Insight: Demographic differences and associations with one-year outcome in schizophrenia and schizoaffective disorder

Background: Insight is increasingly seen as an important variable for study in psychotic illness, particularly in relation to treatment adherence. This study aims to quantify the association of insight with outcome, sociodemographic variables and diagnosis in a large stable patient sample. Method: Data are from a one-year, open-label, international, multicenter trial (n=670) of long-acting risperidone in adult symptomatically stable patients with schizophrenia or schizoaffective disorder. Psychopathology and insight were quantified using the Positive and Negative Syndrome Scale (PANSS). Patients were assessed at four time points over the year of the study. Results: 31.2% of the sample showed clinically significant deficits in insight at baseline. There were no differences based on sex, but significant differences in age and diagnosis, with oldest patients and schizophrenia patients (cf., schizoaffective disorder) showing more deficits. Baseline insight impairment was correlated with change in PANSS score at one year (r=-0.243, p<0.001). Recursive partitioning showed that, of those whose symptoms improved, those whose insight also improved were more likely to complete the trial. Conclusions: Insight is important above and beyond the effects of symptoms for predicting continuation in drug trials. This may have implications for the design and analysis of such trials, as well as suggesting the importance of targeting insight in treatment to increase likelihood of adherence to treatment. There also appear to be small but significant differences in insight based on age and diagnosis within the schizophrenia spectrum.