QUALITY OF DRINKING-WATER AND ITS RELATION TO GASTROINTESTINAL DISEASE IN CHILDREN.

Abstract

Evidence from NZ and overseas suggests that gastrointestinal diseases transmitted by drinking-water contribute significantly to the disease burden in developed countries. However there has been no systematic or strategic research in New Zealand to examine the size of the burden of endemic gastrointestinal disease, risk factors for illness, or sources of infection. Our aim was to test the relationship between acute gastrointestinal disease and quality of drinking-water, as measured by (i) actual levels of contamination with indicator organisms and Campylobacter, (ii) transgression of microbial drinking-water standards, (iii) grading of water supply by the Ministry of Health, and (iv) traditional Maori value of the water supply. The sampling frame comprised all primary schools of more than 90 children that were on municipal water supplies. All schools on poor water quality supplies (ie >5% of routine samples contaminated with E. coli) were asked to participate and were matched by geographic location and socioeconomic status with an equal number of schools that had good water quality supplies (ie. those of Aa–Bb grade and that complied with the current New Zealand Drinking-Water Standards). A daily diary was used to gather information about water usage, selfreported gastrointestinal disease (GID) and several potential confounding factors from a cohort of 1,194 primary school-aged children over a sevenweek period, at which time the drinking-water quality at each school was monitored. Preliminary results indicate that poor quality drinking-water supplies are a significant cause of GID, relative risk of 1.35 (95% CI of 1.09–1.65; P < 0.05) and an attributable risk of 0.075. These results must be viewed with caution as only 15% of the pupils at the selected schools chose to participate in the diary study, giving rise to potentially serious selection bias although once enlisted the response rate was 85%. Selection bias was addressed by a preliminary questionnaire that sought information about exposure of 1,194 participants and 1,685 non-participants to confounding factors, which showed that similar exposures were reported by both groups for most attributes measured. However, participants reported significantly higher exposure than non-participants to animals. Conclusion. The elevated risk of GID attributed to poor quality drinking-water applies to the 11% of the population not connected to a community water supply and the 10% that are connected to d or e grade or ungraded community drinking-water supplies, with a population attributable risk of 1.4%.