Background and aims: There is a dearth of data on the predictors of insight in schizophrenia. This study attempted to assess the predictors of insight in a cohort of first-episode schizophrenia followed up over 5 years.

Methods: Patients diagnosed to have Diagnostic and statistical manual of mental disorders (4th ed.; DSM-IV) schizophrenia (n = 131) were assessed prospectively for insight, psychopathology and explanatory models of illness over a 5-year period using standard instruments. Multiple linear regression and generalized estimating equations (GEE) were employed to assess predictors of insight.

Results: We could follow up 95 (72.5%) patients, 5 years after recruitment. A total of 65 of these patients interviewed at 60 months (68.4%) achieved remission. Cross-sectional evaluations suggest a relationship between insight, psychosis rating and explanatory models of illness with good insight and medical models associated with good outcome. However, baseline and early illness data do not predict insight scores at 5 years. Serial longitudinal assessment of insight is negatively associated with Brief Psychiatric Rating Scale (BPRS) scores and positively associated with the number of nonmedical explanatory models of illness held by patients.

Conclusion: These findings argue that insight and explanatory models of illness are secondary to psychopathology, course and outcome. They are dependent on the trajectory of the person’s illness, are not independent of the condition and call for multifaceted understanding of the issues.