Chlorinated drinking water exposure and bladder and colon-rectum cancers in the province of Modena (Northern Italy): An ecological study

Introduction: Different by-products in chlorinated drinking waters, including cloroform, brominated trihalomethanes and halacetones, are causing concern over human health, due to their potential carcinogenicity. Epidemiological studies have reported exceeding risks for bladder, colon and rectum cancers in subjects living in areas where chlorinated drinking waters are distributed in comparison with subjects living in areas where no chlorinated drinking water is used. However, up to now, major difficulties in assessing past-exposure do not allow to formulate final conclusions. An Italian association on cancer research has recently granted a study with the aim to evaluate the relationship between chlorinated drinking water consumption and the incidence of bladder and colon-rectum cancers in the province of Modena (Northern Italy).Methods: This study has an ecological design: the units are the 47 municipalities in the province of Modena and the outcomes the incidence rates of bladder (IX-ICD:188), colon (IX-ICD:153) and rectum (IX-ICD:154) cancer, observed in these areas in 1988-94 as reported by Modena Cancer Registry. To evaluate each municipality exposure, information on the main characteristics (disinfection treatment, source and quality) of drinking waters were collected asking local waterworks managers to fill in a questionnaire. As in the mountainous municipalities many little springs represent the watersource, many waterplants are operative inside each unit (overall, more than 200 waterplants were identified). In this case municipality exposure was evaluated by combining information collected by each waterwork. A municipality was considered exposed if in 1980-90 drinking waters were disinfected with chorine. Hierarchical Bayesian methods were applied to calculate smoothed standardized incidence rates (SIRs), relative risks (RRs), and 95% CI.Results: The overall incidence of bladder, colon and rectum cancer in Modena in 1988-1994 was 24.6, 30.5 and 15.8 (x104) respectively. Spatial differences were observed mainly for bladder cancer. In subjects living in areas supplied in 1980-1990 by chlorinated drinking waters in comparison with subjects living in areas supplied by untreated drinking waters an increased risk for bladder cancer (RR = 1.17; 95% CI: 1.02-1.37) was observed, while no differences were evidenced for colon (RR = 1.07; 95% CI: 0.96-1.24) and rectum (RR = 1.02; 95% CI: 0.90-1.16) cancer. Ecological covariates, including population density and % of subjects living in the same house in period 1980-1990 in each municipality were included in the model.Conclusion: Nevertheless the difficulties in determining past-exposure and the weaknesses of the study design (ecological fallacy), this study suggests, in accordance with literature data, an exceeding risk for bladder cancer in subjects living in areas where chlorinated drinking waters are supplied.