long-term course of adolescent schizophrenia

Long-Term Course of Adolescent Schizophrenia

Christian Fleischhaker, Eberhard Schulz, Kathrin Tepper,Matthias Martin, Klaus Hennighausen, and HelmutRemschmidt

Send reprint requests to Dr. Christian Fleischhaker, Departmentof Child and Adolescent Psychiatry, Albert Ludwigs UniversityFreiburg, Hauptstr. 8, D–79104 Freiburg, Germany; e-mail:[email protected].

Our study investigated premorbid functioning, course, andoutcome in early-onset schizophrenia. All inpatients withDSM–III–R diagnoses of schizophrenia (n = 101) consec-utively admitted between 1983 and 1988 to the Departmentof Child and Adolescent Psychiatry at the University ofMarburg in Germany were included. To assess premorbidadaptation and precursor symptoms, we administered theInstrument for the Retrospective Assessment of the Onsetof Schizophrenia, which we modified to assess children andadolescents. Symptomatology was measured by the Scalefor the Assessment of Negative Symptoms, the Scale forthe Assessment of Positive Symptoms, and the Brief Psy-chiatric Rating Scale. In addition, the Global Assessmentof Functioning was applied. Followup data for 81 patients(80.2%) were available. The mean duration of schizophre-nia at followup was 9.56 2.2 years. Assessment of the high-est level of adaptive functioning revealed very good or goodoutcome in 19.8 percent of the patients, fair or poor out-come in 38.2 percent, and very poor outcome and gross im-pairment in 42.0 percent. Premorbid adjustment was thebest predictor of outcome in our schizophrenia sample. Apoor prognosis was found in patients with premorbid devel-opmental delays and those who were introverted and with-drawn before their psychotic state.

Key words: Early-onset schizophrenia/course/outcome/premorbid development

Several studies of adult schizophrenia patients haveshown that disability following the disorder can be se-vere, even though the patients manage to stay out ofthe hospital (Johnstone et al. 1991; Beckmann et al.1992; Harrison et al. 2001). In large-scale followup stud-ies of adult schizophrenia patients, outcome was good in25 percent of patients, about 50 percent achieved at leastpartial remission, and only 25 percent remained perma-nently hospitalized or grossly impaired (Ciompi andMuller 1976; Huber et al. 1979; an der Heiden et al.1995; Mason et al. 1995; Harrison et al. 2001).

Analyzing studies on smaller samples, Weiner (1982)concluded that the outcome of schizophrenia beginningin adolescence is less favorable than that starting in adult-hood. Only about 25 percent of the adolescent patientsreached full recovery, 25 percent improved but sufferedfrom continuing symptoms or occasional relapses, andthe remaining 50 percent required continuing residentialcare. Although schizophrenia adolescents were not morelikely to recover compared to adults, more of them (about50%) remained grossly impaired.Werry et al. (1991) followed up 30 children and ado-

lescents with schizophrenia. They noted complete recov-ery in only 23 percent of their patients at followup. Theauthors concluded that schizophrenia in children andadolescents is a chronic or relapsing disorder accompa-nied by considerable disability and significant deteriora-tion in adaptive function from already impairedpremorbid levels (table 1 contains details).Gillberg et al. (1993) reassessed a population-based

sample of 23 adolescents with schizophrenia. They de-scribed a good outcome in only 13 percent of theirpatients, 11 to 17 years after first diagnosis accordingto different kinds of register data. Only 8.7 percent ofthe patients showed an intermediate outcome.In a more recent study, Lay et al. (2000) followed up 65

children and adolescents with schizophrenia more than10 years after the first episode. Serious social disabilitywas found in 66 percent of patients, and no or minimumdysfunction was found in 20 percent. In this study, a lon-ger duration of inpatient stay was shown to be a prereq-uisite for a lower functioning at followup.Schizophrenia psychoses in childhood are of consider-

able importance for child psychiatry but rare within thespectrum of schizophrenia. Eggers and Bunk (1997), whofollowed up 44 children with schizophrenia, noted the rel-atively high rates of complete and partial remission in thelong-term course. They pointed out that none of thepatients with chronic onset remitted completely. Inpatients with schizophrenia psychosis, Remschmidtet al. (1994) described a chronic course in 90.9 percent(10 of 11 patients) when beginning before age 14. Onthe other hand, Asarnow et al. (1994) observed a morefavorable outcome in children with schizophrenia orschizoaffective psychosis with a substantial remissionof symptoms of schizophrenia and a good social adjust-ment in 22 percent and a chronic course in 78 percent. Ina more recent study, Remschmidt et al. (2000) followed

Schizophrenia Bulletin vol. 31 no. 3 pp. 769–780, 2005doi:10.1093/schbul/sbi014Advance Access publication on February 16, 2005

� The Author 2005. Published by Oxford University Press on behalf of the Maryland Psychiatric Research Center. All rights reserved.For permissions, please email: [email protected]. 769

at University of N

ew O

rleans on June 6, 2014http://schizophreniabulletin.oxfordjournals.org/

Dow

nloaded from

http://schizophreniabulletin.oxfordjournals.org/

Table 1. Outcome studies of child- and adolescent-onset schizophrenia

AuthorsSamplesize (n)

Sex Course of Illness/Psychosocial Outcome

Observationperiod (yrs)and diagnosis

Age at onsetof schizophrenia(yrs)

Female% (n)

Male% (n) Criteria Value % n

Lay et al. 2000 10 yrs No dysfunction 12 8Minimum 8 5

ICD–10 11–18 96 43% (41) 57% (55) Socialdisability

Obvious 14 9

Schizophrenia orschizoaffectivedisorder

Serious 30 19Very serious 31 20Maximum

dysfunction5 3

Remschmidtet al. 2000

42 yrs Good 16 65–14 38 39% (15) 61% (23) Global

AssessmentScale

Moderate 24 9

ICD–10 Poor 60 23Complete remission 25 11

Eggers andBunk (1997)

42 yrs 6–14 44 57% (25) 43% (19) Courseof illness

Partial remission 25 11

Clinical Poorly 50 22

Asarnowet al. (1994)

1–7 yrs 6–11 21 29% (6) 71% (15) Courseof illness

Remission ofsymptoms ofschizophrenia

33 6

DSM–III Chronicschizophrenia

67 12

Krausz andMuller-Thomsen(1993)

11–16 yrs 14–18 61 54% (33) 46% (28) Courseof illness

Remission 21 13Partial remission 10 6

Present StateExamination

Episodic 16 10Chronic 43 26No information 10 6

Gillberg et al.(1993)

11–17 yrs Possibly good 13 3

13–19 23 39% (9) 61% (14) Psychosocialoutcome

Intermediate 9 2DSM–III

and ICD–9Extremely poor 78 18

Schmidt andBlanz (1992)

No impairment 189 mo–14.4 yrs No impairment,

but delay15

ICD–9 School andoccupationalsituation (outof 40)

Mild 28Mean = 16.4 40 53% (21) 47% (19) Moderate 20

schizophrenia = 29 Severe 10affective = 11 Complete 10

C.Fleisch

haker

etal.

770

at University of New Orleans on June 6, 2014 http://schizophreniabulletin.oxfordjournals.org/ Downloaded from


Table 1. Continued

AuthorsSamplesize (n)

Sex Course of Illness/Psychosocial Outcome

Observationperiod (yrs)and diagnosis

Age at onsetof schizophrenia(yrs)

Female% (n)

Male% (n) Criteria Value % n

Serious social disability 55 22Werryet al. (1991)

4.3 6 3.2 yrs Remission 23 7

7–17 30 50% (15) 50% (15) DSM–III–R course Subchronic 13 4DSM–III–R Chronic 64 19

Krausz(1990)

5–10 yrs No impairment 14 7

14–18 59 56% (33) 44% (26) DSM–III–Rcourse (outof 51)

Improved 8 4

Present StateExamination

Episodic 29 15

Chronic 49 25

Inoue et al. (1986) 3 yrs Fully employed atprevious level

16 3

DSM–IIIschizophrenia andacute psychoticepisode

10–17 19 47% (9) 53% (10) Occupational situationat followup

Fully employed belowprevious level

21 4

Limited ability 16 3Incapable of learning 26 5Hospitalized 21 4

Kimura et al. (1978) More than 3 yrs Remittent type 30.5 712–17 23 39% (9) 61% (14) Course of illness Fading type 30.5 7

Scanty type 26 6Clinical Persistent type 13 3

771

Adolescen

tSchizo

phren

ia

at University of New Orleans on June 6, 2014 http://schizophreniabulletin.oxfordjournals.org/ Downloaded from


up 38 patients with childhood-onset schizophrenia (age atonset: 5–14 years; followup time: 42 years). According tothe Global Assessment Scale (GAS), Remschmidt et al.(2000) showed a fairly good outcome in only 16 percentof patients with childhood-onset schizophrenia, whilea poor andmoderate outcome was observed in 60 percentand 24 percent of patients, respectively.

In summary, most of the long-term followup studiesin early-onset and very early onset schizophrenia patientsdemonstrate a worse outcome in adolescents comparedto adults. Moreover, outcome in childhood-onsetschizophrenia seems to be worse in comparison to thatin adolescent-onset schizophrenia (Asarnow et al.2001, 2004).

Therefore, the aim of our study was to investigate pre-morbid functioning, course, and outcome in a larger con-secutive sample of early-onset schizophrenia. In addition,we attempted to identify predictors for outcome in adult-hood at an early phase of illness, applying variables de-rived from previously published short-term studies(Remschmidt et al. 1991, 1994; Fleischhaker et al.1998; Schulz 1998).

Methods

Subjects. All inpatients with DSM–III–R (APA 1987)schizophrenia (n = 101) consecutively admitted between1983 and 1988 to the Department of Child and Adoles-cent Psychiatry at the Philipps University in Marburg,Germany, were included in this followup study. Froma total of 1,351 inpatients, 154 patients were selectedwith a suspected clinical diagnosis of schizophrenia.All of these patients with suspected early-onset schizo-phrenia were carefully reevaluated according to theirsymptomatology at first admission and were rediagnosedaccording to the ICD–10 (World Health Organization1993) andDSM–III–R criteria of schizophrenia by a con-sensus rating of two experienced child psychiatrists. Outof the patients with suspected childhood- or adolescent-onset schizophrenia, a sample of 101 patients fulfilledDSM–III–R criteria of schizophrenia. The remaining53 patients were withdrawn from the study because ofsubstance-induced psychotic disorder (292.1) (n= 3), psy-chotic disorder due to general medical condition (293.8x)(n = 5), brief psychotic disorder (298.8) (n = 6), major de-pressive disorder (296.x) (n = 30), personality disorder(301.x) (n = 3), pervasive developmental disorder(299.8) (n = 3), or adjustment disorder (309.x) (n = 3).

The schizophrenia sample included in the followupstudy met the following criteria:

1. They were less than 18 years old when they showed thefirst symptoms of schizophrenia.

2. They were consecutively admitted to our departmentand received treatment as schizophrenia inpatientsaccording to the DSM–III–R criteria.

Followup Assessment. Of the 101 patients (53 males, 48females), 81 (80.2%) were assessed by a semistructuredinterview, either with the patient (n = 58) or with the rel-atives and/or medical staff (n = 23). For 6 other patients(5.9%) who had died by committing suicide, an interviewwith first or second degree relatives and/or the last treat-ing physician was performed to complete the sets of med-ical records. Eleven (10.9%) patients refused a home visit.Three cases (3%) were not traceable. We found no signif-icant differences in the characteristics at initial assess-ment between patients who were followed up anddropouts.

Instruments forFollowupEvaluation. For the assessmentof characteristics of the first episode, including sociode-mographic data, premorbid adaptation, and precursorsymptoms of schizophrenia, we administered the Instru-ment for the Retrospective Assessment of the Onset ofSchizophrenia (IRAOS) (Hafner et al. 1990, 1992).This instrument was modified by our group for investi-gating children and adolescents and their relatives(Remschmidt et al. 1994; Tepper 1998). For this study,the IRAOS was administered by experienced clinicianswho interviewed the patients or relatives directly. In ad-dition, information was obtained from past medicalrecords and a standardized documentation system.To evaluate premorbid symptomatology, we devel-

oped a checklist of premorbid symptoms that could beclassified as either ‘‘internalizing’’ or ‘‘externalizing.’’Examples of internalizing symptoms were mutism, men-tal slowness, social isolation, general anxieties, specificanxieties, and obsessive-compulsive symptoms. The ex-ternalizing dimension comprised items such as hyperac-tive and antisocial behavior, and aggression. A secondpart of the checklist was designed to include developmen-tal retardation. Every kind of developmental retardationin the field of speech and language, motor development,reading, and writing was rated based on all the kinds ofinformation we could gather on premorbid behaviorthrough a careful analysis of the case histories, informa-tion from parents, and school reports. Each item of thechecklist was evaluated as present or absent. For calcu-lations, we classified three dimensions for the checklist asinternalizing, externalizing, and developmental retarda-tion. If one or more of the items of each dimensionwere present within a dimension, it was classified as beingpresent.The evaluation of premorbid symptomatology was

performed independently by two investigators. If consen-sus was not achieved, a third senior researcher in childand adolescent psychiatry (E.S.) was consulted and a finalbest estimate was achieved.Followup interviews were performed based on the

IRAOS (Hafner et al. 1990, 1992) for the assessmentof the age at onset, the characteristics of the first episode,and sociodemographic data. The age of the patients at

772

C. Fleischhaker et al.

at University of N

ew O


Dow

nloaded from


onset of nonspecific psychopathological symptoms wasdefined as the age of the first occurrence of nonpsychoticsymptoms. The age at onset of symptoms of schizophre-nia was defined as the age of the first occurrence of symp-toms of schizophrenia (according to the ICD–10).

The assessment of educational and occupational im-pairment was performed according to the IRAOS. Pos-itive and negative symptoms of schizophrenia wereevaluated using the Scale for the Assessment of NegativeSymptoms (SANS) and the Scale for the Assessment ofPositive Symptoms (SAPS) developed by Andreasen(1982, 1984a, 1984b). The interrater reliability of boththe SANS and the SAPS was found to be good, with jvalues of 0.8 (Andreasen 1982; Moscarelli et al. 1987;Andreasen et al. 1991) with the exception of the SANSitem ‘‘attention’’ (j value 0.67). As described elsewhere(Remschmidt et al. 1991), relating attentional impair-ment to negative symptoms seems to be problematic;therefore, we excluded this item from the rating scale.The summary scores of negative and positive symptomswere calculated according to Andreasen (1982). In addi-tion to the evaluation of positive and negative symptoms,the Brief Psychiatric Rating Scale (BPRS) (Overall andGorham 1962) was used to measure symptomatologyand outcome during the followup investigation. The ap-plication of the BPRS in the followup studies of schizo-phrenia patients is well established (Thiemann et al. 1987;Bell et al. 1992; Deister and Marneros 1993). This scalehas also proved to be effective for the evaluation of out-come in treatment studies of schizophrenia (Kane et al.1988; Meltzer 1991; Beckmann et al. 1992). In our study,the BPRS total score and the BPRS depressive score (in-cluding items 1, 2, 5, and 9) were applied for statisticalanalysis.

To measure psychosocial adaptation, we used theGlobal Assessment of Functioning (GAF) scale(DSM–III; American Psychiatric Association 1980)and the Global Assessment Scale (GAS) (Endicottet al. 1976).

All interviews were conducted by experienced clini-cians who had extensive experience with using theIRAOS, the SANS, the SAPS, the BPRS, the GAS,and the GAF in prior studies (Remschmidt et al. 1991;Schulz 1998). The full package of instruments was intro-duced at two training seminars for principal investiga-tors. During this training, standard videotapes andcase vignettes were used. Each principal investigatorhad to rate three standard videotapes. A deviationfrom the standard rating of one point was allowed foreach item and for the summary scores. Before startingthe followup interviews, the principal investigatorswere trained by holding live interviews with an observerpresent. The observer made independent ratings.

For the evaluation, we compared the values at fol-lowup with the values for the same patient at initialassessment.

Statistical Methods

A comparison was drawn between patients we followedup and dropouts, as well as between genders, using chi-square tests for categorical variables, if expected cell fre-quencies could be regarded as large enough. Otherwise,Fisher’s exact test was applied. For continuous variables,median test or t test was used. The significance level wasfixed at alpha = 0.05. All statistical calculations were per-formed with SAS statistical analysis programs (SAS In-stitute 1989).

Results

Sample Characteristics. After a mean duration ofschizophrenia of 9.5 years (62.2 years; range 4–14 years),81 patients (80.2%) of the schizophrenia sample were in-vestigated.The first nonpsychotic symptoms occurred on average

at age 14.5 years, followed by the first symptoms ofschizophrenia approximately 1.5 years later. The age atfirst admission because of schizophrenia was 16.5 years(62.1 years; range 11–18 years) (table 2).According to the characteristics of the onset of schizo-

phrenia (first presentation, first symptoms of schizophre-nia, first hospitalization for schizophrenia, and age atfollowup), we found no significant differences betweenmales and females (table 3).In table 4, the diagnoses according to DSM–III–R cri-

teria and the course of illness according to ICD–10 cri-teria at followup are presented. There are no

Table 2. Characteristics of the sample of childhood- andadolescent-onset schizophrenia

n % Mean SD Range

Gender

Female 48 47.5 — — —Male 53 52.5 — — —

Age at firstnonpsychoticsymptoms, yrs

14.5 3.7 2–18

Age at first symptomsof schizophrenia,yrs

15.9 2.2 10–18

Age at first admissionbecause ofschizophrenia, yrs

16.5 2.1 11–18

Age at followup, yrs 26.0 2.8 18–34

Course of illness, yrs 9.5 2.2 4–14

No. ofhospitalizations forschizophrenia

3.9 2.6 1–13

Note.—SD = standard deviation.

773

Adolescent Schizophrenia

at University of N

ew O


Dow

nloaded from


significant differences between males and females, butthere was a trend toward a more serious course in males.We observed an episodic course of schizophrenia withrelapses and readmissions in 40 percent (n = 32) ofpatients, with a mean number of 2.9 readmissions.

FamilyHistory. Patients with a family history of schizo-phrenia or schizoaffective disorder (n = 22) as opposed tothose without (n = 56) had a significantly lower GAFscore (table 5) and higher scores of both positive and neg-ative symptoms (SAPS: mean 8.0 vs. 5.5; median test p =0.08; SANS: mean 4.2 vs. 3.4; median test p = 0.32). Afamily history of affective disorder did not increase therisk of poor outcome in this sample of schizophreniapatients.

Premorbid Characteristics. Figure 1 demonstrates pre-morbid symptoms of internalizing, externalizing, and de-velopmental retardation in the schizophrenia sample. Inonly 19 patients (19.6%) did we find no impairment. Pre-morbid symptoms of internalizing were present in 63patients (63.6%) and premorbid symptoms of externaliz-ing in 29 patients (29.3%).

There were no differences with regard to psychosocialadjustment (measured by the GAF) between those indi-

viduals with premorbid internalizing (n = 63) or external-izing (n = 29) symptoms.However, there was a significant difference between

the patients who had premorbid developmental retarda-tion and those who did not in regard to a lower level ofpsychosocial adjustment, measured by the GAF (mediantest: p = 0.0007) (table 5).

Characteristics of the Onset of Schizophrenia andFollowup. Social adjustment in terms of the GAF wasfavorably influenced by the following four factors:

1. An acute course at onset of the first episode of schizo-phrenia (beginning of acute symptoms in <1 month).

2. A shorter duration of first episode (duration of firstepisode <6 months).

3. An older age at onset of schizophrenia (age of onset>14 years).

4. Female sex (table 5).

Psychopathology at Followup

Depressive symptoms. Figure 2 shows depressive symp-toms at followup according to the BPRS. The figure dem-onstrates very clearly that a substantial number of these

Table 3. Age at onset of schizophrenia and gender (n = 101)

First presentation ofsymptoms

First symptoms ofschizophrenia

First hospitalizationfor schizophrenia

Followup (yrs),mean 6 SD

Female (n = 48) 14.5 6 3.6 15.9 6 2.2 16.3 6 2.0 25.8 6 3.0

Male (n = 53) 14.6 6 3.9 16.1 6 2.3 16.7 6 2.2 26.2 6 2.6

Note.—SD = standard deviation.

Table 4. Diagnosis according to DSM–III–R criteria and course of illness according to ICD–10 criteria

Sex

TotalFemale Male

n % n % n %

Diagnosis accordingto DSM–III–R criteria295.1 disorganized type 9 25.0 10 22.2 19 23.5295.3 paranoid type 17 47.2 31 68.9 48 59.3295.7 schizoaffective disorder 10 27.8 4 8.9 14 17.3All patients 36 100.0 45 100.0 81 100.0

Course of illness according to ICD–10 criteria0.00 continuous 10 27.8 12 26.7 22 27.20.01 episodic 4 11.1 7 15.6 11 13.60.02 episodic with stable residual symptoms 8 22.2 7 15.6 14 17.30.03 episodic with interepisode remission 3 8.3 4 8.9 7 8.60.04 partial remission 4 11.1 6 13.3 10 12.30.05 full remission 7 19.4 9 20.0 17 21.0All patients 36 100.0 45 100.0 81 100.0

774


at University of N

ew O


Dow

nloaded from


patients (39.7%; n = 23) revealed severe or moderatesymptoms of depression.

Negative and positive symptoms. The same applies tonegative symptoms as compared to positive ones. Nine-teen of these patients (32.8%) had severe negative symp-toms, 20 patients (34.9%) mild negative symptoms, and19 patients minimal negative symptoms. In contrast, se-vere positive symptoms were found in only 10.3 percentof the patients, whereas most schizophrenia patientsshowed onlymoderate (41.4%) orminimal (48.3%) symp-toms on the SAPS. The severity of positive and negativesymptoms was defined in the following way: scores be-tween 0 and 4 were rated as minimal, scores between 4and 8 as moderate, and scores between 8 and 20 as severe.

Global Outcome. A subdivision into three categorieswas introduced for the GAS. The rating ‘‘poor outcome’’

(GAS score # 40) was given if the first four items (needsconstant supervision, needs some supervision, unable tofunction in almost all areas, major impairment in severalareas) were present, a ‘‘moderate outcome’’ rating wasgiven if the items ranging from 41 to 70 (any serioussymptomatology, moderate symptoms, some mild symp-toms) were present, and a ‘‘good outcome’’ rating wasgiven if the items ranging from 71 to 100 (minimal symp-toms, transient symptoms, good functioning, no symp-toms) were present.The result of this rating is demonstrated in table 6,

which also compares our child and adolescent-onsetschizophrenia sample (n = 81) with a sample of child-hood-onset schizophrenia (n = 38, age at onset 12.7,age at followup 55) from the Department of Child andAdolescent Psychiatry in Marburg (Remschmidt et al.2000). The patients of the very early onset sample hadbeen treated as inpatients between 1920 and 1960.

Table 5. Social adjustment in terms of the Global Assessment of Functioning (DSM–III) and risk factors (separate mediantest for each risk factor)

Global Assessment ofFunctioning (mean 6 SD) p

Course of onset >1 mo (n = 51) 5.31 6 1.10 0.0023 (median test)#1 mo (n = 30) 4.13 6 1.20

Developmentalretardation

None (n = 37) 4.29 6 1.27 0.0007 (median test)Present (n = 44) 5.36 6 1.06

Duration of firstepisode

>6 mo (n = 49) 5.10 6 1.14 0.043 (t test)#6 mo (n = 32) 4.57 6 1.39

Age of onset #14 yrs (n = 14) 5.43 6 0.94 0.034 (t test)>14 yrs (n = 67) 4.76 6 1.30

Sex Female (n = 36) 4.67 6 1.27 0.021 (median test)Male (n = 45) 5.04 6 1.26

Family history ofschizophrenia

None (n = 56) 5.32 6 1.25 0.039 (median test)Present (n = 22) 4.56 6 1.26

Fig. 1. Premorbid symptoms in children and adolescents with schizophrenia (n = 101).

775


at University of N

ew O


Dow

nloaded from


Although the samples differed in age at onset and follow-up period, a tendency toward a worse outcome of the veryearly onset group was found.

Educational Outcome. Prior to the onset of schizophre-nia, all patients attended school. The public school sys-tem in Germany is divided into three branches followingfour years of elementary school education: Hauptschulecontinues up to grade 9, Realschule up to grade 10, andGymnasium up to grade 13.

In table 7, school graduation at followup is shown. Therate of achievement of different graduations amongGermans between the ages of 25 and 30 is compared(Zentrum fur Umfragen, Methoden und Analysen 1995).

With regard to school education, 28 patients (29.5%)had not graduated (3% of the general population), 57patients (60.0%) graduated from either grade 9 or 10within different settings (69% of the general population),and only 10 (10.5%) graduated from secondary school,which is composed of 12/13 years of school attendance(28% of the general population). There was no significantdifference between male and female patients with respect

to school education at followup (chi-square: nonsignifi-cant).

Living Situation/Family Status at Followup. Only 15patients (17%) needed hospitalization at followup,28 patients (31.8%) used semisheltered/shelteredliving, and 45 patients (51.1%) lived in normal housing(table 8).Regarding the occupational status at the time of the

interview, only 25 patients (28.7%) were employed ona nonsheltered basis, 13 patients (14.9%) were employedin a semisheltered labor market, and 33 patients (37.9%)were employed in a clinical setting; 16 patients (18.4%)did not work. The living conditions and occupational sta-tus were similar for males and females, with no significantdifference (table 8; chi-square: nonsignificant).Seventy-five of the patients (84.2%) were single at fol-

lowup or at the time of death, only 5 (5.7%) were married,and 8 patients (9.1%) lived in a relationship. Remarkably,more male patients lived alone than female (table 9; chi-square: nonsignificant).

Level of Social Adjustment at Followup. In table 10, theGAF results are shown. Assessment of the highest level ofadaptive functioning revealed outcome as follows: verygood and good (19.8%), fair and poor (38.3%), andvery poor and grossly impaired (42%). More male thanfemale patients showed a very poor and grossly impairedsocial adjustment at followup (table 10; chi-square p =0.06).

Discussion

This study was designed to provide information on thecharacteristics, course, and outcome of adolescentschizophrenia.

Limitations. A possible methodological limitationrelates to the clinical sample and the study design. Be-cause of the presence of a specialized rehabilitation centerfor early-onset schizophrenia in our area, results may be

Fig. 2. Depressive symptoms according to the Brief PsychiatricRating Scale at followup in the patients with childhood- andadolescent-onset schizophrenia who could be investigatedpersonally (n = 58).

Table6. Socialadjustment in termsof theGlobalAssessmentScale in thegroupofpatientswithchildhood-onset schizophrenia (Remschmidtet al. 2000) in comparison to our sample of predominantly adolescent-onset schizophrenia

Global Assessment ScaleChild and adolescentschizophrenia, n = 81 (%)

Very early onsetschizophrenia, n = 38 (%)

100–71 Good 19.8 15.8

70–41 Moderate 38.2 23.7

40–0 Poor 42.0 60.5

Followupinvestigations

1983–1988 1920–1960

Age at onset ofschizophrenia

10–18 years #14 yrs

776


at University of N

ew O


Dow

nloaded from


biased. However, we did not find differences betweenpatients admitted from home and patients admittedfrom the rehabilitation center.

Eighty-one of 101 patients were assessed in a semistruc-tured interview with the patient, relatives, or medicalstaff. Data were incomplete in 20 patients, so thesepatients were not included in the analysis. However, par-ticipants and nonparticipants did not differ in regard tothe diagnosis, sociodemographic data, and inpatientcourse of illness. Therefore, we think that our sampleis representative.

A further limitation could be said to be that only cross-sectional followup data were given on the outcome sta-tistics of symptomatology and social functioning. As a re-sult, the values reported for positive symptoms were lowand those for negative symptoms were fairly high, be-cause negative symptoms represent a fairly stable dimen-

sion and positive symptoms a dimension subject tofluctuation because of its episodic nature.

Strengths. Because of the low incidence of childhood-and adolescent-onset schizophrenia, most of the previousstudies in the field included relatively few patients (table1). To our knowledge, the present study—with 101 con-secutively admitted patients and a mean followup of 9.5years—is the biggest sample of early-onset schizophreniato date.Schizophrenia psychosis with early manifestation has

a poor prognosis. As demonstrated, course and outcomeare influenced by the patient’s premorbid personality.Withaverygoodorgoodoutcomein20percentofpatients,our study compares well to previous reports by Layet al. (2000), Werry et al. (1991), and Krausz andMuller-Thomsen (1993),which reported 20 to 23percent (table 1).

Table 7. School education/diploma at followup

Sex

TotalFemale Male

School education/diploma at followup n % n % n %

Without qualificationor special needsschool

10 22.2 18 36.0 28 29.5

Basic 21 46.7 19 38.0 40 42.1

Grade 10 9 20.0 8 16.0 17 17.9

Grade 12/13 5 11.1 5 10.0 10 10.5

All patients 45 100.0 50 100.0 95 100.0

Table 8. Living conditions and occupational situation at followup

Sex

TotalFemale Male

n % n % n %

Living conditions at followup

Locked ward 0 0.0 2 4.2 2 2.3Open ward 5 12.2 5 10.4 10 11.4Day or night hospital 2 4.9 1 2.1 3 3.4Temporary housing (sheltered) 6 14.6 8 16.7 14 15.9Sheltered living (flat share) 3 7.3 7 14.6 10 11.4Semisheltered living 2 4.9 2 4.2 4 4.6Normal housing 22 53.7 23 47.9 45 51.1All patients 40 100.0 48 100.0 88 100.0

Occupational situation at followup

Occupied in a clinical setting with no or low payment 14 35.9 19 4.2 33 37.9Semisheltered labor market 4 10.3 9 39.6 13 14.9Occupied on a nonsheltered basis 12 30.8 13 27.1 25 28.7Did not work 9 23.1 7 14.2 16 18.4All patients 39 100.0 48 100.0 87 100.0

777


at University of N

ew O


Dow

nloaded from


Adult long-term investigations have generally foundthat patients have better outcomes; for example, in thelong-term study in Bonn, 56 percent of patients withthe diagnosis of schizophrenia were found to have recov-ered and to be fully employed (Huber et al. 1979). Similarrecovery rates were reported in other adult long-termstudies (an der Heiden et al. 1995; Mason et al. 1995;Harrison et al. 2001). This supports the view of a lesssevere course of illness in later-onset schizophrenia.The better outcome may, in part, reflect a less severecourse of illness in these patients that is at least partlycaused by already established social roles (e.g., schoolgraduate, member of a stable partnership, person livingindependently).

The general outcome found in this study was poor butis still better than in the very early onset study of Asarnowet al. (1994), who used the same measure (GAS) that weused. Twenty percent of our patients, compared with 22percent in Asarnow’s study, had a good outcome; and 42percent of our patients, compared with 60 percent ofAsarnow’s, had a poor outcome. The results in thevery early onset study of Asarnow are similar to ourresults with a very early onset schizophrenia study sam-ple. Part of our very early onset sample (Remschmidt

et al. 2000) was investigated recently by Eggers andBunk (1997), who found a somewhat better outcome.The better outcome in adolescent-onset schizophreniamay, in part, reflect a less severe course of illness inpatients who fall ill during adolescence.Six patients could not be interviewed because they were

deceased, having committed suicide. This mortality rateis far higher than in a comparable age group of the gen-eral population, where a mortality rate of 1.97/1,000 wasfound in 10- to 30-year olds (Statistisches Bundesamt1998).Patients with premorbid developmental delays and

patients who are internalizing and withdrawn beforethe beginning of their psychotic state are at risk ofa poor outcome. More attention to premorbid featuresis essential in the development of preventive measures.

The Authors

Christian Fleischhaker, M.D., is Consultant, Depart-ment of Child and Adolescent Psychiatry, Albert Lud-wigs University, Freiburg, Germany. Eberhard Schulz,M.D., is Professor of Child and Adolescent Psychiatry

Table 9. Family status at followup

Sex

TotalFamily status atfollowup

Female Male

n % n % n %

Single 32 78.1 43 89.6 75 84.2

Married 3 7.3 2 4.2 5 5.7

In a relationship 5 12.5 3 6.3 8 9.1

All patients 40 100.0 48 100.0 88 100.0

Table 10. Level of social adjustment (DSM–III Axis V; Global Assessment of Functioning) at followup (n = 81)

Sex

TotalLevel of socialadjustment (DSM–IIIAxis V) at followup

Female Male

n % n % n %

Excellent 0 0.0 0 0.0 0 0.0

Very good 1 2.8 0 0.0 1 1.2

Good 7 19.4 8 17.8 15 18.5

Fair 7 19.4 8 17.8 15 18.5

Poor 11 30.6 5 11.1 16 19.8

Very poor 8 22.2 22 48.9 30 37.0

Grossly impaired 2 5.6 2 4.4 4 4.9

All patients 36 100.0 45 100.0 81 100.0

778


at University of N

ew O


Dow

nloaded from


and Head, Department of Child and Adolescent Psy-chiatry, Albert Ludwigs University. Kathrin Tepper,M.D., is Physician, Department of Child and Adoles-cent Psychiatry, Philipps University, Marburg, Ger-many. Matthias Martin, M.D., is Professor of Childand Adolescent Psychiatry, Department of Child andAdolescent Psychiatry, Philipps University. KlausHennighausen, M.D., is Consultant, Department ofChild and Adolescent Psychiatry, Albert Ludwigs Uni-versity. Helmut Remschmidt, M.D., Ph.D., is Profes-sor of Child and Adolescent Psychiatry and Head,Department of Child and Adolescent Psychiatry, Phil-ipps University.

References

American Psychiatric Association DSM–III: Diagnostic and Sta-

tistical Manual of Mental Disorders. 3rd ed. Washington,DC: APA, 1980.

American Psychiatric Association DSM–III–R: Diagnostic and

Statistical Manual of Mental Disorders. 3rd ed., revised Wash-ington, DC: APA, 1987.

an der Heiden, W., Krumm, B., Mueller, S., Weber, I., Biehl, H.,and Schaefer, M. Mannheimer Langzeitstudie der Schizophre-nie. Nervenarzt, 66:820–827, 1995.

Andreasen, N.C. Negative symptoms in schizophrenia: Definitionand reliability. Archives of General Psychiatry, 39:784–788,1982.

Andreasen, N.C. The Scale for the Assessment of Negative Symp-toms (SANS). Iowa City, IA: University of Iowa, 1984 a.

Andreasen, N.C. The Scale for the Assessment of Positive Symp-toms (SAPS). Iowa City, IA: University of Iowa, 1984 b.

Andreasen, N.C., Flaum, M., Arndt, S., Alliger, R., and Swayze,V.W. Positive and negative symptoms: Assessment and valid-ity. In: Marneros, A., Andreasen, N.C., and Tsuang, M.T.,eds. Negative Versus Positive Schizophrenia. Berlin, Germany:Springer Verlag, 1991. pp. 28–51.

Asarnow, J.R., Tompson, M.C., and Goldstein, M.J. Childhood-onset schizophrenia: A follow-up study. Schizophrenia Bulletin,20(4):599–617, 1994.

Asarnow, J.R., Tompson, M.C., and Goldstein, M.J. Psychoso-cial factors: The social context of child and adolescent-onsetschizophrenia. In: Remschmidt, H. ed., Schizophrenia in Chil-

dren and Adolescents. Cambridge, U.K.: Cambridge UniversityPress, 2001.

Asarnow, J.R., Tompson, M.C., and McGrath, E.P. Annotation:Childhood-onset schizophrenia: Clinical and treatment issues.Journal of Child Psychology and Psychiatry, and Allied Disci-

plines, 45:180–194, 2004.

Beckmann, H., Fritze, J., and Franzek, E. The influence of neuro-leptics on specific syndromes and symptoms in schizophrenicswith unfavourable long-term course: A 5-year follow-up studyof 50 chronic schizophrenics. Neuropsychobiology, 26:50–58,1992.

Bell, M., Milstein, R., Beam-Goulet, J., Lysaker, P., and Cicchetti,D. The Positive and Negative Syndrome Scale and the BriefPsychiatric Rating Scale: Reliability, comparability, andpredictive validity. Journal of Nervous and Mental Disease,

180:723–728, 1992.

Ciompi, L., and Muller, C. Lebensweg und Alter der Schizophre-nen. Eine katamnestische Langzeitstudie bis ins Senium. Berlin,Germany: Springer, 1976.

Deister, A., and Marneros, A. Long-term stability of subtypes inschizophrenic disorders: A comparison of four diagnostic sys-tems. European Archives of Psychiatry and Neurological Scien-ces, 242:184–190, 1993.

Eggers, C., and Bunk, D. The long-term course of childhood-on-set schizophrenia: A 42-year followup. Schizophrenia Bulletin,23(1):105–117, 1997.

Endicott, J., Spitzer, R.L., Fleiss, J.L., and Cohen, J. The GlobalAssessment Scale: A procedure for measuring overall severityof psychiatric disturbance. Archives of General Psychiatry,33:766–771, 1976.

Fleischhaker, C., Schulz, E., and Remschmidt, H. Depressivesymptoms and plasma monoamine levels as predictors of re-sponse to clozapine treatment in adolescents with schizophre-nia. Zeitschrift fur Kinder- und Jugendpsychiatrie undPsychotherapie, 26:235–243, 1998.

Gillberg, C., Hellgren, L., and Gillberg, C. Psychotic disorders di-agnosed in adolescence. Outcome at age 30 years. Journal ofChild Psychology and Psychiatry, and Allied Disciplines,34:1173–1185, 1993.

Hafner, H., Riecher, A., Maurer, K., Meissner, S., Schmidtke, A.,Fatkenheuer, B., Loffler, W., and an der Heiden, W. Instru-ment for the retrospective assessment of the onset of schizo-phrenia (IRAOS). Zeitschrift fur klinische Psychologie undPsychotherapie, 19:230–255, 1990.

Hafner, H., Riecher-Rossler, A., Hambrecht, M., Maurer, K.,Meissner, S., Schmidtke, A., Fatkenheuer, B., Loffler, W.,and van der Heiden, W. IRAOS: An instrument for the assess-ment and early course of schizophrenia. Schizophrenia Re-search, 6:209–223, 1992.

Harrison, G., Hopper, K., Craig, T., Lask, E., Siegel, C., Wander-ling, J., Dube, K.C., Ganev, K., Giel, R., an der Heiden, W.,Holmberg, S.K., Janca, A., Lee, P.W.H., Leon, C.A., Mal-horta, S., Marsella, A.J., Nakane, Y., Sartorius, N., Shen,Y., Skoda, C., Thara, R., Tsirin, S.J., Varma, V.K., Walsh,D., and Wiersma, D. Recovery from psychotic illness: A 15-and 25-year international follow-up study. British Journal ofPsychiatry, 178:506–517, 2001.

Huber, G., Gross, G., and Schuttler, R. Schizophrenie. Eine ver-laufs- und sozialpsychiatrische Langzeitstudie. Berlin, Germany:Springer, 1979.

Inoue, K., Nakajima, T., and Kato, N. A longitudinal study ofschizophrenia in adolescence: I. The one- to three- year out-come. Japanese Journal of Psychiatry and Neurology, 40:143–151, 1986.

Jarbin, H., Ott, Y., and von Knorring, A-L. Adult outcome of so-cial function in adolescent-onset schizophrenia and affectivepsychosis. Journal of the American Academy of Child and Ado-lescent Psychiatry, 42:176–183, 2003.

Kane, J., Honigfeld, G., Singer, J., and Meltzer, H. Clozapine forthe treatment-resistant schizophrenic. A double-blind compar-ison with chlorpromazine. Archives of General Psychiatry,45:789–796, 1988.

Kimura, S., Asai, S., Wakeno, M., and Aoki, N. On early andmid-onset schizophrenia, Part 1: Phenomenological aspects.Folia Psychiatrica et Neurologica Japonica, 32:41–56, 1978.

Krausz, M. Schizophrenia in adolescents—A follow-up study.Psychiatrische Praxis, 17:107–114, 1990.

Krausz, M., and Muller-Thomsen, T. Schizophrenia with onset inadolescence: An 11-year followup. Schizophrenia Bulletin,19(4):831–841, 1993.

779


at University of N

ew O


Dow

nloaded from


Lay, B., Blanz, B., Hartmann, M., and Schmidt, M.H. The psy-chosocial outcome of adolescent-onset schizophrenia: A 12-year followup. Schizophrenia Bulletin, 26(4):801–816, 2000.

Johnstone, E.C., Frith, C.D., Leary, J., Owens, D.G., Wilkins, S.,and Hershon, H.I. Background, method and general descrip-tion of the sample. British Journal of Psychiatry, 159(Suppl13):7–12, 1991.

Mason, P., Harrison, G., Glazebrook, C., Medeley, I., Dalkin, T.,and Croudace, T. Characteristics of outcome in schizophreniaat 13 years. British Journal of Psychiatry, 167:596–603, 1995.

Meltzer, H.Y. The effect of clozapine and other atypical antipsy-chotic drugs on negative symptoms. In: Marneros, A., Andrea-sen, N.C., and Tsuang, M.T., eds. Negative Versus PositiveSchizophrenia. Berlin, Germany: Springer Verlag, 1991.

Moscarelli, M., Maffei, C., Cesana, B.M., and Boato, P. An in-ternational perspective on assessment of negative and positivesymptoms in schizophrenia. American Journal of Psychiatry.,144:1595–1598, 1987.

Overall, J.E., and Gorham, D.R. The Brief Psychiatric RatingScale. Psychological Reports, 10:799–812, 1962.

Remschmidt, H., Martin, M., Fleischhaker, C., Hennighausen,K., Gutenbrunner, C., and Schulz, E. A forty-two year fol-low-up study of childhood-onset schizophrenia. Implicationsfor the Kraepelian dichotomy, Unpublished manuscript, 2000.

Remschmidt, H., Martin, M., Schulz, E., Gutenbrunner, C., andFleischhaker, C. The concept of positive and negative schizo-phrenia in child and adolescent psychiatry. In: Marneros, A.,Andreasen, N.C., and Tsuang, M.T., eds. Negative Versus Pos-itive Schizophrenia. Berlin, Germany: Springer, 1991. pp. 219–242.

Remschmidt, H.E., Schulz, E., Martin, M., Warnke, A., andTrott, G.E. Childhood-onset schizophrenia: History of theconcept and recent studies. Schizophrenia Bulletin, 20(4):727–745, 1994.

SAS Institute. SAS/STAT User’s Guide, Version 6. Cary, NC:

SAS Institute, 1989.

Schmidt, M.H., and Blanz, B. Behandlungsverlauf und Katamne-

sen bei 122 Psychosen in der Adoleszenz. [Course and follow

up investigation in 122 psychotic adolescents]. In: Nissen, G.

ed., Endogene Psychosyndrome. Gottingen, Germany: Verlag

Hans Huber, 1992. pp. 163–177.

Schulz, E. Verlaufspradiktoren schizophrener Psychosen in derAdoleszenz. Entwicklungspsychopathologische, neurochemi-sche und psychopharmakologische Befunde. Gottingen Ger-

many: Hogrefe. 1998.

Statistisches Bundesamt, ed. Statistical Yearbook 1998 for the

Federal Republic of Germany. Stuttgart Germany: Metzler-

Poschel, 1998.

Tepper, K. Katamnese einer vollstandigen Inanspruchnahmepopu-

lation (1983–1988) schizophren erkrankter Kinder und Jugend-

licher. Marburg, Germany: Gorich and Weiershauser, 1998.

Thiemann, S., Csernansky, J.G., and Berger, P.A. Rating scales in

research. The case of negative symptoms. Psychiatry Research,

20:47–55, 1987.

Weiner, J.S. Child and Adolescent Psychopathology. New York,

NY: Wiley, 1982.

Werry, J.S., McClellan, J.M., and Chard, L. Childhood and ado-

lescent schizophrenic, bipolar and schizoaffective disorders: A

clinical and outcome study. Journal of the American Academy

of Child and Adolescent Psychiatry, 30(3):457–465, 1991.

World Health Organization. ICD-10: Internationale Klassifikation

psychischer Storungen. In: Dilling, H., Mombour, W., and

Schmidt, M.H., eds. Gottingen, Germany: Huber, 1993.

Zentrum fur Umfragen, Methoden und Analysen. Demographic

Data: Anonymized Sample of the German Microcensus 1995

(ZUMA-file). Mannheim Germany: ZUMA, 1995.

780


at University of N

ew O


Dow

nloaded from


long-term course of adolescent schizophrenia

Documents