The Risk of Suicide in Persons With AIDS

To the Editor.— Marzuk et al 1 present a detailed report of the first epidemiologic investigation of suicide in persons with acquired immunodeficiency syndrome (AIDS). While they use careful ascertainment methods and have done an excellent job in reporting the study, the relative risk is expressed as a ratio of the incidence densities (number of events per 100 000 person-years of follow-up) between the patients with AIDS and the general population. The result is a staggeringly high value of 36.30 among men 20 to 59 years of age. While the relative - risk - estimate method is a common statistical tool in epidemiologic studies with long followup periods, its direct application without regard to the underlying assumptions is somewhat naive. The basic premise in using the ratio of incidence densities is that the potential personyears of follow-up in the two groups, if the events being studied did not occur , are comparable. The