Welcome to the Schizophrenia Resource Centre

Welcome, this website is intended for healthcare professionals in EMEA with an interest in the treatment of schizophrenia. By clicking the link below you are declaring and confirming that you are a healthcare professional

You are here

Outcome in early-onset schizophrenia revisited: Findings from the Early Psychosis Prevention and Intervention Centre long-term follow-up study

Schizophrenia Research

Abstract

Objective

To compare the long-term outcome in individuals with early-onset (before age 18) and adult-onset schizophrenia spectrum disorder who were initially diagnosed and treated in the same clinical center.

Method

A prospective follow-up study of 723 consecutive first-episode psychosis patients (age range 14 to 30 years) on average 7.4 years after initial presentation to an early psychosis service, the Early Psychosis Prevention and Intervention Centre in Melbourne, Australia. The outcome measures included the Brief Psychiatric Rating Scale, the Schedule for the Assessment of Negative Symptoms, the Beck Depression Inventory, the Global Assessment of Functioning Scale, the Social and Occupational Functioning Assessment Scale, and the Quality of Life Scale.

Results

Follow-up interviews were conducted on 66.9% (484/723) individuals, of whom 75.6% (366/484) received a schizophrenia spectrum disorder diagnosis at baseline. Early-onset schizophrenia spectrum disorder was observed in 11.2% (41/366). At follow-up, individuals with early-onset reported significantly fewer positive symptoms and were characterised by significantly superior functioning on measures assessing global functioning, social-occupational functioning, and community functioning than individuals with adult-onset. The early-onset group also achieved significantly better vocational outcomes and had a more favourable course of illness with fewer psychotic episodes over the last two years prior to follow-up. Finally, when investigated as a continuous variable, younger age at onset significantly correlated with better symptomatic and functional outcomes.

Conclusions

These results question the assumption that early-onset schizophrenia typically has a poor outcome. Early detection and specialised treatment for the first psychotic episode appear to be more effective at improving long-term functional outcomes in people with early-onset schizophrenia as in those with adult-onset schizophrenia. This possibility and the reasons for it need further investigation.

Keywords: Long-term follow-up, Prospective, Outcome, Early-onset schizophrenia, Adult-onset schizophrenia.

1. Introduction

It has been suggested that early-onset schizophrenia (with onset before age 18) may represent a more severe variant of schizophrenia (Schulz et al, 1998, Hollis, 2000, Kyriakopoulos and Frangou, 2007, Kumra et al, 2008, and Sikich, 2008). Although schizophrenia often commences in adolescence (Loranger, 1984 and Amminger et al, 2006), both the number of long-term follow-up studies in early-onset schizophrenia and the sample sizes investigated are small (AACAP, 2001 and Reichert et al, 2008). Most of these studies report a poorer long-term outcome for schizophrenia with onset in adolescence compared with onset in adulthood (Hollis, 2000, AACAP, 2001, Fleischhaker et al, 2005, and Reichert et al, 2008). However, since early-onset and adult-onset patients are traditionally treated in separate clinical services the notion that outcome in early-onset schizophrenia may be worse than in adult-onset schizophrenia is based on samples that are not drawn from the same population, and may therefore be subject to selection bias. Furthermore, long-term follow-up studies in early-onset schizophrenia largely lack standardised sampling methods or information on recruitment and completeness or the representativeness of the samples investigated (Fleischhaker et al, 2005 and Reichert et al, 2008).

Since the 1990's early detection and specialised intervention programmes for people with first-episode schizophrenia and other psychoses have been implemented in many countries (Edwards and McGorry, 2002). Some of these programmes recruit across the traditional age divide of psychiatric services, offering new opportunities for comparative research. The Early Psychosis Prevention and Intervention Centre (EPPIC) was founded in 1992 in Melbourne, Australia. EPPIC provides a comprehensive, integrated, community based treatment programme for first-episode psychosis patients aged 14–30 years, drawn from a defined urban catchment area in northwestern Melbourne of approximately 850,000. The catchment area is socio-economically disadvantaged, has minimal private sector services, and those that exist typically refer first-episode psychosis cases to EPPIC. It is therefore estimated that the number of new cases accepted into EPPIC (approximately 240 per year) represents close to the entire treated incidence of first-episode psychosis in this age group within the catchment area (Amminger et al., 2006).

The EPPIC long-term follow-up study was designed in an attempt to address methodological shortcomings of previous longitudinal studies in schizophrenia (Henry et al., 2007). The study provides baseline data and a naturalistic, prospective follow-up of a large cohort of 723 consecutive patients experiencing a first episode of psychosis, who had been treated at EPPIC for up to the first two years of their illness, at a median of 7.4 years after their initial presentation. To our knowledge, no previous follow-up study in schizophrenia has investigated long-term (≥ 5 years) outcome across the traditional age divide (< 18 years vs. ≥ 18 years) in individuals who were recruited, diagnosed and treated for their first psychotic episode within the same clinical facility.

The purpose of our study was to compare long-term outcome in individuals with early-onset (onset before age 18) and adult-onset schizophrenia spectrum disorder in an epidemiologically representative cohort of first-episode patients. Based on previous studies (Hollis, 2000, AACAP, 2001, Fleischhaker et al, 2005, and Reichert et al, 2008) we hypothesised that an early-onset would be associated with more severe psychopathology and lower psychosocial functioning at follow-up.

2. Methods

2.1. Study overview and sample

This analysis was carried out within a naturalistic, prospective long-term follow-up of a cohort of 723 consecutive first-episode psychosis patients. All patients were detected and initially treated (for up to two years) at EPPIC or its immediate predecessor, the Aubrey Lewis Unit (McGorry et al., 1996). The baseline characteristics of the cohort and the methodology of the EPPIC long-term follow-up study of first-episode psychosis are detailed elsewhere (Henry et al., 2007). In brief, the inclusion criteria were: being aged between 14 and 30 years, having a DSM-III-R/DSM-IV diagnosis of a psychotic disorder and experiencing a first treated episode of psychosis with less than 6 months of prior neuroleptic medication, and living in the study catchment area. Patients were also required to have an adequate comprehension of the English language, and to give their informed consent for participation in the study. Axis I diagnoses were determined using the Royal Park Multidiagnostic Instrument for Psychosis (McGorry et al., 1990) or the Structured Clinical Interview for DSM (SCID I/P) upon service entry. Exclusion criteria were primary organic mental disorder, intellectual disability, drug and/or alcohol-induced psychosis, and epilepsy.

Long-term follow-up assessments were carried out between 1998 and 2005. Participants were traced in chronological order from the date of their baseline assessment. Participants' written informed consent was obtained for study participation and access to clinical records and informants. Approval was obtained from the relevant Human Research Ethics Committees for the study and the collection of data from medical records and informants for non-interviewed individuals (people who refused, could not be contacted, or were deceased). Seventy-five percent (542/723) of the first-episode cohort received a schizophrenia spectrum diagnosis (i.e., schizophrenia, schizophreniform disorder, schizoaffective disorder, delusional disorder, brief psychotic disorder, and psychosis not otherwise specified) at service entry. Of these individuals, 67.5% (366/542) were available for a long-term follow-up interview; they comprise the present study sample (Fig. 1). Comparisons of interviewed and non-interviewed groups on baseline demographic and clinical characteristics revealed only minor, non-significant differences (Table 1).

gr1

Fig. 1 Flow chart outlining patient participation.

Table 1 Baseline characteristics in individuals with first-episode schizophrenia spectrum disorder by interview status. Values are numbers (%) unless stated otherwise.

Baseline characteristics Interviewed group (n = 366) Non-interviewed group (n = 176) P valuea
Age at service entry, Mean (SD) 21.9 (3.5) 21.9 (3.6) 0.98
Male 258 (70.5) 135 (76.7) 0.13
Educational level
 Some secondary 167 (59.6) 99 (69.7)
 Secondary 42 (15.0) 18 (12.7)
 Post-secondary 71 (25.4) 25 (17.6)
Never married 272 (87.7) 131 (87.3) 0.98
Baseline diagnosis
 Schizophrenia 174 (47.5) 87 (49.4)
 Schizophreniform disorder 102 (27.9) 51 (29.0)
 Schizoaffective disorder 48 (13.1) 21 (11.3)
 Delusional disorder 7 (1.9) 3 (1.7)
 Brief psychotic disorder 7 (1.9) 8 (4.5)
 Psychotic disorder not otherwise specified 28 (7.7) 6 (3.4)

aBased on independent samples t-test or chi-square test.

2.2. Early-onset definition

To best characterise the age of onset and classify types of schizophrenia, past research has used cut-off points (i.e., age-strata) (for reviews see AACAP, 2001 and Kyriakopoulos and Frangou, 2007). We applied age 17 years as our upper cut-off to define early-onset schizophrenia which corresponds with the upper age cut-off in most published studies of child and adolescent psychosis, and is also the lower age cut-off for most adult psychosis studies and clinical services (Hollis, 2000b). Additional empirical evidence for this cut-off point has been provided by Rabinowitz et al. (2006). For consistency with previous studies, and because first admission has been suggested as the clearest evidence of onset that can be obtained (DeLisi, 1992), the age at service entry was used to define onset.

2.3. Long-term follow-up assessments

2.3.1. Demographics

The demographic characteristics of the cohort were recorded using a modified version of the World Health Organization International Study of Schizophrenia (WHO ISoS) information sheets (Sartorius et al., 1996).

2.3.2. Psychopathology

Psychiatric measures included the Brief Psychiatric Rating Scale-Expanded Version (BPRS-E) (Lukoff et al., 1986), the Schedule for the Assessment of Negative Symptoms (SANS) (Andreasen, 1982), and the Beck Depression Inventory (BDI) (Beck and Beck, 1972).

2.3.3. Functioning

Measures of functioning included the Global Assessment of Functioning Scale (GAF) (Endicott et al., 1976), the Social and Occupational Functioning Assessment Scale (SOFAS) (Goldman et al., 1992), and the Quality of Life Scale (QLS) (Heinrichs et al., 1984).

2.3.4. Employment

Vocational status at follow-up was assessed using: 1) the current work status item on the WHO ISoS information sheet (Sartorius et al., 1996) and 2) QLS items 9 and 10 (extent of occupational role functioning; level of role accomplishment).

2.3.5. Treatment

A service and treatment questionnaire was devised to record details about psychiatric treatment at long-term follow-up, treatment type, and recency of latest psychiatric hospital admission. Course of illness over last two years was assessed using the Life Chart Schedule (LCS) (WHO, 1992).

2.3.6. Follow-up diagnosis

The SCID-I/P was used to establish DSM-IV axis I diagnoses at long-term follow-up.

2.3.7. Reliability exercise

Assessments were conducted by trained researchers with a minimum 4-year undergraduate degree in psychology. Inter-rater reliability was established between three raters on 12 participants using a balanced incomplete block design as utilised in the Chestnut Lodge Follow up Study (McGlashan, 1984). The raters were paired in all possible ways and each pair assessed the same number of participants. High intra-class correlation coefficients indicated very good agreement among the raters: 0.97 for BPRS total; 0.94 for QLS; 0.93 for GAF; 0.91 for SANS; and 0.92 for SOFAS. Efforts were made to maintain inter-rater reliability across the entire follow-up. Raters were blind to diagnostic information and clinical ratings from previous assessments (for further details see Henry et al., 2007).

2.4. Statistical analysis

Independent samples t-tests were used to compare groups with early-onset and adult-onset schizophrenia spectrum disorder on continuous psychopathology and functional outcome measures at long-term follow-up. The unequal variances test was used where group variances significantly differed. Analysis of covariance (ANCOVA) assessed whether these results remained robust when the effects of length of follow-up were adjusted for. Where the homogeneity of variances assumption was violated, separate variance estimates were fitted using mixed modelling techniques. A sensitivity analysis was conducted for these measures to assess whether the results for early-onset and adult-onset comparisons were maintained for individuals with schizophrenia. Binary logistic regression, multinomial logistic regression and chi-square tests were used to assess the association of onset status with baseline and long-term categorical outcomes with exact tests used where appropriate. Because a continuous measure provides useful information over and above that provided by the dichotomized variable, we also investigated the linear relationship between the age at service entry (by individual year of age) and the outcome measures. Partial correlation coefficients were used to assess the magnitude of association between age at service entry and outcome measures, adjusted for the length of follow-up. Severity of positive and negative symptoms and depression scores were substantially skewed and were transformed to approximate normality prior to analysis using a log or square root transformation. SPSS version 15.0.1.1 and LogXact version 8.0.0 were used to analyse the data. Statistical tests were two-tailed. P-values of 0.05 or less were considered statistically significant.

3. Results

3.1. Follow-up and characteristics of the sample

The mean age in the study sample (n = 366) at baseline was 21.9 (SD = 3.5) years. Numbers (%) of individuals with schizophrenia, schizophreniform disorder, schizoaffective disorder, delusional disorder, brief psychotic disorder, and psychosis not otherwise specified were 174 (47.5%), 102 (27.9%), 48 (13.1%), 7 (1.9%), 7 (1.9%), and 28 (7.7%), respectively. Mean (SD, range) duration of follow-up was 7.3 (1.2, 4.1–10.4) years. Mean (SD) age at follow-up was 29.2 (3.8). Table 2 shows the demographic characteristics and baseline diagnostic information for the study sample by onset group. With the exception of age at entry to the service (t = − 28.7, df = 245.6, P < 0.001), age at follow-up (t = − 21.8, df = 108.1, P < 0.001), and length of follow-up (t = − 2.2, df = 364, P = 0.027), none of these variables differed significantly between the onset groups, ruling out these potential sources of bias for the long-term outcome finding.

Table 2 Demographic characteristics and baseline diagnosis of individuals with first-episode schizophrenia spectrum disorder by onset group. Values are numbers (%) unless stated otherwise.

Baseline characteristics Early-onset
(n = 41)
Adult-onset (n = 325) P valuea
Male 27 (65.9) 231 (71.1) 0.49
Family history of psychiatric illnessb 18 (69.2) 138 (54.8) 0.16
Migrant (born in a country other than Australia)b 4 (15.4) 50 (19.6) 0.60
Illicit substance useb 18 (69.2) 182 (71.4) 0.82
Problematic alcohol useb 3 (11.5) 38 (14.9) 0.64
Occupational status (employed or student)b 27 (77.1) 236 (79.2) 0.78
Onset within one monthb 11 (42.3) 130 (51.0) 0.40
Baseline diagnosis 0.82
 Schizophrenia 20 (48.8) 154 (47.4)
 Schizophreniform psychosis 11 (26.8) 91 (28.0)
 Schizoaffective disorder 7 (17.1) 41 (12.6)
 Delusional disorder 0 (0.0) 7 (2.2)
 Brief psychotic disorder 0 (0.0) 7 (2.2)
 Psychotic disorder not otherwise specified 3 (7.3) 25 (7.7)
Age at entry (years), Mean (SD) 16.5 (0.8) 22.6 (3.1) < 0.001
Age at follow-up (years), Mean (SD) 23.5 (1.5) 29.9 (3.4) < 0.001
Length of follow-up (years), Mean (SD) 6.9 (1.2) 7.3 (1.2) 0.03
Duration of untreated psychosis (days), Mean (SD)b and c 266.4 (605.4) 230.6 (452.9) 0.51
Median 40.5 62.0

aBased on independent samples t-test, chi-square test or Fisher's exact test.

bSubject numbers vary between n = 26 and n = 35 for the early-onset group and n = 252 and n = 298 for the adult-onset group for these variables.

cLog transformed due to severe skewness, but untransformed data are displayed here.

3.2. Symptomatic and functional outcome

Table 3 shows psychiatric and functional outcome measures at long-term follow-up by onset group. T-tests indicated significant group differences for the BPRS psychotic subscale, SANS, SOFAS, GAF, and QLS scores, with fewer symptoms, better functioning and higher quality of life in the group with early-onset schizophrenia spectrum disorder. Significant group differences remained robust for all outcome variables, except for the SANS, when the effects of length of follow-up were adjusted for using ANCOVA and mixed modelling techniques (Table 3).

Table 3 Psychiatric and functional measures in individuals with first-episode schizophrenia spectrum disorder or with first-episode schizophrenia at long-term follow-up by onset group. Values are means (SDs) unless stated otherwise.

Schizophrenia spectrum disorder Schizophrenia
Early-onset (n = 41) Adult-onset (n = 325) Adjusted mean difference (95% CI) P valuea Early-onset (n = 20) Adult-onset (n = 154) Adjusted mean difference
(95% CI)
P valuea
Brief psychiatric rating scale
 Psychotic subscaleb 6.5 (3.2) 8.3 (4.7) −1.8 (−2.9 to −0.7) 0.01 6.6 (3.3) 9.2 (4.9) −2.5 (−4.7 to -0.3) 0.02
 Total 37.0 (9.2) 39.1 (10.6) −2.2 (−5.7 to 1.3) 0.23 37.1 (9.2) 41.3 (10.2) −3.8 (−8.7 to 1.1) 0.13
Symptomsb 14.9 (12.2) 19.8 (14.4) −4.5 (−9.2 to 0.2) 0.07 15.7 (13.5) 22.9 (14.7) −6.3 (−13.2 to 0.7) 0.06
Beck depression inventoryb 5.4 (5.9) 5.8 (6.0) −0.5 (−2.5 to 1.4) 0.38 5.7 (6.0) 6.3 (6.2) −0.8 (−3.7 to 2.1) 0.48
Global assessment of functioning scale 64.2 (13.1) 56.2 (17.4) 7.5 (2.8 to 12.2) < 0.01 64.1 (13.5) 52.2 (16.4) 10.6 (3.1 to 18.1) < 0.01
Social and occupational functioning assessment scale 66.4 (12.7) 58.4 (17.2) 7.4 (2.8 to 12.0) < 0.01 66.1 (12.5) 54.1 (16.5) 10.6 (3.2 to 18.1) < 0.01
Heinrichs–Carpenter quality of life scale
 Interpersonal relations and social networks 33.0 (9.4) 27.1 (12.7) 5.4 (2.1 to 8.8) < 0.01 33.6 (9.9) 24.7 (12.8) 8.1 (2.2 to 13.9) < 0.01
 Instrumental role functioning 13.5 (8.5) 10.6 (9.1) 2.6 (−0.4 to 5.5) 0.09 11.9 (8.8) 8.6 (8.4) 2.7 (−1.2 to 6.7) 0.17
 Intrapsychic foundations 30.1 (6.7) 26.5 (9.6) 3.0 (0.5 to 5.5) 0.02 30.6 (6.7) 24.8 (9.2) 5.1 (0.9 to 9.2) 0.02
 Common objects and activities 9.1 (1.7) 8.7 (1.9) 0.3 (-0.3 to 0.9) 0.31 9.3 (1.8) 8.4 (2.1) 0.8 (−0.1 to 1.7) 0.09
 Total 85.6 (21.8) 72.8 (29.5) 11.4 (3.5 to 19.2) < 0.01 85.3 (23.0) 66.5 (28.0) 16.7 (3.9 to 29.4) < 0.01

aAdjusted P values were derived from analysis of covariance adjusting for length of follow-up.

bVariables were transformed due to severe or moderate skewedness, but untransformed scores are displayed here.

3.3. Employment

At follow-up, 39.6% (145/366) of the total cohort were employed, 4.1% (15/366) were students, 4.4% (16/366) fulfilled home duties or were working voluntarily, and 51.9% (190/366) solely received government benefit or were unemployed. Binary logistic regression indicated a statistically significant association between employment status and early-onset v adult-onset group (Table 4), with individuals in the early-onset group more likely to report a positive vocational outcome, being either employed or students (58.5% v 41.8%). Individuals in the early-onset group were also twice as likely to report being in “half time or more” employment compared with the adult-onset group (61.0% v 42.8%).

Table 4 Vocational outcome measures in individuals with first-episode schizophrenia spectrum disorder at long-term follow-up by onset group. Values are % of patients unless stated otherwise.

Schizophrenia spectrum disorder
Early-onset (n = 41) Adult-onset (n = 325) Odds ratio (95% CI) P value of odds ratio
Current work status 2.0 (1.0 to 3.8) 0.045
 Employed or student 58.5 41.8
 Solely receiving government benefit or unemployeda 41.5 58.2
Occupational role functioningb 2.1 (1.1 to 4.1) 0.03
 Half-time or more 61.0 42.8
 Less than half-time 39.0 57.2
Occupational accomplishmentc 1.8 (0.9 to 3.6) 0.08
 Very good or generally adequate 65.9 51.4
 None or marginal 34.1 48.6

aThis category also includes home duties and voluntary work.

bQuality of Life Scale item no. 9.

cQuality of Life Scale item no.10.

3.4. Treatment

Over the most recent two year period 45.1% (165/366) individuals were rated as never being actively psychotic, 15.3% (56/366) reported an episodic course (discrete episodes of no longer than six months), 37.4% (137/366) reported a continuous course (psychotic over most of the two year period), and 1.6% (6/366) reported neither episodic nor continuous course. Table 5 displays information on the course of illness over the last two-year period and treatment utilisation at follow-up. Individuals in the early-onset group were three times as likely than individuals in the adult onset group to have never been actively psychotic (63.4% v 43.8%), as opposed to having a continuous course.

Table 5 Course of illness and treatment utilisation in individuals with first-episode schizophrenia spectrum disorder at long-term follow-up by onset group. Values are % of patients unless stated otherwise.

Schizophrenia spectrum disorder
Early-onset (n = 41) Adult-onset (n = 325) Odds ratio (95% CI) P value of odds ratio
Course of illness over last two yearsa
 Never psychotic 63.4 43.8 3.0 (1.3 to 6.9) < 0.01
 Episodic 17.1 15.5 2.3 (0.8 to 6.7) 0.13
 Continuousb 19.5 40.7
Currently receiving psychiatric treatment 65.9 79.4 0.5 (0.2 to 1.0) 0.053
Treatment type
 GP 24.4 23.7 1.0 (0.5 to 2.2) 0.92
 Private psychiatrist 14.6 22.5 0.6 (0.2 to 1.5) 0.26
 Community health care 26.8 37.5 0.6 (0.3 to 1.3) 0.18
 Inpatient psychiatric carec 0.0 1.5 0.9 (0.1 to undefined) > 0.99

aMultinomial logistic regression; 6 subjects rated as ‘neither episodic nor continuous course’ were excluded from analysis.

bReference category.

cExact binary logistic regression.

3.5. Diagnoses at follow-up

The SCID based follow-up diagnoses and information regarding diagnostic stability is displayed in Table 6. Diagnostic stability between baseline and long-term follow-up diagnosis was not significantly different between the onset groups. The likelihood of receiving a current diagnosis of a psychotic disorder, indicating an episode within the last year, was significantly lower in the early-onset schizophrenia group (40.0% v 64.3%).

Table 6 Long-term follow-up diagnoses in individuals who were diagnosed at baseline with first-episode schizophrenia spectrum disorder or first-episode schizophrenia by onset group. Values are % of patients unless stated otherwise.

Schizophrenia spectrum disorder Schizophrenia
Early-onset (n = 41) Adult-onset (n = 325) Odds ratio (95% CI) P value Early-onset (n = 20) Adult-onset (n = 154) Odds ratio (95% CI) P value
Follow-up diagnosisa
 Schizophrenia 56.1 72.0 80.0 80.5
 Schizoaffective disorder 9.8 8.3 1.5 (0.4 to 4.9) 0.66 0.0 3.2 1.2 (undefined to 9.2) > 0.99
 Affective psychosisb 19.5 9.8 2.5 (0.9 to 6.5) 0.08 10.0 5.8 1.7 (0.2 to 9.4) 0.08
 Other psychotic disorderc 14.6 9.8 1.9 (0.6 to 5.3) 0.30 10.0 10.4 1.0 (0.1 to 4.8) > 0.99
Diagnostic status
 Current diagnosis 36.6 52.0 0.5 (0.3 to 1.0) 0.07 40.0 64.3 0.4 (0.1 to 1.0) 0.04
 Lifetime diagnosis 63.4 48.0 60.0 35.7
Diagnostic stabilityd
 Yes 80.5 90.2 0.5 (0.2 to 1.2) 0.12 90.0 94.2 0.6 (0.1 to 5.7) 0.74
 No 19.5 9.8 10.0 5.8

aExact multinomial logistic regression; reference category against which other follow-up diagnoses are compared is ‘Schizophrenia’.

bAffective psychosis includes bipolar disorder and major depression with psychotic features.

cOther psychotic disorder includes delusional disorder, brief psychotic disorder and psychosis not otherwise specified.

dExact binary logistic regression; diagnostic stability between baseline and long-term follow-up diagnosis.

3.6. Age at service entry and long-term outcome

Table 7 shows partial correlations between the age at service entry (by individual year of age) and outcome measures, adjusted for length of follow-up. Younger age at service entry significantly correlated with less severe psychopathology (SANS and BDI) and better functional outcome (SOFAS, QLS, and GAF) at follow-up.

Table 7 Partial correlations adjusted for length of follow-up between age at service entry and measures at long-term follow-up.

Schizophrenia spectrum disorder (n = 366) Schizophrenia (n = 174)
Variable R P value R P value
Brief psychiatric rating scale
 Psychotic subscalea 0.07 0.16 0.12 0.11
 Total 0.10 0.06 0.15 0.048
Scale for the assessment of negative symptomsa 0.13 0.01 0.17 0.03
Beck depression inventorya 0.12 0.03 0.14 0.08
Global assessment of functioning scale −0.14 0.01 −0.24 < 0.01
Social and occupational functioning assessment scale −0.13 0.01 −0.23 < 0.01
Heinrichs–Carpenter quality of life scale
 Interpersonal relations and social networks −0.16 < 0.01 −0.24 < 0.01
 Instrumental role functioning −0.08 0.12 −0.08 0.30
 Intrapsychic foundations −0.12 0.02 −0.18 0.02
 Common objects and activities −0.04 0.47 −0.12 0.10
 Total −0.14 < 0.01 −0.20 < 0.01

aVariables were transformed due to severe or moderate skewedness.

For ‘Scale for the Assessment of Negative Symptoms’ and ‘Beck Depression Inventory’ subject numbers vary between n = 338 and n = 342 for the schizophrenia spectrum disorder group and n = 164 and n = 165 for the schizophrenia group.

4. Discussion

The present study is the first direct, within-the-same-service comparison of long-term outcome in groups with early-onset and adult-onset schizophrenia spectrum disorder. Contrary to our expectations, and in contrast to the literature (Hollis, 2000, AACAP, 2001, Fleischhaker et al, 2005, and Reichert et al, 2008), we found that individuals with an early onset had significantly fewer positive symptoms and significantly superior functioning on measures assessing global, social/occupational, and community functioning compared to patients with adult-onset disorder. Consistent with these findings, early-onset individuals achieved significantly better vocational outcomes and had a more favourable course of illness with fewer psychotic episodes in the two years prior to follow-up. The onset group differences were corroborated by a series of correlations indicating significant associations linking younger age at service entry (investigated as a continuous variable) to better symptomatic and functional outcomes. Overall, these findings suggest that early detection and specialised treatment for the first psychotic episode may specifically improve long-term functional outcome and to some extent symptomatic outcome in people with early-onset schizophrenia as compared to adult-onset schizophrenia.

Psychosocial functioning was broadly assessed in the present study. The SOFAS differs from the GAF scale in that it focuses on the individual's level of social and occupational functioning and is not directly influenced by symptom severity. The QLS assesses community functioning on four subscales that address interpersonal relations, instrumental functioning, intrapsychic functioning (i.e., sense of purpose, motivation, curiosity, and ability to experience pleasure), and the use of objects and participation in activities. The observed findings indicate that the early-onset group was significantly more successful in performing the tasks of daily life, engaging in social relationships, and at meeting the needs of the community in which they live.

It has been suggested that earlier onset of schizophrenia may be associated with more severe consequences because it interrupts cognitive and social development at an earlier stage (Hafner and Nowotny, 1995). One possible explanation for the better functional outcome in the early-onset group in the present study could be that maintenance of one's school career and support in achieving adolescent developmental milestones are highly promoted by the EPPIC therapeutic model for those under the age of 18 (McGorry et al, 1996 and Edwards and McGorry, 2002). Psychological interventions and the intensive case management provided to this cohort for up to the first two years of their treatment could be relatively more effective in the younger individuals through interaction with a range of different biological, psychological or social variables. There is preliminary evidence to support this view, showing age-related differences in response to psychological and routine treatments in people with recent-onset psychosis (Haddock et al., 2006). Younger patients who received a psychological intervention experienced greater improvement than those who received treatment as usual, while this pattern was not found in adult patients.

4.1. Strengths and weaknesses

Important strengths that distinguish the present study from previous long-term follow-up studies in early-onset schizophrenia include the use of a large comparison group which was recruited from the same clinical service; all individuals received standardised treatment for up to the first two years of their illness; the study sample was drawn from an epidemiologically representative cohort of individuals with first episode psychosis assembled from a front-line public psychiatric service with a geographically defined catchment area. (Henry et al, 2007 and Henry et al, 2010); and baseline comparison of onset groups ruled out potential sources of bias for the long-term outcome finding. Unlike previous studies which also included 18- or 19-year olds (Hafner and Nowotny, 1995, Schmidt et al, 1995, Jarbin et al, 2003, and Ballageer et al, 2005), the age categories in our study stringently followed the convention of defining early-onset as before 18 years (AACAP, 2001) to avoid ambiguity. This cut-off is also consistent with the traditional age divide of psychiatric services (Hollis, 2000b), increasing the practical relevance of the findings. Other strengths include the application of standardised assessment instruments; that all assessments were conducted by experienced raters with high inter-rater reliability; and the robustness of key findings in the sensitivity analysis including only those patients who received a diagnosis of schizophrenia at baseline.

An important limitation of the present study is that the mechanisms of efficacy remain untested. Upon leaving the early psychosis programme, the ongoing treatment of participants was not controlled, and thus consisted of a variable degree of exposure to standard public and/or private sector care. This also limits the degree to which we can comment upon the impact of the early psychosis programme upon outcome. Similar limitations apply to previous long-term follow-up studies of early-onset schizophrenia samples, none of which factored out the influence of treatment (AACAP, 2001). However, these limitations should not compromise the validity of the group comparisons in the present study, especially as those in both groups received similar treatment at the time of first admission. Another limitation of our study is that its age range is 14 to 30 years, i.e. censored at both the lower and the upper ends. This age range is consistent with the EPPIC inclusion criteria, which aim to cover the peak onset phase of schizophrenia. It may be that older adults have a better outcome than younger adults and the rare cases with a very early onset (before the age of 14 years) a more chronic course of illness. A final limitation of the study is the cross-sectional nature of some of the outcome variables.

4.2. Comparison with previous studies

The few long-term follow-up studies available which have been conducted in early-onset schizophrenia describe the typical course of illness as chronic or relapsing (Hollis, 2000, AACAP, 2001, Fleischhaker et al, 2005, and Reichert et al, 2008). These largely non-comparative data have been widely interpreted as suggestive of a poorer outcome for schizophrenia with onset in adolescence compared to in adulthood (AACAP, 2001). To our knowledge, only one study with a follow-up of more than 5 years has compared outcome in early- and adult-onset schizophrenia, although the groups came from different services (Schmidt et al., 1995). This study showed that impairment, in particular of social function, was much greater in the younger group. The authors concluded that their findings, consistent with the body of non-comparative studies (Gillberg et al, 1993 and Cawthron et al, 1994), support the assumption that schizophrenic psychoses starting in adolescence may have a poorer outcome than those beginning later in life. However, the negative outcomes in prior reports may be related to the traditional reluctance of child and adolescent psychiatrists to assign severe psychiatric diagnoses to minors (Krausz and Muller-Thomsen, 1993 and Ballageer et al, 2005), leading to an overrepresentation of more severely ill, chronic cases, and denying people with early-onset schizophrenia evidence-based treatment, especially antipsychotic medication, until relatively late in the course of their illness.

The present findings of more benign outcomes in early-onset schizophrenia are partially supported by two studies. Firstly, a short-term follow-up study by Pencer et al. (2005) found that patients treated in adolescence (age 15–19) had similar symptomatic and functional outcomes at two-year follow-up compared to those treated in adulthood (age 26–50). Secondly, a large file audit study from our own service has revealed no significant differences in outcome, including psychosocial functioning, remission of positive symptoms, or employment status between early-onset and adult-onset psychosis patients at discharge from EPPIC (Schimmelmann et al., 2007). These studies, both from early psychosis programmes detecting and treating adolescent and adult patients under the same conditions, together with the present long-term data, also suggest that previous early-onset samples may not have been fully representative for first-episode schizophrenia in this age group and may have been focused on more chronic illness and the associated poor outcomes.

5. Conclusions

First-episode services treating adolescents and young adults offer an ideal opportunity to study age-related differences in people with early-onset and adult-onset schizophrenia. The present study draws a much more optimistic picture of the long-term outcome in early-onset schizophrenia spectrum disorder than has been previously seen, questioning the validity of the assumption of a generally poor outcome in early-onset cases. Our results have the potential to destigmatise this patient group and promote a more recovery-oriented therapeutic approach in this age group. The findings also imply that individuals with early-onset schizophrenia may specifically benefit from early intervention and require treatment which includes developmentally orientated, psychosocial, cognitive-behavioural interventions adapted to the needs of this group.

Role of funding source

This study was supported by a series of grants from the Australian National Health and Medical Research Council (grant number 350241) and the Victorian Health Promotion Foundation (grant number 91-0084C), together with generous funding from the Colonial Foundation to the Orygen Research Centre. The sponsors did not participate in the design or conduct of this study; in the collection, management, analysis, or interpretation of data; in the writing of the manuscript; or in the preparation, review, approval, or decision to submit this manuscript for publication.

Contributors

GP-A designed the study and wrote the first draft of the manuscript. LP-H and SM-H contributed to the design of the study and critically revised the manuscript. PD-M, HJ-J, H-H, MG-H, and MA-J supervised the research project and critically revised the manuscript. All authors contributed to and have approved the final manuscript.

Conflict of interest

The authors report no additional financial or other affiliation relevant to the subject of this article.

Acknowledgements

The authors particularly wish to thank all of the young people and family members who took part in the study, and the dedicated research team who conducted the interviews. We also thank Dr Sherilyn Goldstone for editing the final manuscript.

References

  • AACAP, 2001 AACAP. Practice parameter for the assessment and treatment of children and adolescents with schizophrenia. J. Am. Acad. Child. Adolesc. Psychiatry. 2001;40(7 Suppl):4S-23S
  • Amminger et al., 2006 G.P. Amminger, M.G. Harris, P. Conus, M. Lambert, K.S. Elkins, H.P. Yuen, P.D. McGorry. Treated incidence of first-episode psychosis in the catchment area of EPPIC between 1997 and 2000. Acta Psychiatr. Scand.. 2006;114(5):337-345 Crossref
  • Andreasen, 1982 N.C. Andreasen. Negative symptoms in schizophrenia. Definition and reliability. Arch. Gen. Psychiatry. 1982;39(7):784-788 Crossref
  • Ballageer et al., 2005 T. Ballageer, A. Malla, R. Manchanda, J. Takhar, R. Haricharan. Is adolescent-onset first-episode psychosis different from adult onset?. J. Am. Acad. Child. Adolesc. Psychiatry. 2005;44(8):782-789 Crossref
  • Beck and Beck, 1972 A.T. Beck, R.W. Beck. Screening depressed patients in family practice. A rapid technic. Postgrad. Med.. 1972;52(6):81-85
  • Cawthron et al., 1994 P. Cawthron, A. James, J. Dell, V. Seagroatt. Adolescent onset psychosis. A clinical and outcome study. J. Child. Psychol. Psychiatry. 1994;35(7):1321-1332 Crossref
  • DeLisi, 1992 L.E. DeLisi. The significance of age of onset for schizophrenia. Schizophr. Bull.. 1992;18(2):209-215 Crossref
  • Edwards and McGorry, 2002 J. Edwards, P.D. McGorry. Implementing Early Intervention in Psychosis. (Martin Dunitz Publisher, 2002)
  • Endicott et al., 1976 J. Endicott, R.L. Spitzer, J.L. Fleiss, J. Cohen. The global assessment scale. A procedure for measuring overall severity of psychiatric disturbance. Arch. Gen. Psychiatry. 1976;33(6):766-771 Crossref
  • Fleischhaker et al., 2005 C. Fleischhaker, E. Schulz, K. Tepper, M. Martin, K. Hennighausen, H. Remschmidt. Long-term course of adolescent schizophrenia. Schizophr. Bull.. 2005;31(3):769-780 Crossref
  • Gillberg et al., 1993 I.C. Gillberg, L. Hellgren, C. Gillberg. Psychotic disorders diagnosed in adolescence. Outcome at age 30 years. J. Child. Psychol. Psychiatry. 1993;34(7):1173-1185 Crossref
  • Goldman et al., 1992 H.H. Goldman, A.E. Skodol, T.R. Lave. Revising axis V for DSM-IV: a review of measures of social functioning. Am. J. Psychiatry. 1992;149(9):1148-1156
  • Haddock et al., 2006 G. Haddock, S. Lewis, R. Bentall, G. Dunn, R. Drake, N. Tarrier. Influence of age on outcome of psychological treatments in first-episode psychosis. Br. J. Psychiatry. 2006;188:250-254 Crossref
  • Hafner and Nowotny, 1995 H. Hafner, B. Nowotny. Epidemiology of early-onset schizophrenia. Eur. Arch. Psychiatry Clin. Neurosci.. 1995;245(2):80-92 Crossref
  • Heinrichs et al., 1984 D.W. Heinrichs, T.E. Hanlon, W.T. Carpenter Jr. The quality of life scale: an instrument for rating the schizophrenic deficit syndrome. Schizophr. Bull.. 1984;10(3):388-398 Crossref
  • Henry et al., 2007 L.P. Henry, M.G. Harris, G.P. Amminger, H.P. Yuen, S.M. Harrigan, M. Lambert, P. Conus, O. Schwartz, A. Prosser, S. Farrelly, R. Purcell, H. Herrman, H.J. Jackson, P.D. McGorry. The EPPIC long-term follow-up study of first-episode psychosis: methodology and baseline characteristics. Early Interv. Psychiatry. 2007;1(1):49-60 Crossref
  • Henry et al., 2010 L.P. Henry, G.P. Amminger, M.G. Harris, H.P. Yuen, S.M. Harrigan, A.L. Prosser, O.S. Schwartz, S.E. Farrelly, H. Herrman, H.J. Jackson, P.D. McGorry. The EPPIC follow-up study of first-episode psychosis: longer-term clinical and functional outcome 7 years after index admission. J. Clin. Psychiatry. 2010;71(6):716-728 Crossref
  • Hollis, 2000 C. Hollis. Adult outcomes of child- and adolescent-onset schizophrenia: diagnostic stability and predictive validity. Am. J. Psychiatry. 2000;157(10):1652-1659 Crossref
  • Jarbin et al., 2003 H. Jarbin, Y. Ott, A.L. Von Knorring. Adult outcome of social function in adolescent-onset schizophrenia and affective psychosis. J. Am. Acad. Child. Adolesc. Psychiatry. 2003;42(2):176-183 Crossref
  • Krausz and Muller-Thomsen, 1993 M. Krausz, T. Muller-Thomsen. Schizophrenia with onset in adolescence: an 11-year followup. Schizophr. Bull.. 1993;19(4):831-841 Crossref
  • Kumra et al., 2008 S. Kumra, J.V. Oberstar, L. Sikich, R.L. Findling, J.M. McClellan, S. Vinogradov, S.C. Schulz. Efficacy and tolerability of second-generation antipsychotics in children and adolescents with schizophrenia. Schizophr. Bull.. 2008;34(1):60-71
  • Kyriakopoulos and Frangou, 2007 M. Kyriakopoulos, S. Frangou. Pathophysiology of early onset schizophrenia. Int. Rev. Psychiatry. 2007;19(4):315-324 Crossref
  • Loranger, 1984 A.W. Loranger. Sex difference in age at onset of schizophrenia. Arch. Gen. Psychiatry. 1984;41(2):157-161 Crossref
  • Lukoff et al., 1986 D. Lukoff, R.P. Liberman, K.H. Nuechterlein. Symptom monitoring in the rehabilitation of schizophrenic patients. Schizophr. Bull.. 1986;12(4):578-602
  • McGlashan, 1984 T.H. McGlashan. The chestnut lodge follow-up study. I. Follow-up methodology and study sample. Arch. Gen. Psychiatry. 1984;41(6):573-585 Crossref
  • McGorry et al., 1990 P.D. McGorry, B.S. Singh, D.L. Copolov, I. Kaplan, C.R. Dossetor, R.J. van Riel. Royal Park Multidiagnostic Instrument for Psychosis: Part II. development, reliability, and validity. Schizophr. Bull.. 1990;16(3):517-536 Crossref
  • McGorry et al., 1996 P.D. McGorry, J. Edwards, C. Mihalopoulos, S.M. Harrigan, H.J. Jackson. EPPIC: an evolving system of early detection and optimal management. Schizophr. Bull.. 1996;22(2):305-326 Crossref
  • Pencer et al., 2005 A. Pencer, J. Addington, D. Addington. Outcome of a first episode of psychosis in adolescence: a 2-year follow-up. Psychiatry Res.. 2005;133(1):35-43 Crossref
  • Rabinowitz et al., 2006 J. Rabinowitz, S.Z. Levine, H. Hafner. A population based elaboration of the role of age of onset on the course of schizophrenia. Schizophr. Res.. 2006;88(1–3):96-101 Crossref
  • Reichert et al., 2008 A. Reichert, S. Kreiker, C. Mehler-Wex, A. Warnke. The psychopathological and psychosocial outcome of early-onset schizophrenia: preliminary data of a 13-year follow-up. Child Adolesc. Psychiatry Ment. Health. 2008;2(1):6 Crossref
  • Sartorius et al., 1996 N. Sartorius, W. Gulbinat, G. Harrison, E. Laska, C. Siegel. Long-term follow-up of schizophrenia in 16 countries: A description of the international study of schizophrenia conducted by the World Health Organization. Soc. Psychiatry Psychiatr. Epidemiol.. 1996;31:249-258
  • Schimmelmann et al., 2007 B.G. Schimmelmann, P. Conus, S. Cotton, P.D. McGorry, M. Lambert. Pre-treatment, baseline, and outcome differences between early-onset and adult-onset psychosis in an epidemiological cohort of 636 first-episode patients. Schizophr. Res.. 2007;95(1–3):1-8 Crossref
  • Schmidt et al., 1995 M. Schmidt, B. Blanz, A. Dippe, T. Koppe, B. Lay. Course of patients diagnosed as having schizophrenia during first episode occurring under age 18 years. Eur. Arch. Psychiatry Clin. Neurosci.. 1995;245(2):93-100 Crossref
  • Schulz et al., 1998 S.C. Schulz, R.L. Findling, L. Friedman, J.T. Kenny, A.L. Wise. Treatment and outcomes in adolescents with schizophrenia. J. Clin. Psychiatry. 1998;59(Suppl 1):50-54 discussion 55–56
  • Sikich, 2008 L. Sikich. Efficacy of atypical antipsychotics in early-onset schizophrenia and other psychotic disorders. J. Clin. Psychiatry. 2008;69(Suppl 4):21-25
  • WHO, 1992 WHO. Life Chart Schedule. (World Health Organization, Geneva, 1992)

Footnotes

a Orygen Youth Health Centre for Youth Mental Health, 35 Poplar Road, Parkville, VIC 3052, Australia

b Department of Child and Adolescent Psychiatry, Medical University Vienna, Währinger Gürtel 18-20, 1090 Vienna, Austria

c School of Population Health, The University of Queensland, Queensland Centre for Mental Health Research, Wacol, QLD 4076, Australia

d Department of Psychology, The University of Melbourne, VIC 3010, Australia

Corresponding author at: Orygen Youth Health Research Centre, 35 Poplar Road (Locked Bag 10), Parkville, Victoria 3052, Australia. Tel.: +61 3 9342 2800; fax: +61 3 9342 2941.