Risky decision-making in suicide attempters, and the choice of a violent suicidal means: an updated meta-analysis

Studies showed disadvantageous decision-making in suicide attempters. The present meta-analysis aims to examine the stability of these findings and related questions.

EMBASE and Pubmed databases were searched for studies published between 01/01/2000 and 01/01/2020 with an additional search through bibliographical references. English or French articles published in peer-reviewed journals, reporting quantitative task-based measures of decision-making in suicide attempters were included: 3,582 records were identified, 33 full-text articles screened, and 21 articles finally included.

All studies were conducted in mood disorders; 18 used the Iowa Gambling Task (IGT) and 3 the Cambridge Gamble Task (CGT). With the IGT, suicide attempters showed riskier choices than patient controls (Hedges' g=-0.28 95%CI (-0.44 - -0.12)) and healthy controls (g=-0.54 (-0.83 - -0.25)) with no significant difference between control groups. The difference between suicide attempters and patient controls was not related to age group, mood disorder type, author, or research center while an effect of time of publication was found (p=0.006). Poorer performance was also found in suicide attempters compared to patient controls when using the CGT (g=-0.57 95%CI (-0.82 - -0.31)). Suicide attempters who used a violent means showed poorer IGT performance than those who used a non-violent means (3 studies).

Limited number of studies outside mood disorders. No data to calculate a gender effect.

The present meta-analysis confirmed riskier decision-making in suicide attempters. Although group differences appear to be of modest effect size in general, they were particularly marked in the subgroup of those who used a violent suicidal means.