Background: Suicide in children is a pressing public health concern. The increasing number of deaths by suicide and emergency visits for suicidal ideation and self-harm in children might not be fully representative; it is likely that many more children are in distress but do not seek out help. We conducted a systematic review and meta-analysis of existing studies to quantify the prevalence of suicidal ideation and self-harm behaviours among children in the community aged 12 years and younger.

Methods: In this systematic review and meta-analysis, we searched PsycINFO, MEDLINE, and Web of Science via OVID from database inception to Feb 28, 2022, for articles published in French or English that reported estimates of prevalence of suicidal ideation (including suicide planning) and self-harm behaviours (namely, self-harm, suicide attempts, and non-suicidal self-injury) in children aged 12 years and younger. Reference lists were also searched; case studies, qualitative studies, and health-care visit studies were excluded. The outcomes were suicidal ideation, suicide plan or attempts, and self-harm. We used a random-effects model to calculate the overall pooled prevalence of suicidal ideation and self-harm behaviours for all timeframes combined and for ever versus the past 12 months for suicidal ideation. We used the Joanna Briggs Institute Critical Appraisal tool to evaluate the risk of bias in each study. This study is registered with PROSPERO, CRD42020179041.

Findings: 28 articles, encompassing 30 studies overall, met the inclusion criteria, aggregating findings from 98 044 children (of whom 46 980 [50·5%] were girls and 46 136 [49·5%] were boys; six articles did not report sex or gender) aged 6-12 years. The pooled prevalence estimate was 7·5% (95% CI 5·9-9·6) for suicidal ideation from 28 studies and 2·2% (2·0-2·5) for suicide planning from three studies. The pooled prevalence was 1·4% (0·4-4·7) for self-harm from four studies, 1·3% (1·0-1·9) for suicide attempt from six studies, and 21·9% (6·2-54·5) for non-suicidal self-injury from two studies. The prevalence of suicidal ideation was higher in studies that included child-reported outcomes (10·9% [95% CI 8·1-14·5] for child only and 10·4% [6·8-15·5] for child and parent combined) than for parent-only reported outcomes (4·7% [3·4-6·6]; p=0·0004). The prevalence of suicidal ideation and self-harm behaviours was similar in boys and girls (suicidal ideation, 7·9% [95% CI 5·2-12·0] for boys vs 6·4% [3·7-10·7] for girls; self-harm behaviours, 3·5% [1·6-7·2] for boys vs 3·0% [1·4-6·4%] for girls). Detailed ethnicity data were not available. High heterogeneity was identified across estimates (I²>90%), which was not well explained by the characteristics of the studies.

Interpretation: A high number of children in the general population can experience suicidal ideation and self-harm behaviours, thus underlining the need for more research on childhood suicide, including developmentally appropriate preventive strategies, such as youth-nominated support teams or dialectical behavioural therapy.

Funding: Canada Research Chair in Youth Mental Health and Suicide Prevention.

Declaration of interests We declare no competing interests.