New Tool for Predicting Suicidal Behavior Combines Clinical and Biochemical Markers

New Tool for Predicting Suicidal Behavior Combines Clinical and Biochemical Markers

Posted: September 22, 2015
New Tool for Predicting Suicidal Behavior Combines Clinical and Biochemical Markers

Story highlights


A research team has identified biochemical and clinical signs that may help predict suicide attempts, estimated to cause one death every 40 seconds worldwide.

The team, led by 2002 and 2005 NARSAD Young Investigator grantee Alexander Niculescu, M.D., Ph.D., of the Indiana University School of Medicine, examined gene expression patterns  in blood cells of men with psychiatric diagnoses at the Indianapolis Veterans Affairs Medical Center. They correlated patterns of gene expression with suicidal thoughts. Such thoughts often do not always lead to an actual attempt, but their frequency correlates with risk for suicidal behavior.

Publishing their results online August 18 in Molecular Psychiatry, Dr. Niculescu and colleagues focused on genes whose activation or suppression most strongly corresponded with suicidal thoughts.  They then prioritized  these genes based on genetic evidence from past research, or if they had been shown to be changed in expression in brain tissue studied from individuals who committed suicide.  Lastly, the team validated the genes by showing even stronger changes in expression in blood samples from suicide completers.

The genes the team ultimately identified have a range of functions. Some affect the function of the body’s immune system, while others have roles in growth regulation, stress responses, and a signaling mechanism called the m-TOR pathway that is important in controlling the cell cycle.  That pathway has been implicated in recent research on depression, and in the mechanism of action of the experimental fast-acting depression treatment ketamine.

Importantly, some of the genes whose activity in the study was correlated with suicide risk are involved in disorders that can heighten the chance of suicide —mood disorders, anxiety, psychosis.

Another goal of the study was to test clinical measures for predicting suicidal behavior. The team looked at two tests, called CFI-S (Convergent Functional Information for Suicide) and SASS (Simplified Affective State Scale). Both can be taken by at-risk individuals using smart-phone apps. CFI-S tests factors that may influence suicide attempts including mental, physical and environmental health, cultural risks, age and gender. SASS, in contrast, assesses mood and anxiety levels.

Combining the genetic results with scores on the CFI-S and SASS tests, the researchers were able to predict suicidal thoughts across all psychiatric diagnoses with levels of accuracy that, depending on the specific indicator, they characterized as “good” or “very good” (over 90% accuracy).  They were somewhat less accurate in predicting  future hospitalizations for suicidality (over 70% accuracy).  The combined predictive tool worked best for people with mood disorders, such as bipolar disorder and depression.

The researchers say future work should extend these analyses to women — typically underrepresented at veterans’ hospitals  — and people who may be at risk for suicide without having a psychiatric diagnosis.  They hope such tools will come into widespread use for early intervention and prevention.

Read the paper.

Support research now. Donate today!