Flat affect and social functioning: A 10year follow-up study of first episode psychosis patients

Page 1

Schizophrenia Research 139 (2012) 99–104

Contents lists available at SciVerse ScienceDirect

Schizophrenia Research

j ourna l homepage: www.e lsev ie r .com/ locate /schres

Flat affect and social functioning: A 10 year follow-up study of first episodepsychosis patients

Julie Evensen a,b,⁎, Jan Ivar Røssberg a,b, Helene Barder a,b, Ulrik Haahr j,k, Wenche ten Velden Hegelstad c,Inge Joa c, Jan Olav Johannessen c,i, T.K. Larsen c,d, Ingrid Melle a,b, Stein Opjordsmoen a,b,Bjørn Rishovd Rund g,h, Erik Simonsen j,k, Per Vaglum f, Thomas McGlashan e, Svein Friis a,b

a Division of Mental Health and Addiction, Oslo University Hospital, N-0407, Oslo, Norwayb Institute of Clinical Medicine, University of Oslo, P.O. 1130 Blindern, 0318 Oslo, Norwayc Regional Centre for Clinical Research in Psychosis, Psychiatric Division, Stavanger University Hospital, Postboks 8100, 4068 Stavanger, Norwayd Institute of Psychiatry, University of Bergen, Jonas Lies vei 65, 5021 Bergen, Norwaye Department of Psychiatry, Yale University School of Medicine, Yale Psychiatric Research at Congress Place, 301 Cedar St., New Haven, CT 06519, USAf Department of Behavioral Sciences in Medicine, University of Oslo, P.O. 1130 Blindern, 0318 Oslo, Norwayg Institute of Psychology, University of Oslo, P.O. 1130 Blindern, 0318 Oslo, Norwayh Vestre Viken Hospital Trust/Department of Psychology, 3004 Drammen, Norwayi Faculty of Social Sciences, University of Stavanger, Norwayj Psychiatric Research Unit, Zealand Region, Psychiatry Roskilde, Roskilde University and University of Copenhagen, Smedegade 10‐16, 4000 Roskilde, Denmarkk Early Psychosis Intervention Center, Psychiatry Roskilde, Region Zealand, Smedegade 10, 4000 Roskilde, Denmark

⁎ Corresponding author at: Division of Mental HealthHospital, Ullevaal, N-0407 Oslo, Norway. Tel.: +47 99 4

E-mail address: [email protected] (J. Even

0920-9964/$ – see front matter © 2012 Elsevier B.V. Alldoi:10.1016/j.schres.2012.04.019

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 10 February 2012Received in revised form 30 March 2012Accepted 30 April 2012Available online 23 May 2012

Keywords:Negative symptomsFlat affectOutcomeFirst episode psychosis

Background: Affective flattening has been described as enduring, but long term follow-up studies of first ep-isode psychosis patients are lacking.Objective: The aim of this study was to follow the symptom development of flat affect (FA), over a 10 yearfollow-up period, with focus on prevalence, predictors and outcome factors including social functioning.Methods: Three-hundred-and-one patients with FEP were included at baseline, 186 participated in the10 year follow-up. These were followed on PANSS item N1 (FA) from baseline through 5 follow-up assess-ments over 10 years. Patients were grouped as having never-present, improving, deteriorating, fluctuatingor enduring FA. The groups were compared on baseline variables, variables at 10 year follow-up, and socialfunctioning throughout the follow-up period.Results: Twenty nine percent never displayed FA, 66% had improving, deteriorating or fluctuating FA, while

5% of patients had enduring FA. Premorbid social function predicted enduring FA. The patients with enduring,fluctuating and deteriorating FA did poorer on all outcome variables, including remission and recovery rates.The enduring FA group did significantly poorer in social functioning over the 10 year period.Conclusions: FA is expressed at some point of time in the majority of FEP patients in a 10 year follow-up pe-riod, and appears more fluctuant than expected from the relevant literature. FA is associated with poorer out-come after 10 years, and enduring FA to poorer social function at all points of assessment.

© 2012 Elsevier B.V. All rights reserved.

1. Introduction

The negative symptom dimension of psychotic disorders adverselyinfluences functioning and outcome (Bowie et al., 2006; Ventura et al.,2009; White et al., 2009). The NIMH-MATRICS consensus statement re-flects increased interest in negative symptoms andencourages the studyof individual negative symptoms to facilitate treatment development(Kirkpatrick et al., 2006). Negative symptoms include avolition/apathy,asociality, anhedonia, inattention, alogia and flat affect. Factor analyses

and Addiction, Oslo University1 07 52; fax: +47 22 11 84 23.sen).

rights reserved.

of the negative symptomdimension have indicated two separate factors:avolition/apathy and flat affect (Blanchard and Cohen, 2006; Messingeret al., 2011). Recent studies have emphasized the role of apathy inshort and long-term outcomes including general functioning (Kiang etal., 2003; Faerden et al., 2009, 2010; Foussias et al., 2009; Foussias andRemington, 2010). Flat affect (FA) has received less specific attention.

Flat affect is described as unchanging facial expression, paucity ofgestures and affective non-responsiveness (Kay et al., 1987). Studiesof patients with first episode psychosis (FEP) have found FA to be pre-sent at the onset of illness (Shtasel et al., 1992), and predicted bymalegender, a diagnosis of schizophrenia and length of prodromal period(Malla et al., 2002). Kelley et al. (2008) investigated FA, avolition,alogia and anhedonia separately using the Scale for the Assessment

Page 2

100 J. Evensen et al. / Schizophrenia Research 139 (2012) 99–104

of Negative Symptoms (SANS) at baseline and four points of follow-upover one year in both a chronic and a first episode schizophrenia sample.The study concluded that levels of FA were largely stable over 1 year, incontrast with the other negative symptoms that followed a more fluctu-ating course.

Long-term longitudinal studies of schizophrenia and FEP popu-lations generally focus on the negative symptom dimension ratherthan individual symptoms, and report cross-sectional results ratherthan longitudinal symptom profiles (Eaton et al., 1995; Milev et al.,2005; Siegel et al., 2006; Bertelsen et al., 2009; White et al., 2009).Harrow and colleagues, however, studied the longitudinal symptomtrajectory of 42 young patients with schizophrenia and schizoaffectivedisorder and found that 19% had enduring, 41% episodic and 40% nonegative symptoms over a 10 year follow-up period. Enduring symptomswas defined as being above threshold level (>1 on≥1 item, Pogue-Geileand Harrow's Negative Symptom Scale) at 3 follow-up assessments(Herbener andHarrow, 2001). This study did not look at specific negativesymptoms. To our knowledge no FEP study has followed the symptomdevelopment of FA for longer than 1 year.

Studies have shown a significant association between negativesymptoms and reduced social functioning over time (Shtasel et al.,1992; Ho et al., 1998; Lysaker and Davis, 2004). A 1 year follow-upstudy of patients with schizophrenia found that patients with FA hadpoorer social outcome compared to patients without FA (Gur et al.,2006).We do not know the relationship between FA and social function-ing beyond one year.

This study aims to explore symptomdevelopment of PANSS rated FAin FEP patients. We followed patients over a 10 year period and identi-fied patients with never-present, improving, deteriorating, fluctuatingand enduring FA. We wished to address the following questions:

1. What is the prevalence and stability of FA in a FEP sample followedover a 10 year period?

2. Do the FA trajectory groups differ with regard to baseline variables,and do any of these variables predict enduring flat affect?

3. Do patients with different FA trajectories differ on outcomemeasuresincluding remission and recovery?

4. Do the FA trajectory groups differ in social functioning over the10 year follow-up period?

2. The TIPS study

The TIPS (Early Treatment and Intervention in Psychosis) project is alarge, longitudinal study of consecutively admitted FEP patients. Theoverall study design, samples, and assessment instruments are detailedin other reports (Larsen et al., 2001;Melle et al., 2004). Briefly, the studywas designed to identify and follow-up clinical, epidemiologic samplesof FEP patients from four Scandinavian catchment sites. Patients wereassessed at baseline, 3 months, 1, 2, 5 and 10 years.

2.1. Study participants

The study was carried out within the specialist psychiatric healthcareservices of four Scandinavian health care sectors (North and South sector,Rogaland County, Norway, Ullevaal Sector, Oslo, Norway, and Fjordenmid-sector, Roskilde, Denmark).

Inclusion criteria were as follows:

1) Afirst episode psychosis (PANSS score≥4 onone ormore of positivesubscale items 1, 3, 5 or 6 or on general subscale 9 for ≥7 days).

2) Meeting the DSM-IV criteria for schizophrenia, schizophreniformdisorder, schizoaffective disorder, brief psychotic episode (BPE), de-lusional disorder (DD), affective psychosis with mood-incongruentpsychotic features (MDE) or psychotic disorder not otherwise spec-ified (Psychosis NOS).

3) Age 18–65 years (15–65 in Rogaland).4) IQ >70.

The exclusion criteria were having received adequate prior antipsy-chotic treatment (antipsychoticmedication>3.5 haloperidol equivalentsfor >12 weeks or until psychotic symptom remission) and an organic- orsubstance-induced psychosis.

Written informed consent was obtained from all subjects. Alto-gether 301 patients were included from 1997 to 2000. Of these 186were interviewed at ten year follow-up. Drop-out analyses of 281 pa-tients in the original sample (aged >18 years) showed no statisticaldifference in inclusion age, GAF, diagnostic distribution, PANSS symp-tom levels or drug abuse between patients lost versus followed up at10 years (Hegelstad et al., 2012). PANSS negative component itemN1was used to assess and follow the development of FA. Only patientswho participated in the 10 year follow-upwere described in this study(N=186). Two patients were excluded as they lackedmore than 2 as-sessments. For patients who lacked≤2 scores amean scorewas calcu-lated as the mean of the scores preceding and following the missingscores. We used a PANSS score of ≥3 as threshold for clinically signif-icant FA. Patients were grouped according to FA symptom trajectory:

1. Never-present (item score ≤2 on all assessments).2. Improving (starts ≥3, ends ≤2).3. Deteriorating (starts ≤2, ends ≥3).4. Fluctuating (≥2 fluctuations across threshold for FA).5. Enduring (item scores≥3 on all assessments, a separate score of 2,

but not 1, allowed).

The diagnostic distribution at 10 year follow-upwas schizophrenia 96(52%), schizophreniform disorder 9 (5%), schizoaffective disorder 34(19%), BPE 3 (2%), DD 7 (4%), psychosis NOS 11 (6%), andMDE 24 (13%).

2.2. Instruments and measures

The structured clinical interview for the DSM-IV (SCID) was usedfor diagnostic purposes (Spitzer et al., 1992). Duration of untreatedpsychosis (DUP) was measured as the time in weeks from the firstpositive psychotic symptoms (PANSS score ≥4 on Positive scaleitems 1, 3, 5 or 6 or General scale item 9) to the start of the first ade-quate treatment of psychosis. Premorbid functioning was measuredby the Premorbid Assessment of Functioning Scale (PAS) (Larsen etal., 2004). Symptom levels were measured by the Positive and Nega-tive Syndrome Scale (PANSS) (Kay et al., 1987), and symptom do-mains were represented by the corresponding PANSS components(positive, negative, excitative, cognitive and depressive) (Bentsen etal., 1996). Global functioning was measured by the Global Assessmentof Functioning Scale (GAF-F) (Endicott et al., 1976). Employment andability to live independently were measured with the Strauss Carpen-ter Level of Functioning Scale (SCLFS) (Strauss and Carpenter, 1974).Objective social functioning was measured using Lehman's Quality ofLife Interview (L-QoLI) (Lehman, 1988). Antipsychotic medication,Defined Daily Dose (DDD), was measured using the World Health Or-ganization Collaborating Center for Drugs statistics methodology(WHO, Collaborating Centre for Drug Statistics Methodology, 2008).Symptom remissionwas defined according to international standardizedcriteria (Andreasen et al., 2005), including PANSS score b4 on Positivescale items 1, 2, 3, 5 and 6, Negative scale items 1, 4 and 6, and Generalscale items 5 and 9. Recovery was operationalized as a combination ofsymptom remission and 3 functional dimensions from the SCLFS: Inde-pendent living, Role functioning (work, academic, or full-time home-making) and Social Interaction. A score of 0 indicated poor, and 4 indicat-ed good functioning. Patients in recovery had, during the last 12 months,fulfilled the symptom remission criteria above and scored 4 on all func-tional dimensions.

The above test-batterywas used at baseline and repeated at 1, 2, 5 and10 years, excluding DUP and PAS. At 3 months only PANSS and GAFwereused.

Good reliability for major variables (GAF, DUP and diagnosis) hasbeen documented for earlier assessments (Friis et al., 2003). For the

Page 3

Note: -Flat affect=Positive and Negative Syndrome Scale Score (PANSS) item N1

-Total N=184

1

1.5

2

2.5

3

3.5

4

4.5

Baseline 3 months 1 year 2 years 5 years 10 years

Never-present (29%) Improving (10%)

Deteriorating (16%) Fluctuating (40%)

Enduring (5%)

Fig. 1. Development of flat affect in 5 trajectory groups, baseline to 10 years.

101J. Evensen et al. / Schizophrenia Research 139 (2012) 99–104

10 year follow-up 26 video-taped patient interviews were rated by anexperienced psychologist not involved in the project and blind to all rat-ings. For GAF the ICCs were=.83 (symptoms) and 0.88 (function). Forthe five PANSS components the ICCs ranged from 0.61 to 0.82 with amedian of 0.67. ICC for PANSS item N1 was .76.

2.3. Statistical analysis

The analyses were performed with the SPSS Statistical Program (ver-sion 18: SPSS Inc., Chicago, IL, USA). Mean and standard deviations arereported for continuous variables and percentages for categorical vari-ables. DUP had amarkedly left skewed distribution andwas transformedto its natural logarithm (ln(DUP+1)). One-way ANOVA was used tocompare the FA groups on parametric data (Tukey post hoc test), andKruskal–Wallis test was used on non-parametric and unevenly distribut-ed data. Binary logistic regression was performed to assess baseline

Table 1Baseline variables and PAS, differences between the 5 flat affect trajectory groups.

1. Never-present(N=53)Mean (S.D.)

2. Improving(N=18)Mean (S.D.)

Males (%) 45 67Age baseline (years) 28.5 (9.37) 29.2 (8.72)SZP spectrum (%) 55 61Years of education 12.5 (2.46) 11.8 (2.53)DUP (weeks, median, range) 4 (0–235) 18 (0–450)PAS soc. childhood .70 (.98) 1.35 (1.31)PAS soc. last score 1.12 (1.30) 1.88 (1.14)PAS ac. childhood 1.54 (1.05) 2.00 (1.12)PAS ac. last score 2.38 (1.52) 2.68 (1.50)PANSS pos. comp. 14.1 (4.27) 15.7 (3.03)PANSS neg. comp. 16.0 (4,52) 26.7 (7.80)PANSS N1 1.09 (.30) 3.6 (.78)PANSS dep. comp. 11.1 (3.84) 14.2 (3.92)PANSS cog. comp. 7.2 (3.09) 8.3 (3.44)PANSS exc. comp. 10.5 (4.84) 10.3 (4.00)GAF-F 31.9 (11.20) 29.7 (8.50)

Note: Stat. sign. diff. p=b.05.PAS social last score: Kruskal–Wallis test: pb .05.PANSS neg. comp.: 1b2, 4 and 5.PANSS N1: 1 and 3b2, 4 and 5.PANSS dep. comp.: 1b2.SZP spectrum disorder equals a diagnosis of schizophrenia, schizophreniform disorder or scDUP = duration of untreated psychosis.PAS = Premorbid Assessment of Functioning Scale.PANSS = Positive and Negative Syndrome Scale Score.GAF-F = Global Assessment of Functioning Scale, functional component.

predictors of enduring FA. Themodel contained 3 independent variables:1. PANSS baseline positive, depressive, cognitive and excitative compo-nents, 2. baseline GAF-F, and 3. PAS last social function. Baseline variableswhere the FA groups showed statistically significant differences werechosen. Independent sample t-test was used to compare the enduringgroup to the other groups on objective social functioning from baselineto 10 year follow-up.

3. Results

Out of a total of 184 patients 10 patients (5%) showed enduring flataffect (FA) over the course of the ten year follow up period, 53 (29%) pa-tients never displayed FA, while 18 (10%) patients had initial FA that re-solved over the course of follow-up (the improving group). Twenty-ninepatients (16%) developed FA in the follow-up period (the deterioratinggroup). The fluctuating group was the largest group, containing 74 pa-tients (40%). This group was heterogeneous: approximately 1/3 movedfrom above to below to above threshold FA, 1/3 moved from below toabove to below threshold FA, and 1/3 experienced several fluctuationsin FA. A total of 76 out of 184 patients (41%) had clinically significantFA at 10 years. Fig. 1 describes the symptom development in FA in the5 groups over the 10 year follow-up period.

Table 1 describes the 5 FA trajectory groups on baseline characteris-tics and premorbid function (PAS). Onmost clinical and functional vari-ables the never-present group had the best scores, while the enduringgroup scored poorest. This trend was most pronounced for premorbidsocial function (last score), and PANSS negative symptoms where thedifferences between the groups reached statistical significance.

Binary logistic regression was performed to assess possible base-line predictors of enduring FA (Table 2). The model contained threeblocks of independent variables: 1. PANSS baseline positive, depressive,cognitive and excitative components, 2. baseline GAF-F, and 3. PAS lastsocial function. The model as a whole explained between 8.7% (Coxand Snell square) and 26.7% (Nagelkerke R Square) of the variance. Asshown in Table 2 the only significant contributors to the model wereGAF-F and PAS last social function. The strongest predictor of enduringFA was PAS last social function (OR of 1.9, Wald 7.17, p=.007).

3. Deteriorating(N=29)Mean (S.D.)

4. Fluctuating(N=74)Mean (S.D.)

5. Enduring(N=10)Mean (S.D.)

55 62 6028.1 (10.70) 26.2 (9.07) 28.6 (9.86)66 64 8011.9 (2.00) 12.1 (2.51) 12.6 (4.21)5 (0–140) 6 (0–520) 10 (1–87).91 (1.08) 1.11 (1.22) 1.17 (.61)1.66 (1.57) 2.02 (1.53) 3.17 (1.52)1.83 (1.21) 1.64 (1.10) 2.22 (2.02)2.16 (1.18) 2.21 (1.34) 2.61 (1.14)15.2 (4.01) 15.69 (4.40) 17.40 (5.13)17 (5.01) 22.24 (9.68) 30.4 (10.41)1.17 (.38) 2.23 (1.41) 3.75 (1.32)12.10 (3.82) 11.9 (3.95) 12.10 (5.26)6.8 (3.27) 7.1 (3.62) 8.4 (3.27)8.1 (2.60) 8.72 (3.89) 11.5 (7.44)34.0 (11.20) 32.8 (10.55) 23.9 (8.63)

hizoaffective disorder.

Page 4

Table 2Binary logistic regression analysis with enduring flat affect vs. not enduring flat affect as dependent variable and baseline scores as independent variables.

Model variable O.R. 95% C.I. p value S.E. Wald

Stp. 1PANSS Pos. 1.1 .93–1.3 .257 .08 1.28PANSS Cog. .93 .73–1.2 .579 .13 .31PANSS Dep. 1.01 .85–1.21 .891 .09 .02PANSS Exc. 1.02 .87–1.19 .836 .08 .08

Stp. 2: GAF-F .90 .87–1.0 .042 .05 4.12Stp. 3: PAS social last score 1.9 1.19–3.03 .007 .24 7.17

Logistic regression model for enduring flat affect (n=10) vs. not (n=174). For overall model including PANSS pos., cogn. dep. and exc. components, GAF-F and PAS social last score :χ2=16.5, df=6, p=. 01, Nagelkerke R2=26.7.PAS = Premorbid Assessment of Functioning Scale.PANSS = Positive and Negative Syndrome Scale Score.GAF-F = Global Assessment of Functioning Scale, functional component.

102 J. Evensen et al. / Schizophrenia Research 139 (2012) 99–104

Table 3 describes differences between the 5 trajectory groups inPANSS symptom components at 10 years. The never-present groupscored significantly lower than the deteriorating and the fluctuatinggroups on both the positive and cognitive components. On the generalnegative component the never-present and the improving groupsscored statistically significantly lower than the remaining groups. Nostatistically significant differences were found on the depressive or ex-citative components. In the enduring group all patients fulfilled criteriafor schizophrenia spectrum disorder at 10 year follow-up.

Table 3 also shows that the fluctuating, deteriorating and enduringgroups were psychotic for statistically significantly longer during thefollow-up period, and fewerwere in remission or recovery after 10 years.These trajectory groups also had poorer employment status, reducedability to live independently and lower GAF-F general function score.Fig. 2 delineates the development of objective social function in thefollow-up period. Statistically significant differenceswere found between

Table 3Variables at 10 year follow-up, differences between the 5 flat affect trajectory groups.

Never-present(N=53)Mean (S.D.)

Improving(N=18)Mean (S.D.

Time psychotic (%) 12.2 27In remission (%) 77.4 61.1Recovered (%) 52.8 27.8Unable to live independently (%) 3.8 11.1Employment 2.38 (1.76) 1.33 (1.53)Obj. social function 3.6 (.78) 3.6 (1.0)GAF-F 61.8 (16.0) 55.0 (14.5)PANSS pos. comp. 7.4 (3.5) 8.1 (3.04)PANSS neg. comp. 11.5 (2.57) 12.3 (2.4)PANSS N1 1.1 (.30) 1.2 (.38)PANSS dep. comp. 8.5 (3.64) 9.2 (3.47)PANSS cog. comp. 3.8 (1.5) 4.4 (1.7)PANSS exc. comp. 7.0 (2.8) 7.6 (2.2)Antipsychotic medication (DDD) .31 (.64) .57 (.61)SZP spectrum (%) 57 78

Note: Stat. sign. diff. p=b.05.Time psychotic (% baseline to ten years): Kruskal–Wallis test: pb .05.In remission (%): Kruskal–Wallis test: pb .05.Recovered (%): Kruskal–Wallis test: pb .05.Unable to live independently (%) (SCLFS): 1 and 2b3, 4 and 5.Employment (SCLFS): 1>3, 4 and 5.Obj. social function (L-QoLI): 1 and 2>3 and 5.GAF-F: 1>3, 4 and 5.PANSS pos. comp.: 1b3 and 4.PANSS neg. comp.: 1 and 2b3, 4 and 5.PANSS N1: 1 and 2b3, 4 and 5.PANSS cog. comp.: 1b3 and 4.SZP spectrum (%): Kruskal–Wallis test pb .05.DDD: 1b3 and 4.GAF-F = Global Assessment of Functioning Scale, functional component.PANSS = Positive and Negative Syndrome Scale Score.DDD = Defined daily Dose (WHO criteria).SZP spectrum disorder equals a diagnosis of schizophrenia, schizophreniform disorder and

the enduring group and the remaining groups at all points of measure(pb .05 on all points).

To assess if the inclusion of non-schizophrenia spectrum patientshad influenced the results we repeated all the above analyses on theschizophrenia spectrum patients only (N=139). We found little changein group distribution, and only very minor differences with regard tobaseline variables and outcome measures at 10 years compared to anal-yses of the full sample. Social premorbid function remained themost im-portant predictor of enduring FA, and social function from baseline to10 years remained significantly poorer in the enduring FA group com-pared to the other groups.

4. Discussion

Of a total of 184 patients 71% had clinically significant flat affect(PANSS N1 ≥3) at least one point of measure over the 10 year follow-

)

Deteriorating(N=29)Mean (S.D.)

Fluctuating(N=74)Mean (S.D.)

Enduring(N=10)Mean (S.D.)

46.6 40.9 48.731 39.2 206.9 13.5 048.3 23 60.90 (1.47) 1.08 (1.43) .70 (1.49)2.5 (1.14) 3.1 (1.00) 2.0 (.85)43.1 (10.84) 47.6 (14.41) 41.5 (8.73)11.1 (4.46) 11.1 (6.04) 9.7 (4.99)22.9 (9.43) 17.6 (6.36) 25.2 (6.36)3.3 (.61) 2.2 (1.24) 3.6 (.52)10.45 (3.89) 9.5 (3.7) 8.9 (4.25)5.7 (2.6) 5.2 (2.76) 6.0 (3.02)8.5 (3.16) 8.2 (4.0) 6.9 (1.66)1.32 (.90) 1.37 (1.31) 1.06 (.46)86 81 100

schizoaffective disorder.

Page 5

Note:- Objective social function (L-QoLI)- Total N=184

Baseline 1 year 2 years 5 years 10 years

Never-present ImprovingDeteriorating Fluctuating

Enduring

1

1,5

2

2,5

3

3,5

4

Fig. 2. Longitudinal development of objective social functioning in 5 flat affect trajectorygroups, baseline to 10 years.

103J. Evensen et al. / Schizophrenia Research 139 (2012) 99–104

up period. Only 5% of the sample had enduring FA throughout thefollow-up period. The instability of FA found in our study differs fromprevious literature that emphasizes the stability of negative symptomsover positive symptoms (Pogue-Geile and Harrow, 1985; Fenton andMcGlashan, 1992) and FA over other negative symptoms (Malla et al.,2004; Kelley et al., 2008). Though most studies emphasize the stabilityof negative symptoms over time, some studies indicate a more fluidsymptom picture (Edwards et al., 1999). A recent 3 year long follow-up study of FEP patients found enduring negative symptoms in 23.7%in the last year of the follow-up period, but only in 6.5% of patients inthe first year of follow-up (Chang et al., 2011). By following patientslongitudinally from the first psychotic episode we were able to showthat the degree of flat affect changes over time to amuch greater extentthan anticipated.

The prevalence of enduring FA found in our study is lower thanthe reported 15% prevalence of deficit psychopathology in FEP sam-ples (Kirkpatrick et al., 2001). The definition of deficit psychopathol-ogy emphasizes the stability of negative symptoms and includeshaving a diagnosis of schizophrenia (Kirkpatrick et al., 1989;Kirkpatrick and Galderisi, 2008). In our study the enduring FA groupwas similar to patients with deficit psychopathology in that they hadpoorer premorbid function. Also, by 10 year follow-up, all the patientsin the enduring group had a diagnosis within the schizophreniaspectrum.

Identifying patients with enduring FA early seems important asthis group appears particularly vulnerable. We found that the endur-ing group was similar to the remaining patients on most baseline var-iables, but differed significantly in premorbid social functioning.Reduced display of affect has been found in the children who later de-veloped schizophrenia compared to their siblings who did not devel-op schizophrenia (Walker et al., 1993). The combination of FA andpoor social functioning could potentially be present from early stagesin life of a child vulnerable to psychotic illness.

Earlier studies of the longitudinal course of negative symptoms inschizophrenia have found increasing prevalence and severity of neg-ative symptoms over the course of follow-up (Fenton and McGlashan,1991). We found an increase from below to above clinical thresholdin 16% of the sample. Most of the patients that developed FA (deteri-orating group) did so in the latter part of the follow-up period, and by10 years had levels of FA similar to those of the enduring group. Fur-thermore, the deteriorating group was similar to the never-presentgroup on symptom and functional measures at baseline. However,by 10 year follow-up these patients had developed symptom and

functional levels close to those of the enduring group. We also identi-fied a somewhat smaller group (10%) whose FA remitted during thefollow-up. Most of these patients showed no sign of FA after 1 year,and in this group FA must be considered as secondary.

We found a clear association between FA and poor functioning.The patients with enduring FA performed poorest on all functionaloutcome variables, while the fluctuating and deteriorating FA groupsscored closer to the enduring group at 10 years. In the improving anddeteriorating groups we found that as FA decreased functioning in-creased and vice versa. These findings could imply that FA is a possi-ble marker of active illness, reflecting a more profound and chronicdisease process in the enduring group. Alternatively, FA could be di-rectly related to functioning, and to social functioning in particular.FA may be viewed as an expressive deficit with a specific impact onsocial communication and interaction. Laboratory studies of patientswith schizophrenia have reported reduced facial expressions duringsocial interaction (Krause et al., 1989) and watching emotional films(Berenbaum and Oltmanns, 1992) and cartoons (Dworkin, 1992). Afailure to respond to social stimuli could reduce the likelihood of esta-blishing and maintaining friendships and other relationships. FA mayalso be related to poor social cognition. Social cognition is defined as theability to process and apply social information, and includes the abilityto recognize facial expressions of emotion. Impaired performance on af-fect perception tasks is an established finding in patients with schizo-phrenia (Kohler et al., 2010). Though results are not unequivocal,most studies have found that negative symptoms correlate with affectperception (Kohler et al., 2010), and facial expression recognition andunfamiliar facematching have been found to be poorer in schizophreniapatients with FA compared to patients without FA (Gur et al., 2006). Afailure to respond to and interpret verbal and nonverbal stimuli is likelyto affect social abilities and contacts over time.We found that the endur-ing FA group scoredpoorer than the remaining patients on social functionat baseline and throughout the follow-up period. Our findings indicatethat in patients where FA is more trait than state, social function ismost likely to be negatively affected.

In a recently published article we reported on clinical apathy in thepatient sample described here (Evensen et al., 2012). We found thatnearly 30% of patients displayed clinical apathy at 10 year follow-up,i.e. a lower prevalence than FA at 10 years. Similar to the currentstudy we found that apathy was significantly related to functioning.The correlation, however, was stronger between apathy and generalfunctioning (r=.49), than between apathy and social contacts(r=.30). We assessed apathy, using the self-report Apathy EvaluationScale (AES-S-apathy), only at the 10 year follow-up, and thus couldnot describe the same symptom trajectory groups as in the current arti-cle. Thismakes comparisons between the studies difficult. However, thedifference in prevalence, and to some degree symptom correlates, sup-ports the two factor model of the negative symptom construct.

This study is unique in that it explores a large and representativegroup of FEP patients 10 years after their first psychotic episode witha particular focus on different FA trajectories. The study was notdesigned to separate primary from secondary negative symptoms. Pre-vious studies on FEP samples have, however, found that FA is not signif-icantly influenced by positive and depressive symptoms (Malla et al.,2002) or medication (Kelley et al., 2008). A further limitation of thestudy was the use of a single item measure of FA. A strength of theTIPS study, however, is the strict focus on reliability testing throughoutthe follow-up period (Friis et al., 2003; Hegelstad et al., 2012). The as-sessment of PANSS item N1 (FA) came out favorably (ICC=.76).

In conclusion, this study indicates that flat affect is a more fluctuantsymptom than anticipated. FA, especially when enduring, is related topoorer functioning, and particularly to poorer social functioning, bothpremorbidly and throughout the 10 year follow-up period. Identifyingpatients with enduring or developing FA challenges clinicians to target,track, and engage those particularly vulnerable groups within a firstepisode psychosis sample.

Page 6

104 J. Evensen et al. / Schizophrenia Research 139 (2012) 99–104

Role of funding sourceThe project has been approved by the Regional Committee for Medical Research

Ethics Health Region II (# S-95189) and the Regional Committee for Medical ResearchEthics Health Region East (# 1.2007.2177).

Data Inspectorate (License # 96/3017-2 and # 2003/2052).Biological data collection was approved by Norwegian Directory of Health

(# 200403453) and the Regional Committee for Medical Research Ethics Health RegionEast (# 493-03-01179). The Regional Committee for Science Ethics region Sjælland,Denmark (# 1-01-83-0002-07).

Supported by Health West (# 911369), Norway (Wenche ten Velden Hegelstad);supported by the Norwegian National Research Council (# 133897/320 and # 154642/320), theNorwegianDepartment ofHealth andSocial Affairs, theNational Council forMentalHealth/Health and Rehabilitation (# 1997/41 and # 2002/306), Rogaland County and OsloCounty (Drs Vaglum, Johannessen, Friis, Larsen, Melle, Opjordsmoen). Also funded by theTheodore andVada Stanley Foundation, theRegionalHealth Research Foundation for EasternRegion, Denmark; Roskilde County, Helsefonden, Lundbeck Pharma, Eli Lilly and Janssen-Cilag Pharmaceuticals, Denmark (Drs Simonsen andHaahr). Also supported by aNational Al-liance for Research on Schizophrenia and Depression (NARSAD) Distinguished InvestigatorAward and NIMH grant MH-01654 (Dr. McGlashan) and a NARSAD Young InvestigatorAward (Dr. Larsen). Health South East (# 2008001), Health West (# 200202797-65) (IngeJoa) and # 911313 (Regional Centre for Clinical Research in Psychosis).

ContributorsAuthors SF, TM, IM, PV, SO, BRR, JIR, JOJ, TKL, IJ, ES and UH took part in designing the

study. Authors JE,WTVH and UH collected the data. Authors JHE, JIR,WTVH and SF under-took the statistical analysis. Author JE wrote the first draft of the manuscript.

All authors contributed to and have approved of the manuscript.

Conflict of interestAll authors declare that they have no conflicts of interest.

AcknowledgementsMany thanks to all the patients who have contributed to this study.

References

Andreasen, N.C., Carpenter Jr., W.T., Kane, J.M., Lasser, R.A., Marder, S.R., Weinberger,D.R., 2005. Remission in schizophrenia: proposed criteria and rationale for consen-sus. Am. J. Psychiatry 162 (3), 441–449.

Bentsen, H., Munkvold, O.G., Notland, T.H., Boye, B., Bjoerge, H., Lersbryggen, A.B.,Oskarsson, K.H., Berg-Larsen, R., Malt, U.F., 1996. The interrater reliability of the Posi-tive and Negative Syndrome Scale (PANSS). Int. J. Methods Psychiatr. Res. 6, 227–235.

Berenbaum, H., Oltmanns, T.F., 1992. Emotional experience and expression in schizo-phrenia and depression. J. Abnorm. Psychol. 101 (1), 37–44.

Bertelsen, M., Jeppesen, P., Petersen, L., Thorup, A., Øhlenschlaeger, J., Le Quach, P.,Østergaard Christensen, T., Krarup, G., Jørgensen, P., Nordentoft, M., 2009. Schizophr.Res. 107 (2–3), 173–178.

Blanchard, J.J., Cohen, A.S., 2006. The structure of negative symptoms within schizo-phrenia: implications for assessment. Schizophr. Bull. 32 (2), 238–245.

Bowie, C.R., Reichenberg, A., Patterson, T.L., Heaton, R.K., Harvey, P.D., 2006. Determinantsof real-world functional performance in schizophrenia subjects: correlations withcognition, functional capacity, and symptoms. Am. J. Psychiatry 163 (3), 418–425.

Chang, W.C., Hui, C.L., Tang, J.Y., Wong, G.H., Lam, M.M., Chan, S.K., Chen, E.Y., 2011. Per-sistent negative symptoms in first-episode schizophrenia: a prospective three-yearfollow-up study. Schizophr. Res. 133 (1–3), 22–28.

Dworkin, R.H., 1992. Affective deficits and social deficits in schizophrenia: what'swhat? Schizophr. Bull. 18 (1), 59–64.

Eaton,W.W., Thara, R., Federman, B.,Melton, B., Liang, K.Y., 1995. Structure and course of pos-itive and negative symptoms in schizophrenia. Arch. Gen. Psychiatry 52 (2), 127–134.

Edwards, J., McGorry, P.D., Waddell, F.M., Harrigan, S.M., 1999. Enduring negativesymptoms in first-episode psychosis: comparison of six methods using follow-updata. Schizophr. Res. 40 (2), 147–158.

Endicott, J., Spitzer, R.L., Fleiss, J.L., Cohen, J., 1976. The global assessment scale. A pro-cedure for measuring overall severity of psychiatric disturbance. Arch. Gen. Psychi-atry 33, 766–771.

Evensen, J., Røssberg, J.I., Barder, H., Haahr, U., Hegelstad, W.T., Joa, I., Johannessen, J.O.,Larsen, T.K., Melle, I., Opjordsmoen, S., Rund, B.R., Simonsen, E., Sundet, K., Vaglum,P., Friis, S., McGlashan, T., 2012. Apathy in first episode psychosis patients: a tenyear longitudinal follow-up study. Schizophr. Res. 136 (1–3), 19–24.

Faerden, A., Friis, S., Agartz, I., Barrett, E.A., Nesvag, R., Finset, A., Melle, I., 2009. Apathyand functioning in first episode psychosis. Psychiatr. Serv. 60 (11), 1495–1503.

Faerden, A., Finset, A., Friis, S., Agartz, I., Barrett, E.A., Nesvåg, R., Andreassen, O.A.,Marder, S.R., Melle, I., 2010. Apathy in first episode psychosis patients: one yearfollow up. Schizophr. Res. 116 (1), 20–26.

Fenton, W.S., McGlashan, T.H., 1991. Natural history of schizophrenia subtypes. II. Positiveand negative symptoms and long-term course. Arch. Gen. Psychiatry 48 (11), 978–986.

Fenton, W.S., McGlashan, T.H., 1992. Testing systems for assessment of negative symp-toms in schizophrenia. Arch. Gen. Psychiatry 49 (3), 179–184.

Foussias, G., Remington, G., 2010. Negative symptoms in schizophrenia. Avolition andOccam's Razor. Schizophr. Bull. 36 (2), 359–369.

Foussias, G., Mann, S., Zakzanis, K.K., van Reekum, R., Remington, G., 2009. Motivationaldeficits as the central link to functioning in schizophrenia: a pilot study. Schizophr.Res. 115 (2–3), 333–337.

Friis, S., Larsen, T.K., Melle, I., Opjordsmoen, S., Johannessen, J.O., Haahr, U., Simonsen,E., Rund, B.R., Vaglum, P., McGlashan, T., 2003. Methodological pitfalls in early de-tection studies — the NAPE Lecture 2002. Nordic Association for Psychiatric Epide-miology. Acta Psychiatr. Scand. 107, 3–9.

Gur, R.E., Kohler, C.G., Ragland, J.D., Siegel, S.J., Lesko, K., Bilker, W.B., Gur, R.C., 2006.Flat affect in schizophrenia: relation to emotion processing and neurocognitivemeasures. Schizophr. Bull. 32 (2), 279–287.

Hegelstad, W.V., Larsen, T.K., Auestad, B., Evensen, J., Haahr, U., Joa, Inge, Johannesen, J.O.,Langeveld, J., Melle, I., Opjordsmoen, S., Rossberg, J.I., Rund, B.R., Simonsen, M.D.,Sundet, K., Vaglum, P., Friis, S., McGlashan, T., 2012. Long-term follow-up of the TIPSearly detection in psychosis study: effects on ten-year outcome. Am. J. Psychiatry169 (4), 374–380.

Herbener, E.S., Harrow, M., 2001. Longitudinal assessment of negative symptoms inschizophrenia/schizoaffective patients, other psychotic patients, and depressedpatients. Schizophr. Bull. 27 (3), 527–537.

Ho, B.C., Nopoulos, P., Flaum, M., Arndt, S., Andreasen, N.C., 1998. Two-year outcome infirst-episode schizophrenia: predictive value of symptoms for quality of life. Am. J.Psychiatry 155 (9), 1196–11201.

Kay, S.R., Fiszbein, A., Opler, L.A., 1987. The positive and negative syndrome scale(PANSS) for schizophrenia. Schizophr. Bull. 13, 261–276.

Kelley, M.E., Haas, G.L., van Kammen, D.P., 2008. Longitudinal progression of negativesymptoms in schizophrenia: a new look at an old problem. Schizophr. Res. 105 (1–3),188–196.

Kiang,M., Christensen, B.K., Remington, G., Kapur, S., 2003. Apathy in schizophrenia: clinicalcorrelates and association with functional outcome. Schizophr. Res. 63 (1–2), 79–88.

Kirkpatrick, B., Galderisi, S., 2008. Deficit schizophrenia: an update. World Psychiatry 7(3), 143–147.

Kirkpatrick, B., Buchanan, R.W., McKenney, P.D., Alphs, L.D., Carpenter Jr., W.T., 1989.The Schedule for the Deficit syndrome: an instrument for research in schizophre-nia. Psychiatry Res. 30 (2), 119–123.

Kirkpatrick, B., Buchanan, R.W., Ross, D.E., Carpenter Jr., W.T., 2001. A separate diseasewithin the syndrome of schizophrenia. Arch. Gen. Psychiatry 58 (2)m, 165–171.

Kirkpatrick, B., Fenton, W.S., Carpenter Jr., W.T., Marder, S.R., 2006. The NIMH-MATRICSconsensus statement on negative symptoms. Schizophr. Bull. 32, 214–219.

Kohler, C.G., Walker, J.B., Martin, E.A., Healey, K.M., Moberg, P.J., 2010. Facial emotion per-ception in schizophrenia: a meta-analytic review. Schizophr. Bull. 36 (5), 1009–1019.

Krause, R., Steimer, E., Sänger-Alt, C., Wagner, G., 1989. Facial expression of schizo-phrenic patients and their interaction partners. Psychiatry 52 (1), 1–12.

Larsen, T.K., McGlashan, T.H., Johannessen, J.O., Friis, S., Guldberg, C., Haahr, U.,Horneland, M., Melle, I., Moe, L.C., Opjordsmoen, S., Simonsen, E., Vaglum, P.,2001. Shortened duration of untreated first episode of psychosis: changes in pa-tient characteristics at treatment. Am. J. Psychiatry 158, 1917–1919.

Larsen, T.K., Friis, S., Haahr, U., Johannessen, J.O., Melle, I., Opjordsmoen, S., Rund, B.R.,Simonsen, E., Vaglum, P.V., McGlashan, T.H., 2004. Premorbid adjustment in first-episode non-affective psychosis: distinct patterns of pre-onset course. Br. J. Psychiatry185, 108–115.

Lehman, A.F., 1988. A quality of life interview for the chronically mentally ill. Eval. Pro-gram Plann. 11, 51–62.

Lysaker, P.H., Davis, L.W., 2004. Social function in schizophrenia and schizoaffectivedisorder: associations with personality, symptoms, and neurocognition. HealthQual. Life Outcomes 2, 15.

Malla, A.K., Takhar, J.J., Norman, R.M., Manchanda, R., Cortese, L., Haricharan, R., Verdi,M., Ahmed, R., 2002. Negative symptoms in first episode non-affective psychosis.Acta Psychiatr. Scand. 105 (6), 431–439.

Malla, A.K., Norman, R.M., Takhar, J., Manchanda, R., Townsend, L., Scholten, D.,Haricharan, R., 2004. Can patients at risk for persistent negative symptoms be identi-fied during their first episode of psychosis? J. Nerv. Ment Dis. 192 (7), 455–463.

Messinger, J.W., Trémeau, F., Antonius, D., Mendelsohn, E., Prudent, V., Stanford, A.D.,Malaspina, D., 2011. Avolition and expressive deficits capture negative symptomphe-nomenology: implications for DSM-5 and schizophrenia research. Clin. Psychol. Rev.31 (1), 161–168.

Milev, P., Ho, B.C., Arndt, S., Andreasen, N.C., 2005. Predictive values of neurocognitionand negative symptoms on functional outcome in schizophrenia: a longitudinalfirst-episode study with 7-year follow-up. Am. J. Psychiatry 162 (3), 495–506.

Pogue-Geile, M.F., Harrow, M., 1985. Negative symptoms in schizophrenia: their longi-tudinal course and prognostic importance. Schizophr. Bull. 11 (3), 427–439.

Shtasel, D.L., Gur, R.E., Gallacher, F., Heimberg, C., Gur, R.C., 1992. Gender differences inthe clinical expression of schizophrenia. Schizophr. Res. 7 (3), 225–231.

Siegel, S.J., Irani, F., Brensinger, C.M., Kohler, C.G., Bilker, W.B., Ragland, J.D., Kanes, S.J.,Gur, R.C., Gur, R.E., 2006. Prognostic variables at intake and long-term level of func-tion in schizophrenia. Am. J. Psychiatry 163 (3), 433–441.

Spitzer, R.L., Williams, J.B., Gibbon, M., First, M.B., 1992. The Structured Clinical Inter-view for DSM-III-R (SCID). I: History, rationale, and description. Arch. Gen. Psychi-atry 49 (8), 624–629.

Strauss, J.S., Carpenter, W.T., 1974. The prediction of outcome in schizophrenia II. Rela-tionships between predictor and outcome variables: a report from the WHO inter-national pilot study of schizophrenia. Arch. Gen. Psychiatry 31 (1), 37–42.

Ventura, J., Hellemann, G.S., Thames, A.D., Koellner, V., Nuechterlein, K.H., 2009. Symp-toms as mediators of the relationship between neurocognition and functional out-come in schizophrenia: a meta-analysis. Schizophr. Res. 113 (2–3), 189–199.

Walker, E.F., Grimes, K.E., Davis, D.M., Smith, A.J., 1993. Childhood precursors of schizo-phrenia: facial expressions of emotion. Am. J. Psychiatry 150 (11), 1654–1660.

White, C., Stirling, J., Hopkins, R.,Morris, J.,Montague, L., Tantam,D., Lewis, S., 2009. Predictorsof 10-year outcome of first-episode psychosis. Psychol. Med. 39 (9), 1447–1456.

WHO, Collaborating Centre for Drug Statistics Methodology, 2008. ATC Index with DDD's.