Association between depression and compliance to treatment of tuberculosis and its outcomes: a systematic review and meta-analysis

Depression is common among people with tuberculosis (TB) and may undermine adherence and end-of-treatment success. We quantified the association between depression and adverse TB outcomes, including non-compliance. Following PRISMA 2020, we searched PubMed, Embase, PsycINFO, Web of Science, Scopus, and WHO Global Index Medicus till August 2025 with no language limits. Eligible observational studies assessed depression (validated scale or clinical diagnosis) in patients on anti-TB therapy and reported treatment outcomes. Risk of bias was appraised using the Newcastle-Ottawa Scale. Odds ratios (ORs) were pooled with random-effects (DerSimonian-Laird), heterogeneity quantified with I² and 95% prediction intervals, small-study effects examined by DOI plot/LFK index, and robustness evaluated with leave-one-out analyses. We also synthesized adjusted odds ratios (aORs) when reported for non-compliance. 13 studies were included. Depression was associated with substantially higher odds of unfavourable end-of-treatment outcomes (failure, loss to follow-up, death): pooled OR = 3.66 (95%CI: 1.51–8.90; I²=81.2%; prediction interval: 0.20-65.99). Leave-one-out analyses yielded pooled ORs 1.31–4.79. DOI plot showed major asymmetry (LFK = 6.96). For non-compliance (7 studies), depression increased risk (OR = 2.85; 95%CI: 1.48–5.46; I²=80.6%; prediction interval: 0.34–23.79) with minor asymmetry (LFK = 1.85); leave-one-out estimates ranged 1.99–2.67. Using adjusted estimates for non-compliance (4 studies), the association strengthened and heterogeneity disappeared (aOR = 3.48; 95%CI: 2.51–4.45; I²=0%). Depression is strong, independent predictor of non-compliance and unfavourable TB outcomes. Routine depression screening and integrated, stepped mental-health care within TB services are likely to improve adherence and treatment success; prospective trials of such models are warranted. Not applicable.