Risk estimates of arsenic related skin lesions in two large villages in Rajshahi Division, Bangladesh

Abstract

In rural Bangladesh people drink water from an estimated 10-11 million tube wells, many with arsenic concentrations well above the national standard of 50µg/L. Characteristic skin lesions and more serious health effects are probable. The research reported here was designed to determine the relation between arsenic concentration and skin lesions on hands and feet controlling for education, use of tube well water, protein consumption and body mass index (BMI). The investigation included a well survey, a prevalence survey and a nested case-referent analysis. Studies were conducted in two villages (population 11670) in northern Bangladesh. In the well survey 1509 wells in use were identified and measurements of arsenic concentration made in 1422 (94%). Three estimates were made for each well, which yielded correlation among repeats = 0.93. The overall median was 47mg/L and the highest recorded 1760 mg/L. Paramedics examined soles and palms of 11087 individuals for skin lesions and identified 168 cases (1.5%). In the third phase, cases (over 16 years) were matched by age, gender and village to referents (target of 3) without skin lesions. Cases and referents were interviewed about protein in diet, use of well water, education and residential history. Conditional logistic regression was used to assess the effects of arsenic concentration, controlling for confounders. Subjects with lesions had a higher median concentration (250µg/L) than those without (47µg/L). Prevalence increased with both age and arsenic concentration, reaching 26.7% in those over 50 years of age and >500 µg/L. Of the 160 cases (≥16 years) 137 were interviewed, 127 with arsenic concentration measured in well water, together with their 504 referents. Cases were somewhat more likely to have lived in the villages throughout their lives and less likely to report using tube well water for cooking. The final model confirmed the high risk of lesions associated with arsenic concentrations. Using ≤50µg/L for comparison, an odds ratio of 15.2 (95%CI 7.2-32.2) was observed for those using tube wells with concentrations >500 µg/L, adjusting for use of tube well water in cooking and lifetime residence in the villages. The results from this research provide continuing support for the use of 50µg/L as a useful national standard. While the enforcement of this standard has immediate value, it cannot be considered final without comparable information on more serious health risks.