Monsoon Flooding and Undernutrition Among Children in Bangladesh: A Population-Level Analysis of the 2017 Monsoon...

Flooding puts young children at risk of adverse nutritional outcomes. With climate change expected to increase the frequency and intensity of flooding, it is important to understand relationships between flooding and childhood undernutrition to inform policy design.

This study aims to examine the relationship between flood exposure and undernutrition in children younger than age 5 years in Bangladesh; investigate heterogeneity in this relationship by geography, child sex, and child age; and test whether socioeconomic status, water access, and repeated flood exposure moderate the relationship between flood exposure and undernutrition.

Satellite flood inundation data for Bangladesh's severe monsoon flooding in August 2017 were combined with nationally representative, cross-sectional household survey data from the Bangladesh Demographic and Health Survey. Surveys were collected between October 2017 and March 2018. Households were classified as flood-exposed if their home location overlapped with the flood inundation map.

The sample includes 6638 children younger than age 5 years in Bangladesh. After applying survey weights, the adjusted sample size is 6620.

Nutritional outcomes include height-for-age and weight-for-age z scores, along with indicators for stunting and underweight. Indicator variables for stunting and underweight describe children ≥2 SD below the reference median height-for-age (stunting) or weight-for-age (underweight) z score.

Logistic and linear regressions with individual-level, household-level, and cluster-level controls, and district and month fixed effects were used to compare nutritional outcomes among children who were flood-exposed with those who were not. In heterogeneity analyses, models were subset by geography, child sex, and child age. Interaction models were used to examine the moderating characteristics of socioeconomic status, water access, and repeated flood exposure.

Flood-exposed children had 1.30 times the odds of experiencing stunting (95% CI, 1.10 to 1.54) and 0.14 lower height-for-age z scores (95% CI, -0.24 to -0.03) compared with children who were not exposed to floods. These relationships were more pronounced in coastal districts, where flood exposure was associated with 1.60 times the odds of stunting (95% CI, 1.08 to 2.35), 0.21 lower height-for-age z scores (95% CI, -0.41 to -0.00), and 0.21 lower weight-for-age z scores (95% CI, -0.36 to -0.07). Among children aged 2 years and younger, flood exposure had a significant association with all measures of undernutrition. Finally, interaction models point to water access as a moderating factor in the relationship between flood exposure and undernutrition.

Monsoon flooding is associated with undernutrition in children, particularly in coastal areas and among very young children. Addressing underlying population-level influences of undernutrition, such as the development of high-quality water sources near the home, may improve the nutritional status of children, especially in areas vulnerable to floods.