Considering Alternate Pathways of Drinking-Water Contamination: Evidence of Risk Substitution from Arsenic Mitigation Programs in Rural Bangladesh

Estimating effects of monsoon flooding on household water access

The importance of climate in water resources management is well recognized, but less is known about how climate affects water access at the household level. Understanding this is crucial for identifying vulnerable households, reducing health and well-being risks, and finding equitable solutions. Using difference-in-differences regression analyses and relying on temporal variation in interview timing from multiple, cross-sectional surveys, we examine the effects of monsoon riverine flooding on household water access among 34 000 households in Bangladesh in 2011 and 2014. We compare water access, a combined measure of both water source and time for collection, among households living in flood-affected and non-flood-affected districts before and after monsoon flooding events. We find that households in monsoon flood-affected districts surveyed after the flooding had between 2.27 and 4.42 times higher odds of experiencing low water access. Separating geographically, we find that while households in coastal districts have lower water access than those in non-coastal districts, monsoon flood exposure is a stronger predictor of low water access in non-coastal districts. Non-coastal districts were particularly burdened in 2014, when households affected by monsoon flooding had 4.71 times higher odds of low water access. We also find that household wealth is a consistent predictor of household water access. Overall, our results show that monsoon flooding is associated with a higher prevalence of low water access; socioeconomically vulnerable households are especially burdened.

Deep tubewell use is associated with increased household microbial contamination in rural Bangladesh: Results from a prospective cohort study among households in rural Bangladesh

Deep tubewells are important sources of arsenic mitigation in rural Bangladesh. Compared to commonly available shallow tubewells, deep tubewells tap into deeper low-arsenic aquifers and greatly reduce exposure to arsenic in drinking-water. However, benefits from these more distant and expensive sources may be compromised by higher levels of microbial contamination at point-of-use (POU). This paper examines differences in microbial contamination levels at source and POU among households using deep tubewells and shallow tubewells, and investigates factors associated with POU microbial contamination among deep tubewell users. We assessed a prospective longitudinal cohort of 500 rural households in Matlab, Bangladesh, across 135 villages. Concentration of Escherichia coli (E. coli) in water samples at source and POU using Compartment Bag Tests (CBTs) was measured across rainy and dry seasons. We employed linear mixed-effect regression models to measure the effect of different factors on log E. coli concentrations among deep tubewell users. CBT results show that log E. coli concentrations are similar at source and at POU during the first dry and rainy season, but are significantly higher at POU among deep tubewell users during the second dry season. Log E. coli at POU among deep tubewell users is positively associated with both presence (exponentiated beta exp(b) = 2.52, 95% Confidence Interval (CI) = 1.70, 3.73) and concentration of E. coli (exp(b) = 1.36, 95% CI = 1.19, 1.54) at source, and walking time to the tubewell source (exp(b) = 1.39, 95% CI = 1.15, 1.69). Drinking-water during the second dry season is associated with reduced log E. coli (exp(b) = 0.33, 95% CI = 0.23, 0.57) compared to the rainy season. These results suggest that while households that use deep tubewells have lower arsenic exposure, they may be at higher risk of consuming microbially contaminated water compared to households that use shallow tubewells.