Drinking Water in Rural Saskatchewan: Public Perception of Water Quality and Health Risks, and Direct and Indirect Effects of Drinking Water Quality on Chronic Disease

Abstract

In rural areas of Saskatchewan, tap water supplied to houses is not typically tested for quality and safety with the same intensity as public supplies that serve larger populations. Consequently, rural residents might be at greater risk of exposure to poor quality water and any resulting health impacts. The overall objective of this study was to investigate if poor quality water in rural areas was directly or indirectly associated with increased occurrence of diabetes and cardiovascular disease. Perceptions about water quality and safety and how these relate to choices about drinking water among rural Saskatchewan residents were also investigated. Existing water surveillance data from both public water supplies and private wells in SK were summarized using a combination of principal components analysis and geostatistics. The summarized water data were used to estimate regional water quality exposure indicators for use in Bayesian hierarchical models examining ecological associations with health outcomes derived from administrative health data. A quarter of 2065 respondents to a questionnaire sent to rural SK residents reported being unsatisfied with the aesthetic quality of their water, although fewer (12%) believed their water was unsafe to drink. Of the respondents, 31% reported drinking primarily bottled water, while 61% reported drinking tap water at least daily and 48% reported treating their water in the home. The type of water supply along with past experiences and familiarity with the water were consistently associated with risk perception. As expected, perception of quality and risk were important predictors of drinking water choices; aspects of familiarity, experience, and type of water supply were also important. The parameters listed under health standards and aesthetic objectives grouped differently for public water supply and private well data following the application of principal components analysis, suggesting residents using different types of water supplies may be exposed to different patterns of contaminants. Summarizing water quality data through geostatistical models resulted in attenuation of extreme concentrations recorded in the observed data, but appeared to predict trends in water quality that could be useful for prioritizing monitoring efforts and public health messaging about water testing for private wells. Overall, poor groundwater quality was not associated with increased occurrence of diabetes or cardiovascular disease. An increase in principal component scores for public water supplies, characterized mainly by the presence of high levels of hardness and magnesium, was associated with a decrease in the prevalence of ischemic heart disease. This finding was consistent with previously reported results in other regions, and raises the question of whether the in-home treatment of water to remove high mineral content could inadvertently increase the risk of cardiovascular disease. Studies with individual-level exposure measures are recommended to more definitively characterize potential associations between water quality and chronic disease. This study used innovative methods to address gaps in knowledge about perceptions of water quality and risk and drinking water choices for people living in rural SK, summarized water quality over a large region of the province, and investigated associations between water quality and the occurrence of important chronic diseases in rural Saskatchewan.