Health Stream Literature Summary - Issue 58 - June 2010

Enteric illness risks before and after water treatment improvements.
Frost FJ, Tollestrup K, Roberts M, Kunde TR, Craun GF and Harter L. (2009) Journal of Water and Health, 7(4); 581-589.

Drinking water regulators in the US and elsewhere are concerned about both endemic and outbreak waterborne disease risks in susceptible subpopulations such as the elderly and young children. There are new requirements for unfiltered surface water supplies, even if they are in protected watersheds, which may involve water utilities filtering the water unless they can consistently maintain a very high water quality. The health benefits from new technologies such as filtration and ozonation at water treatment plants have not been extensively studied, even though they can effectively remove and inactivate Giardia and Cryptosporidium oocysts. Results from randomised household intervention studies evaluating home drinking water filtration systems have been conflicting.

This study was conducted to evaluate whether acute gastrointestinal illness risks declined after filtration and ozonation were added to an unfiltered but chlorinated surface water source which came from well-protected watersheds with no evidence of human sewage contamination. The only human pathogens that were known to be present in the source water were Cryptosporidium and Giardia probably from wild animals as there was no livestock grazing in the catchment.

This prospective cohort study was conducted in two geographic areas (control and intervention) in a north-western U.S. city during two time periods: phase 1 from June to November 2000 and phase 2 from June to November 2001. The intervention and control sites were in the same city and were served by one water utility. Each site however was supplied by a different unfiltered, chlorinated surface water source. Both surface water sources were from well-protected watersheds with no source or evidence of human sewage contamination. Treated water from both sources was stored in uncovered, secured and monitored distribution system reservoirs. Both water sources were chlorinated over the whole study period and drinking water met the current US Environmental Protect Agency (EPA) Safe Drinking Water Standards for coliform bacteria. Before the beginning of phase 2, a new water treatment facility was completed at the intervention site. Treatment included ozone for primary disinfection, coagulation/flocculation, high-rate granular filtration and chlorine as a secondary disinfectant. Fluoride and chemicals for corrosion control were added also. Following filtration, turbidity in the intervention site declined. The control site had no change in treatment.

Families were recruited from both sites and were eligible to participate if: 1) a household member included either a child aged 2 to 10 years or an adult at least 65 years of age; 2) the household drank municipal water and did not have a home filtration system; and 3) family members had lived in the residence for at least six months and planned to stay in the community for the next two years. The family needed to live in an area where the water came from only one source. A designated person from each family recorded illness for each member in a daily illness diary for each 6-month period in phase 1 and phase 2. The daily diary recorded whether each person was out-of-town, was well or had diarrhoea (loose/watery stools), vomiting, nausea or other symptoms of colds etc. There were four categories of illness defined: diarrhoea (at least two episodes of soft or loose stools), gastrointestinal illness (having nausea, any vomiting or abdominal cramps), highly credible gastrointestinal illness (HCGI) (at least one of the following: vomiting or liquid diarrhoea or nausea or diarrhoea combined with abdominal cramps) and other illness (fever, chills, headache or cold without enteric symptoms). Symptoms occurring within five days of each other were considered to be related and counted as one episode, if 6 or more days had elapsed without diarrhoea or gastrointestinal symptoms then it was counted as a new episode.

There were 906 participants from the intervention site and 471 from the control site. The analysis included 711 participants from the intervention site and 361 from the control sites who successfully completed both study phases. For both the intervention and control sites, the mean number of reported illnesses in each category per diary-year declined for all age groups during phase 2, except for a slight increase in HCGI in participants aged 20-64 in the intervention group. The decline in illness rates was greater in those less than 20 years and in those 65 years and older. There was only a relatively small decline in gastrointestinal and diarrhoea illness rates for the 20-64 year age intervention group. Unadjusted incidence density ratios (IDRs) for phase 2 were calculated comparing the control and intervention sites for the four categories of illness. For any of the four illness categories there was no significant elevation of the IDRs. Unadjusted IDRs were calculated for the age groups and compared with participants age 2-19 years and both older age groups (20-64 years and 65 plus years) were at higher risk of diarrhoea in phase 2 (IDRs of 1.99 and 1.45). Participants aged 20-64 were also at a slightly higher risk of HCGI compared with the younger age group (IDR 1.03). Participants having three or more diarrhoea episodes in phase 1 had a fivefold higher risk of diarrhoea episodes compared to those who had less than three episodes in phase 1. A similar magnitude of increased risk was also found for gastrointestinal illness (IDR 4.90) and HCGI (IDR 4.32) and other illness (IDR 2.73) was also increased.

Adjusted IDRs were calculated to examine the effect of the intervention as both age and number of illness episodes were related to an elevated IDR. The effect of the intervention was examined for participants with less than three episodes of illness in phase 1 and for participants with three or more episodes of illness in phase 1. IDRs were not significantly elevated or reduced for any of the four types of illness comparing the intervention site and the control site for participants with less than three illness episodes in phase 1. However, for participants with three or more episodes of illness in phase 1, IDRs were elevated in the intervention area for all four types of illness (although not significantly for gastrointestinal and other illnesses) and were 1.16 for other illnesses, 1.86 for gastrointestinal illnesses, 2.23 for diarrhoea and 2.67 for HCGI.

Overall, no differences were found in the risk of any of the illnesses after the new water treatment plant was completed. This study suggests that prior gastrointestinal illness history or the propensity to report the illness may be a significant risk factor for increased risk of current reported gastrointestinal events. The risk of the occurrence of future gastrointestinal, diarrhoea and other illness episodes therefore may be strongly dependent on individual characteristics and/or susceptibility such as age or prior illness rates. Other studies have hypothesised that improved water quality may actually reduce the protective immunity to Cryptosporidium and result in higher rates of illness with gastrointestinal symptoms and diarrhoea. Evidence from this study is insufficient to conclude that elevated risks in phase 2 were due to decreased levels of protective immunity and further studies using high-risk immunosuppressed populations need to be conducted. Also, exposure to Cryptosporidium oocysts and other enteric pathogens can occur from sources other than drinking water (water recreation, raw fruit and vegetable consumption) and these exposures may help to explain the increased risks.

© Copyright Water Quality Research Australia Limited http://www.wqra.com.au/
Health Stream articles may be reproduced and communicated to third parties provided WQRA is acknowledged as the source. Literature summaries are derived in part from copyright material by a range of publishers. Original sources should be consulted and acknowledged.