Risk of dying unnaturally among people aged 15-35 years who have harmed themselves and inflicted violence on others: a national nested case-control study.

BACKGROUND

Self-harm and violent criminality have overlapping causes, but people who engage in these behaviours are typically studied as two discrete populations. In this study, we aimed to examine the risk of unnatural death (ie, death from external causes such as accidents, suicide, and undetermined causes) among people with a history of self-harm and violent crime, focusing specifically on those with co-occurring behaviours.

METHODS

For this population-based nested case-control study, we used national interlinked Danish registers. Individuals aged 35 years or younger, who were alive and residing in the country on their 15th birthday, and who died from external causes (cases) were matched by age and gender to living people (controls). We compared risks of suicide, accidental death, and any death by external causes among those with a history of hospital-treated self-harm, violent criminality, or both behaviours with those in individuals without histories of either behaviour. We estimated incidence rate ratios (IRRs), adjusted for age and gender, to compare risks.

FINDINGS

We identified 2246 individuals who died from external causes, whom we matched to 44 920 living controls. 1499 (66·7%) of 2246 individuals died from accidental causes and 604 (26·9%) died by suicide. The risk of unnatural death was elevated for individuals with a history of violence (IRR 5·19, 95% CI 4·45-6·06) or self-harm (12·65, 10·84-14·77), but the greatest risk increase was among those with histories of both behaviours (29·37, 23·08-37·38). Substance misuse disorders, particularly multiple drug use, was more prevalent among individuals with co-occurring self-harm and violence than among those engaging in just one of these behaviours. Psychiatric disorders seemed to account for some of the excess risk of unnatural death among people with dual-harm histories, but excess risk, particularly of accidental death, persisted in the multivariable models.

INTERPRETATION

Among individuals with co-occurring self-harm and violence, the risk of accidental death, particularly accidental self-poisoning, should be considered to be as important as the risk of suicide. People with a history of both behaviours who also have a substance misuse disorder are at particularly high risk of dying from external causes. Strategies should be designed to be accessible for people facing turbulent lives with multiple problems. Individuals in this group with both behaviours are likely to be treated by health-care services for self-harm and have contact with criminal justice services, providing multiple opportunities for proactive intervention.

FUNDING

European Research Council.