Depression and violent crime: is there a relationship?

What is already known on this topic

Major mental disorders, such as schizophrenia and personality disorder, have been associated with a higher risk of committing violent crime.1 However, the relationship between depression and violent crime is still unclear. Interestingly, some authors have recently presented evidence that this relationship occurs in the opposite direction: individuals with major depression present a higher risk of having homicidal death. For instance, in Sweden, a nationwide cohort found an increased risk of homicidal death in individuals with major depression than in general population (HR=2.61; 95% CI 1.58 to 4.33).2

Methods of the study

To clarify this issue, Fazel et al performed two longitudinal studies. In the first study, outpatients with at least two episodes of major depression, according to the International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10), were identified from the National Patient Register between 2001 and 2009 and were included in the study (n=47 158). Exclusion criteria were suffering from schizophrenia, schizophrenia-spectrum and bipolar disorder and having had inpatient episodes of depression. Each individual with depression had 20 controls from the general population, matched by birth year and sex and without any diagnosis of depression. Unaffected half-siblings and full-siblings of patients were used for comparison; in addition, each one of those had 20 general population controls. The mean follow-up time was 3.2 years. The second study was a cohort of twins composed of 15 298 twin pairs (5574 monozygotic and 9724 dizygotic pairs). Differently from the first study, depressive symptoms were measured using a short form of the Center for Epidemiologic Studies Depression scale. The mean follow-up time was 5.4 years. The main outcome for both studies was violent crime (homicide, attempted homicide and all forms of assault), as registered in the National Crime Register. The authors addressed the following covariates: gender, marital status, immigrant status, family disposable income at age 15, drug/alcohol use disorders and history of self-harm and criminality.

What does this paper add

• This is the first study to use two family-based studies to investigate violent crime as a possible consequence of depression.

• Individuals with depression were at threefold increased odds of violent crime compared with general population controls (OR 3.0, 95% CI 2.8 to 3.3) and had a twofold increased odds of violent crime risk compared with their unaffected siblings (ratio of OR 2.1, 95% CI 1.8 to 2.4).

• The risk of violent crime significantly increased in those individuals with more depressive symptoms. In analyses of the association between depression and violent crime in co-twin analyses using standard Cox regression, the HR (HR; 95% CI) for all twins was 1.09; (1.06 to 1.13).

Limitations

• There is a lack of information about whether patients were depressed at the time they committed a violent crime or whether they had recovered after antidepressant treatment, which may have diluted the associations observed. However, a sensitivity analysis looked specifically at time after diagnosis and there was a slightly stronger association when diagnoses were within 3 months of first diagnosis (adjusted OR 3.6, 95% CI 2.8 to 4.5).

• The generalisability of the results is limited to a set of conditions: high-income country with low levels of criminality and to patients with more severe depression.

• Considering that major determinants of violent behaviour are sociodemographic and economic factors and familial, social and environmental contexts,3 some key-variables may have not been evaluated in the study. However, the sibling analyses accounted for familial and early environmental factors (for instance, parental disorders, schooling, etc) the authors did explicitly adjust for family disposable income.

What next in research

Future studies should consider: (1) the use of directed acyclic graphs to identify variables that need to be adjusted for in estimating environmental conditions such as exposures to neighbourhood and familial characteristics and their effects on the outcome4; (2) multilevel analysis using data on geographic clustering to evaluate the role of contextual factors (eg, neighbourhood deprivation or inequalities) in determining the observed association; and (3) investigation in settings with higher rates of violent crime, particularly in low-income middle-income countries, to examine whether the association is similar in such settings.

Do these results change your practices and why?

The authors found a statistically significant association between depression and violent crime, with an effect size even stronger than that found between depression and self-harm. However, implications for changes in clinical practices and guidelines are still limited, since generalisability of findings is restricted to patients with more severe symptoms living in a high-income population with low level of criminality. Furthermore, possible impact on self-stigma and public stigma still strongly associated with mental illness, including depression, need to be evaluated in order to modify guidelines.5