Predicting suicide risk using health data at scale
27 June 2019
More than three-quarters of a million people globally take their own lives every year, to devastating personal, social and economic cost. Whilst past research has shown links between suicide and mental health, environmental and biological causes, a team at the University of Glasgow, part of HDR UK, has now identified some specific DNA risk markers involved.
Dr Rona Strawbridge, HDR UK Fellow, said: “The long-term ambition is for genetics researchers to work closely with social scientists and psychologists to develop useful models of suicidal behaviour, informed by evidence from biological, social, psychological and medical research. Building these networks and having greater access to datasets is made possible by HDR UK and will help us create a more unified research environment.”
A global issue
According the World Health Organisation, close to 800,000 people die due to suicide every year and for every suicide there are many more people who attempt suicide.
The team, co-led by HDR UK-funded Dr Strawbridge, sought to identify genetic variations across the whole genome associated with a broad range of suicidal thoughts and behaviours, using UK Biobank data from nearly 500,000 people. They also analysed information including physical measurements, lifestyle, family history and personal medical records, in parallel.
Approximately six years after the study began, UK Biobank asked 157,000 participants (from the 500,000 people in the initial study), with and without mental health problems, to complete a comprehensive thoughts and feelings questionnaire. In the largest study to date, the team assessed the severity of suicidal tendencies of the respondents and analysed their DNA to determine if there were any genetic variants associated with suicidality across the genome.
Joining up the data
The survey of suicidal behaviours – from thoughts about self-harm and ending life to suicide attempts – was analysed to provide an order of increasing severity. Three regions of DNA containing genetic variations that were associated with suicidal thoughts and behaviour were identified.
By linking this with death certification records, it was possible to show that more of the genetic variants associated with suicidal tendencies was observed in individuals who died by suicide. They also found some evidence of overlap between the genetic variants for suicidal behaviours and the risk of developing other mental illnesses, most notably major depression and anxiety disorders.
The team is planning its next study to look for further predictors of suicidal behaviours. This will include detailed analysis of the three regions of the genome described above to understand their biological effects. The next question is whether genetic variants will prove to be useful in predicting suicidal behaviour alone, or if this is only achieved when combined with lifestyle, social and negative life events.
Science abstract: Suicidality is complex and includes a broad range of thoughts and behaviours. Genetic factors increasing the severity of suicidality were assessed in 123,000 individuals from the UK Biobank cohort. Three risk markers were identified for suicidality. These showed moderate-to-strong overlap with a range of other psychiatric disorders, most notably depression. Future work will explore how these risk markers, combined with non-genetic risk factors, can be used in suicide risk prediction and prevention.
Dr Rona Strawbridge
UKRI Innovation Fellow at University of Glasgow
Dr Rona Strawbridge completed her BSc in Biochemistry with Medical Biochemistry at Cardiff University (UK) in 2003. Subsequently, Rona moved to the Karolinksa Institute (Sweden) to undertake her...
Using AI to make medical records more accurate and help hospitals plan services
13 April 2023
With only 20% of patient health information previously recorded digitally, researchers realised they needed a system that could make health records more useful to hospitals to improve care.