The association between violence exposure and general and cause-specific mortality in people using mental health services: cohort study

TL;DR

People with mental illness who experience violence have higher mortality rates from both natural and external causes.

Contribution

This study quantifies the increased mortality risk in mental health service users exposed to violence.

Findings

Violence exposure is associated with a 2.14 times higher risk of all-cause mortality.

Mortality risks are elevated for natural causes, suicide, and alcohol-related deaths.

Weapon-related assault admissions are linked to increased all-cause mortality.

Abstract

Many studies have observed a link between mortality and mental illness, although the contribution of violence exposure to mortality in people with mental illness remains under-researched.

To examine the association of violence exposure, such as being physically assaulted, with general and cause-specific mortality in a population using mental health services.

We assembled a cohort study using electronic health records from a mental health and substance use treatment provider in south-east London. Records were linked to acute medical admission and emergency department presentation data, as well as to a national mortality register with death certificates for deaths registered in England and Wales. Cox regressions estimated the associations of binary and cumulative violence exposure, as indicated by assault admission and presentation to emergency departments for violence-related reasons. Mortality was adjusted for sociodemographic and clinical potential confounders.

The hazard ratio for assault admission with all-cause mortality was 2.14 (95% CI: 1.93–2.36) following covariate adjustment. Adjusted associations were also found with mortality from the following causes: internal (natural) (hazard ratio 1.72, 95% CI: 1.50–1.98), external (hazard ratio 1.94, 95% CI: 1.51–2.48), suicide (hazard ratio 2.20, 95% CI: 1.38–3.52), respiratory (hazard ratio 2.01, 95% CI: 1.41–2.85), circulatory (hazard ratio 1.71, 95% CI: 1.27–2.28), diabetes-related (hazard ratio 2.86, 95% CI: 1.20–6.86) and alcohol-related (hazard ratio 1.56, 95% CI: 1.10–2.22). Results for cumulative assault were consistent with these in both direction and magnitude. There was evidence for an association of weapon-related assault admission with all-cause mortality (hazard ratio 1.58, 95% CI: 1.14–2.18).

People with mental illness, who are exposed to assault, experience greater mortality than those who are not exposed. Excess mortality attributable to violence exposure in people with mental illness was related to deaths from natural and external causes.

Figures