Suicide Mortality During the Perinatal Period

Abstract

Importance: The US has the highest maternal mortality rate among developed countries. The Centers for Disease Control and Prevention deems nearly all of these deaths preventable, especially those attributable to mental health conditions. Coordination between US health care and social service systems could help further characterize circumstances and risks associated with perinatal suicide mortality.

Objective: To examine contextual and individual precipitating circumstances and risks associated with perinatal suicide.

Design, setting, and participants: This cross-sectional observational study used a convergent mixed methods design to explore factors contributing to maternal suicides and deaths of undetermined intent (hereinafter, undetermined deaths) identified in National Violent Death Reporting System (NVDRS) data for January 1, 2003, to December 31, 2021. Analyses included decedents who were aged 10 to 50 years and pregnant or post partum at death (collectively, the perinatal group) and demographically matched female decedents who were not pregnant or recently pregnant (nonperinatal group) at death. Analyses were performed between December 2022 and December 2023.

Exposures: Pregnancy status at death (perinatal or nonperinatal).

Main outcomes and measures: The main outcomes included contributing circumstances associated with suicides and undetermined deaths cited in coroner, medical examiner, or law enforcement case narratives. The study examined quantitative differences between groups using a matched analysis and characterized key themes of salient suicide circumstances using qualitative content analysis.

Results: This study included 1150 perinatal decedents identified in the NVDRS: 456 (39.6%) were pregnant at death, 203 (17.7%) were pregnant within 42 days of death, and 491 (42.7%) were pregnant within 43 to 365 days before death, yielding 694 postpartum decedents. The nonperinatal comparison group included 17 655 female decedents aged 10 to 50 years. The mean (SD) age was 29.1 (7.4) years for perinatal decedents and 35.8 (10.8) years for nonperinatal decedents. Compared with matched nonperinatal decedents, perinatal decedents had higher odds of the following identified contributing circumstances: intimate partner problems (IPPs) (odds ratio [OR], 1.45 [95% CI, 1.23-1.72]), recent argument (OR, 1.33 [95% CI, 1.09-1.61]), depressed mood (OR, 1.39 [95% CI, 1.19-1.63]), substance abuse or other abuse (OR, 1.21 [95% CI, 1.03-1.42]), physical health problems (OR, 1.37 [95% CI, 1.09-1.72]), and death of a family member or friend (OR, 1.47 [95% CI, 1.06-2.02]). The findings of the qualitative analysis emphasized the importance of mental health and identified 128 decedents (12.4%) with postpartum depression.

Conclusions and relevance: This study provides insights into complex factors surrounding maternal suicide, and it highlights opportunities for further research to understand long-term consequences of perinatal mental health. These findings also underscore the need for targeted evidence-based interventions and effective policies targeting mental health, substance use, and IPPs to prevent maternal suicide and enhance maternal health outcomes.

Figure 1.. Study Flowchart

The National Violent Death Reporting System (NVDRS) pregnancy category “pregnant, not otherwise specified” (NOS) (n = 42) was evaluated via manual review of text narratives. The following were included in the counts, when applicable: current pregnancy (n = 5), no mention of recent or current pregnancy (n = 26), not pregnant but pregnant 43 days to 1 year before death (n = 7), not pregnant but pregnant within 42 days of death (n = 2), and recent pregnancy (timing unknown) (n = 2). NA indicates not applicable.

Figure 2.. Odds Ratios (ORs) of Circumstances Contributing to Suicide

A, Circumstances contributing to suicide among postpartum decedents relative to pregnant decedents (reference group). B, Circumstances contributing to suicide among perinatal decedents relative to nonperinatal decedents (reference group). Both forest logistic regression models and associated forest plots adjusted for age quartile, race and ethnicity, educational attainment, marital status, and other covariates (eg, injured at decedent’s home, autopsy, and in labor force).