Exposure to arsenic causes keratosis, hyperpigmentation, and hypopigmentation and seemingly also diabetes mellitus, at least in subjects with skin lesions. Here we evaluate the relations of arsenical skin lesions and glucosuria as a proxy for diabetes mellitus. Through existing measurements of arsenic in drinking water in Bangladesh, wells with and without arsenic contamination were identified. Based on a questionnaire, 1595 subjects > or = 30 years of age were interviewed; 1481 had a history of drinking water contaminated with arsenic whereas 114 had not. Time weighted mean arsenic concentrations and mg-years/l of exposure to arsenic were estimated based on the history of consumption of well water and current arsenic concentrations. Urine samples from the study subjects were tested by means of a glucometric strip. People with positive tests were considered to be cases of glucosuria. A total of 430 (29%) of the exposed people were found to have skin lesions. Corresponding to drinking water with < 0.5, 0.5-1.0, and > 1.0 mg/l of arsenic, and with the 114 unexposed subjects as the reference, the prevalence ratios for glucosuria, as adjusted for age and sex, were 0.8, 1.4, and 1.4 for those without skin lesions, and 1.1, 2.2, and 2.6 for those with skin lesions. Taking exposure as < 1.0, 1.0-5.0, > 5.0-10.0 and > 10.0 mg-years/l of exposure to arsenic the prevalence ratios, similarly adjusted, were 0.4, 0.9, 1.2, and 1.7 for those without and 0.8, 1.7, 2.1, and 2.9 for those with skin lesions. All series of risk estimates were significant for trend, (p < 0.01). The results suggest that skin lesions and diabetes mellitus, as here indicated by glucosuria, are largely independent effects of exposure to arsenic although glucosuria had some tendency to be associated with skin lesions. Importantly, however, glucosuria (diabetes mellitus) may occur independently of skin lesions.
