ABSENCE OF RELATIONSHIP BETWEEN HEALTH-EFFECTS DUE TO TAP WATER-CONSUMPTION AND DRINKING-WATER QUALITY PARAMETERS

An 18-month prospective epidemiological study of gastrointestinal illnesses was conducted on 300 families consuming conventionally treated tap water and 300 consuming the same water after further treatment by reverse-osmosis. Drinking water met current bacteriological and physicochemical quality standards, but was found to be associated with a significant level of gastrointestinal illnesses: a reduction of 30% of the gastrointestinal illnesses was observed in the group consuming the filtered water. The presence or absence of total coliforms or fecal coliforms was not indicative of the health effects observed. The heterotrophic plate counts at 20-degrees-C in the distribution system were weakly associated with the duration of the symptoms when the data was analyzed by subregion. Several approaches to clustering of the family data to the nearest sampling site were attempted: no association could be demonstrated. Even if consumers of reverse-osmosis water experienced, on the average, less gastrointestinal illnesses were significantly associated with the number of bacteria growing at 35-degrees-C on medium R2A. The problems associated with the predictive value of the bacterial content of a water sample and in particular, the major differences between water quality at the tap and in the distribution system are presented. For example, standard procedures require analysis of the water after flushing the tap for several minutes: this is not however typical of the water that is consumed. Water that comes out of the tap has stagnated for long periods in household pipes and regrowth of bacterial contaminants can easily occur. The bacteria growing in this water might thus be responsible for some of the health observed in tap water.