Summary: A large Swedish registry study finds an association between deliberate self-harm and an increased risk of violent criminal conviction.

Source: Karolinska Institutet

Self-harm Is Linked with Higher Risk of Violent Criminality, Swedish Registry Study Finds

Researchers at Karolinska Institutet report a significant association between deliberate self-harm and subsequent violent criminal convictions in a population-based study published in JAMA Psychiatry. Using nationwide Swedish registries that cover the whole population, the team examined whether individuals who received clinical care for self-harm were more likely to be convicted of violent offences.

Key findings

• The study followed all Swedish citizens born between 1982 and 1998 from age 15, encompassing 1,850,525 individuals.
• During the follow-up, 55,185 people received clinical care for self-harm, 66,561 were convicted of a violent crime, and 8,155 experienced both events.
• Those who ever received clinical care for self-harm had about a five-fold crude risk of being convicted for a violent crime compared with those who never received such care (crude hazard ratio 4.9).
• After adjusting for psychiatric comorbidity and socioeconomic factors, the increased risk remained, with an adjusted hazard ratio around 1.8—nearly double the risk.
• The association was present for both sexes. Women who self-harmed and had comorbid substance abuse showed a particularly elevated risk—about seven times greater—compared with women who never received clinical care for self-harm.

What the study examined

The cohort comprised all Swedish citizens born between 1982 and 1998 who were at least 15 years old during the follow-up. Individuals who emigrated before age 15 or who immigrated after age 13 were excluded to ensure complete registry coverage during adolescence and early adulthood. The observational period ran from January 1, 1997, through December 31, 2013, with data analysis performed in 2016.

Exposure was defined as having received clinical care associated with deliberate self-harm (nonfatal), and the main outcome was conviction for a violent crime as recorded in Swedish criminal registers. The investigators adjusted analyses for relevant psychiatric diagnoses and socioeconomic status to reduce confounding.

Interpretation

Lead author Hanna Sahlin and colleagues emphasize that the findings point to a shared vulnerability that may underlie both self-harming behaviour and violent acts toward others, rather than a direct causal pathway from self-harm to violence. When the researchers reversed the analysis—examining the risk of self-harm among people convicted of violent crimes—they observed a similar association, supporting the idea of common risk factors driving both outcomes.

The results underscore that clinicians assessing self-harm should also screen for aggressive or violent behaviours, and conversely that assessments of violent offenders should include evaluation of suicide risk and self-harming behaviours. Addressing these overlapping risks may improve both risk management and tailored interventions.

Clinical and public health implications

The study highlights the importance of integrated assessment in mental health and criminal justice settings. Because the association persisted after accounting for psychiatric comorbidity and socioeconomic status, self-harm itself signals an elevated risk profile that merits attention in prevention, treatment planning, and risk assessment protocols.

Funding and publication

The research was financed by the Swedish Research Council, the Marcus and Amalia Wallenberg Memorial Foundation, Forte, Stockholm County Council and Karolinska Institutet. The full research report, titled “Association Between Deliberate Self-harm and Violent Criminality,” is published in JAMA Psychiatry (online April 5, 2017) and lists Hanna Sahlin and colleagues as authors.

Abstract (condensed)

A population-based longitudinal cohort study of 1,850,525 individuals born 1982–1998 and followed from age 15 found that nonfatal self-harm requiring clinical care was associated with a higher risk of conviction for violent crime. The crude hazard ratio was 4.9; after adjustment for psychiatric comorbidity and socioeconomic factors the hazard ratio was 1.8. The association was present in both sexes and was particularly pronounced among women with comorbid substance abuse. The authors conclude that self-harm and violent criminality may reflect a common underlying vulnerability and recommend assessing risks of both violence and self-harm in affected individuals.

For clinicians and policymakers, these findings suggest a need for cross-disciplinary screening and interventions that address the overlapping risks of self-harm and violent behaviour.