A lthough the puerperium has been associated with higher risks of depression and psychosis (1), pregnancy has been considered a relatively quiescent emotional period (2). However, pregnant women experience higher rates of depressive, anxiety, and psychosomatic symptoms and lower levels of social adjustment than nonpregnant women of childbearing age (3, 4), and they might be expected to have a higher risk of suicide. Nonetheless, case series and death certificate reviews (3, 5, 6) have found that very few pregnant women commit suicide.Only two studies have calculated age-adjusted standardized mortality ratios for pregnant women. A British study (7) found that pregnant women were 1/20th as likely to commit suicide as nonpregnant women of childbearing age (standardized mortality ratio=0.05). Another study, in New York City (8), found a standardized mortality ratio of 0.62, which, although suggestive of lower risk, was not statistically significant. Therefore, we manually reviewed medical examiner case files, having hypothesized that pregnancy would be associated with a substantially lower risk of suicide. This finding would be intriguing because it may suggest that pregnancy could be a model for exploring protective factors against suicide in women in general.
METHODAll cases of suicide among female residents of New York City, 10-44 years old, from 1990 to 1993 were studied. To be considered a case, an individual had to have been certified as a suicide (diagnoses E950-959 in ICD-9) by the New York City Chief Medical Examiner. Pregnancy was determined by complete autopsy, including examination of the uterine lining and contents, and by forensic investigation, including interviews with witnesses and treating physicians as well as review of medical records. For the few women who did not undergo autopsy, pregnancy was assessed by external examination and forensic investigation. Suicide rates, expressed per 100,000 person-years, were calculated in 5-year age groups from age 10 to age 44 for each of four racial-ethnic groups for each year. Population denominators for these calculations were obtained from the 1990 U.S. Census (9). The numbers of live births, spontaneous abortions, and induced abortions were provided by the New York City Health Department for each of these demographic strata for each year for city residents.To calculate the expected number of suicides of pregnant women, the number of person-years of pregnancy was determined by using the numbers of both live births and abortions. We assumed that women who had live births during the study period were pregnant for 38 of the 52 weeks in a year. Because 90% of spontaneous and induced abortions occur during the first trimester (9), we assumed that women who had either type of abortion had been pregnant for 12 of the 52 weeks of the year. To guard against overestimating the expected number of suicides of pregnant women, we also calculated the number of person-years of pregnancy solely on the basis of the number of live births. The expected number of suicides of p...