Suicide and dementia: Systematic review and meta-analysis

Introduction: To analyse the possible relationship between dementia in the elderly and the subsequent development of suicide ideation, attempts and / or completed suicides.

Methods: Systematic review and meta-analysis.

Selection criteria: studies that analysed the relationship between dementia and suicide.

Search strategy: i) in PubMed, EMBASE, CINAHL, IME and Lilacs until December 2018; ii) manual search of the bibliography of selected articles; iii) contact with leading authors. Article selection and data extraction according to a predefined protocol, including bias risk assessment, were performed by independent peer reviewers. The effect size index was calculated using Odds Ratio (OR) and its 95% confidence interval (random-effects model). Heterogeneity was evaluated with forest plots, Cochran's Q and I² index. Assessment of publication bias using funnel plots ("trim-and-fill" method) and the Egger test. The analysis of moderating variables was performed using a multiple meta-regression under a mixed-effects model.

Results: 37 studies and 47 basic units of study were identified. Effect size of the association of dementia with: Suicidal Ideation OR = 1.37 (95% CI: .78-2.39); Suicide Attempt: OR = 2.24 (95% CI: 1.01-4.97); and Completed Suicide: OR = 1.28 (95% CI: .77-2.14). Possible publication bias was ruled out.

Conclusions: A trend towards suicidal events is identified, especially suicide attempts in people with dementia. Greater attention and care are recommended after a recent diagnosis of dementia, especially with adequate assessment of comorbidities, which could influence the occurrence and outcome of suicidal events.