Suicide is known to vary according to season, with peaks in the spring and troughs in the winter. The presence of psychopathology is a significant predictor of suicidality, and it is possible that the seasonal variation of suicide completion may be related to seasonality in the manifestation of psychiatric disorders common to suicide completers. In the current study, we evaluated 115 French-Canadian male suicide completers from the Greater Montreal Area for DSM-IV psychiatric disorders using proxy-based diagnostic interviews. Subjects were assessed for seasonal differences in the prevalence of DSM-IV psychiatric diagnoses just before their deaths. Diagnoses of major depressive disorder (MDD) without comorbid cluster B personality disorders, and schizophrenia were differently distributed between seasons. Most (63.4%) subjects with MDD committed suicide in the spring/summer (P = .038). However, closer examination revealed that depressed suicides with comorbid cluster B personality disorders did not show seasonality, while 83.3% of depressed suicides without comorbid cluster B personality disorders committed suicide in the spring/summer (P = .019). 87.5% of those suicides with schizophrenia committed suicide in the fall/winter (P = .026), and the only suicide with schizophrenia who died in the spring/summer was also the only one without positive symptomology. Our study is limited to male suicide completers, and results should not be generalized to women. We conclude that seasonal variation in suicide manifests itself differently in patients with different psychopathology. These findings indicate that assessment of suicide risk may need to include consideration of possible seasonal effects, depending on psychopathology.
THE SEASONAL variation of suicide completion, with a trend in the general population of a peak in the spring/summer and a trough in the winter, is a phenomenon that has been observed by many different groups.1, 2, 3 and 4 Studies have also revealed variations within annual rhythmicity when subjects are stratified by gender,1, 3, 5 and 6 age,1 and 7 and method (violent/nonviolent) of suicide.1, 3, 5, 6, 7, 8 and 9
Previous studies have proposed different social, environmental, biological, and psychological factors as possible causes of the seasonal fluctuations in suicide. As early as the 1890s, Durkheim attributed the seasonality of suicide to seasonal changes in social life. More recent studies have suggested that the modality of suicide can vary according to social7 and marital status,10 and even type and social status of employment.7 and 11 The seasonal variation in availability of health services could also be put forward as a potential sociological basis for the circ-annual rhythmicity of suicide completion. Other studies have proposed a role for environmental factors, reporting significant association with suicide and temperature, humidity, or weather conditions4 and 12 in seasonal suicide. Seasonal fluctuation at the neurochemical level has also been associated with suicide completion and behavior. For example, levels of serum cholesterol, which have been linked to behaviors related to suicide, such as violence, impulsivity, and aggression,13 exhibit seasonal variation14 that is similar to the observed circ-annual changes in suicide rate.15, 16, 17, 18, 19 and 20 Similarly, studies have implicated seasonality in serotonin function in seasonal suicide.21 and 22 Finally, it has been hypothesized that the seasonal variation of suicide is linked to the seasonal variation of violent behavior as measured by the number of violent incidents,23 and a comparison of nonviolent to violent suicide completers found that only violent suicide completions exhibit seasonality.9 These possible explanations of seasonality in suicide are not necessarily mutually exclusive and correlations between some factors, such as weather conditions and biological rhythms, have been suggested.12
Presence of psychopathology is known to be the largest single predictor of suicidality, and it is possible that the spring/summer peak and winter trough of suicides may be related to seasonal variation in incidence, severity, or clinical features associated with the psychiatric disorders common to suicide completers. If this were true, the psychopathology of suicide completers (just before their deaths) would vary according to season. In the present study, we investigated this hypothesis by analyzing the seasonal distribution of psychiatric diagnoses in a sample of consecutively recruited French-Canadian male suicide completers.
