbach, J. S. (2001). Mental representation, severe psychopathology, and the

therapeutic process. Journal of the American Psychoanalytic Association, 49, 113–159.

Blatt, S. J., Stayner, D. A., Auerbach, J. S., & Behrends, R. S. (1996). Change in object and self-

representations in long-term, intensive, inpatient treatment of seriously disturbed adolescents

and young adults. Psychiatry: Interpersonal and Biological Processes, 59, 82–107.

Blatt, S. J., Wein, S. J., Chevron, E., & Quinlan, D. M. (1979). Parental representations and

depression in normal young adults. Journal of Abnormal Psychology, 88, 388–397.

Blatt, S. J., & Zuroff, D. C. (2005). Empirical evaluation of the assumptions in identifying evidence

based treatments in mental health. Clinical Psychology Review, 25, 459–486.

Blomberg, J., Lazar, A., & Sandell, R. (2001). Long-term outcome of long-term psychoanalytically

oriented therapies: First findings of the Stockholm outcome of psychotherapy and

psychoanalysis study. Psychotherapy Research, 11, 361–382.

Bordin, E. S. (1979). The generalizability of the psychoanalytic concept of the working alliance.

Psychotherapy: Theory, Research and Practice, 16, 252–260.

Brockmann, J., Schluter, T., & Eckert, J. (2006). Langzeitwirkungen psychoanalytischer und

verhaltenstherapeutischer langzeitpsychotherapien: Eine vergleichende studie aus der praxis

niedergelassener psychotherapeuten [Long-term effects of long-term psychoanalytic and long-

term behavior therapy. A comparative study from the general practices of psychotherapists].

Psychotherapeut, 51, 15–25.

Burlingame, G. M., Fuhriman, A., & Mosier, J. (2003). The differential effectiveness of group

psychotherapy: A meta-analytic perspective. Group Dynamics: Theory, Research, and

Practice, 7, 3–12.

Clarkin, J. F., & Levy, K. N. (2004). The influence of client variables on psychotherapy. In M. J.

Lambert (Ed.), Bergin and Garfield’s handbook of psychoterapy and behavior change

(pp. 194–226). New York: Wiley.

Crits-Christoph, P., & Barber, J. P. (2000). Long-term psychotherapy. In C. R. Snyder & R. E. Ingram

(Eds.), Handbook of psychological change: Psychotherapy processes and practices for the

21st century (pp. 455–473). New York: Wiley.

Derogatis, L. R. (1994). Symptom-Checklist-90-R: Administration, scoring and procedures

manual (3rd ed., revised). Minneapolis: National Computer Systems.

Derogatis, L. R., Rickels, K., & Rock, A. F. (1976). The SCL-90 and the MMPI: A step in the validation

of a new self-report scale. British Journal of Psychiatry, 128, 280–289.

DeRubeis, R. J., & Feeley, M. (1990). Determinants of change in cognitive therapy for depression.

Cognitive Therapy and Research, 14, 469–482.

Diamond, D., Blatt, S. J., Steyner, D., & Kaslow, N. (1995). Differentiation-relatedness of self and

object representations. Revised scale. Unpublished research manual, New Haven, CT.

Doidge, N. (1997). Empirical evidence for the efficacy of psychoanalytic psychotherapies and

psychoanalysis: An overview. Psychoanalytic Inquiry, 17(Suppl.), 102–150.

40 Annika Lindgren et al.

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

Dreher, U. (2002). What are the aims of psychoanalysis? In M. Leuzinger-Bohleber & M. Target

(Eds.), Outcomes of psychoanalytic treatment: Perspectives for therapists and researchers

(pp. 17–29). London: Whurr Publishers.

Emde, R. N. (1985). From adolescence to midlife: Remodelling the structure of adult development.

Journal of the American Psychoanalytic Association, 33, 59–112.

Escoll, P. J. (1987). Psychoanalysis of young adults: An overview. Psychoanalytic Inquiry, 7, 5–30.

Evans, C., Margison, F., & Barkham, M. (1998). The contribution of reliable and clinically

significant change methods to evidence-based mental health. Evidence-Based Mental Health,

1, 70–72.

Feaster, D. J., Newman, F. L., & Rice, C. (2003). Longitudinal analysis when the experimenter does

not determine when treatment ends: What is dose–response? Clinical Psychology and

Psychotherapy, 10, 352–360.

Feeley, M., DeRubeis, R. J., & Gelfand, L. A. (1999). The temporal relation of adherence and

alliance to symptom change in cognitive therapy for depression. Journal of Consulting and

Clinical Psychology, 67, 578–582.

Fonagy, P., Roth, A., & Higgitt, A. (2005). Psychodynamic psychotherapies: Evidence-based

practice and clinical wisdom. Bulletin of the Menninger Clinic, 69, 1–58.

Fridell, M., Zvonomir, C., Johansson, M., & Malling Thorsen, S. (2002). SCL-90. Svensk normering,

standardisering och validering av symtomskalan. (Swedish standardization and

validation of SCL-90). Stockholm: Statens Institutionsstyrelse.

Fuhriman, A., & Burlingame, G. M. (1994). Group psychotherapy: Research and practice. In

A. Fuhriman & G. M. Burlingame (Eds.), Handbook of group psychotherapy: An emprical and

clinical synthesis (pp. 3–40). New York: Wiley.

Gruen, R. J., & Blatt, S. J. (1990). Change in self- and object representation during long-term

dynamically oriented treatment. Psychoanalytic Psychology, 7, 399–422.

Heidrich, J., Liese, A. D., Lowel, H., & Kiel, U. (2002). Self-rated health and it’s relation to all-cause

and cardiovascular mortality in southern Germany. Annual Epidemiology, 12, 338–345.

Hjalmdahl, Y., Claesson, M., Werbart, A., & Levander, S. (2001). Bedomning av differentiering-

relaterande i sjalv-och objektrepresentationer I: En validitetsstudie. [Differentiation-

relatedness of self and object relations I: Validity of the Swedish version]. No. 20. Stockholm:

Institute of Psychotherapy and Psychotherapy Section at Karolinska Institutet.

Horowitz, L. M., Rosenberg, S. E., Baer, B. A., Ureno, G., & Villasenor, V. S. (1988). Inventory of

interpersonal problems: Psychometric properties and clinical applications. Journal of

Consulting and Clinical Psychology, 56, 885–892.

Horvath, A. O., & Symonds, B. (1991). Relation between working alliance and outcome in

psychotherapy: A meta-analysis. Journal of Counseling Psychology, 38, 139–149.

Jacobson, N. S., Roberts, L. J., Berns, S. B., & McGlinchey, J. B. (1999). Methods for defining and

determining the clinical significance of treatment effects: Description, application, and

alternatives. Journal of Consulting and Clinical Psychology, 67, 300–307.

Jacobson, N. S., & Truax, P. (1991). Clinical significance: A statistical approach to defining

meaningful change in psychotherapy research. Journal of Consulting and Clinical

Psychology, 59, 12–19.

Johansson, L., & Werbart, A. (2009). Patients’ view of therapeutic action in psychoanalytic group

psychotherapy. Group Analysis, 42, 120–142.

Kazdin, A. E. (2008). Evidence-based treatment and practice: New opportunities to bridge clinical

research and practice, enhance the knowledge base, and improve patient care. American

Psychologist, 63, 146–159.

Kopta, S. W., Howard, K. I., Lowry, J. L., & Beutler, L. E. (1994). Patterns of symptomatic recovery

in psychotherapy. Journal of Consulting and Clinical Psychology, 62, 1009–1016.

Lambert, M. J., Bergin, A. E., & Garfeild, S. L. (2004). Introduction and historical overview.

In M. J. Lambert (Ed.), Bergin and Garfeild’s handbook of psychotherapy and behavior

change (pp. 3–15). New York: Wiley.

Long-term outcome for young adults 41

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

Larsson, D., Hemmingsson, T., Allebeck, P., & Lundberg, I. (2002). Self-rated health and mortality

among young men: What is the relation and how may it be explained? Scandinavian Journal

of Public Health, 30, 259–266.

Leichsenring, F., Biskup, J., Kreische, R., & Staats, H. (2005). The Gottingen study of

psychoanalytic therapy: First results. International Journal of Psychoanalysis, 86, 433–455.

Leichsenring, F., & Rabung, S. (2006). Change norms: A complementary approach to the issue of

control groups in psychotherapy outcome research. Psychotherapy Research, 16, 594–605.

Leichsenring, F., & Rabung, S. (2008). Effectiveness of long-term psychodynamic psychotherapy:

A meta-analysis. JAMA, 300, 1551–1565.

Leichsenring, F., Rabung, S., & Leibing, E. (2004). The efficacy of short-term psychodynamic

psychotherapy in specific psychiatric disorders: A meta-analysis. Archives of General

Psychiatry, 61, 1208–1216.

Leuzinger-Bohleber, M., Stuhr, U., Ruger, B., & Beutel, M. (2003). How to study the ‘quality of

psychoanalytic treatments’ and their long-term effects on patients’ well-being: A

representative, multi-perspective follow-up study. International Journal of Psychoanalysis,

84, 263–290.

Lilliengren, P., & Werbart, A. (2005). A model of therapeutic action grounded in patients’ view of

curative and hindering factors in psychoanalytic psychotherapy. Psychotherapy: Theory,

Research, Practice, Training, 42, 324–339.

Lorig, K., Stewart, A., Ritter, P., Gonzalez, V., Laurent, D., & Lynch, J. (1996). Outcome measures

for health education and other health care interventions. Thousand Oaks, CA: Sage

Publications.

Luborsky, L., Barber, J. P., Siqueland, L., Johnson, S., Najavits, L. M., Frank, A., et al. (1996). The

revised helping alliance questionnaire (Haq-II). Journal of Psychotherapy Practice and

Research, 5, 260–270.

Martin, D. J., Garske, J. P., & Davis, M. (2000). Relation of the therapeutic alliance with outcome

and other variables: A meta-analytic review. Journal of Consulting and Clinical Psychology,

68, 438–450.

Mossey, J. M., & Shapiro, E. (1982). Self-rated health: A predictor of mortality among the elderly.

American Journal of Public Health, 72, 800–808.

Noren, K., Lindgren, A., Hallstrom, T., Thormahlen, B., Vinnars, B., Wennberg, P., et al. (2007).

Psychological distress and functional impairment in patients with personality disorders.

Nordic Journal of Psychiatry, 61, 260–270.

Orlinsky, D. E., Ronnestad, M. H., & Willutzki, U. (2004). Fifty years of psychotherapy process-

outcome research: Continuity and change. In M. J. Lambert (Ed.), Bergin and Garfield’s

handbook of psychotherapy and behavior change (pp. 307–389). New York: Wiley.

Ottosson, H., Bodlund, O., Ekselius, L., Grann, M., von Knorring, L., Kullgren, G., et al. (1998).

DSM-IV and ICD-10 personality disorders: A comparison of a self-report questionnaire (DIP-Q)

with a structured interview. European Psychiatry, 13, 246–253.

Ottosson, H., Bodlund, O., Ekselius, L., von Knorring, L., Kullgren, G., Lindstrom, E., et al. (1995).

The DSM-IV and ICD-10 personality questionnaire (DIP-Q): Construction and preliminary

validation. Nordic Journal of Psychiatry, 49, 285–291.

Perry, J. C., Bond, M., & Roy, C. (2007). Predictors of treatment duration and retention in a study of

long-term dynamic psychotherapy: Childhood adversity, adult personality and diagnosis.

Journal of Psychiatric Practice, 13, 221–232.

Philips, B., Wennberg, P., Werbart, A., & Schubert, J. (2006). Young adults in psychoanalytic

psychotherapy: Patient characteristics and therapy outcome. Psychology and Psychotherapy:

Theory, Research and Practice, 79, 89–106.

Psykologiforlaget (2002). Inventory of interpersonal problems: IIP. Swedish manual. Stockholm:

Psykologiforlaget AB.

Roos, J., Wennberg, P., Philips, B., & Werbart, A. (2009). Young adults who drop-out from long-

term psychoanalytic psychotherapy; results from a prospective longitudinal study.

Manuscript submitted for publication.

42 Annika Lindgren et al.

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

Rosander, M., & Stiwne, D. (1997). Inventory of interpersonal problems, IIP: Konstruktion och

normering av svensk kortversion. [Inventory of interpersonal problems, IIP: Construction

and norms of a Swedish short version]. Unpublished manuscript, Linkopings universitet,

Sweden.

Roseborough, D. J. (2006). Psychodynamic psychotherapy: An effectiveness study. Research on

Social Work Practice, 16, 166–175.

Safran, J. D., Muran, J. C., Samstag, L. W., & Stevens, C. (2002). Repairing alliance ruptures. In J. C.

Norcross (Ed.), Psychotherapy relationships that work: Therapist contributions and

responsiveness to patients (pp. 235–254). New York: Oxford University Press.

Sandell, R., Blomberg, J., Lazar, A., Carlsson, J., Broberg, J., & Schubert, J. (2000). Varieties of long-

term outcome among patients in psychoanalysis and long-term psychotherapy: A review of

findings in the Stockholm outcome of psychoanalysis and psychotherapy project (STOPP).

International Journal of Psychoanalysis, 81, 921–942.

Schneider, W., Buchheim, P., Cierpka, M., Dahlbender, R. W., Freyberger, H. J., Grande, T, et al.

(2002). Operationalized psychodynamic diagnostics: A new approach in psychodynamic

psychotherapy. In L. E. Beutler & M. L. Malik (Eds.), Rethinking the DSM: A psychological

perspective. Decade of behavior (pp. 177–200). Washington, DC: American Psychological

Association.

Schwartz, J. C. (1993). The mixed-model anova: The truth the computer packages, the books. The

American Statistician, 47, 48–59.

Shadbolt, B., Barresi, J., & Craft, P. (2002). Self-rated health as a predictor of survival among

patients with advanced cancer. Journal of Clinical Oncology, 20, 2514–2519.

Statistics Sweden (2006). Health and medical care 1980–2005 (No. 113).

Weinryb, R. M., Gustavsson, J. P., Hellstrom, C., Andersson, E., Broberg, A., & Rylander, G. (1996).

Interpersonal problems and personality characteristics: Psychometric studies of the Swedish

version of the IIP. Personality and Individual Differences, 20, 13–23.

Westen, D., Novotny, C. M., & Thompson-Brenner, H. (2004). The empirical status of empirically

supported psychotherapies: Assumptions, findings and reporting in controlled clinical trials.

Psychological Bulletin, 130, 631–663.

Received 24 January 2008; revised version received 15 May 2009

Long-term outcome for young adults 43