Diabetes Mellitus Associated with Arsenic Exposure in Bangladesh

Size: px

Start display at page:

Download "Diabetes Mellitus Associated with Arsenic Exposure in Bangladesh"

Benjamin French
6 years ago
Views:

American Journal of Epidemiology Copyright 99 by The Johns Hopkins University School of Hygiene and Public Health All rights reserved Vol., No. Printed in U.S.A. Diabetes Mellitus Associated with

1 American Journal of Epidemiology Copyright 99 by The Johns Hopkins University School of Hygiene and Public Health All rights reserved Vol., No. Printed in U.S.A. Diabetes Mellitus Associated with Arsenic Exposure in Bangladesh Mahfuzar Rahman, Martin Tondel, Sk Akhtar Ahmad, and Olav Axelson The objective of this study was to assess whether arsenic exposure is a risk factor for diabetes mellitus as indicated in a few earlier studies. Arsenic in drinking water is known to occur in western Bangladesh, and in 996, two of the authors conducted a survey of the prevalence of diabetes mellitus among 6 subjects with keratosis taken as exposed to arsenic and unexposed individuals. Diabetes mellitus was determined by history of symptoms, previously diagnosed diabetes, glucosuria, and blood sugar level after glucose intake. The crude prevalence ratio for diabetes mellitus among keratotic subjects exposed to arsenic was. (9% confidence interval.-.) and increased to. (9% confidence interval.-.) after adjustment for age, sex, and body mass index. On the basis of a few earlier measurements of arsenic concentrations in drinking water by the authorities in Bangladesh and another new ad hoc analyses, approximate time-weighted exposure levels to arsenic in drinking water could be estimated for each subject. Three time-weighted average exposure categories were created, i.e., less than.,.-., and more than. mg/liter. For the unexposed subjects, the corresponding prevalence ratios were.,.6,.9, and., representing a significant trend in risk (p <.). The result corroborates earlier studies and suggests that arsenic exposure is a risk factor for diabetes mellitus. Am J Epidemiol 99;:9-. body mass index; ecology; environment; keratosis; prevalence; water An increased risk for diabetes mellitus has been reported among people exposed to arsenic through drinking water (). Indications of a relation between arsenic exposure and diabetes mellitus have also been obtained in studies of copper smelter workers () as well as in the art glass industry (). This study is a further investigation of whether or not these earlier observations can be reproduced with regard to arsenic exposure through drinking water. Occupational arsenic exposure seems to cause lung cancer and possibly also other cancers (). Skin disorders such as hyperkeratosis, hyperpigmentation, and depigmentation have been described in relation to arseniccontaining drinking water, and there are dermatologic manifestations in children (). Noncirrhotic portal fibrosis is another disorder reported after consumption of Received for publication January,99, and accepted for publication December, 99. Abbreviations: BMI, body mass index; Cl, confidence interval; MH-PR, Mantel-Haenszel weighted prevalence ratio. Division of Occupational and Environmental Medicine, Department of Health and Environment, Faculty of Health Sciences, Linkoping University, Linkoping, Sweden. Department of Occupational and Environmental Health, National Institute of Preventive and Social Medicine (NIPSOM), Mohakhali, Dhaka, Bangladesh. Reprint requests to Prof. Olav Axelson, Division of Occupational and Environmental Medicine, Department of Health and Environment, Faculty of Health Sciences, Linkoping University, Linkoping, Sweden. water containing arsenic (6). Peripheral vascular disease and cardiovascular disease have been associated with environmental exposure to inorganic arsenic in drinking water (-) as well as occupational exposures (-). Exposure to arsenic through drinking water is a serious problem reported to occur in various countries, including Argentina, Chile, Taiwan, and the United States (). Particularly high arsenic levels have been reported in West Bengal, an Indian province bordering Bangladesh (9-). West Bengal and Bangladesh form a geologic continuity, and the occurrence of arsenic in drinking water seems to depend on arsenic-rich sediments. Withdrawal of a large amount of ground water may decrease the moisture in the soil, thereby facilitating penetration of oxygen into the ground with subsequent oxidation of arsenic-containing minerals and release of arsenic into the aquifers (), but other mechanisms have also been discussed (). In view of the earlier observations (-) indicating a possible association between arsenic exposure and diabetes mellitus, the occurrence of this disorder was examined in the present study among people in western Bangladesh, who were drinking tubewell water containing arsenic. MATERIALS AND METHODS Study design This study was designed as a comparison of the prevalence of diabetes mellitus among subjects living 9 Downloaded from on January

2 Diabetes Mellitus and Arsenic in Drinking Water 99 in areas with and those without exposure to arsenic through drinking water. Extensive measurements of arsenic in tubewell water were not economically possible, and therefore, only subjects with keratosis were enrolled to ensure definite arsenic exposure. The prevalence of diabetes mellitus was compared for this group versus a population with uncontaminated water supply. Exposed subjects Subjects with keratosis were included in the study as the exposed and recruited from six districts with arsenic-contaminated drinking water (table ). Keratosis was considered to be present when signs were found of ) spotted keratosis on palms and soles, often associated with spotted melanosis; ) toad skin on dorsum of hands and feet; or ) hardened patches of skin on the extremities. The recruitment procedure began with one patient with keratosis from Razarampur village, who was treated at the Rajshahi Medical College Hospital. This patient was interviewed at home, and through him, we heard about more keratosis patients in that village. From the same hospital and the local private practitioners in Nawabgong district, we learned about other keratosis patients. The names of another seven villages known to have arsenic-contaminated drinking water were obtained from the Public Health Department. The villages were surveyed on a door-to-door basis to identify keratosis patients for the study. Ten women, but no men, refused participation; exposed keratotic subjects were willing to participate in the study and were interviewed in their homes. Because there were no cases of diabetes mellitus in the age group below years, exposed subjects in that age group were primarily excluded from the analyses. There were three keratotic subjects with diabetes mellitus in one family, but no other family had a TABLE. Arsenic in drinking water in areas affected by arsenical keratosis and Dhaka, Bangladesh, 996 Districts Nawabgong Razbari Khustia Narayongong Faridpur Bagerhat Dhaka Areas Razarampur Barughuria Surzanagar Courtpara Barakhada Pallpara Devinagar Lakpur ' Analyses performed in Bangladesh. No. Of samples 6 Arsenic concentrations (mg/liter) *.-..-o..-.*.* <. history of diabetes mellitus. These three subjects were excluded from some of the analyses to ensure that they were not driving any increase in risk of diabetes mellitus. Unexposed subjects As there were no resources available for an extensive arsenic analysis program, it was impossible to clearly identify unexposed subjects in the villages. We therefore had to arbitrarily recruit unexposed individuals through a door-to-door visit in four suburbs of Dhaka. The Dhaka population is not known to be exposed to arsenic in drinking water, which is provided from the Water and Sewage Agency. The unexposed reference population involved individuals aged years or more. They were interviewed in the same way as the keratotic, exposed subjects. Questionnaire interview and examinations A structured personal interview was carried out or supervised by two of the authors (M. R. and A. A.). Information obtained included resident history on the source and duration of the use of drinking water in the villages along with sociodemographic characteristics and possible symptoms of diabetes. Positive answers regarding suspected symptoms of diabetes, especially polyuria, were followed by a urine sample. These samples were taken early in the morning and tested with a glucometric strip (BM Redia-Test Glucose, Boehringer Mannheim GmbH, Mannheim, Germany), and those who had glucosuria were further sampled in the evening. All subjects who had two positive samples were also examined for hyperglycemia after overnight fasting and hours after a -g glucose intake. These examinations were all undertaken in nearby hospitals in Dhaka as well as in the study areas with contaminated water. Standing height and body weight were measured with the subjects wearing light clothes and not wearing shoes. Individual body mass index was calculated as [weight (kg)]/[height (m)]. Exposure assessment Information about arsenic concentration in drinking water was obtained from an unpublished report by the Department of Occupational and Environmental Health, National Institute of Preventive and Social Medicine, Bangladesh. Furthermore, water samples from eight villages were taken ad hoc and brought to Sweden for analysis at the Division of Occupational and Environmental Medicine, Linkoping University, Linkoping, Sweden (table ). Am J Epidemiol Vol., No., 99 Downloaded from on January

3 Rahman et al. Approximate time-weighted mean arsenic exposure levels were calculated over the lifetime of each subject as y -(a,- CjyZjdj, where a } is the number of years a well with arsenic concentration c y - was used, assuming that the current levels of arsenic in the well water were also representative of the past. Arsenic levels may have been different in the past, however, but there is no reliable information for this. The resulting estimates were then categorized as I, II, and HI, i.e., low, medium, and high, corresponding to arsenic concentrations of less than.,.-., and more than. mg/liter. Data analysis The data were stratified according to age (-, -9, and ^6 years), sex, and body mass index (BMI); the BMI categories were less than 9, 9-, and >. Mantel-Haenszel weighted prevalence ratios (MH-PR) with 9 percent confidence intervals and a test for trend were calculated by means of the Epi-Info package (). RESULTS A diagnosis of diabetes mellitus could be confirmed in persons among the exposed and among the unexposed. Twenty of the cases among the unexposed subjects in Dhaka were known diabetics and were taking medication containing insulin or sulphonylurea. Among the exposed, only the aforementioned family with three diabetic members knew about their disease. The crude overall prevalence ratio for diabetes mellitus was. (9 percent confidence interval (CI).-.), and after adjustment for age, sex, and BMI, the MH-PR was.9 (9 percent CI.9-.6) (table ). However, in subsequent analyses, we excluded the three keratotic and diabetic cases from the same family as possibly representing a hereditary clustering; maintaining adjustment for age, sex, and BMI, the prevalence ratio remained fairly high (MH-PR =., 9 percent CI.-.). With adjustment for age and sex only, the MH-PR was. (9 percent CI.-6.), whereas the prevalence ratio increased to 6. (9 percent CI.-9.) with adjustment for BMI and sex. Thus, BMI appeared as a negative confounder, as also shown in table, whereas age exerted a weak positive confounding for developing diabetes mellitus. As can be seen in table, the prevalence ratios showed a dose-response pattern with a significant trend in relation to the exposure categories (p <.). The numbers behind the rather high prevalence ratio of. in the highest exposure category were small, however. TABLE. Distribution of diabetes mellitus cases and healthy subjects according to age and body mass index (BMI),* Bangladesh, 996 Age (years) and BMI - <9 9- > -9 <9 9- > 6 <9 9- > Crude prevalence ratio MH-PRt 9% Clt / healthy subjects Unexposed Exposed * Sex was stratified for in the analyses but is not shown in the table. t MH-PR, Mantel-Haenszel weighted prevalence ratio; CI, confidence interval. DISCUSSION In this study, a significantly increased prevalence of diabetes mellitus was found among subjects with keratosis compared with subjects who did not have keratosis. A significant trend in risk between an approximate, time-weighted arsenic exposure and the prevalence of diabetes mellitus strengthens the possibility of a causal association. However, the lack of a comprehensive, systematic, long-term sampling of the water supplies in the study area is a limitation of the study because directly measured individual exposure data over time would have been desirable. The calculated, approximate, time-weighted arsenic exposure obviously cannot take into account any time trends in the historical exposure or the likely fluctuations in exposure depending on precipitation. This limitation causes an uncertainty about the arsenic exposure estimates on the basis of the assumption that the current arsenic concentrations were also those of the past. It is Am J Epidemiol Vol., No., 99 Downloaded from on January

4 Diabetes Mellitus and Arsenic in Drinking Water TABLE. Bangladesh, 996 Distribution of diabetes mellitus cases and healthy subjects according to age, sex, and exposure categories,* Age (years) UnexposedI I Exposure category II III - subjects subjects 9 CM CO 6 6 subjects 9 CM CM subjects MH-PRt MH-PR 9% Clt MH-PR 9% Cl MH-PR 9% Cl ' Body mass index was not considered as exerting negative confounding, and three exposed males were excluded, t MH-PR, Mantel-Haenszel weighted prevalence ratio; Cl, confidence Interval. % tor trend, 6.; p <.. nevertheless reasonable to believe that the available water measurements assessed arsenic exposure properly enough for creating the broad exposure categories used in the analysis of trend. The screening for diabetes mellitus by first considering symptoms and then testing for glucosuria was necessary for the practical and economic circumstances under which the study was carried out. The procedure might have led to limited sensitivity but has probably affected the exposed subjects in the same way as those who were unexposed and should be expected to bias the prevalence ratio toward the null. Furthermore, there was also a tendency for the unexposed subjects to have a higher BMI than those who were exposed, resulting in a fairly strong negative confounding. There is a lack of information regarding other potential confounding factors than age, sex, and BMI. It is not known, for example, whether other contaminants, such as trace metals or pesticides, might have occurred in the drinking water and played a role. A possible confounder could also be dietary habits, although they were taken care of indirectly to some extent by the adjustment for BMI. Since there was indication of a negative confounding from BMI, it is unlikely that unknown confounding from diet would have led to increased risks, but rather the reverse. It was obvious from the direct contacts with the subjects in the survey that the reference population had a somewhat better educational, occupational, and socioeconomic status, since they lived in the Dhaka district. This difference in background would most likely mean a greater chance to discover diabetes mellitus in the Dhaka population as well as longer survival of diabetic subjects. As already mentioned, there were also proportionally fewer new cases discovered in the Dhaka population than among the exposed subjects. Again, this would mean negative confounding and, therefore, could not explain the observed increased risks. Another aspect is that the procedure for recruiting subjects could somehow have favored the participation of those who suffered from diabetes mellitus. If so, any difference in this respect between the exposed and the unexposed areas would rather have favored the appearance of cases in the Dhaka population, and such a phenomenon would hardly have produced a doseresponse relation. In ground water, arsenic occurs in tri- and pentavalent forms. The ultimate source and mechanism for the contamination of tubewell water with arsenic in Bangladesh can be ascertained only by further studies. It is noteworthy in this context that the lifetime risk of dying from cancer of the liver, lung, kidney, or bladder from drinking liter per day of water containing. mg of arsenic per liter (that is, the current standard by the United States Environmental Protection Agency) could be as high as per, persons (6). A recent risk analysis by the Environmental Protection Agency predicts an increased lifetime risk of skin cancer on the order of or per, from chronic exposure to. mg/liter of inorganic arsenic in drinking water (). A need for revision of the standard to even less than. mg/liter was therefore discussed. The levels displayed in table may be considered against these Am J Epidemiol Vol., No., 99 Downloaded from on January

5 Rahman et al. risk estimates regarding cancer. It is currently unclear whether the risk of contracting diabetes mellitus also operates at the low exposure levels that may cause cancer, and a mechanism through which inorganic arsenic would induce diabetes mellitus is unknown. To investigate further the likely diabetogenic effect of arsenic, various sources of exposure might be considered for future studies. Other health effects of arsenic probably also need consideration in Bangladesh, as ingested arsenic might induce blackfoot disease and ischemic heart disease, probably through a direct effect on the atherogenic process involving endothelial cells, smooth muscle cells, platelets, and macrophages (). Diabetes mellitus is an important determinant for peripheral vascular disease and ischemic heart disease, so such disorders could also be caused indirectly via arsenicinduced diabetes mellitus. Little attention has been paid to this possibility as yet, however, because the likely connection between long-term exposure to arsenic and diabetes mellitus has only recently attracted interest. The situation regarding arsenic in drinking water as so far reported in Bangladesh and West Bengal is alarming (, ). To our knowledge, this is the first report to indicate health effects of arsenic-contaminated drinking water in Bangladesh. The source and mechanism of the arsenic problem has to be identified clearly and remedial measures taken. There is already enough knowledge about the adverse effects of arsenic exposure, however, to make clear the urgent need for a technical solution of how to provide good-quality water to large populations both in Bangladesh and elsewhere. ACKNOWLEDGMENTS This study was partially supported by a planning grant from SIDA (Swedish International Development Cooperation Agency). It was approved by the ethical committee of Bangladesh Medical Research Council, and participation was voluntary. The authors thank Dr. Ireen Akhter, International Center for Diarrhoel Disease Research in Bangladesh, Dr. Pappu and Mr. Dipu, Bangladesh Institute of Research and Promotion of Essential Reproductive Health and Technologies, and Dr. Nurul for their support. REFERENCES. Lai MS, Hsueh YM, Chen CJ, et al. Ingested inorganic arsenic and prevalence of diabetes mellitus. Am J Epidemiol 99; 9:-9.. Rahman M, Axelson O. Diabetes mellitus and arsenic exposure: a second look at case-control data from a Swedish copper smelter. Occup Environ Med 99;:-.. Rahman M, Wingren G, Axelson O. Diabetes mellitus among Swedish art glassworkers an effect of arsenic exposure? Scand J Work Environ Health 996;:6-9.. International Agency for Research on Cancer. IARC monographs on the evaluation of the carcinogenic risk of chemicals to humans: some metals and metallic compounds. Vol.. Lyon, France: International Agency for Research on Cancer, 9:9-.. World Health Organization. Environmental health criteria : arsenic. Geneva, Switzerland: World Health Organization, 9:-. 6. Datta DV, Kaul MK. Arsenic content of drinking water in villages in Northern India. A concept of arsenicosis. J Assoc Physicians India 96;: Tseng WP. Effects and dose-response relationships of skin cancer and blackfoot disease with arsenic. Environ Health Perspect 9; 9:9-9.. Zaldivar R. Arsenic contamination of drinking water and foodstuffs causing endemic chronic poisoning. Beitr Pathol 9;:-. 9. Tseng CH, Chong CK, Chen CJ, et al. Dose-response relationship between vascular disease and ingested inorganic arsenic among residents in blackfoot disease endemic villages in Taiwan. Atherosclerosis 996;:-.. Chen CJ. Blackfoot disease. (Letter). Lancet 99;6:.. Wu MM, Kuo TL, Huang YH, et al. Dose-response relation between arsenic concentration in well water and mortality from cancers and vascular diseases. Am J Epidemiol 99; :-.. Chen CJ, Hsueh YM, Lai MS, et al. Increased prevalence of hypertension and long-term arsenic exposure. Hypertension 99;:-6.. Chen CJ, Chiou HY, Chiang MH, et al. Dose-response relationship between ischemic heart disease mortality and longterm arsenic exposure. Arterioscler Thromb Vase Biol 996; 6:-.. Tseng CH, Chong CK, Chen CJ, et al. Abnormal peripheral microcirculation in seemingly normal subjects living in blackfoot-disease-hyperendemic villages in Taiwan. Int J Microcirc Clin Exp 99;:-.. Axelson O, Dahlgren E, Jansson CD, et al. Arsenic exposure and mortality: a case-referent study from a Swedish copper smelter. Br J Ind Med 9;:-. 6. Lagerkvist BE, Linderholm H, Nordberg GF. Arsenic and Raynaud's phenomenon: vasospastic tendency and excretion of arsenic in smelter workers before and after the summer vacation. Int Arch Occup Environ Health 9;6:6-.. Gustavsson P, Gustavsson A, Hogstedt C. Excess mortality among Swedish chimney sweeps. Br J Ind Med 9;: -.. Wingren G, Axelson O. Mortality in the Swedish glassworks industry. Scand J Work Environ Health 9;: Datta DV, Mitra SK, Chuttani PN, et al. Chronic oral arsenic intoxication as a possible aetiological factor in idiopathic portal hypertension (non-cirrhotic portal fibrosis) in India. Gut 9;:-.. Garai R, Chakraborty AK, Dey SB, et al. Chronic arsenic poisoning from tube-well water. J Ind Med Assoc 9;:-.. Guha Mazumder DN, Guptas JP, Chakraborty AK, et al. Environmental pollution and chronic arsenicosis in South Calcutta. Bull World Health Organ 99;:-.. Guha Mazumder DN, Chakraborty AK, Ghose A, et al. Chronic arsenic toxicity from drinking tubewell water in rural West Bengal. Bull World Health Organ 9;66: Mandal BK, Chowdhury TR, Samanta G, et al. Arsenic in groundwater in seven districts of West Bengal, India the biggest arsenic calamity in the world. Current Sci 996;: Bagla P, Kaiser J. India's spreading health crisis draws global arsenic experts. Science 996;:-. Am J Epidemiol Vol., No., 99 Downloaded from on January

6 Diabetes Mellitus and Arsenic in Drinking Water. Dean JA, Dean AG, Burton A, et al. EPI INFO. Atlanta, GA:. Brown KG, Guo HR, Kuo TL, et al. Skin cancer and Centers for Disease Control, 9. inorganic arsenic: uncertainty-status of risk. Risk Anal 6. Smith AH, Hopenhayn-Rich C, Bates MN, et al. Cancer risks 99;:-. from arsenic in drinking water. Environ Health Perspect 99;. Ross R. The pathogenesis of atherosclerosis: an update. 9:9-6. N Engl J Med 96;:-. Am J Epidemiol Vol., No., 99 Downloaded from on January

Relations between exposure to arsenic, skin lesions, and glucosuria

Relations between exposure to arsenic, skin lesions, and glucosuria Occup Environ Med 1999;56:277 281 277 Division of Occupational and Environmental Medicine, Department of Health and Environment, Faculty of Health Sciences, Linköping University, 581 85 Linköping, Sweden