Behavioral and intellectual markers for schizophrenia in apparently healthy male adolescents

Citation
M. Davidson et al., Behavioral and intellectual markers for schizophrenia in apparently healthy male adolescents, AM J PSYCHI, 156(9), 1999, pp. 1328-1335
Citations number
62
Categorie Soggetti
Psychiatry,"Clinical Psycology & Psychiatry","Neurosciences & Behavoir
Journal title
AMERICAN JOURNAL OF PSYCHIATRY
ISSN journal
0002953X → ACNP
Volume
156
Issue
9
Year of publication
1999
Pages
1328 - 1335
Database
ISI
SICI code
0002-953X(199909)156:9<1328:BAIMFS>2.0.ZU;2-0
Abstract
Objective: Subtle behavioral and intellectual abnormalities are often prese nt in apparently healthy adolescents who later develop schizophrenia. The a uthors investigated whether these abnormalities can predict vulnerability f or schizophrenia before the first psychotic manifestation. Method: The stud y consisted of linking the Israeli Draft Board Registry with the National P sychiatric Hospitalization Case Registry, The draft board tests measure int elligence, social functioning, organizational ability, interest in physical activity, and individual autonomy. Patients (N=509) were compared to nonpa tients, i,e,, adolescents not appearing in the National Psychiatric Registr y (N=9,215), matched to patients by age, gender, and school attended at tim e of testing. Results: Healthy male adolescents who were later hospitalized for schizophrenia had significantly lower test scores on all measures than adolescents not reported to the National Psychiatric Registry. The stronge st predictors for schizophrenia were deficits in social functioning, organi zational ability, and intellectual functioning. When patients were compared to matched nonpatients, the prediction model had a 75% sensitivity, a 100% specificity, a positive predictive value of 72%, and an overall rate of co rrect classification of 87.5%. Applied to the Israeli Draft Board Registry, the model yielded a sensitivity of 74.7%, a validated specificity of 99.7% , and a positive predictive value of 42.7%. Conclusions: This study demonst rated that simple assessment tools can predict predisposition to schizophre nia in healthy male adolescents, The model's predictive ability does not ch ange as a function of the time elapsed between testing and first hospitaliz ation. This suggests that the model identifies apparently healthy individua ls who will manifest the disease later who are not prodromal to psychosis. Easily applied tools allowing early identification of schizophrenia or vuln erability to it may enable early intervention.