Danish national study reveals new factors linked to suicide

A new study from the Boston University of Public Health (USA) reveals that physical illnesses and injuries increase the risk of suicide in men, but not in women. The results of this research have been published in the journal JAMA Psychiatry.

This first-of-its-kind study used machine learning and health data from the entire Danish population to create gender-specific suicide risk profiles, highlighting the complex factors that can predict the suicide. Researchers have come up with a plethora of other ideas about the complex factors that can increase a person’s risk of suicide. This study is the first to use data from the entire Danish population, and to analyze them using a system of machine learning (a learning system where the machine learns from its mistakes, editor’s note) in order to identify risk factors for suicide.

Predictable men face suicide, not women

“Suicide is very difficult to predict, because every suicide death is the result of multiple interacting risk factors in a person’s life,” insists Dr. Jaimie Gradus, associate professor of epidemiology at the University of Boston Public Health and lead author of the study. The interest of carrying out this study on Denmark, is that the country has a national health care system in which all the health information of the population is compiled in government registers.

This wealth of information and statistics allowed Jaimie Gradus and her colleagues to examine thousands of factors in the medical histories of the 14,103 people who committed suicide in Denmark between 1995 and 2015. At the same time, the researchers analyzed the medical histories of 265,183 other Danes over the same period, using a machine learning system, to identify trends.

The results confirmed previously identified risk factors, such as psychiatric disorders and related prescriptions, but the researchers also uncovered new potential risk factors. First, they found that physical health diagnoses were particularly important in predicting suicide in men, but had no impact in women. Secondly, the research team realized that diagnoses and prescriptions four years before a suicide were more important for prediction than diagnoses and prescriptions six months before the act.

According to Jaimie Gradus, the results of this study do not make it possible to create a model capable of perfectly predicting the risk factors for suicide, knowing in particular that the tendencies which lead to suicide may be different outside of Denmark. Finally, according to her, the results indicate new factors to be taken into account to prevent this persistent public health problem.