psychiatric comorbidity in first episode schizophrenia: a 2 year, longitudinal outcome study

JOURNAL OF

PSYCHIATRIC

RESEARCHJournal of Psychiatric Research 40 (2006) 656–663

www.elsevier.com/locate/jpsychires

Psychiatric comorbidity in first episode schizophrenia:A 2 year, longitudinal outcome study q

Kang Sim a,b,*, Thiam Hee Chua a, Yiong Huak Chan c,Rathi Mahendran a, Siow Ann Chong b

a Department of Adult Psychiatry, Woodbridge Hospital/Institute of Mental Health, 10, Buangkok View, Singapore 539747, Singaporeb Department of Early Psychosis Intervention, Woodbridge Hospital/Institute of Mental Health, Singapore

c Biostatistics Unit, Yong Loo Lin School of Medicine, National University of Singapore, Singapore

Received 23 January 2006; received in revised form 21 June 2006; accepted 26 June 2006

Abstract

Objective: We have previously documented a high prevalence of Axis I psychiatric comorbidity in our patients with first episode psycho-sis. This study sought to determine the longitudinal impact of Axis I psychiatric comorbidity on patients with first episode schizophrenia(FES) and we hypothesised that patients with psychiatric comorbidity were associated with poorer clinical and functional outcomes.Method: One hundred and forty two consecutively hospitalized FES patients were included. Socio-demographic information wasobtained and the PANSS, SUMD, GAF, WHOQOL-Bref were used to assess psychopathology, insight, social/occupational functioningand quality of life respectively at baseline and at 6, 12, 18 and 24 months after discharge.Results: Over time and compared with baseline scores, patients with Axis I psychiatric comorbidity (n = 46, 32.4%) had significantly lessreduction of their PANSS total and subscale scores, less improvement in their awareness of their psychiatric illnesses and symptoms at12, 18 and 24 months and poorer insight into the consequences of their illness at 18 and 24 months. Poor insight at baseline was cor-related positively with PANSS negative symptom subdomain, and negatively with GAF at 24 months.Conclusion: Axis I Psychiatric comorbidity was associated with worse prospective outcomes in hospitalized patients with first episodeschizophrenia, and this highlights a greater need towards the early recognition and management of these conditions.� 2006 Elsevier Ltd. All rights reserved.

Keywords: Comorbidity; Psychiatric; First episode; Schizophrenia

1. Introduction

Recent follow up studies of patients with first episodenon-affective psychotic disorders including first episodeschizophrenia (FES) had found variable clinical or func-tional outcomes including non-deteriorating course interms of psychopathology (Mason et al., 1996), improved

0022-3956/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jpsychires.2006.06.008

q Previous presentation: Part of this paper was presented at the FourthInternational Conference on Early Psychosis held at Vancouver, Canadafrom 28th September to 1st October 2004.

* Corresponding author. Address: Department of Adult Psychiatry,Woodbridge Hospital/Institute of Mental Health, 10, Buangkok View,Singapore 539747, Singapore. Tel.: +65 63892000; fax: +65 63855900.

E-mail address: [email protected] (K. Sim).

positive (Gupta et al., 1997) or enduring positive (Edwardset al., 1999) symptoms, enduring negative symptoms(Edwards et al., 1999), changes in the level of awarenessof illness (Drake and Lewis, 2003; Kemp and Lambert,1996), considerable rates of readmission (Sipos et al.,2001), poor employment (Gupta et al., 1997; Marwahaand Johnson, 2004), improved (Addington et al., 2003) orreduced (Green et al., 2001; Priebe et al., 2000) subjectivequality of life (QOL), deterioration of social adjustment(Mason et al., 1996) and good three year overall outcomein symptoms and global functioning (Singh et al., 2000).In addition, these outcomes in the extant literature had dif-fering clinical correlates such as the association of animprovement of the awareness of illness with a reductionof negative symptoms (Kemp and Lambert, 1996) and level

mailto:[email protected]

K. Sim et al. / Journal of Psychiatric Research 40 (2006) 656–663 657

of depression (Drake et al., 2004; Smith et al., 1998),changes in the quality of life and its association with theduration of untreated psychosis, negative symptoms as wellas premorbid adjustment (Browne et al., 2000; Malla et al.,2004).

Multiple clinical factors may affect the prospective out-comes of patients with first episode schizophrenia includ-ing clinical factors such as age, age of onset of illness(Coldham et al., 2002), education, premorbid adjustment(Malla et al., 2004), duration of untreated psychosis(Browne et al., 2000), psychopathology (Edwards et al.,1999) and the level of insight (Dickerson et al., 1997).We argue that psychiatric comorbidity needs to be consid-ered early in the management of FES patients as well.This is supported by the evidence of relatively high prev-alence rates of Axis I psychiatric comorbidity in patientswith first break schizophrenia in our own cohort ofpatients and similar findings in other studies (Craiget al., 2002; Poyurovsky et al., 1999; Sim et al., 2004),as well as their complex relationships with clinical corre-lates such as quality of life and insight (Sim et al.,2005). However, despite this, data are sparse specifically,on the longitudinal impact of Axis I psychiatric comor-bidity on the clinical and functional outcomes of patientswith FES. Hence, we sought to examine the longitudinalimpact of comorbid Axis I psychiatric conditions onFES patients and hypothesised that patients with FESand Axis I psychiatric comorbidity were associated withpoorer longitudinal outcomes, especially in terms of moresevere psychopathology, poorer insight and QOL com-pared to those without Axis I psychiatric comorbidity.

2. Method

2.1. Study design and participants

The study population consisted of 142 consecutive sub-jects with FES enrolled in the National Early PsychosisIntervention Program at the Institute of Mental Health/Woodbridge Hospital in Singapore from March 2001 toMarch 2003 fulfilling the following criteria: (1) age between18 and 40 years, (2) English speaking, (3) presentation witha first episode schizophrenia, and had no previous psychi-atric hospitalisation or antipsychotic treatment. The Insti-tute of Mental Health/Woodbridge Hospital is the onlystate psychiatric hospital in the country, which also servesas a principal treatment and follow up facility for patientssuffering from severe psychotic illnesses such as schizophre-nia. The subjects were excluded in this study if the psy-chotic symptoms: (1) were secondary to acuteintoxication or withdrawal from alcohol or other psycho-active substances (Strakowski et al., 1995) or (2) resultedentirely from a medical illness as determined by a compre-hensive medical and neurological evaluation. The study hasthe approval of the Hospital Ethics Committee and all par-ticipants provided written, informed consent after a fullexplanation of the nature of the study.

All new psychiatric admissions were reviewed daily andat index hospitalisation, a total of 240 potential study par-ticipants were identified, of whom 162 (67.5%) met inclu-sion criteria, and 142 (59.2%) provided written, informedconsent for the study. All efforts were taken to ensure com-pleteness of follow up data and this was assisted by the useof case management approach within the clinical cumresearch program to enhance regularity with reviewappointments. Nevertheless, default rates of different indi-viduals at different points in time occurred at 6 (14.8%), 12(17.6%), 18 (19.0%) and 24 (23.9%) months and were takeninto account in the repeated measurement analyses. Therewere no significant differences in the basic demographicvariables (age and sex) of patients who were willing ornot willing to participate in the study.

2.2. Diagnostic assessment

Axis I psychiatric diagnoses were assessed by psychia-trists at the index hospitalisation and at 24 months bymeans of the Structured Clinical Interview for DSM IV-Axis I Disorders, Patient Edition (SCID-P) (First et al.,1994). In completing the SCID-P, combined symptominformation was obtained from various sources includingclinical interviews with the subjects as well as significantothers, medical records and other primary, treating clini-cians whenever necessary. There was good inter-rater reli-ability for principal Axis I diagnoses (k = 0.94),performed by the psychiatrists. Psychiatric comorbiditywas defined as the presence of any antecedent or concur-rent DSM-IV axis I disorder. The SCID hierarchy wasmaintained for the primary psychotic disorder but not forthe comorbid diagnoses. Syndromes which were clearly sec-ondary to the principal diagnosis were not assessed ascomorbid disorders (Sim et al., 2004).

2.3. Symptom assessment

The Positive and Negative Symptoms Scale (PANSS)(Kay et al., 1987), Scale to Assess Unawareness of MentalDisorders (SUMD) (Amador et al., 1994) and the GlobalAssessment of Functioning Scale (GAF, SCID Axis V)were used to assess the severity of psychopathology, levelof insight and psychosocial functioning respectively.Inter-rater reliability was obtained by rating 16 subjectsby both raters with intraclass coefficient (ICC) above 0.80for all the observer rated scales.

The subjective QOL at baseline was assessed using the26-item World Health Organization Quality of Life-BrefScale (WHOQOL-BREF), an abbreviated version of theWHOQOL-100 assessment instrument (The WHOQOLGroup, 1998). It is a valid and reliable self-rated instrumentwhich is sensitive to the health-related QOL in subjectswith psychotic illnesses (Herrman et al., 2002) and has beenwell validated in Asian subjects (Saxena et al., 2001; Yaoet al., 2002). The 26 items produce scores for four domainsrelated to QOL, namely physical health (activities of daily

Table 1Demographic and clinical characteristics of patients with (Group 1) andwithout (Group 2) psychiatric comorbidity at baselinea

Characteristic Group 1(N = 46)

Group 2(N = 96)

P valueb

GenderMale 31 (67.4) 49 (51.0) 0.07Female 15 (32.6) 47 (49.0)

Age, mean (SD), years 26.9 (5.9) 28.4 (6.9) 0.24c

EthnicityChinese 31 (67.4) 68 (70.8) 0.72Malay 13 (28.3) 22 (22.9)Indian 2 (4.3) 4 (4.2)Others 0 (0.0) 2 (2.1)

Marital statusSingle 33 (71.7) 67 (69.8) 0.06Married 12 (26.1) 19 (19.8)Separated/divorced 1 (2.2) 10 (10.4)

Education 0.85Less than 10 years 26 (56.5) 55 (57.3)More than 10 years 20 (43.5) 41 (42.7)

Employment 0.49Unemployed 20 (43.5) 40 (41.7)Employed 26 (56.5) 56 (58.3)

Living arrangementsLives alone 0 (0.0) 5 (5.2) 0.16Lives with immediate family 46 (100.0) 86 (89.6)Lives with others 0 (0.0) 5 (5.2)

DUP, mean (SD), months 14.2 (17.1) 13.7 (22.2) 0.35c

Medication, CPZ eq, mg/day 123.7 (89.6) 130.7 (149.5) 0.35c

Abbreviation: DUP, duration of untreated psychosis.a Data are presented as No. (%) unless otherwise noted.b P values derived from v2 test unless otherwise noted.c P values derived from Mann–Whitney test.

658 K. Sim et al. / Journal of Psychiatric Research 40 (2006) 656–663

living, dependence on medical treatment, energy and fati-gue, mobility, pain and discomfort, sleep, work capacity),psychological health (bodily image and appearance, nega-tive feelings, positive feelings, self esteem, spirituality, con-centration), social relationships (personal relationships,social support, sexual activity), and environment (finances,physical safety, access to health services, home environ-ment, opportunities to acquire new information, leisureactivities, physical environment, transport). In addition, italso includes an item each on the overall QOL and generalhealth.

The duration of untreated psychosis (DUP) was definedas the duration between the onset of psychotic symptomsand the time that treatment was initiated. Basic socio-demographic data were also collected.

2.4. Outcome assessment

Follow up assessments were scheduled for 6, 12, 18 and24 months after their index hospitalisation, interviews wereconducted with the patient, best informant and medicalrecords were also reviewed at periodic intervals after theindex hospitalisation. Clinical and symptomatic outcomeswere evaluated using PANSS, SUMD, and other dataincluding number and total duration of rehospitalisationfollowing the index hospitalisation and the medication dos-age in daily chlorpromazine equivalents. Functional out-come was assessed in the area of occupationalfunctioning as well as the GAF and WHOQOL-Brefscores. To improve the validity of the diagnosis and clinicalmeasures, ‘best estimate’ meetings were held at 12 monthsand after the completion of the 24 month follow up visit(Leckman et al., 1982) and involved the review of all assess-ments from index hospitalisation, follow up evaluationsand available clinical records. The comorbid Axis I psychi-atric diagnoses were also reviewed at each of the time pointof follow up at 6, 12, 18 and 24 months to assess whetherthey still met criteria for these conditions identified atbaseline.

2.5. Statistical analysis

Data was analysed using the Statistical Package forSocial Sciences (SPSS)-PC version 11.0 (SPSS Inc, Chi-cago, IL) and Statistical Analysis System 8.02 Proc Mixed(SAS Institute Inc, Cary, NC). The respective rating scoresover time were subjected to repeated measures analyses,using diagnosis (FES patients with and without psychiatriccomorbidity) as between-group factor and the differentclinical ratings over time as within-group factors. Signifi-cant interactions (diagnosis vs. time) were then exploredwith post-hoc change score analyses. Normality of quanti-tative data was also checked using the Kolmogorov–Smir-nov 1-sample test. Differences between the groups weretested by t-test and Mann–Whitney U-test for normaland non-normal continuous variables respectively and v2

test or Fisher exact test for categorical variables whenever

appropriate. Correlations for normally distributed datawere made with linear regression (Pearson’s r), and non-normally distributed data were correlated with a rank-method (Spearman’s rs). Multivariate linear regressionanalyses were also performed to determine significant base-line predictors of clinical and functional outcomes. Covar-iates entered into the multivariate linear regression analysesincluded age, gender, marital status, duration of untreatedpsychosis and specific Axis I psychiatric comorbidities. Sta-tistical significance was set at p < 0.05 (2-tailed).

3. Results

3.1. Demographic and clinical characteristics

Table 1 compares the basic demographic and clinicalcharacteristics of FES patients with and without Axis Ipsychiatric comorbidity. There were no significant baselinedifferences in terms of age, gender, marital status, employ-ment status, education, DUP, between patients with andwithout Axis I psychiatric comorbidity. Of the 142patients, psychiatric comorbidity was present in 32.4%(n = 46) of the patients: majority (97.9%, n = 43) had oneand 2.1% (n = 3) had two comorbid psychiatric diagnosis.


The comorbid Axis I psychiatric diagnoses includeddepression (n = 19, 13.4%), obsessive compulsive disorder(n = 9, 6.3%), social phobia (n = 8, 5.6%) and drug andalcohol abuse or dependence (n = 13, 9.1%).

At baseline, patients with psychiatric comorbidity hadlower PANSS total, positive subdomain and hostility(1.9 ± 1.3 vs. 2.8 ± 1.8, p < 0.05), grandiosity (1.2 ± 0.9vs. 1.7 ± 1.6, p < 0.05) symptom item scores (Table 2).Patients with psychiatric comorbidity had better insight(lower scores on SUMD) in the awareness of mental illness,consequences of the mental illness as well as effects of med-ications. However, patients with psychiatric comorbiditywere found to have poorer QOL scores in the overalldomain as well as physical (activities of daily living, workcapacity) psychological (positive feelings, bodily imageand appearance), social (personal relationships and socialsupport) and environment (physical safety and security,opportunities for acquiring new information and skills)subdomains (see Table 2).

Table 2Comparison of baseline SUMD, PANSS, GAF, WHOQOL-Bref scores inpatients with (Group 1) and without psychiatric comorbidity (Group 2)

Characteristica Group 1(N = 46)

Group 2(N = 96)

P-valued

SUMD (current awareness)b

Mental disorder 2.2 ± 0.7 2.7 ± 0.6 <0.001 (z = �3.49)Social consequences 2.2 ± 0.7 2.6 ± 0.6 0.001 (z = �3.43)Treatment efficacy 2.0 ± 0.8 2.3 ± 0.9 0.043 (z = �2.02)

PANSS

Total score 63.9 ± 19.7 70.3 ± 17.1 0.016 (z = �2.42)Positive symptom 17.2 ± 5.4 20.5 ± 6.3 0.004 (z = �2.84)Negative symptom 13.4 ± 7.7 15.1 ± 8.5 0.27General symptom 34.59 ± 13.10 33.9 ± 7.9 0.49GAF 43.8 ± 12.4 41.6 ± 14.9 0.31

WHOQOL-Brefc

Physical health 46.6 ± 13.7 53.8 ± 15.0 0.032 (z = �2.14)ADL 2.9 ± 1.2 3.6 ± 1.2 0.02 (z = �2.32)Working capacity 2.8 ± 1.2 3.6 ± 1.1 0.003 (z = �2.94)

Psychological health 49.0 ± 17.6 58.4 ± 16.7 0.031 (z = �2.16)Positive feelings 2.6 ± 1.2 3.4 ± 1.2 0.011 (z = �2.55)Bodily image 3.0 ± 1.1 3.8 ± 1.2 0.009 (z = �2.63)

Social relationships 39.2 ± 16.9 48.7 ± 18.7 0.015 (z = �2.44)Personal relationships 2.8 ± 1.2 3.6 ± 1.2 0.005 (z = �2.78)Social support 2.8 ± 1.2 3.6 ± 1.1 0.012 (z = �2.52)

Environment 49.9 ± 20.8 62.5 ± 16.1 0.015 (z = �2.42)Information 2.5 ± 1.2 3.5 ± 1.1 0.04 (z = �2.06)Transport 2.9 ± 1.2 3.8 ± 1.1 0.003 (z = �2.94)

Abbreviations: SUMD, Scale to Assess Unawareness of Mental Disorders;PANSS, Positive and Negative Symptoms Scale; GAF, Global Assessmentof Functioning; WHOQOL-Bref, World Health Organization Quality ofLife-Bref Scale, ADL, activities of daily living.

a Values reported as mean ± SD.b Lower scores indicate better insight.c Lower scores indicate poorer QOL, and included scores for the four

domains, overall quality of life, total score and also indicating which sub-items within each of the four domains were statistically significant in andof themselves.

d P values derived from Mann–Whitney test.

3.2. Clinical and functional outcomes

Over time, significant differences between the twogroups with and without psychiatric comorbidity (groupvs. time interactions) involved PANSS total scores(F = 154.13, p < 0.001), positive (F = 190.63, p < 0.001)and negative (F = 13.02, p < 0.001) subdomain scores aswell as awareness of psychiatric illness (F = 3.54,p < 0.01) and its consequences (F = 2.84, p < 0.05) (referFigs. 1–7).

Compared with baseline, patients with Axis I psychiatriccomorbidity had less reduction of their PANSS total scoresat 12 (�26.76 vs. �33.21, z = �2.26, p < 0.05), 18 (�24.00vs. �32.82, z = �2.29, p < 0.05), 24 months (�26.71 vs.�34.33, z = �2.19, p < 0.05); PANSS positive subdomainscores at 6 (�9.32 vs. �11.78, z = �2.15, p < 0.05), 12(�9.10 vs. �11.71, z = �2.33, p < 0.05), 18 (�8.68 vs.�11.27, z = �2.30, p < 0.05), 24 months (�9.09 vs.�11.36, z = �2.05, p < 0.05); and negative subdomainscores at 18 months (�2.38 vs. �7.32, z = �2.19,p < 0.05), indicating more severe psychopathology over

-45

-40

-35

-30

-25

-20

-15

-10

-5

0

0 6 12* 18* 24*

Time (months)

PAN

SS

to

tal s

core

(C

han

ge

fro

m b

asel

ine) Patients with

psychiatriccomorbidity

Patientswithoutpsychiatriccomorbidity

Fig. 1. Change in PANSS total scores between the two groups over 24months. (*<0.05).

-16

-14

-12

-10

-8

-6

-4

-2

0

0 6* 12* 18* 24*

Time (months)

PAN

SS

po

siti

ve s

ymp

tom

su

bsc

ale

(Ch

ang

e fr

om

bas

elin

e) Patients withpsychiatriccomorbidity


Fig. 2. Change in PANSS positive symptom subscale scores between thetwo groups over 24 months. (*<0.05).

-12

-10

-8

-6

-4

-2

0

2

0 6 12 18* 24

Time (months)

PAN

SS

neg

ativ

e sy

mp

tom

su

bsc

ale

(Ch

ang

e fr

om

bas

elin



Fig. 3. Change in PANSS negative symptom subscale scores between thetwo groups over 24 months. (*<0.05).

-18

-16

-14

-12

-10

-8

-6

-4

-2

0

0 6 12 18 24

Time (months)

PAN

SS

gen

eral

psy

cho

pat

ho

log

y su

bsc

ale

(Ch

ang

e fr

om

bas

elin



Fig. 4. Change in PANSS general psychopathology subscale scoresbetween the two groups over 24 months.

-1.2

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0 6 12* 18* 24*

Time (months)

Aw

aren

ess

of

psy

chia

tric

illn

ess

(Ch

ang

e fr

om

bas

elin

e)

Patients withpsychiatriccomorbidity

Patients withoutpsychiatriccomorbidity

Fig. 5. Change in insight scores (awareness of psychiatric illness, SUMD1) between the two groups over 24 months. (*<0.05).

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0 6 12 18* 24*

Time (months)

Aw

aren

ess

of

con

seq

uen

ces

(Ch

ang

e fr

om

bas

elin



Fig. 6. Change in insight scores (awareness of consequences of psychiatricillness, SUMD 2) between the two groups over 24 months. (*<0.05).

0

10

20

30

40

0 6 12 18 24

Time (months)

GA

F

(Ch

ang

e fr

om

bas

elin



Fig. 7. Change in GAF scores between the two groups over 24 months.


time and when compared to patients without psychiatriccomorbidity. In addition, patients with psychiatric comor-bidity had less reduction of their insight scores at 12(awareness of mental illness, �0.09 vs. �0.68, z = �2.36,

p < 0.05), 18 (awareness of mental illness, �0.21 vs.�0.74, z = �2.15, p < 0.05; and its consequences, �0.17vs. �0.62, z = �2.15, p < 0.05) and 24 months (awarenessof mental illness, �0.12 vs. �0.62, z = �2.52, p < 0.05and its consequences, �0.04 vs. �0.53, z = �2.66,p < 0.01) when compared with baseline measures andpatients without comorbidity, indicating a generally poorerlevel of insight over time. There were no significant differ-ences between the two groups with and without Axis I psy-chiatric comorbidity in terms of total duration ofhospitalization as well as daily CPZ mg equivalents at base-line and at 6 monthly intervals.

In terms of correlations, within the group of patientswith psychiatric comorbidity, the duration of untreatedpsychosis was significantly correlated with chlorpromazineequivalents at baseline (rs = 0.35, p < 0.05), PANSS nega-tive symptom subdomain scores at 24 months (rs = 0.34,p < 0.05); age was negatively correlated with QOL physical(rs = �0.70, p < 0.05), psychological (rs = �0.79, p < 0.01),social (rs = �0.86, p < 0.01), environment (rs = �0.70,p < 0.05) subdomain scores at 18 months; and the pooreroverall level of insight at baseline was correlated with


negative symptoms (rs = 0.43, p < 0.05), and GAF at 24months (rs = �0.45, p < 0.05).

On multivariate regression analyses, clinical outcome interms of greater severity of psychopathology at 24 monthswas predicted by a longer duration of untreated psychosis(p < 0.05). Significant predictors of a better level of psycho-social functioning at 24 months were a lower SUMD score(better insight) (p < 0.05) and lower level of psychopathol-ogy (p < 0.01) at baseline.

4. Discussion

This study highlighted a few findings: first, better insightbut a poorer subjective QOL at baseline amongst patientsadmitted with FES and Axis I psychiatric comorbidity.Second, there was a significant reduction in the level ofinsight and non-reduction in the severity of psychopathol-ogy over time. Third, in patients with Axis I psychiatriccomorbidity, DUP was positively correlated with baselinemedication dosages as well as negative symptom scores at24 months and poor insight at baseline was correlated withnegative symptoms and inversely correlated with GAF at24 months. Fourth, a longer DUP was a significant predic-tor of greater psychopathology longitudinally and signifi-cant predictors of a lower level of psychosocialfunctioning were poorer insight and more severe illness atbaseline.

At baseline, the better level of insight in FES patientscould be related to these patients having a lower level ofpsychopathology in the PANSS total and positive subdo-main scores (Debowska et al., 1998). This is consistent withfindings of previous studies which have found relationshipsbetween decreased positive symptomatology and betterinsight (Amador et al., 1994; Fennig et al., 1996; Mintzet al., 2003). However, presence of poorer QOL in the pres-ence of better insight may suggest that patients with psychi-atric comorbidity tend to evaluate their QOL moreadversely or that they have inherently poorer quality livesat baseline. This may be a reflection of the sense of self-vul-nerability towards the onset of a highly stressful event suchas a psychotic illness (Bechdolf et al., 2002) as suggested bythe lower WHOQOL item scores on personal sense ofsafety and security; or the poorer underlying self constructand less adaptive coping strategies adopted by these indi-viduals to deal with their psychotic illness and psychiatriccomorbidity (Ruggeri et al., 2001; Zissi et al., 1998) as sug-gested by the lower WHOQOL item scores on sense ofpositive feelings, body image and ability to access newinformation or skills.

Longitudinally, patients with psychiatric comorbidityshowed less improvement of their psychopathology scorescompared to those patients without, suggesting the adverseimpact that comorbid psychiatric disorders may have onthe primary psychotic illness. In particular, the non-reduc-tion in positive and negative symptom subdomains in theseindividuals may further affect their awareness of their men-tal illness, its consequences as well as the need for treat-

ment as was also suggested by the correlation analysis.The reduction of insight over time is in contrast to thestudy by Cuesta et al. (2000) and Smith et al. (2004) whofound an improvement in the level of insight across thesubjects generally over time, although they studied patientswho had chronic illness, who had no comorbid psychiatricdiagnosis and who were only followed up for 6 months.Conversely, the lower level of insight in the patients mayalso interact with the severity of psychopathology overtime (Fennig et al., 1996; Sanz et al., 1998), and negativelyinfluence the subjective quality of life (Ho et al., 1998).

In our study, DUP was found to be a significant predic-tor of greater psychopathology, and worse symptoms andpoorer insight were predictive of poorer psychosocial func-tioning at 24 months. The inter-relationships betweenDUP, insight, symptomatic and functional outcomes arecomplex. An increased DUP was not only associated withincomplete symptom remission and recovery (Robinsonet al., 2004) but also poorer insight (Drake et al., 2000;Mintz et al., 2003) and social relationships (Malla et al.,2004). In FES patients with psychiatric comorbidity, theDUP was correlated with chlorpromazine equivalents atbaseline and the severity of negative symptoms which wasknown to affect negatively the subjective quality of life(Browne et al., 2000). The association of psychiatriccomorbidity, increased DUP with increased daily chlor-promazine equivalents at baseline may provide further sup-port of the neurotoxic potential of a longer period ofuntreated psychotic illness (Wyatt, 1991), requiringincreased medication dosage for treatment. As a caveat,although recent systematic reviews and meta-analytic stud-ies have also found evidence for modest associationsbetween DUP and outcome (Marshall et al., 2005; Perkinset al., 2005), most of the data are correlational in natureand hence firm conclusions regarding causality are not pos-sible. In this aspect, there is a need for longer term followup studies and results from non-correlational studies, suchas quasi-experimental designs could be potentially valuableand provide stronger evidence regarding causality (Nor-man et al., 2005). With regard to awareness of illness, thedecreased level of insight may be attributed to underlyingcognitive deficits in multiple domains suggestive of dys-function of neural networks involved in memory, learning,and executive functions (Drake and Lewis, 2003; Keshavanet al., 2004) and which involved the frontal (Laroi et al.,2000) and temporal cortices (Ha et al., 2004). Furthermore,we found that the level of psychopathology at baselineaffected longitudinal functional outcomes. The poorerinsight may also be related to more severe illness (Mintzet al., 2003; Williams and Collins, 2002) and both these fac-tors may, in turn, further impact negatively on treatmentadherence (Coldham et al., 2002; Smith et al., 1999). Thecorrelation between poorer insight at baseline and lowerGAF scores at 24 months was consistent with that of pre-vious studies (Schwartz et al., 1997) which also found alower level of interpersonal and social functioning (Francisand Penn, 2001; Lysaker et al., 1998) prospectively in their


cohort of patients. The lack of correlation between theobserver rated GAF and subjective QOL is not surprisingas there were often noted discrepancies between observerratings of functional activity and subjective ratings of thesense of well being by patients with severe mental illness(Sainfort et al., 1996; Voruganti et al., 1998). The associa-tion of an increased age with a poorer QOL should alertclinicians towards the need to identify a subset of patientswho may be disadvantaged and more predisposed to nega-tive perceptions of their sense of well-being and who maybenefit from case management (Jin et al., 2003), supportivetherapy or other forms of psychosocial intervention (Rits-ner et al., 2000).

There are several limitations in this study. First, thelength of follow up is limited to two years and improve-ments in the clinical or functional domains may occur attime points after the first two years. Second, we onlyincluded inpatients with first episode psychosis between18 and 40 years of age, thus disallowing the results to begeneralisable to all first episode cases including those whowere not admitted to a psychiatric facility and who fell out-side the age inclusion criteria. Third, we did not parse outassociations of particular comorbid conditions with specificoutcomes due to the small numbers when taking each sub-group. Fourth, other outcome variables such as cognitivefunction and contributors towards longitudinal outcomessuch as personality factors and coping resources were notstudied.

In conclusion, clinicians should be aware of the signifi-cant rates of Axis I psychiatric comorbidity in patients pre-senting with first episode schizophrenia which are notconfounded by the effects of treatment or illness chronicity.Its association with worse longitudinal outcomes calls forearlier identification and appropriate management of theseindividuals with the hope of improving not only their psy-chopathology but also their insight into the illness as wellas subjective sense of health.

Acknowledgement

This study was supported by National HealthcareGroup Research Grant (K.S.; STP/02003).

References

Addington J, Young J, Addington D. Social outcome in early psychosis.Psychological Medicine 2003;33:1119–24.

Amador XF, Flaum M, Andreasen NC, et al. Awareness of illness inschizophrenia and schizoaffective and mood disorders. Archives of theGeneral Psychiatry 1994;51:826–36.

Bechdolf A, Schultze-Lutter F, Klosterkotter J. Self-experienced vulner-ability, prodromal symptoms and coping strategies preceding schizo-phrenic and depressive relapses. European Psychiatry 2002;17:384–93.

Browne S, Clarke M, Gervin M, et al. Determinants of quality of life atfirst presentation with schizophrenia. British Journal of Psychiatry2000;176:173–6.

Coldham EL, Addington J, Addington D. Medication adherence ofindividuals with a first episode of psychosis. Acta PsychiatricaScandinavia 2002;106:286–90.

Craig T, Hwang MY, Bromet EJ. Obsessive-compulsive and panicsymptoms in patients with first-admission psychosis. American Journalof Psychiatry 2002;159:592–8.

Cuesta MJ, Peralta V, Zarzuela A. Reappraising insight in psychosis.Multi-scale longitudinal study. British Journal of Psychiatry 2000;177:233–40.

Debowska G, Grzywa A, Kucharska-Pietura K. Insight in paranoidschizophrenia – its relationship to psychopathology and premorbidadjustment. Comprehensive Psychiatry 1998;39:255–60.

Dickerson FB, Boronow JJ, Ringel N, et al. Lack of insight amongoutpatients with schizophrenia. Psychiatric Services 1997;48:195–9.

Drake RJ, Lewis SW. Insight and neurocognition in schizophrenia.Schizophrenia Research 2003;62:165–73.

Drake RJ, Haley CJ, Akhtar S, et al. Causes and consequences ofduration of untreated psychosis in schizophrenia. British Journal ofPsychiatry 2000;177:511–5.

Drake RJ, Pickles A, Bentall RP, et al. The evolution of insight, paranoiaand depression during early schizophrenia. Psychological Medicine2004;34:285–92.

Edwards J, McGorry PD, Waddell FM, et al. Enduring negativesymptoms in first-episode psychosis: comparison of six methods usingfollow-up data. Schizophrenia Research 1999;40:147–58.

Fennig S, Everett E, Bromet EJ, et al. Insight in first-admission psychoticpatients. Schizophrenia Research 1996;22:257–63.

First MB, Spitzer RL, Gibbon M, et al. Structured Clinical Interviewfor DSM IV Axis I Disorders, Patient Edition (SCID-P), version2. New York: New York State Psychiatric Institute, BiometricsResearch; 1994.

Francis JL, Penn DL. The relationship between insight and social skill inpersons with severe mental illness. Journal of Nervous and MentalDisorders 2001;189:822–9.

Green CA, Fenn DS, Moussaoui D, et al. Quality of life in treated andnever-treated schizophrenic patients. Acta Psychiatrica Scandinavia2001;103:131–42.

Gupta S, Andreasen NC, Arndt S, et al. The Iowa Longitudinal Study ofRecent Onset Psychosis: one-year follow-up of first episode patients.Schizophrenia Research 1997;23:1–13.

Ha TH, Youn T, Ha KS, et al. Gray matter abnormalities in paranoidschizophrenia and their clinical correlations. Psychiatry Research2004;132:251–60.

Herrman H, Hawthorne G, Thomas R. Quality of life assessment inpeople living with psychosis. Social Psychiatry and PsychiatricEpidemiology 2002;37:510–8.

Ho BC, Nopoulos P, Flaum M, et al. Two-year outcome in first-episodeschizophrenia: predictive value of symptoms for quality of life.American Journal of Psychiatry 1998;155:1196–201.

Jin H, Folsom DP, Lindamer L, et al. Patterns of public mental healthservice use by age in patients with schizophrenia. American Journal ofGeriatric Psychiatry 2003;11:525–33.

Kay SR, Fiszbein A, Opler LA. The positive and negative syndromescale (PANSS) for schizophrenia. Schizophrenia Bulletin 1987;13:261–76.

Kemp RA, Lambert TJ. Insight in schizophrenia and its relationship topsychopathology. Schizophrenia Research 1996;18:21–8.

Keshavan MS, Rabinowitz J, DeSmedt G, et al. Correlates of insight infirst episode psychosis. Schizophrenia Research 2004;70:187–94.

Laroi F, Fannemel M, Ronneberg U, et al. Unawareness of illness inchronic schizophrenia and its relationship to structural brain measuresand neuropsychological tests. Psychiatry Research 2000;100:49–58.

Leckman JF, Sholomskas D, Thompson WD, et al. Best estimate oflifetime psychiatric diagnosis: a methodological study. Archives of theGeneral Psychiatry 1982;39:879–83.

Lysaker PH, Bell MD, Bryson GJ, et al. Insight and interpersonalfunction in schizophrenia. Journal of Nervous and Mental Disorders1998;186:432–6.

Malla AK, Norman RM, McLean TS, et al. Determinants of quality oflife in first-episode psychosis. Acta Psychiatrica Scandinavia 2004;109:46–54.


Marshall M, Lewis S, Lockwood A, et al. Association between durationof untreated psychosis and outcome in cohorts of first-episodepatients: a systematic review. Archives of the General Psychiatry2005;62:975–83.

Marwaha S, Johnson S. Schizophrenia and employment – a review. SocialPsychiatry and Psychiatric Epidemiology 2004;39:337–49.

Mason P, Harrison G, Glazebrook C, et al. The course of schizophreniaover 13 years. A report from the International Study on Schizophrenia(ISoS) coordinated by the World Health Organization. British Journalof Psychiatry 1996;169:580–6.

Mintz AR, Dobson KS, Romney DM. Insight in schizophrenia: a meta-analysis. Schizophrenia Research 2003;61:75–88.

Norman RM, Lewis SW, Marshall M. Duration of untreated psychosisand its relationship to clinical outcome. British Journal of Psychiatry(Supplement) 2005;48:s19–23.

Perkins DO, Gu H, Boteva K, et al. Relationship between duration ofuntreated psychosis and outcome in first-episode schizophrenia: acritical review and meta-analysis. American Journal of Psychiatry2005;162:1785–804.

Poyurovsky M, Fuchs C, Weizman A. Obsessive-compulsive disorder inpatients with first-episode schizophrenia. American Journal of Psychi-atry 1999;156:1998–2000.

Priebe S, Roeder-Wanner UU, Kaiser W. Quality of life in first-admittedschizophrenia patients: a follow-up study. Psychological Medicine2000;30:225–30.

Ritsner M, Modai I, Endicott J, et al. Differences in quality of lifedomains and psychopathologic and psychosocial factors in psychiatricpatients. Journal of Clinical Psychiatry 2000;61:880–9.

Robinson DG, Woerner MG, McMeniman M, et al. Symptomatic andfunctional recovery from a first episode of schizophrenia or schizoaf-fective disorder. American Journal of Psychiatry 2004;161:473–9.

Ruggeri M, Bisoffi G, Fontecedro L, et al. Subjective and objectivedimensions of quality of life in psychiatric patients: a factor analyticalapproach: The South Verona Outcome Project 4. British Journal ofPsychiatry 2001;178:268–75.

Sainfort F, Becker M, Diamond R. Judgments of quality of life ofindividuals with severe mental disorders: Patient self-report versusprovider perspectives. American Journal of Psychiatry 1996;153:497–502.

Sanz M, Constable G, Lopez-Ibor I, et al. A comparative study of insightscales and their relationship to psychopathological and clinicalvariables. Psychological Medicine 1998;28:437–46.

Saxena S, Carlson D, Billington R. The WHO quality of life assessmentinstrument (WHOQOL-Bref): the importance of its items for cross-cultural research. Quality Life Research 2001;10:711–21.

Schwartz RC, Cohen BN, Grubaugh A. Does insight affect long-termimpatient treatment outcome in chronic schizophrenia? Comprehen-sive Psychiatry 1997;38:283–8.

Sim K, Swapna V, Mythily S, et al. Psychiatric comorbidity in firstepisode psychosis: the Early Psychosis Intervention Program (EPIP)experience. Acta Psychiatrica Scandinavia 2004;109:23–9.

Sim K, Mahendran R, Chong SA. Health related-quality of life andpsychiatric comorbidity in first episode psychosis. ComprehensivePsychiatry 2005;46:278–83.

Singh SP, Croudace T, Amin S, et al. Three-year outcome of first-episodepsychoses in an established community psychiatric service. BritishJournal of Psychiatry 2000;176:210–6.

Sipos A, Harrison G, Gunnell D, et al. Patterns and predictors ofhospitalisation in first-episode psychosis. Prospective cohort study.British Journal of Psychiatry 2001;178:518–23.

Smith TE, Hull JW, Santos L. The relationship between symptoms andinsight in schizophrenia: a longitudinal perspective. SchizophreniaResearch 1998;33:63–7.

Smith TE, Hull JW, Goodman M, et al. The relative influences ofsymptoms, insight, and neurocognition on social adjustment inschizophrenia and schizoaffective disorder. Journal of Nervous andMental Disorders 1999;187:102–8.

Smith TE, Hull JW, Huppert JD, et al. Insight and recovery frompsychosis in chronic schizophrenia and schizoaffective disorderpatients. Journal of Psychiatric Research 2004;38:169–76.

Strakowski SM, Keck Jr PE, McElroy SL, et al. Chronology of comorbidand principal syndromes in first-episode psychosis. ComprehensivePsychiatry 1995;36:106–12.

The WHOQOL Group. Development of the World Health OrganizationWHOQOL-BREF quality of life assessment. Psychological Medicine1998;28:551–8.

Voruganti L, Heslegrave R, Awad AG, et al. Quality of life measurement inschizophrenia: reconciling the quest for subjectivity with the question ofreliability. Psychological Medicine 1998;28:165–72.

Williams CC, Collins A. Factors associated with insight among outpatientswith serious mental illness. Psychiatric Services 2002;53:96–8.

Wyatt RJ. Early intervention with neuroleptics may decrease the long-term morbidity of schizophrenia. Schizophrenia Research 1991;5:201–2.

Yao G, Chung CW, Yu CF, et al. Development and verification ofvalidity and reliability of the WHOQOL-BREF Taiwan version.Journal of the Formos Medical Association 2002;101:342–51.

Zissi A, Barry MM, Cochrane RA. Mediational model of quality of lifefor individuals with severe mental health problems. PsychologicalMedicine 1998;28:1221–30.

psychiatric comorbidity in first episode schizophrenia: a 2 year, longitudinal outcome study

Documents