PT - JOURNAL ARTICLE AU - Miriam J Haviland AU - Ali Rowhani-Rahbar AU - Frederick P Rivara TI - Age, period and cohort effects in firearm homicide and suicide in the USA, 1983–2017 AID - 10.1136/injuryprev-2020-043714 DP - 2020 Jul 30 TA - Injury Prevention PG - injuryprev-2020-043714 4099 - http://injuryprevention.bmj.com/content/early/2020/07/30/injuryprev-2020-043714.short 4100 - http://injuryprevention.bmj.com/content/early/2020/07/30/injuryprev-2020-043714.full AB - Background Rates of firearm homicide and suicide have varied over time. These variations are due to a number of factors including temporal trends, age, birth year and gender. We sought to conduct an age–period–cohort analysis to understand the intersection of these factors with firearm homicide and suicide.Methods We used data on firearm homicide and suicide for the years 1983–2017 from the Centers for Disease Control and Prevention’s Web-based Injury Statistics Query and Reporting System for this analysis. We restricted our analysis on firearm homicide to persons aged 10–44 years and our analysis on firearm suicide to persons aged 50–84 years, as these age groups are most at risk of each outcome. We calculated annual incidence rates for both outcomes per 100 000 population, overall and by gender.Results Across all age groups, rates of firearm homicide increased dramatically in the late 1980s and early 1990s. The peak age for firearm homicide varied across cohorts, although it was generally between ages 15 and 29 years. Rates of firearm homicide were substantially higher among men than women, regardless of age, period or cohort. Firearm suicide rates varied significantly by gender. Among men, older cohorts had higher firearm suicide rates, although the rate of firearm suicide increased with age across all cohorts. Among women, firearm suicide rates were also highest among older cohorts; however, firearm suicide rates decreased or remained relatively constant with age.Conclusion There are important differences in rates of firearm homicide and suicide with respect to gender, age, period and cohort.