Neuropsychological Deficits in Adolescent-Onset Schizophrenia Compared With Attention Deficit Hyperactivity Disorder

A total of 69 adolescents participated in this study: 19 with a DSM-III-R diagnosis of a schizophrenic disorder, 20 with a DSM-III-R diagnosis of ADHD, and 30 healthy volunteer comparison subjects. The schizophrenic group consisted of five female and 14 male patients whose mean age was 16.2 years (SD=1.1). The ADHD group consisted of 20 male patients whose mean age was 14.1 years (SD=1.5), and the normal group consisted of 14 female and 16 male subjects whose mean age was 15.7 years (SD=1.6). The ADHD group was significantly younger than the two other groups (F=11.52, df=2, 66, p<0.001). Eleven of the subjects with schizophrenia and all of the ADHD subjects were recruited from the National Centre for Child and Adolescent Psychiatry in Oslo. Eight subjects with schizophrenia were recruited from other hospitals in the region. Fifteen of the subjects with schizophrenia were inpatients. They had been hospitalized approximately 12 weeks (mean=11.7, SD=15.6). Four outpatients with schizophrenia had never been inpatients. All of the ADHD subjects were outpatients. The subjects in the normal group attended regular school classes at normal grade level; these subjects received a small compensation to cover their traveling costs. The Child Behavior Checklist (14) was used to screen for mental problems, and individuals with a total raw score higher than 45 were excluded. The checklist is a general questionnaire to be filled out by parents, generating eight syndrome scale scores.

The diagnosis of schizophrenia and the subtype were assigned on the basis of clinical interviews by senior clinicians (including B.R.R.) and the patients’ case records. Consensus research diagnoses were made for a subgroup of 13 patients, and two senior psychologists (one of whom was B.R.R.) agreed on the schizophrenia diagnosis in 12 (92%) of the cases. Schizophrenic patients received the following subtype diagnoses: disorganized (N=11), paranoid (N=4), undifferentiated (N=1), schizophreniform disorder (N=1), schizoaffective disorder (N=1), and delusional disorder (N=1). The patients were reassessed 1 year later, and the diagnosis was confirmed in each case.

Diagnostic evaluations for the ADHD group were made according to semistructured clinical interviews and standardized rating scales. The adolescents met at least eight of the DSM-III-R criteria for the condition. There were marked attention problems both at home and at school. In addition, all of these subjects had shown substantial hyperactivity, impulsivity, and inattention between ages 6 and 10 as assessed by the Parent’s Rating Scale (15). None of these patients had a history of psychosis. Patients received the following comorbidity diagnoses: oppositional defiant disorder (N=9), developmental reading disorder (N=2), both oppositional defiant disorder and developmental reading disorder (N=3).

Eleven patients in the schizophrenic group and eight patients in the ADHD group were drug-naive. Five of the patients with schizophrenia were receiving standard neuroleptic medication (perphenazine, N=3; thioridazine, N=1; zuclopenthixol, N=1) at the time of testing. Three of the patients with schizophrenia were drug-free during testing and for a period of at least 5 days before testing. The mean defined daily dose of neuroleptic medication was 0.7 (SD=0.3) (defined daily dose is the assumed average maintenance dose per day for a drug used for its main indication in adults [Drug Consumption in Norway, Norwegian Medical Depot, 1993]). The neuroleptic-naive patients had been hospitalized for a mean of 6.4 weeks (SD=6.2), while the 12 patients who used medication or had used medication 5 days before testing had been hospitalized for a mean of 17.5 weeks (SD=21.8). Twelve of the subjects with ADHD received stimulant medication (methylphenidate, N=11; dextroamphetamine, N=1). The medication was discontinued for at least 24 hours before testing. One patient with ADHD received a small dose of haloperidol (1 mg/day) because of tics.

Scores on the Brief Psychiatric Rating Scale (BPRS) (16) were obtained for 11 of the patients with schizophrenia within 1 week of testing. Interrater reliability (intraclass correlation coefficient 1, 2 [17]) for the BPRS total score was 0.99. A positive symptom score and a negative symptom score based on a factor analysis conducted by Ventura et al. (unpublished paper, 1995) were extracted from the BPRS scores. The scores showed that the patients were characterized by psychotic symptoms at the time of testing and by more positive than negative symptoms. The following numbers of these 11 patients had a score of more than 3 on variables in the positive symptom score: unusual thought content, N=9; grandiosity, N=3; hallucinations, N=7; conceptual disorganization, N=3; bizarre behavior, N=4; suspiciousness, N=7; disorientation, N=2. The mean total score on the BPRS in the schizophrenic group was 61.1 (SD=19.3, range=31–90), the mean positive symptom score was 21.2 (SD=7.5, range=9–30), and the mean negative symptom score was 4.6 (SD=3.4, range=3–9). The mean score on the Global Assessment Scale (GAS) (18) was 41.7 (SD=12.0) in the schizophrenic group and 50.8 (SD=7.5) in the ADHD group. The patients with schizophrenia scored significantly lower on the GAS than the patients with ADHD (F=8.10, df=1, 36, p=0.007). The mean raw score on the Child Behavior Checklist was 61.8 (SD=32.8) in the schizophrenic group, 60.3 (SD=17.7) in the ADHD group, and 13.4 (SD=10.2) in the normal comparison group. The scores on the Child Behavior Checklist showed that the level of psychopathology in the patient groups was significantly above that in the normal comparison group (F=45.80, df=2, 62, p<0.001). (For more detailed information on the BPRS, GAS, and Child Behavior Checklist, see Rund et al. [12].)

All subjects were screened for a history of head injury with loss of consciousness, substance abuse, a general medical condition that was likely to affect CNS functions, and mental retardation. The estimated average IQ, as determined by the similarities, block design, digit span, and digit symbol subtests of the Wechsler Intelligence Scale for Children—Revised (WISC-R) (19), was 98.22 (SD=11.38) in the schizophrenic group, 98.62 (SD=11.24) in the ADHD group, and 110.54 (SD=9.49) in the normal group; the patient groups scored significantly lower than the normal group (F=11.17, df=2, 66, p<0.001). One subject in the schizophrenic group, three subjects in the ADHD group, and one subject in the normal group were left-handed.

All subjects were tested with a comprehensive neuropsychological test battery designed to assess the following functional domains: abstraction-flexibility, spatial organization, visual memory, verbal memory and learning, speeded visual-motor processing and attention, motor function, auditory processing-distractibility, sustained attention, apprehension, and selective attention. The test battery included (in order of administration) the Wisconsin Card Sorting Test (20), the digit span subscale of the WISC-R (19), dichotic listening (21, 22), the digit symbol subscale of the WISC-R, digit symbol location (23), the California Verbal Learning Test, trials 1–5 (24), Kimura recurring figures (25), the Seashore rhythm test (26), the digit span distractibility test (27), the California Verbal Learning Test, long delay and recognition (24), the Trail Making Test, parts A and B (26), a partial-report span of apprehension test (Asarnow and Nuechterlein, unpublished), the similarities subscale of the WISC-R, the degraded stimulus Continuous Performance Test (1, 12), the block design subscale of the WISC-R, the Sustained Attention Test (28), and the grooved pegboard (26).

After a complete description of the study to the subjects, written informed consent was obtained. The consent form was signed by both the adolescent and a parent. The subjects were tested individually. All subjects received the same battery of tests in a fixed order. Total time for the test battery was about 4 hours, including breaks.

Raw test scores (Table 1) were transformed to standard equivalents (z scores) on the basis of the means and standard deviations of the normal comparison group. Test data were reduced to 10 summary measures by averaging each subject’s z scores on tests assessing the same functional domain. The organization of tests indicated in Table 2 is modified according to Saykin et al. (4, 5).

Where high scores indicated impairment, scores were transformed (direction reversed) so that high scores always indicated better cognitive function. The 10 summary measures were used as dependent variables in a multivariate analysis of covariance (MANCOVA) with diagnostic group (schizophrenic, ADHD, and normal) serving as the between-group factor, covarying for age because the ADHD group was significantly younger than the two other groups. This was followed by univariate analyses of covariance (ANCOVAs) with age as the covariate for each summary measure. Significant univariate effects were followed by post hoc comparisons (Scheffé test) between each pair of groups.

When the Levene test for homogeneity of variance (29) revealed significant differences between groups, nonparametric tests (Kruskal-Wallis, Spearman rank correlation) were performed on the variables, in addition to standard ANCOVAs. Results are commented on if they altered the findings.

A within-subject contrast on each summary measure was conducted. For each contrast, the score for a particular summary measure was compared with the mean of the remaining summary measures in the profile. This procedure permitted determination of areas of selective deficit, as opposed to generalized impairment equally affecting all functions, by testing for between-group differences in profile shape (4). For significance testing of the contrasts, we used pairwise t tests.

Pearson’s product-moment correlations and Spearman’s rank correlations were used to test associations between the summary scores and age and between the summary scores and sex.

To assess the number of adolescent patients with clinically significant impairment, a threshold of 2 SD from the normal subjects’ mean was used.