(P23-03-26) Understanding Heterogeneity in Exclusive Breastfeeding Interventions in South Asia: A Systematic Review and Meta-analysis of Community Randomized Trials

Objectives: Exclusive breastfeeding (EBF) for the first six months reduces infant morbidity and mortality; however, intervention effects across randomized trials in South Asia vary widely. This study aimed to synthesize evidence from community-based randomized controlled trials (RCTs) and explore sources of heterogeneity using classical and machine learning meta-regression.

Methods: We conducted a systematic review and meta-analysis of peer-reviewed community-based RCTs (including cluster designs) published before January 2025, evaluating infant and young child feeding (IYCF) and maternal–child health interventions among pregnant and lactating women in South Asia. Random-effects models estimated pooled relative risks (RRs) for EBF at 3, 5, and 6 months. Heterogeneity was quantified using I², publication bias assessed using Egger’s test, and subgroup and meta-regression analyses explored moderators including intervention type, outcome focus, follow-up design, country, and duration. Random Forest and Gradient Boosting models identified key moderators and nonlinear associations.

Results: Interventions significantly increased EBF at 3 months (RR = 1.43, 95% CI: 1.15–1.77; 6 studies), 5 months (RR = 3.14, 95% CI: 1.49–6.63; 5 studies), and 6 months (RR = 1.37, 95% CI: 1.10–1.71; 15 studies) compared with controls, with substantial heterogeneity (I² >94%). No publication bias was detected (Egger p = 0.23). Effects were stronger in IYCF-focused and integrated community-based programs. Multivariable meta-regression showed that longitudinal follow-up (RR = 1.54, 95% CI: 1.02–2.34; p = 0.042) and IYCF as a primary outcome (RR = 1.50, 95% CI: 1.05–2.15; p = 0.028) explained about 35% of between-study heterogeneity. Machine learning models confirmed follow-up design and outcome focus as key moderators.

Conclusions: Community-based maternal and child health interventions substantially improve EBF in South Asia, but effects vary considerably. Study design features - particularly longitudinal assessment and explicit IYCF outcome measurement - partly explain this heterogeneity.

Funding Sources: No external funding