Estimating exposure to drinking water quality on a national level - Danish Design Drinking Water Epidemiology

Abstract

Denmark has long been considered a model country for conducting epidemiological studies, because the entire population can be evaluated as a cohort due to comprehensive health and administrative registers. We present a nationwide exposure model for drinking water quality linked to longitudinal residential locations and health outcomes. Using our approach, we can now link drinking water exposure variables to the entire Danish population. The model was first introduced in 2014 and has since been updated so that today we can estimate drinking water quality for every resident of Denmark during the past 40 years and has been applied in numerous epidemiological studies on a range of outcomes, from mental health to cancer.Drinking water samples have been analyzed by certified laboratories and registered in a centralized national monitoring database, covering both public supplies and private wells for several decades. In this presentation, we focus on the details for the exposure modelling: linking approx. 9 million drinking water quality samples in space and time to individuals using a geographical information system (GIS). We digitized more than 3500 modern and historical water supply areas and linked them to monitoring data from more than 4500 public waterworks. Users of the more than 60,000 private wells generally have significantly lower water quality than users of public supplies. It was therefore crucial to identify private well users, which we did by proximity analyses.We will present examples that highlight how a detailed spatial and longitudinal exposure assessment can be critical in epidemiological studies of drinking water quality, e.g. compound-specific exposure patterns or closures of waterworks. We will also give an overview of completed and ongoing epidemiological studies based on this exposure model and discuss opportunities for future studies.