Population attributable fractions of psychiatric disorders for suicide in low and middle-income countries: a systematic review and meta-analysis

Background Low and middle-income countries (LMICs) have a high burden of suicide with limited resources. The population attributable fraction (PAF) is a valuable tool to help policymakers prioritise suicide prevention activities. Objectives To estimate the PAFs of psychiatric morbidity in suicide in LMICs. Methods We searched MEDLINE, Embase, and PsycINFO for empirical studies assessing the prevalence of psychiatric disorders among suicide in LMICs. This is an update and extension of a previous review. We meta-analysed estimates of associations, and calculated PAFs with 95% confidence intervals (CI) for different psychiatric disorders in relation to suicide. Publication bias was assessed using contour plots. Results 14 studies which met our eligibility criteria were included. Five studies were published from China, 13 were case-control studies, and all the studies gathered data between 2000 and 2021. The studies assessed 2132 suicides. Estimates from 10 reasonable quality studies contributed to our meta-analysis. The PAF was 62.1% (95% CI 48.5%-73.6%) for any psychiatric disorder, 28.4% (12.7%-50.1%) for mood disorder, and 23.0% (9.7%-42.8%) for depression. There was substantial heterogeneity in association estimates for these exposures between included studies. PAFs ranged between 1.6% and 7.6% for all other psychiatric disorder categories identified. Contour plots indicated publication bias. Conclusions Treating psychiatric disorders could reduce the incidence of suicide by more than a half in some LMICs. Our findings need to be interpreted with caution due to data not being available from 94% of LMICs and evidence of publication bias which would lead to overestimations in PAF estimates.