Depression diagnosis associated with higher risk of committing violent crime

25 February 2015

A diagnosis of depression is associated with a threefold increase in the risk of committing violent crime, compared to the general population, according to a new study based on nearly 50,000 people in Sweden diagnosed with depression.

The study found that 3.7% (641) of depressed men and 0.5% (152) of depressed women committed a violent crime after their depression diagnosis. This compares to 1.2% (4097) of men and 0.2% (1059) of women in the general population who had committed a violent crime.

‘We wanted to determine whether there was an increased risk of violence in individuals with clinical depression, and without other factors which are known to contribute to this risk. One important finding was that the vast majority of depressed persons were not convicted of violent crimes, and that the rates reported are below those for schizophrenia and bipolar disorder, and considerably lower than for alcohol or drug abuse,’ said study author Professor Seena Fazel at the Department of Psychiatry at Oxford. 

‘Quite understandably, there is considerable concern about self-harm and suicide in depression. We demonstrate that the rates of violent crime are at least as high, but they don’t receive the same level of attention in clinical guidelines or mainstream clinical practice’.

Researchers at Oxford University tracked the medical records and conviction rates of 47,158 people diagnosed with depression, and then compared them to the records of 898,454 people matched for age and gender, but with no diagnosis of depression, following them up for an average of three years.  They found that people who had been diagnosed with depression had a higher risk of both harm to others and self-harm (five to six times higher risk than the general population).

The study did not include people who had been hospitalized with depression, since hospitalization can be due to the threat of harm to others, which might elevate the level of risk calculated by the study.

When researchers controlled for a previous history of violence, self-harm, psychosis, and substance misuse – all of which are known to increase an individual’s risk of violence, with or without depression – they still found an increased risk of violent crime among people with depression.  However, the risk of violence when these factors were excluded was smaller, suggesting that risk of violence should be included in clinical risk assessments for people who have a previous history of violence, self-harm, psychosis, or substance misuse.

Violent crime was defined as a conviction for any of the following offences: homicide or attempted homicide; aggravated or common assault; robbery; arson; sexual offences (including indecent exposure); and illegal threats or intimidation.

Violent crime and depression levels in Sweden are similar to other western countries such as the USA, suggesting that the results may be broadly applicable in other countries.

Sweden maintains a register connecting every person born in Sweden after 1932 to their parents. The research team used this register and medical databases to track the family history of a large group of people who had been diagnosed with depression. Compared to their siblings who did not have depression, depressed people were still twice as likely to be violent.

The smaller difference in comparison to siblings versus the general population suggests that family-related factors – including shared genetic heritage and mistreatment during childhood – may contribute to the increased risk of violence by people with depression.

The study did not examine whether treatment for depression made any difference to the risk of violent crime, although this will be explored in further research planned by the authors.

The research was funded by the Wellcome Trust, and it is published in the journal The Lancet Psychiatry. The research paper will be freely available via open access once it is published.

For more information contact: Professor Seena Fazel, [email protected], 44 (0)1865 2239041

Alternatively contact the Oxford University News Office on +44 (0)1865 270046 or email [email protected]

For embargoed copies of the full paper, please contact The Lancet Psychiatry press office: [email protected], +44 (0) 207 424 4249

Notes to editors: