Prevention of suicidal behavior in older people: A systematic review of reviews.

Older people have the highest rates of suicide, yet the evidence base on effective suicide preventions in late-life is limited. This systematic review of reviews aims to synthesize data from existing reviews on the prevention and/or reduction of suicide behavior in late-life and evidence for effectiveness of interventions. A systematic database search was conducted in eight electronic databases from inception to 4/2020 for reviews targeting interventions among adults ≥ 60 to prevent and/or reduce suicide, suicide attempt, self-harm and suicidal ideation. Four high quality reviews were included and interventions categorized as pharmacological (antidepressant use: 239 RCTs, seven observational studies) and behavioral (physical activity: three observational studies, and multifaceted primary-care-based collaborative care for depression screening and management: four RCTs). The 2009 antidepressant use review found significant risk reduction for suicide attempt/self-harm (OR = 0.06, 95% CI 0.01-0.58) and suicide ideation (OR = 0.39, 95% CI 0.18-0.78) versus placebo. The 2015 review found an increased risk of attempts with antidepressants versus no treatment (RR = 1.18, 95% CI 1.10-1.27) and no statistically significant change in suicides versus no treatment (RR = 1.06, 95% CI 0.68-1.66) or ideation versus placebo (OR = 0.52, 95% CI 0.14-1.94). Protective effects were found for physical activity on ideation in 2 out of 3 studies when comparing active versus inactive older people. Collaborative care demonstrated significantly less attempts/ideation (OR = 0.80, 95% CI 0.68-0.94) in intervention group versus usual care. The results of this review of reviews find the evidence inconclusive towards use of antidepressants for the prevention of suicidal behavior in older people, thus monitoring is required prior to start, dosage change or cessation of antidepressants. Evidence to date supports physical activity and collaborative management for reduction of suicide ideation, but additional trials are required for a meta-analysis. To build on these findings, continued high-quality research is warranted to evaluate the effectiveness of interventions in late life.