Winter birth excess in schizophrenia and in non-schizophrenic psychosis: sex and birth-cohort differences

Abstract

Objective: Despite the fact that association between winter birth excess and schizophrenia in the northern Hemisphere is well established, possible sex or birth-cohort differences in this winter birth excess remain unclear. We aimed to evaluate sex and birth-cohort differences in the seasonal birth distribution of patients with schizophrenia or non-schizophrenic psychosis.

Method: The sample included 321 ICD-10 schizophrenia and 294 non-schizophrenic psychosis patients consecutively admitted into a psychiatric hospitalization unit in Granada, southern Spain, during a nine-year period (1998-2006). The distribution of births among the general population born over the same period as the patients was calculated.

Results: Among schizophrenia males (n=258), it was possible to demonstrate that the observed proportion of winter birth (December, January or February) was significantly higher than expected. Among schizophrenia females (n=63), although proportions were as in males and the effect size of the difference between observed and expected winter births was not lower than for men, only a statistical trend could be demonstrated. Among patients with non-schizophrenic psychosis, the observed proportion of winter birth was significantly higher than expected in women, but not in men. The sex-adjusted proportion of winter birth among schizophrenia patients born in the 1940's (a time period characterized by poor economy and widespread food restrictions because of the Spanish post-civil-war period) was significantly higher than among those born later; a difference that does not occur among patients with a non-schizophrenic psychosis.

Conclusions: Among schizophrenia patients born in winter there appear to be slight sex-differences and strong birth-cohort differences, possibly due to epidemiological factors such as poverty or maternal nutritional deprivation. Epidemiological findings related to winter birth excess among patients with schizophrenia must be identified in longitudinal studies.