Self-harmers are three times as likely to die prematurely

The LancetSep 18 2012

People who have a history of self-harm are more than three times as likely to die prematurely as the general population, and not just from the obvious causes, with deaths due to natural causes at least two times greater than anticipated and the risk also much higher for individuals living in socially deprived areas, according to a UK study published Online First in The Lancet.

Led by Keith Hawton from the University of Oxford Centre for Suicide Research, the study looked at over 30 000 individuals who attended emergency departments in Oxford, Manchester, and Derby after self-poisoning or self-injury between 2000 and 2007. The causes of premature death and years of life lost (YLL) were assessed and compared with the general population. Associations with socioeconomic deprivation were also tested using residential postcodes.

Roughly 6% (1832) of patients died during the median 6 year course of follow up. Death from both natural and external causes (suicides, accidental poisonings, and accidents other than poisoning) was substantially increased for both men and women, equating to an average of at least 30 YLL by each individual.

Accidental poisoning was the most common cause of premature death followed by suicide. However, the researchers found that deaths due to natural causes were 2 to 7.5 times greater than expected, with diseases of the digestive (largely alcohol related) and circulatory systems and mental and behavioural disorders (87% of these due to psychoactive substance abuse) the largest contributors.

What is more, the risk of early death from natural causes (but not external causes) was closely linked with socioeconomic status, increasing with greater economic deprivation.

“Our study confirms that the increase in premature death among people who self-harm is not limited to suicide or other external causes, but includes dying prematurely from a wide variety of natural causes such as diseases of the circulatory and digestive systems which accounted for a third of deaths in our study”,  says Hawton. “Our findings have significant public health implications, and emphasise the importance of assessing physical health as well as psychosocial problems as part of standard checks when individuals present with self-harm.”

Writing in a linked Comment, Eric Caine from the University of Rochester Medical Center in the USA says, “[This research highlights that] although review of risk and protective factors for imminent, potentially fatal suicide attempts is essential, assessment of the nature of people’s lives—ie, appraisal of the broad contexts of their actions, social and interpersonal struggles, behaviours, and basic medical problems—is equally important…[and] should encourage policy makers to look for new models of service delivery to meet patients’ diverse needs.”