Abstract
<title>Abstract</title> <p>Background Mental-health prevalence studies in Nigeria are dispersed across conditions, instruments and high-risk settings, limiting their use for research planning. This review aimed to synthesise current and point prevalence for ten common mental-health condition families in Nigerian populations. Methods PubMed/MEDLINE and Europe PMC formed the core bibliographic search; African Journals Online and the first retrievable Google Scholar results were supplementary. Searches were run on 6 August 2026. Eligible reports used validated current or < = 30-day measures, with approved condition-specific exceptions kept secondary. JBI prevalence criteria informed risk-of-bias judgments. Logit-transformed inverse-variance random-effects models with REML between-study variance were fitted in R; screening and diagnostic estimates were separated. Results One hundred three reports contributed 133 estimates. Pooled screening prevalence was 30.2% (95% CI 24.6%-36.5%) for depression and 34.8% (25.9%-44.9%) for anxiety; diagnosed depression was 20.2% (15.4%-26.0%). The current prevalence of substance-use disorder was 16.4% (11.9%-22.3%), prevalence of diagnosed dementia was 2.2% (0.7%-6.9%), prevalence of cognitive impairment screening was 24.1% (21.7%-26.6%), prevalence of psychosis was 4.2% (1.3%-12.7%), prevalence of antenatal depression was 16.7% (8.4%-30.4%), prevalence of postpartum screens was 18.5% (9.0%-34.4%), prevalence of diagnosed PTSD was 12.2% (4.5%-29.3%), current prevalence of suicidal ideation was 20.4% (11.6%-33.4%), prevalence of current attempts 2.3% (0.7%-6.9%), and prevalence of ADHD was 8.5% (5.3%-13.4%). PTSD screening in trauma-exposed samples was 58.8% (37.1%-77.5%). No eligible bipolar point-prevalence estimate was found. Heterogeneity was usually substantial (I² 58.0%-99.3%). Conclusion These estimates are planning values for populations resembling the contributing studies, not national population rates. Researchers should match p to the target population and measurement method; sparse and highly heterogeneous outcomes require conservative sensitivity calculations.</p>