[The risk of suicide in cancer patients derived from a cancer registry]

The incidence of suicide among cancer patients has been examined in several previous epidemiological studies, which have suggested an increased risk especially in the period following diagnosis. The role of factors which may modulate the risk, such as individual patient characteristics or that of cancer (site, prognosis or type of treatment), have not been systematically studied and available results are often contradictory. In Geneva, the existence of a tumor registry, which follows patients continuously until the cause of death is known, has enabled us to conduct a cohort study on all incident cases of cancer registered for the years 1970 to 1983. The risk of suicide (SMR) was found to be significantly elevated for both sexes in the first two years following diagnosis. Thereafter, the risk diminished, and for the entire 15 year period considered, the risk remained significantly elevated only for women. The effect of variables which may have modulated the risk was examined through a logistic regression analysis comparing cases and controls selected from the same cohort. The risk was significantly lower for married patients. No difference between sex was observed, but the risk of suicide appeared to be increased in the elderly, as well as in foreigners, and to be reduced in upper socio-economic classes. Risk could be higher with neoplasms of the digestive tract and lower with those of the breast. Because incidence of suicide is relatively low, studies may lack statistical power to detect a moderately increased risk. Nonetheless, these results confirm part of previous results and contribute further to the study of this problem.