Summary
A preprint analysis of Bangladesh’s 2025 household survey estimates that about 9 million people live in households whose tested water source exceeded an electrical-conductivity threshold of 1,500 µS/cm. The high readings were concentrated in coastal districts.
A medRxiv preprint analysing Bangladesh’s 2025 Multiple Indicator Cluster Survey estimates that 5.3% of households had tested source water with electrical conductivity (EC) above 1,500 microsiemens per centimetre (µS/cm). That corresponds to about 9 million people; an estimated 1.8 million lived in households whose tested source exceeded 3,000 µS/cm.
The survey measured source water at 6,040 households in 64 districts between February and June. The analysis used survey-weighted estimates. In 14.5% of households, the tested source was not the household’s main drinking-water source, an important distinction when interpreting the results as an indicator of people’s exposure.
High readings were concentrated in coastal districts
Coastal-zone districts accounted for 26% of households but 70% of households with tested source water above 1,500 µS/cm. The share of households above that threshold varied little by wealth or by whether a household included young children or pregnant women.
The well-type comparisons depended on geography. Nationally, deep tube wells more often had high EC, in part because most are in coastal areas. Within the coastal zone, the high-EC shares were similar: 16.1% for deep wells and 17.9% for shallow wells. After adjustment for district, measured EC was about 18% lower in deep wells (geometric mean ratio 0.82; 95% confidence interval 0.73–0.92). In a restricted comparison within upazilas, the ratio was 0.93 (0.80–1.07). The authors also report that the estimate depended on how implausibly low readings were handled.
Arsenic above 10 parts per billion was more common in samples with high EC, but the association weakened after adjustment for district. The authors recommend measuring EC alongside arsenic and documenting meter calibration.
What the measurements indicate
EC measures how readily water conducts electricity; dissolved ions contribute to conductivity, making EC an indicator of salinity. The study classified readings above 1,500 and 3,000 µS/cm as risk categories, but these cut-offs are not health-based drinking-water standards. They should therefore be read as survey thresholds for mapping source-water conditions, not as direct counts of people drinking water shown to be unsafe.
The analysis is a cross-sectional study of water sources, not a measurement of participants’ water consumption or health outcomes. Its findings map where high-EC sources were recorded and how readings varied across household and geographic characteristics. The preprint’s background notes concern about links between saline drinking water and raised blood pressure; this survey itself did not assess blood pressure or other health effects.