Social patterning of elevated blood lead among children aged 12–59 months in Bangladesh: a national cross-sectional analysis of MICS7 2025

Lead has no safe level in children. The 2025 Bangladesh Multiple Indicator Cluster Survey (MICS7) provides the first nationally representative blood lead data for the country. Using the WHO Commission on Social Determinants of Health framework, we examined the social patterning of elevated blood lead (≥ 5 µg/dL).

Methods

Of 18,905 age-eligible children, 10,622 had a valid blood lead measurement and 9,717 complete covariates. We fitted survey-weighted logistic (aOR) and modified Poisson (aPR) models with 20 covariates across six domains. Biomarker non-response (43.8%), as expected for pediatric venous sampling, was addressed through inverse-probability-of-selection weighting, multiple imputation ( m = 20), selection-bias scenarios and 500 Rao–Wu bootstrap replicates; the IPW-adjusted prevalence exceeded the naive estimate, suggesting the naive estimate may be conservative under the specified selection model, although this direction is model-dependent.

Results

Weighted prevalence was 38.5% (95% CI 37.2–39.9) at ≥ 5 µg/dL and 66.2% at the CDC reference value (≥3.5 µg/dL). Prevalence rose from 30.2% in the poorest to 51.9% in the richest wealth quintile (aOR 1.24 per quintile; 1.13–1.37). Maternal education (aOR 0.83 per level; 0.79–0.88) and ownership of agricultural land (0.77; 0.67–0.88), livestock (0.77; 0.67–0.88) and the dwelling (0.54; 0.43–0.69) were inversely associated. The wealth association differed across divisions (Wald , p < 0.001), strongest in Dhaka (aOR 1.70; 1.52–1.90). The wealth × maternal-education product term was significant (aOR 0.945; 0.913–0.977), and the wealth association remained positive at higher maternal education (per-quintile aPR 1.09; 1.02–1.16 at higher secondary). The wealth association was robust to most selection-bias scenarios, but attenuated under two adverse asymmetric scenarios and crossed the null under a symmetric worst case (aOR 0.915; 0.833–1.006).

Conclusions

Elevated blood lead follows a reverse socioeconomic gradient: wealthier children are more exposed, while maternal education and agrarian assets are inversely associated. This is a social pattern, not a source attribution; market-mediated pathways remain hypotheses. Surveillance should extend beyond poverty-targeted frameworks while continuing to protect the poorest.