Duration of untreated psychotic illness

Abstract

A lengthy delay often exists between the onset of psychotic symptoms and the start of appropriate treatment. However, the causes of this long delay remain poorly understood, and there is a need to search for the factors involved in such a delay in order to reduce the time of untreated psychosis. This study aimed at examining the influence of premorbid social networks on the duration of untreated psychotic illness. One hundred subjects with a first episode of schizophrenia or related psychotic disorders never treated with antipsychotics made up the study sample. Social support was assessed by means of the Sturtees's social support scale that comprises two subscales measuring close and diffuse social support. Duration of untreated illness was assessed according to three definitions: duration of untreated unspecific symptoms, duration of untreated psychotic symptoms, and duration of untreated continuous psychotic symptoms. We also examined the effect of putative confounding factors such as gender, residence (urban-rural), age at illness onset, years of education, and parental socio-economic status. Correlational analysis showed that poor diffuse social support, but not poor close social support, predicted long duration of untreated illness according to the three definitions; this association being mainly due to poor work/academic support. Logistic regression analysis confirmed such an association, but it was limited to duration of continuous psychotic symptoms (unadjusted OR=3.44, 95% CI=1.51-7.83); this association persisted after adjusting for the confounding variables (adjusted OR=3.39, 95% CI=1.39-8.29). We also found that subjects with low socio-economic status were depending on the definition of duration of untreated illness considered, between 2.7 and 4.3 times more likely to present with a long duration of untreated illness. Both poor diffuse social support and a low socio-economic status seem to be relevant factors of a prolonged duration of untreated psychosis.