Drinking water nitrate estimation at household-level in Danish population-based long-term epidemiologic studies

This study presents an approach making drinking water quality data from the Danish national geodatabase Jupiter available for epidemiological studies on long-term health effects of drinking water quality. Drinking water quality was assigned to all Danish households using spatial methods, and the related uncertainty due to sampling density was identified. Nitrate was used as an example contaminant. From 1978, 98% of all Danish households were geocoded with their precise location, yielding a total of 69 million evaluated household-years. Households supplied by a private well (4% of all household-years) were identified; the remaining majority was supplied by public waterworks. Water supply areas were connected to the Jupiter geodatabase to estimate drinking water quality at publicly supplied households. For privately supplied households, the exposure estimate had a substantially higher uncertainty, with 52% of household-years never being sampled, compared with 1% of publicly supplied household-years. For the 18% of all household-years with no nitrate samples, concentrations were estimated by interpolation and an uncertainty score based on the closest nitrate measurement was introduced. While nitrate was used as an example contaminant, the proposed approach is generic and can be used for all monitored drinking water parameters, after preprocessing and validation of the specific data and assumptions.