Prevalence and associated behavioral symptoms ofdepression in mild cognitive impairment and dementia dueto Alzheimer’s disease

Stefan Van der Mussele1,2, Kim Bekelaar3, Nathalie Le Bastard1, Yannick Vermeiren1, Jos Saerens3, Nore Somers3,Peter Mariën3, Johan Goeman3, Peter P. De Deyn1,3,4,5 and Sebastiaan Engelborghs1,3

1Reference Center for Biological Markers of Dementia (BIODEM), Laboratory of Neurochemistry and Behavior, Institute Born-Bunge,University of Antwerp, Antwerp, Belgium2Department of Nursing and Midwifery Sciences, Faculty of Medicine and Health Sciences, University of Antwerp, Antwerp, Belgium3Department of Neurology and Memory Clinic, Hospital Network Antwerp (ZNA), Middelheim and Hoge Beuken, Antwerp, Belgium4Department of Health Care Sciences, Artesis University College Antwerp, Antwerp, Belgium5Department of Neurology and Alzheimer Research Center, University Medical Center Groningen, Groningen, The NetherlandsCorrespondence to: S. Engelborghs, MD, PhD, E-mail: Sebastiaan.Engelborghs@ua.ac.be

Background:Mild cognitive impairment (MCI) is a clinical concept that categorizes subjects who are inan intermediate cognitive state between normal aging and dementia. The aims of this study are todetermine the prevalence of significant depressive symptoms in MCI and Alzheimer’s disease (AD)patients and to characterize the behavior associated with significant depressive symptoms in MCI andAD patients.

Methods: A cross-sectional analysis of baseline data from a prospective, longitudinal study on behavioralsymptoms of dementia and MCI was performed. The study population consisted of 270 MCI and 402AD patients. Behavioral assessment was performed by means of Middelheim Frontality Score, BehavioralPathology in Alzheimer’s Disease Rating Scale (Behave-AD) and Cohen-Mansfield Agitation Inventory.The presence of significant depressive symptoms was defined as a Cornell Scale for Depression in Dementiatotal score >7.

Results: The prevalence of significant depressive symptoms in AD patients (25%) was higher compared withMCI patients (16%) (p=0.005). Patients with significant depressive symptoms showed an increased severityof frontal lobe symptoms, behavioral symptoms and agitation (Middelheim Frontality Score, Behave-AD andCohen-Mansfield Agitation Inventory total scores; p< 0.001). Also, most of the individual frontal lobe andbehavioral symptoms were more prevalent and severe, resulting in higher Behave-AD global scores. Mildcognitive impairment patients with depressive symptoms showed more severe behavioral symptoms andmore severe verbally agitated behavior than AD patients without depressive symptoms (p< 0.001).

Conclusions: Frontal lobe and behavioral symptoms are more prevalent and severe in MCI and ADpatients with significant depressive symptoms as compared with patients without depressive symptoms.Copyright # 2012 John Wiley & Sons, Ltd.

Key words: depression; depressive symptoms; behavioral symptoms; frontal lobe symptoms; mild cognitive impairment;Alzheimer’s disease; dementia; Cornell Scale

History: Received 14 May 2012; Accepted 24 October 2012; Published online 17 December 2012 in Wiley Online Library(wileyonlinelibrary.com)DOI: 10.1002/gps.3909

Introduction

Mild cognitive impairment (MCI) is a clinical conceptthat identifies subjects who are in an intermediate state

between normal aging and dementia. MCI is a syn-drome characterized by an impairment of memoryor other cognitive decline, which does not affect aperson’s basic activities of daily living, whereas the

Copyright # 2012 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2013 28: 947–958

RESEARCH ARTICLE

complex instrumental functions may be minimallyimpaired. The clinical presentation, etiology andoutcome of MCI are heterogeneous (Petersen, 2004;Winblad et al., 2004). MCI can be divided into twosubtypes: an amnestic subtype with memory deficitsand a non-amnestic subtype with a cognitive declineother than memory. The subtypes can be furtherspecified on the basis of cognitive impairment in a‘single domain’ or in ‘multiple domains’.

The prevalence rate of MCI in population-basedstudies ranges from 3% to 19%, and higher rates areassociated with increasing age and lower education(Frisoni et al., 2000; Ritchie et al., 2001; Hanninenet al., 2002; Fisk et al., 2003; Tognoni et al., 2005;Das et al., 2007). Most of the amnestic MCI patientshave Alzheimer’s disease (AD) in a predementia stage(Albert et al., 2011) and progress to dementia at a rateof 10–15% yearly in comparison with approximately1–2% a year in healthy older adults (Petersen et al.,2001). Mitchell and Shiri-Feshki (2009) confirmed thesedata in ameta-analysis, but different criteria and differentsampling and assessment methods explain the few andinconsistent findings (Portet et al., 2006).

Patients with amnestic MCI are likely to havesymptomatic, predementia AD and are prone to prog-ress to dementia (Albert et al., 2011), whereas theoutcomes of non-amnestic MCI appear to be moreheterogeneous, including vascular dementia, fronto-temporal dementia and dementia with Lewy bodies(Petersen, 2004). However, not all MCI patientsprogress to dementia, and some recover to normalcognition (Winblad et al., 2004). This can at leastpartially be explained by the fact that older adults withdepression and cognitive symptoms were diagnosed asMCI in some studies (Dierckx et al., 2007).

Although behavioral symptoms are common in MCIwith reported prevalence ranging from 35% to 85%, be-havioral symptoms are not included in the concept ofMCI (Monastero et al., 2009). There is evidence that be-havioral features, particularly depression, may representrisk factors for MCI or predictors for the progression ofMCI to dementia due to AD (Monastero et al., 2009).

The prevalence of depression or depressive symp-toms in MCI and AD varies as a result of differentdiagnostic criteria and different patient selection andassessment procedures. The reported prevalence ofdepression in ADpatients clusters around approximately30% (Olin et al., 2002; Lee and Lyketsos, 2003). Themedian prevalence of depression in MCI patients is16% (3–63%) in population-based and 44% (9–83%)in hospital-based studies (Panza et al., 2010).

Research linking depression in dementia to otherbehavioral symptoms is limited (Menon et al., 2001;

Starkstein et al., 2005; Prado-Jean et al., 2010). Toour knowledge, the relation between depression inMCI and other behavioral symptoms is not investi-gated. Therefore, we performed an analysis of baselinedata of an ongoing large prospective, longitudinalstudy on behavioral and psychological signs andsymptoms of dementia and MCI aiming to determinethe prevalence of significant depressive symptoms inMCI and AD patients and to characterize behavioraldisturbances in MCI and AD patients with and with-out significant depressive symptoms.

Methods

Study population and diagnostic criteria

The study population consisted of MCI patients(n= 270) and AD patients (n= 402).

Patients were included at the moment of their diag-nostic work-up in our memory clinic. The diagnosticwork-up consisted of a general physical and neurologicalexamination, blood screening, structural neuroimagingconsisting of brain magnetic resonance imaging or, ifnot feasible, brain computerized tomography, functionalbrain imaging (single photon emission computedtomography), standard electroencephalogram and anextensive time-linked (�3months) neuropsychologicalexamination with adjustment for age and education,comprising amongst others the Mini Mental StateExamination (Folstein et al., 1975) and the WechslerMemory Scale III (The Psychological Corporation,1998), Hierarchic Dementia Scale (Cole and Dastoor,1987) and/or Repeatable Battery for the Assessment ofNeuropsychological Status (Randolph et al., 1998).

To diagnoseMCI, Petersen’s diagnostic criteria (2004)were applied. As all cognitive domains of subjects weretested in an extensive time-linked (�3months) neuro-psychological examination, all patients received adiagnosis of amnestic/non-amnestic, single/multipledomain MCI.

Probable AD was diagnosed according to NationalInstitute of Neurological and Communicative Disor-ders and Stroke/Alzheimer’s Disease and RelatedDisorders Association (NINCDS/ADRDA) criteria(McKhann et al., 1984), although all patients as wellfulfilled the DSM-IV criteria (American PsychiatricAssociation, 1994).

Clinical and neuropsychological follow-up of in-cluded patients and autopsy in deceased consentedpatients during follow-up period did contribute tothe diagnostic accuracy of the subjects in this study(Le Bastard et al., 2012).

948 S. Van der Mussele et al.

Copyright # 2012 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2013 28: 947–958

Staging of cognitive deterioration was assessed bymeans of the Global Deterioration Scale (Reisberget al., 1982). Age at disease onset was estimated bythe clinician following an interview with the patient’smain caregiver. In case a non-professional caregiverwas lacking, the patient’s main professional caregiverwas contacted and interviewed.

The local ethics committee approved this study. Allpatients and/or patients’ caregivers gave written in-formed consent. All patients were of Caucasian origin.

Depressive symptoms

Depressive symptoms were assessed by means of theCornell Scale for Depression in Dementia (CSDD), a19-item depression scale (Alexopoulos et al., 1988a).Item scores range from 0 (absent) to 2 (severe), witha maximum total score of 38 points. The items areclustered in five groups: (i) Mood-related signs: anxi-ety, sadness and lack of reactivity to pleasant events;(ii) Behavioral disturbances: agitation, retardation,multiple physical complaints and loss of interest; (iii)Physical signs: appetite loss, weight loss and lack ofenergy; (iv) Cyclic functions: diurnal variation of mood,difficulty falling asleep, multiple awakenings during sleepand early morning awakening; and (v) Ideationaldisturbances: suicide, poor self-esteem, pessimism andmood-congruent delusions. The presence of significantdepressive symptoms was defined as a total CSDD score>7 (Burns et al., 2004). Studies have shown the CSDDto be valid for screening depression in non-dementedpatients too (Alexopoulos et al., 1988b).

Behavioral assessment

All subjects underwent behavioral assessment at inclu-sion (baseline). Behavioral assessment, by interviewingboth patient and caregiver, covered a period of 2weeksprior to inclusion, using a battery of behavioral assess-ment scales: Middelheim Frontality Score (MFS),Behavioral Pathology in Alzheimer’s Disease Rating Scale(Behave-AD), Cohen-Mansfield Agitation Inventory(CMAI) and CSDD.

The MFS is a validated clinical and behavioralassessment scale that measures frontal lobe features andreliably discriminates frontotemporal dementia fromAD patients with a sensitivity and specificity of almost90% and with good inter-rater and intra-rater reliability(De Deyn et al., 2005; Aries et al., 2010). According tothe Instructions for Administration and Scoring, theMFS was rated by the clinician or researcher and wasobtained by summating scores in a standardized fashionon 10 items. Each item was scored either zero (absent)

or one (present), yielding a total maximal score of 10.The 10 items scored are noted in Figure 1.

The Behave-AD is a 25-item scale that measures be-havioral symptoms in seven clusters (Figure 2), scoredon a four-point scale of increasing severity (Reisberget al., 1987). A psychosis cluster score is calculated bysummating the scores of the paranoid/delusionalideation (delusions) and hallucinations clusters. Besidesa total score, a global score on a four-point scale of in-creasing severity is provided, reflecting how troubling tothe caregiver or dangerous to the patient the behavioralsymptoms are, from not troubling or dangerous (score0) to severe (score 3). The total severity of behavioralsymptomswas divided into three categories: no symptoms(Behave-AD total score 0), mild symptoms (Behave-ADtotal scores 1–7) and moderate to severe symptoms(Behave-AD total score ≥8) (Engelborghs et al., 2005).

The CMAI assesses 29 agitated behaviors on a seven-point scale of increasing frequency (1=never; 7= severaltimes an hour) (Cohen-Mansfield, 1996). CMAI clusterscores include aggressive behavior (10 items), physicallynon-aggressive behavior (11 items) and verbally agitatedbehavior (8 items); a total score is provided as well.

Dutch versions of the MFS and CMAI have been val-idated (de Jonghe and Kat, 1996; De Deyn et al., 2005).The MFS has been developed in a Dutch setting. TheBehave-AD and CSDD are not used to interview, butrather as a guidance to detect behavioral symptoms.

Statistical analyses

The behavioral assessment scales provide semi-continuousvariables. Therefore, non-parametric statistics were used:Mann–Whitney U test was applied for (semi) continuousvariables and Chi-square statistics for categorical data.

To reduce the type I error rate, probability levels of0.01 were considered significant. Statistical analyses werecarried out using SPSS (IBM Corporation, Armonk, NY)Statistics 17.0.

Results

Demographic, clinical and neuropsychologicalcharacteristics

The MCI group consisted of 53 amnestic single domain,150 amnestic multiple domain, 29 non-amnestic singledomain and 38 non-amnestic multiple domain MCIpatients. The AD group consisted of 35 definite and367 probable AD patients.

In AD patients, the male–female ratio was higher inthe subgroup with depressive symptoms.

949Depressive symptoms in MCI and Alzheimer’s disease

Copyright # 2012 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2013 28: 947–958

In bothMCI and AD patients, the Global DeteriorationScale score was higher in the group with depressivesymptoms.

Mild cognitive impairment patients with depres-sive symptoms were more frequently treated withhypnotics, sedatives, anxiolytics and antiparkinso-nian agents (Table 1).

Prevalence of depressive symptoms

The prevalence of depressive symptoms in AD patients(25%) was higher than the prevalence of depressivesymptoms in MCI patients (16%) (p=0.005). No differ-ence in prevalence of depressive symptoms was foundamong the four subtypes of MCI (p=0.582) (Table 1).

Figure 1 Prevalence (%) of frontal lobe symptoms in MCI and AD patients. Chi-square statistics were used, comparing the groups with and withoutdepressive symptoms within the MCI and AD groups. Significant differences are mentioned as follows: *p< 0.01, **p< 0.001. Abbreviations: MFS,Middelheim Frontality Score; MCI, mild cognitive impairment; AD, Alzheimer’s disease.

950 S. Van der Mussele et al.

Copyright # 2012 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2013 28: 947–958

Behavior in mild cognitive impairment patients with de-pressive symptoms

Middelheim Frontality Score. The prevalence and se-verity of frontal lobe symptoms in MCI patients withand without depressive symptoms are displayed inFigure 1 and Table 2.

The severity of frontal lobe symptoms (as measuredby the MFS total score) was higher in MCI patientswith depressive symptoms, and the following MFS

items were significantly more prevalent: impaired con-trol of emotions, euphoria and emotional bluntness;restlessness; and disinhibition.

Behavioral Pathology in Alzheimer’s Disease RatingScale. The prevalence and severity of behavioralsymptoms in MCI patients with and without depres-sive symptoms are displayed in Figure 2 and Table 2.

In 45% of MCI patients with depressive symptoms,moderate to severe behavioral symptoms were present,

Figure 2 Prevalence (%) of behavioral symptoms in MCI and AD patients. Chi-square statistics were used, comparing the groups with and withoutdepressive symptoms within the MCI and AD groups. Significant differences are mentioned as follows: *p< 0.01, **p< 0.001. Abbreviations: MCI, mildcognitive impairment; AD, Alzheimer’s disease.

951Depressive symptoms in MCI and Alzheimer’s disease

Copyright # 2012 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2013 28: 947–958

Table1

Dem

ograph

ic,clin

ical

andne

urop

sycholog

ical

data

MCI

AD

Total

Dep

ressivesymptoms

Total

Dep

ressivesymptoms

No

Yes

No

Yes

n=27

0n=22

8n=42

n=40

2n=30

3n=99

%84

.4%

15.6%

75.4%

24.6%

Male/female

120/15

010

5/12

315

/27

132/27

090

/213

42/57

Ageat

inclus

ion(yea

rs)

75.5�8.1

(50–94

)75

.7�7.9

(50–94

)74

.7�9.0

(54–90

)80

.1�7.6

(51–97

)80

.3�7.1

(51–97

)79

.5�8.9

(54–97

)

Ageat

ons

et(yea

rs)

73.2�8.3

(48–90

)73

.3�8.1

(48–90

)72

.5�9.5

(53–88

)76

.8�8.4

(48–96

)76

.9�7.9

(48–96

)76

.3�9.8

(50–95

)

Disea

seduration(yea

rs)

2.2�1.7

(0–11

)2.2�1.8

(0–11

)2.1�1.5

(0–7)

3.3�2.5

(0–16

)3.3�2.5

(0–16

)3.2�2.4

(0–12

)

Global

Deterioratio

nSca

le(1–7

)3.0�0.7

(1–5)

3.0�0.7

(1–5)

3.3�0.7

(2–5)

5.1�1.0

(2–7)

5.1�1.0

(2–7)

5.3�0.9

(3–7)

MiniM

entalS

tate

Exa

minationsc

ore

(/30

)25

.6�2.9

(16–30

)25

.7�2.9

(16–30

)25

.5�2.9

(18–30

)15

.2�6.3

(0–30

)15

.5�6.2

(0–30

)14

.3�6.7

(0–26

)

Freeofpsych

otrop

icmed

ication(%

)52

.055

.632

.5*

16.7

18.2

12.2

Antidep

ressan

ts(%

)25

.122

.738

.538

.836

.346

.5

Antipsych

otic

s(%

)5.0

3.7

12.2

45.3

42.5

53.5

Hyp

notic

s,se

datives

andan

xiolytics(%

)29

.125

.846

.3*

23.6

22.4

27.3

Cho

lines

terase

inhibito

rs(%

)3.4

3.6

2.4

36.8

38.4

32.0

Antiparkins

onian

agen

ts(%

)2.7

1.4

9.8*

1.5

1.7

1.0

Antiepileptic

s(%

)2.3

2.7

0.0

1.5

1.7

1.0

AD,A

lzhe

imer’sdisease;

MCI,mild

cogn

itive

impairmen

t.Dataaregivenas

ratio

,percentageor

mean�SD

with

rang

esrepresen

tedbetw

eenbrackets.

Sign

ificant

diffe

rences

comparing

thegrou

pswith

andwith

outd

epressivesymptom

swith

intheMCIgrou

pandwith

intheAD

grou

paremen

tione

das

follo

ws:

*p<0.01

,**p<

0.00

1.Fo

rcomparisonof

male–femaleratio

sandpercen

tages,Chi-squ

arestatistic

swereused.

Forothe

rcomparisons,M

ann–

Whitney

Utestwas

applied.

952 S. Van der Mussele et al.

Copyright # 2012 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2013 28: 947–958

and in 55%, mild symptoms occurred that was in contrast(p< 0.001) toMCI patients without depressive symptomsin whom behavioral symptoms were less prevalent (mod-erate to severe: 7%; mild: 74%; no symptoms: 19%).

Behavioral symptoms were more prevalent andsevere in MCI patients with depressive symptoms,except for delusions and hallucinations. The three mostprevalent behavioral symptoms were anxiety/phobias,aggressiveness and diurnal rhythm disturbances.

In addition to the prevalence and severity of behav-ioral symptoms, the Behave-AD global score, indicat-ing how troubling to the caregiver and/or dangerousto the patient the behavioral symptoms are, is higherin MCI patients with depressive symptoms. Consider-ing the MCI patients with depressive symptoms, 63%is at least mildly troubling or dangerous, as comparedwith 29% in MCI patients without depressivesymptoms.

Table 2 Severity of behavioral symptoms

MCI AD

(n=270) (n=402)

Depressive symptoms Depressive symptoms

No Yes No Yes

n=228 n=42 n=303 n=99

MFS 1.8�1.4 3.3�1.8 2.9�1.7 4.5�1.7Total score (0–8) (0–7)** (0–9) (0–8)**Behave-AD 0.2�0.9 0.4�0.9 1.1�2.1 2.0�3.4Delusions (0–9) (0–4) (0–13) (0–20)*Behave-AD 0.2�0.8 0.1�0.5 0.3�0.9 0.6�1.3Hallucinations (0–6) (0–2) (0–7) (0–6)*Behave-AD 0.4�1.5 0.5�1.2 1.3�2.4 2.6�4.2Psychosis (delusions+ hallucinations) (0–15) (0–6) (0–13) (0–26)**Behave-AD 0.2�0.6 0.5�0.9 1.2�1.7 2.6�2.3Activity disturbances (0–3) (0–4)* (0–8) (0–9)**Behave-AD 1.1�1.5 2.6�2.3 1.8�2.4 3.7�2.8Aggressiveness (0–7) (0–9)** (0–9) (0–9)**Behave-AD 0.3�0.5 1.1�1.0 0.3�0.6 1.0�1.0Diurnal rhythm disturbances (0–3) (0–3)** (0–3) (0–3)**Behave-AD 0.6�1.0 1.9�1.1 0.5�1.0 1.7�1.5Affective disturbances (0–5) (0–5)** (0–5) (0–6)**Behave-AD 0.5�0.9 1.8�1.7 0.4�0.8 1.0�1.3Anxiety/phobias (0–4) (0–8)** (0–5) (0–5)**Behave-AD 3.0�2.7 8.4�4.7 5.6�5.2 12.6�7.5Total score (0–17) (2–20)** (0–24) (1–47)**Behave-AD 0.4�0.7 1.0�0.9 1.0�0.9 1.7�0.9Global score (0–3) (0–3)** (0–3) (0–3)**CMAI 10.0�0.2 10.3�1.1 11.0�3.7 13.1�6.6Aggressive behavior (10–13) (10–16) (10–38) (10–45)**CMAI 11.9�2.0 14.4�4.5 15.4�5.7 21.3�8.2Physically non-aggressive behavior (11–24) (11–28)** (11–38) (11–48)**CMAI 10.4�3.2 15.3�7.4 11.4�4.9 17.2�6.8Verbally agitated behavior (8–30) (8–36)** (8–40) (8–34)**CMAI 32.3�4.1 40.0�11.5 37.7�10.5 51.6�15.5Total score (29–60) (29–72)** (29–102) (29–99)**Cornell Scale for Depression 3.1�2.1 12.1�5.2 3.3�2.3 11.0�3.1Total score (0–7) (8–31)** (0–7) (8–25)**

AD, Alzheimer’s disease; Behave-AD, Behavioral Pathology in Alzheimer’s Disease Rating Scale; CMAI, Cohen-Mansfield Agitation Inventory;MCI, mild cognitive impairment; MFS, Middelheim Frontality Score.Data are given as ratio, percentage or mean�SD with ranges represented between brackets.Significant differences comparing the groups with and without depressive symptoms within the MCI group and within the AD group are mentionedas follows:*p< 0.01,**p< 0.001.For all comparisons, Mann–Whitney U test was applied.

953Depressive symptoms in MCI and Alzheimer’s disease

Copyright # 2012 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2013 28: 947–958

Cohen-Mansfield Agitation Inventory. The frequency ofagitated behavior was higher in MCI patients with depres-sive symptoms, especially for the clusters ‘physically non-aggressive’ and ‘verbally agitated’ behavior (Table 2).

Importance of psychotropic drug intake. Mild cognitiveimpairment patients were categorized as untreatedpatients when no psychotropic drugs were taken, ex-cept for cholinesterase inhibitors.

The prevalence of depressive symptoms was higher(p=0.005) in the treated MCI subgroup (n=120) thanin the untreated subgroup (n=134), respectively 23%and 10%.

Treated (n=27) and untreated (n=13)MCI patientswith depressive symptoms did not differ with regard tofrontal lobe (MFS total score; p=0.418) and behavioral(Behave-AD total score; p=0.955) symptoms, agitatedbehavior (CMAI total score; p=0.648) and being trou-bling to the caregiver and/or dangerous to themselves(Behave-AD global score; p=0.566).

Behavior in Alzheimer’s disease patients with depressivesymptoms

Middelheim Frontality Score. The prevalence andseverity of frontal lobe symptoms in AD patients withand without depressive symptoms are displayed inFigure 1 and Table 2.

The severity of frontal lobe symptoms (as measuredby the MFS total score) was higher in AD patients withdepressive symptoms, and the following MFS items weresignificantly more prevalent: restlessness; impairedcontrol of emotions, euphoria and emotional bluntness;disinhibition; aspontaneity; and stereotyped behavior.

Behavioral Pathology in Alzheimer’s Disease RatingScale. The prevalence and severity of behavioralsymptoms in AD patients with and without depressivesymptoms are displayed in Figure 2 and Table 2.

In 74% of AD patients with depressive symptoms,moderate to severe behavioral symptoms were present,and in 26%,mild symptoms occurred that was in contrast(p< 0.001) toMCI patients without depressive symptomsin whom behavioral symptoms were less prevalent (mod-erate to severe: 28%; mild: 53%; no symptoms: 19%).

All behavioral symptoms were more severe and moreprevalent in AD patients with depressive symptoms. Thethreemost prevalent behavioral symptomswere aggressive-ness, activity disturbances and diurnal rhythmdisturbances.

In addition to the prevalence and severity of behav-ioral symptoms, the Behave-AD global score is higherin AD patients with depressive symptoms. Considering

the AD patients with depressive symptoms, 90% is atleast mildly troubling or dangerous, as compared with67% in MCI patients without depressive symptoms.

Cohen-Mansfield Agitation Inventory. The frequency ofagitated behavior was higher in AD patients with de-pressive symptoms, also for ‘aggressive’ behavior,which was not the case in MCI patients (Table 2).

Importance of psychotropic drug intake. Alzheimer’sdisease patients were categorized as untreated patientswhen no psychotropic drugs were taken, except forcholinesterase inhibitors.

There was no difference (p= 0.079) in prevalence ofdepressive symptoms between the treated (n= 276)and untreated (n= 126) AD subgroups with respec-tively 27% and 19%.

Treated (n= 75) and untreated (n= 24) AD patientswith depressive symptoms did not differ with regardto frontal lobe (MFS total score; p= 0.464) and behav-ioral (Behave-AD total score; p= 0.641) symptoms,agitated behavior (CMAI total score; p= 0.362) andbeing troubling to the caregiver and/or dangerous tothemselves (Behave-AD global score; p= 0.282).

Comparison of associated behavior across subgroups

Alzheimer’s disease patients with depressive symptomsdisplayed more severe frontal lobe symptoms, behav-ioral symptoms and agitated behavior (aggressive andphysically non-aggressive) as compared with MCIpatients with depressive symptoms (p< 0.001). ADpatients without depressive symptoms displayed moresevere frontal lobe symptoms, behavioral symptomsand agitated behavior as compared with MCI patientswithout depressive symptoms (p< 0.001) (Figure 3).

Remarkably, we found that MCI patients with depres-sive symptoms showed more severe behavioral symp-toms than AD patients without depressive symptoms(p< 0.001). This finding remained statistically significanteven after removing the cluster score affective distur-bance from the Behave-AD total score (p< 0.009).MCI patients with depressive symptoms showed alsomore severe verbally agitated behavior than AD patientswithout depressive symptoms (p< 0.001).

Discussion

Strengths and limitations of the study

Only few studies dealt with the difference in behav-ioral symptoms between dementia patients with and

954 S. Van der Mussele et al.

Copyright # 2012 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2013 28: 947–958

without depressive symptoms (Menon et al., 2001;Starkstein et al., 2005; Prado-Jean et al., 2010). Onlyone of these studies focused on AD patients (Starksteinet al., 2005), whereas the others investigated dementiapatients in general, diagnosed by DSM-III (Menonet al., 2001) and DSM-IV (Prado-Jean et al., 2010) crite-ria. To our knowledge, this study is the first to investigatethe difference in behavior in MCI patients with andwithout depressive symptoms.

This study has additional strengths. First, the study in-cluded large and well-characterized MCI and AD patientgroups that have been diagnosed following strict applica-tion of clinical diagnostic criteria. Second, as the datacame from a prospective, longitudinal study, the follow-up of included subjects and AD autopsy confirmation(n=35) contributed to increased diagnostic accuracy.Third, all patients were diagnosed by clinicians in thesame center in order to preserve homogeneity in the di-agnostic groups. Fourth, two previous studies also usedthe CSDD for quantification of depressive symptoms,which facilitates comparing our results in dementiapatients (Menon et al., 2001; Prado-Jean et al., 2010).Lastly, we repeatedly found homogeneous behavioramong the four MCI subtypes of this MCI population(Van der Mussele et al., 2012). However, the absence ofdifferences in behavior comparing MCI subtypes mightbe due to limited statistical validity, because of small sub-groups of non-amnestic MCI patients included.

Three limitations to the study are known to theauthors. First, our study population was recruited ina memory clinic, which might have introduced a selec-tion bias as behavioral symptoms might have contrib-uted to referral. Second, this study dealt withsignificant depressive symptoms by applying a CSDDcut-off of >7, whereas a diagnosis of major depression

by a psychiatrist, using the DSM-IV criteria, mighthave been more accurate. Third, behavioral assess-ment raters were not blinded for the subjects belong-ing to the AD, MCI or control group.

Prevalence of depressive symptoms

Our figures on the prevalence of depressive symptomscorrespond with previous research on the prevalenceof depression in MCI and AD. With a prevalence of25% in AD patients, our result approximates the30% prevalence of depression of previous studies(Olin et al., 2002; Lee and Lyketsos, 2003). With aprevalence of 16% in MCI patients, our outcomematches the 16% median prevalence in population-based studies (Panza et al., 2010) despite the fact thatrecruitment took place in a memory clinic, includinginpatients and outpatients.

Associated behavioral symptoms

Our preceding research showed a high prevalence offrontal lobe and behavioral symptoms in MCI andAD (Van der Mussele et al., 2012). Moderate to severebehavioral symptoms were observed in 13% of theMCI patients and in 39% of the AD patients. Thepresent study demonstrated an even higher prevalenceof moderate to severe behavioral symptoms within thegroups of MCI and AD patients with depressivesymptoms as compared with the subgroups withoutsignificant depressive symptoms. Moreover, severalfrontal lobe and behavioral symptoms were moreprevalent and more severe in the diagnostic groupswith depressive symptoms.

Figure 3 Severity of frontal lobe symptoms, behavioral symptoms and agitation in MCI and AD patients. For all comparisons, Mann–Whitney U testwas applied. Abbreviations: MCI, mild cognitive impairment; AD, Alzheimer’s disease. (a) MCI patients without depressive symptoms, (b) MCI patientswith depressive symptoms, (c) AD patients without depressive symptoms and (d) AD patients with depressive symptoms.

955Depressive symptoms in MCI and Alzheimer’s disease

Copyright # 2012 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2013 28: 947–958

These findings correspond to former studies on therelationship between depression and behavioral symp-toms in dementia. Starkstein et al. (2005) showed thatcertain non-characteristic symptoms were increasedamong AD patients with depression. Menon et al.(2001) found that aggression is closely linked withsevere depression in dementia patients. Prado-Jeanet al. (2010) showed that the prevalence of all Neuro-psychiatric Inventory items was higher in dementiapatients with depression. The behavioral symptoms‘agitation’, ‘anxiety’ and ‘irritability’ were mostprevalent, which corresponds partially to our findingswith the highest prevalence of ‘aggressiveness’, ‘activitydisturbances’ and ‘diurnal rhythm disturbances’ in theAD group with depressive symptoms.

Associated frontal lobe symptoms

As diurnal rhythm disturbances is one of the coresymptoms in patients with depression (American Psy-chiatric Association, 1994; Fava, 2004; Kupfer, 2006;Quera-Salva et al., 2010) and anxiety and depressionfrequently co-exist (Devane et al., 2005), it is not sur-prising to find a higher prevalence of diurnal rhythmdisturbances and anxiety in MCI and AD patients withdepressive symptoms. Nonetheless, the higher preva-lence of aggressiveness and activity disturbances inboth MCI and AD patients with depressive symptomsis more difficult to understand. Furthermore, also theprevalence of frontal lobe symptoms is increased inpatients with depressive symptoms, especially restless-ness and disinhibition.

The following hypothesis might explain theincreased prevalence and severity of frontal lobe symp-toms, including aggressiveness and activity disturbancesin depressed MCI and AD patients: decreasedserotonergic activity has been related to depression(Coppen, 1967), and dopaminergic neurons are modu-lated by serotonergic innervation (Di Giovanni et al.,2010). Indeed, the serotonergic system appears to havean inhibitory effect on the dopaminergic function(Engelborghs et al., 2008; Di Giovanni et al., 2010).Furthermore, ascending dopaminergic pathways are partof the frontal–subcortical circuitry (Bonelli andCummings, 2007), and the dopaminergic system is oneof the important modulators of frontal lobe function(Goldman-Rakic et al., 1990). Consequently, frontal–subcortical circuit dysfunction leads to impairedexecutive functions, apathy and impulsivity (Bonelliand Cummings, 2007), and might as well explain theprevalence and severity of frontal lobe symptoms inpatients with depressive symptoms. In our study,

disinhibition and restlessness are closely related to impul-sivity. Furthermore, the MFS item ‘impaired control ofemotions, euphoria and emotional bluntness’ involvesemotional bluntness and regulation of emotional control,including euphoria. The emotional control meant by thisitemmay be impaired because of depression.Nevertheless,the emotional bluntness in this item is closely related toapathy. Accordingly, our study found a relation betweenimpulsivity, apathy and depression as a probable conse-quence of the frontal–subcortical circuit dysfunction.

Psychotropic drug intake and associated behavior indepression

First of all, there is no difference in associated behaviorbetween the treated (psychotropic drugs) and un-treated depressed MCI or AD patients. With un-treated, we mean no intake of psychotropic drugs,except for cholinesterase inhibitors. The prevalenceof depressive symptoms was the same in treated anduntreated AD patients. However, the prevalence of de-pressive symptoms was higher in MCI patients treatedwith psychotropic drugs compared with untreatedMCI patients. Furthermore, most psychotropic drugsin our study population were prescribed for treatingassociated behavioral symptoms with exception forcholinesterase inhibitors that were only prescribed assymptomatic treatment in case of biological evidencefor AD. Indeed, assessing both the need for and effi-cacy of these medications would be outside the scopeof this current study, even though it is an interestingresearch question. Second, psychotropic drug intakerates vary among countries, and Belgium is known tohave a high prevalence of psychotropic drug utiliza-tion in community-dwelling older adults and nursinghome residents (Azermai et al., 2011). This also par-tially explains the high prevalence of psychotropicdrug intake in MCI and AD patients in this study.

Implications for practice

A systematic behavioral screening policy in inpatientand outpatient services could increase the quality ofcare for patients with cognitive deterioration as behav-ioral symptoms are frequent in MCI and AD (Van derMussele et al., 2012), and behavioral and psychologicaldementia-related problems such as aggression, depres-sion and hallucinations are consistent predictors forinstitutionalization (Luppa et al., 2008; Gaugler et al.,2009). Specific tools and training might be valuablein the implementation of a systematic behavioralscreening policy.

956 S. Van der Mussele et al.

Copyright # 2012 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2013 28: 947–958

(Non-) pharmacological treatment for depressionmight also decrease associated behavioral symptoms.Nonetheless, also the opposite approach, the treat-ment of associated behavioral symptoms might reducedepression in patients with cognitive deterioration.Both approaches could reduce eventually the risk ofinstitutionalization.

Implications for research

Further research would assist in understanding of thepossible inter-relation of associated behavioral symp-toms in depression. Our results suggest, for instance,a depressive syndrome including anxiety, aggressive-ness and diurnal rhythm disturbances.

Conclusion

The prevalence of significant depressive symptoms inAD patients was higher than the prevalence of signifi-cant depressive symptoms in MCI patients, respec-tively 25% and 16%.

Patients with significant depressive symptoms inthe MCI and AD groups displayed more and alsomore severe frontal lobe and behavioral symptoms.Depression in AD is accompanied with even moreprevalent and more severe frontal lobe and behavioralsymptoms than in MCI. And MCI patients withsignificant depressive symptoms showed more severebehavioral symptoms and more severe verballyagitated behavior than AD patients without depressivesymptoms (p< 0.001).

Key points

• Depressive symptoms are more prevalent in ADthan in MCI patients.

• AD and MCI patients with depressive symptomsdisplay associated behavioral symptoms.

• Depression in AD is accompanied by even morebehavioral symptoms than depression in MCI.

• Further research would assist in understandingof the possible inter-relation of associatedbehavioral symptoms in depression.

Acknowledgements

This research was supported by the Special ResearchFund of the University of Antwerp; the Foundationfor Alzheimer Research (SAO–FRMA); the Institute

Born-Bunge; an unrestricted research grant fromLundbeck NV (Belgium); the agreement between theInstitute Born-Bunge and the University of Antwerp;the central Biobank facility of the Institute Born-Bunge/University of Antwerp; Neurosearch Antwerp;the Thomas Riellaerts Research Fund; the Fund forScientific Research—Flanders (FWO-F); the Interuni-versity Attraction Poles (IAP) program P7/16 of theBelgian Science Policy Office; the Foundation forAlzheimer Research (SAO–FRMA); the Methusalemexcellence program of the Flemish Government; theResearch Foundation Flanders (FWO); the SpecialResearch Fund of the University of Antwerp; and theInstitute Born-Bunge, Antwerp, Belgium. The authorsacknowledge Prof. Dr. M. Elseviers and Mr. M. VanBuggenhout (University of Antwerp), Mrs. J. Luyckx(Institute Born-Bunge), the administrative assistanceof S. Hicketick, W. Wittebolle, A. Eyckens and clinicalstaff involved (Hospital Network Antwerp).

Conflict of interest

The authors report no conflicts of interest.

References

Albert MS, Dekosky ST, Dickson D, et al. 2011. The diagnosis of mild cognitive im-pairment due to Alzheimer’s disease: recommendations from the National Insti-tute on Aging–Alzheimer’s Association workgroups on diagnostic guidelines forAlzheimer’s disease. Alzheimers Dement 7: 270–279.

Alexopoulos GS, Abrams RC, Young RC, Shamoian CA. 1988a. Cornell Scale for De-pression in Dementia. Biol Psychiatry 23: 271–284.

Alexopoulos GS, Abrams RC, Young RC, Shamoian CA. 1988b. Use of the Cornellscale in nondemented patients. J Am Geriatr Soc 36: 230–236.

American Psychiatric Association. 1994. DSM-IV: Diagnostic and Statistical Manual ofMental Disorders. American Psychiatric Association: Washington, DC.

Aries MJ, Le Bastard N, Debruyne H, et al. 2010. Relation between frontal lobe symp-toms and dementia severity within and across diagnostic dementia categories. Int JGeriatr Psychiatry 25: 1186–1195.

Azermai M, Elseviers M, Petrovic M, Van Bortel L, Vander Stichele R. 2011. Geriatricdrug utilisation of psychotropics in Belgian nursing homes. Hum PsychopharmacolClin Exp 26: 12–20.

Bonelli RM, Cummings JL. 2007. Frontal–subcortical circuitry and behavior.Dialogues ClinNeurosci 9: 141–151.

Burns A, Lawlor B, Craig S. 2004. Assessment Scales in Old Age Psychiatry. MartinDunitz: London.

Cohen-Mansfield J. 1996. Conceptualization of agitation: results based on the Cohen-Mansfield Agitation Inventory and the Agitation Behavior Mapping Instrument.Int Psychogeriatr 8 Suppl 3: 309–315.

Cole MG, Dastoor DP. 1987. A new hierarchic approach to the measurement ofdementia. Accurate results within 15 to 30minutes. Psychosomatics 28: 298–301,304.

Coppen A. 1967. The biochemistry of affective disorders. Br J Psychiatry 113: 1237–1264.Das SK, Bose P, Biswas A, et al. 2007. An epidemiologic study of mild cognitive im-

pairment in Kolkata, India. Neurology 68: 2019–2026.De Deyn PP, Engelborghs S, Saerens J, et al. 2005. The Middelheim Frontality Score: a

behavioural assessment scale that discriminates frontotemporal dementia fromAlzheimer’s disease. Int J Geriatr Psychiatry 20: 70–79.

Devane CL, Chiao E, Franklin M, Kruep EJ. 2005. Anxiety disorders in the 21st cen-tury: status, challenges, opportunities, and comorbidity with depression. Am JManag Care 11: S344–S353.

Di Giovanni G, Esposito E, Di Matteo V. 2010. Role of serotonin in central dopaminedysfunction. CNS Neurosci Ther 16: 179–194.

957Depressive symptoms in MCI and Alzheimer’s disease

Copyright # 2012 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2013 28: 947–958

Dierckx E, Engelborghs S, De Raedt R, De Deyn PP, Ponjaert-Kristoffersen I. 2007.Mild cognitive impairment: what’s in a name? Gerontology 53: 28–35.

Engelborghs S, Maertens K, Nagels G, et al. 2005. Neuropsychiatric symptoms of de-mentia: cross-sectional analysis from a prospective, longitudinal Belgian study. IntJ Geriatr Psychiatry 20: 1028–1037.

Engelborghs S, Vloeberghs E, Le BN, et al. 2008. The dopaminergic neurotransmittersystem is associated with aggression and agitation in frontotemporal dementia.Neurochem Int 52: 1052–1060.

Fava M. 2004. Daytime sleepiness and insomnia as correlates of depression. J ClinPsychiatry 65 Suppl 16: 27–32.

Fisk JD, Merry HR, Rockwood K. 2003. Variations in case definition affect prevalencebut not outcomes of mild cognitive impairment. Neurology 61: 1179–1184.

Folstein MF, Folstein SE, McHugh PR. 1975. “Mini-mental state”. A practicalmethod for grading the cognitive state of patients for the clinician. J PsychiatrRes 12: 189–198.

Frisoni GB, Fratiglioni L, Fastbom J, et al. 2000. Mild cognitive impairment in thepopulation and physical health: data on 1,435 individuals aged 75 to 95. J GerontolA Biol Sci Med Sci 55: M322–M328.

Gaugler JE, Yu F, Krichbaum K, Wyman JF. 2009. Predictors of nursing home admis-sion for persons with dementia. Med Care 47: 191–198.

Goldman-Rakic PS, Lidow MS, Gallager DW. 1990. Overlap of dopaminergic,adrenergic, and serotoninergic receptors and complementarity of their subtypesin primate prefrontal cortex. J Neurosci 10: 2125–2138.

Hanninen T, Hallikainen M, Tuomainen S, Vanhanen M, Soininen H. 2002. Preva-lence of mild cognitive impairment: a population-based study in elderly subjects.Acta Neurol Scand 106: 148–154.

de Jonghe JF, Kat MG. 1996. Factor structure and validity of the Dutch version of theCohen-Mansfield Agitation Inventory (CMAI-D). J Am Geriatr Soc 44: 888–889.

Kupfer DJ. 2006. Depression and associated steep disturbances: patient benefits withagomelatine. Eur Neuropsychopharmacol 16: S639–S643.

Le Bastard N, Coart E, Vanderstichele H, et al. 2012. Comparison of two analyticalplatforms for the clinical qualification of Alzheimer’s disease biomarkers inpathologically-confirmed dementia. J Alzheimers Dis. doi: 10.3233/JAD-2012-121246. [Epub ahead of print]

Lee HB, Lyketsos CG. 2003. Depression in Alzheimer’s disease: heterogeneity andrelated issues. Biol Psychiatry 54: 353–362.

Luppa M, Luck T, Brahler E, Konig HH, Riedel-Heller SG. 2008. Prediction of insti-tutionalisation in dementia. A systematic review. Dement Geriatr Cogn Disord 26:65–78.

McKhann G, Drachman D, Folstein M, et al. 1984. Clinical diagnosis of Alzheimer’sdisease: report of the NINCDS–ADRDA Work Group under the auspices of De-partment of Health and Human Services Task Force on Alzheimer’s Disease. Neu-rology 34: 939–944.

Menon AS, Gruber-Baldini AL, Hebel JR, et al. 2001. Relationship between aggressivebehaviors and depression among nursing home residents with dementia. Int J GeriatrPsychiatry 16: 139–146.

Mitchell AJ, Shiri-Feshki M. 2009. Rate of progression of mild cognitive impairmentto dementia—meta-analysis of 41 robust inception cohort studies. Acta PsychiatrScand 119: 252–265.

Monastero R, Mangialasche F, Camarda C, Ercolani S, Camarda R. 2009. A systematicreview of neuropsychiatric symptoms in mild cognitive impairment. J AlzheimersDis 18: 11–30.

Olin JT, Katz IR, Meyers BS, Schneider LS, Lebowitz BD. 2002. Provisional diagnosticcriteria for depression of Alzheimer disease: rationale and background. Am J GeriatrPsychiatry 10: 129–141.

Panza F, Frisardi V, Capurso C, et al. 2010. Late-life depression, mild cognitive im-pairment, and dementia: possible continuum? Am J Geriatr Psychiatry 18: 98–116.

Petersen RC. 2004. Mild cognitive impairment as a diagnostic entity. J Intern Med256: 183–194.

Petersen RC, Doody R, Kurz A, et al. 2001. Current concepts in mild cognitive im-pairment. Arch Neurol 58: 1985–1992.

Portet F, Ousset PJ, Visser PJ, et al. 2006. Mild cognitive impairment (MCI) in med-ical practice: a critical review of the concept and new diagnostic procedure. Reportof the MCI Working Group of the European Consortium on Alzheimer’s Disease.J Neurol Neurosurg Psychiatry 77: 714–718.

Prado-Jean A, Couratier P, Druet-Cabanac M, et al. 2010. Specific psychological andbehavioral symptoms of depression in patients with dementia. Int J Geriatr Psychi-atry 25: 1065–1072.

Quera-Salva MA, Lemoine P, Guilleminault C. 2010. Impact of the novel antidepres-sant agomelatine on disturbed sleep–wake cycles in depressed patients. HumPsychopharmacol 25: 222–229.

Randolph C, Tierney MC, Mohr E, Chase TN. 1998. The Repeatable Battery for theAssessment of Neuropsychological Status (RBANS): preliminary clinical validity.J Clin Exp Neuropsychol 20: 310–319.

Reisberg B, Ferris SH, de LeonMJ, Crook T. 1982. The Global Deterioration Scale (GDS)for assessment of primary degenerative dementia. Am J Psychiatry 139: 1136–1139.

Reisberg B, Borenstein J, Salob SP, et al. 1987. Behavioral symptoms in Alzheimer’sdisease: phenomenology and treatment. J Clin Psychiatry 48 Suppl: 9–15.

Ritchie K, Artero S, Touchon J. 2001. Classification criteria for mild cognitive impair-ment: a population-based validation study. Neurology 56: 37–42.

Starkstein SE, Jorge R, Mizrahi R, Robinson RG. 2005. The construct of minor andmajor depression in Alzheimer’s disease. Am J Psychiatry 162: 2086–2093.

The Psychological Corporation. 1998. WAIS-III/WMS-III Technical Manual. Har-court Brace & Co: London.

TognoniG, Ceravolo R,Nucciarone B, et al. 2005. Frommild cognitive impairment to de-mentia: a prevalence study in a district of Tuscany, Italy. Acta Neurol Scand 112: 65–71.

Van der Mussele S, Le Bastard N, Vermeiren Y, et al. 2012. Behavioral symptoms inmild cognitive impairment as compared with Alzheimer’s disease and healthy olderadults. Int J Geriatr Psychiatry. doi: 10.1002/gps.3820; Epub ahead of printing.

Winblad B, Palmer K, Kivipelto M, et al. 2004. Mild cognitive impairment—beyondcontroversies, towards a consensus: report of the International Working Group onMild Cognitive Impairment. J Intern Med 256: 240–246.

958 S. Van der Mussele et al.

Copyright # 2012 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2013 28: 947–958