Association between residential greenness and anemia risk among women of reproductive age in low- and middle-income countries.

BACKGROUND

Anemia is a major public health burden in low- and middle-income countries (LMICs), particularly among women of reproductive age (WRA). While biological causes are established, evidence on environmental factors like residential greenness is limited.

METHODS

We analyzed data from 1,074,790 WRA across 40 LMICs using Demographic and Health Surveys (2005-2019), linked with Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI) across multiple buffer sizes (500-3,000 m). Associations between greenness and anemia were assessed using a generalized linear mixed-effects model with adjustment for demographic, socioeconomic, and environmental confounders. Under a hypothetical scenario, we further estimated the fraction and number of anemia cases associated with greenness exposure.

RESULTS

High greenness was consistently associated with lower anemia risk across all buffer sizes of NDVI and EVI. Linear models suggested that the odds ratio (OR) of anemia was 0.932 [95% confidence interval (95% CI): 0.929-0.936] for each 0.1-unit increase in NDVI-1000 m. Non-linear models further indicated that the inverse association became more pronounced at higher greenness levels. Stratified analyses showed that the effect size of greenness varied across subgroups defined by BMI, breastfeeding status, wealth index, residence, and education. Under the hypothetical greenness improvement scenario, 2.25% (95% CI: 1.20-3.30%) of anemia cases among WRA were associated with greenness levels relative to the target NDVI-1000 m level, with substantial variation across countries and exposure buffers.

CONCLUSION

Higher residential greenness was associated with lower anemia risk among WRA in LMICs. These findings suggest that residential greenness may represent a potentially relevant environmental factor in future research and public health strategies addressing anemia-related environmental determinants.