New tool could predict suicide risk

Pitt study maps score for behavior in youth

In the face of the large and growing problem of suicide — there were nearly 45,000 such deaths in the U.S. in 2016 — researchers from the University of Pittsburgh School of Medicine reported Wednesday a new approach to predicting suicidal behavior.

Suicide is the second-leading cause of death for people 10 to 34 years of age, according to the federal Centers for Disease Control and Prevention. The Pitt research, published in JAMA Psychiatry, concluded that the severity and variability of depression symptoms in this age group stood out, of all symptoms, as the strongest indicators that a high-risk young person would attempt suicide.

The researchers recommended that doctors pay attention to those two characteristics of depression

in high-risk young adults. By monitoring and treating the symptoms, they said, it will lower the chance of a suicide attempt.

“Predicting suicidal behaviors is a most challenging task,” said senior author Nadine Melhem, Ph.D., Pitt associate professor of psychiatry and researcher at UPMC Western Psychiatric Hospital.

Previously, she said, all of the predictions so far were shown to be only slightly better than chance.

The new Prediction Risk Score was developed after following 663 young adults whose parents had been diagnosed with mood disorders.

“We followed them because we knew they were at risk of suicide,” Ms. Melhem said.

The average age was 24. Over the course of 12 years, the parents and children went through regular evaluations, being assessed for psychiatric diagnoses and symptoms of depression, hopelessness, irritability, impulsivity, aggression and impulsive aggression.

The most accurate predictors were if patients had severe depressive symptoms and large fluctuations in them over time. These changes were combined with five other factors for the score: younger age (30 years or younger), mood disorders, childhood abuse, and personal and parental history of suicide attempts.

Participants in the study knew researchers were looking at psychiatric symptoms, but didn’t learn their prediction score. Results were 87 percent accurate at predicting someone who is likely to attempt a suicide.

The score predicted a suicide-attempt risk, on average, within 30 to 45 weeks, Ms. Melhem said. Among the 663 young people, 71 had suicide attempts during the study period.

Ms. Melhem said the risk score would be updated as a person’s risk changes — depending on past and present depressive symptoms, their variability and the other factors.

The score is a good tool for clinicians, she explained, but isn’t necessarily useful for patients.

Doctors could focus on the risk factors and the predictive score, without any additional testing or cost, she said. Patients at risk could be identified and more closely monitored, she said.

Colleagues in the research included David Brent, Pitt professor of psychiatry, and John Mann, professor of psychiatry at Columbia University.

Ongoing research, using a larger group of study subjects and adding other factors, aims to improve the power of the Prediction Risk Score, Ms. Melhem said. Biological components are being included, she said, such as indicators of stress response and inflammation.

“Any time clinicians are in touch with high-risk teenagers and young adults, they should be assessed for suicide risk,” Ms. Melhem said. “We know that 90 percent of the people who do attempt suicide do have the highest risk.”