“Younger” and “Older” Old Adults Who Die by Suicide: A Comparison Study and Cluster Analysis

Abstract

Objectives: To compare sociodemographic, clinical, and suicide-specific characteristics between younger old (aged 65–74) and older old (aged 75+) adults who died by suicide in 2015, and to identify clinically relevant subgroups in the entire study cohort 65+ through cluster analysis. Design: Retrospective cohort study. Setting: All health care units across Sweden. Participants: Individuals aged 65+ with at least one physician contact in the year preceding suicide (N = 277; aged 65-74 n = 145, 75+ n = 132). Measurements: Variables retrieved from medical records and the Swedish Cause of Death Register. Results: There were no differences between age groups, except being widowed, which was documented to a higher proportion in those aged 75+ (10% vs 28%, p < 0.001). Nearly half of the total cohort had a prescription for antidepressants at the time of death, but increased suicide risk was noted in only 13%. Two groups emerged via cluster analysis. One large group (n = 197) was characterized by male sex and comparatively low proportions with notations related to mental ill-health. The other group (n = 80) was characterized by high rates of mental illness including suicidal ideation and prescribed psychoactive medication. Conclusions: We identified no clinically meaningful age group differences. Cluster analysis revealed a large, predominantly male group in which one third were prescribed antidepressants. Otherwise, there was little documentation related to mental health, suggesting other suicidal precipitants, underdiagnosis, and/or underestimation of the severity of mental illness in that group. This points to a need for data sources that go beyond medical records to inform targeted prevention efforts.