Studdert · Annals of internal medicine 2022 · Retrospective cohort study · n=17569096

Homicide Deaths Among Adult Cohabitants of Handgun Owners in California, 2004 to 2016 : A Cohort Study.

Cited 42 times in the scientific literature.

Level 3 - non-randomized controlled study

Retrospective controlled cohort study.

PubMed 35377715 · doi:10.7326/M21-3762 · record verified 2026-08-31

What was done

A retrospective cohort study linked longitudinal voter registration records with mortality and lawful handgun transaction data across California from October 18, 2004, through December 31, 2016 (up to 12 years and 2 months of follow-up). The cohort consisted of 17,569,096 adult voter registrants aged 21 years or older who did not own handguns themselves. Researchers compared homicide rates (overall, firearm-specific, and in-home intimate partner homicides) between individuals who began residing with a lawful handgun owner (595,448 individuals) and those living with nonowners.

What was found

Among the cohort, 737,012 individuals died, including 2,293 by homicide. Two-thirds of individuals who began living with handgun owners were women. Cohabitants of handgun owners had significantly higher overall homicide rates compared to cohabitants of nonowners (adjusted hazard ratio [aHR], 2.33; 95% CI, 1.78 to 3.05). This elevation was primarily driven by firearm homicides (aHR, 2.83; 95% CI, 2.05 to 3.91). For homicides occurring at home, cohabitants of handgun owners had a sevenfold higher rate of fatal shooting by a spouse or intimate partner (aHR, 7.16; 95% CI, 4.04 to 12.69), and 84% of these victims were female.

Why it matters

This study demonstrates that the homicide risks associated with household firearm ownership extend significantly to non-owning adult cohabitants, disproportionately impacting women through intimate partner violence.

Limits

The study cohort was restricted to registered voters aged 21 or older in California, which may limit generalizability. Handgun exposure was based solely on lawful transactions, meaning some participants classified as unexposed may have resided with illicit firearms or non-handgun firearms. Additionally, residual confounding from unobserved household- or individual-level traits associated with homicide risk could not be fully ruled out.

Cited by