Background:  To reduce mother-to-child HIV transmission in resource-poor settings, the World Health Organization (WHO) recommends exclusive breastfeeding for 6 months followed by rapid weaning when alternatives are not safe or sustainable. The Kisumu Breastfeeding Study (KiBS) is an ongoing trial of maternal ART from 34 weeks’ gestation to 6 months postpartum in conjunction with exclusive breastfeeding and weaning in the sixth month. Preliminary data from the first 18 months (August 2003 to January 2005) indicated that infant rates of diarrheal adverse events rose sharply during the peri-weaning period (age 6 to 8 months). This was attributed to increased exposure to pathogens through introduction of weaning foods. In late 2005 we instituted a safe water system intervention to all participants comprising hygiene education, point-of-use water chlorination, and safe water storage vessels. We assessed effectiveness of safe water system in reducing diarrhea.

Methods:  We determined rates and overall relative risk (RR) of diarrhea, resulting in clinic visits or hospitalizations, among infants 0 to 12 months of age enrolled before and after safe water system distribution. We also tested residual chlorine in water vessels in homes of a subset of post-safe water system infants to assess compliance with water chlorination.

Results:  The 230 pre-safe water system and 254 post-safe water system live-born infants provided 2800 and 2857 observation months, respectively. Overall diarrhea incidence was significantly lower post-safe water system (4.0 episodes/100 infant-months observation) vs pre-safe water system (5.5) [RR = 0.72 (0.57 to 0.91)]. Pre-weaning diarrhea rates were lower for infants in the post- vs pre-safe water system group, but peaked to similar levels in both groups in the peri-weaning period (age 7 months) and after. Detectable chlorine residuals in stored drinking water confirmed treatment in 88% of 796 visits to 138 post-safe water system homes. Number of clinic visits or hospitalizations for diarrhea per 100 infant-months of observation, by month of age, is shown in the table.

Conclusions:  Overall, KiBS infant diarrhea rates were significantly lower after safe water system implementation; most of the effect was seen before weaning. Rapid weaning from exclusive breastfeeding at 6 months was associated with increased diarrhea risk among KiBS infants regardless of safe water system use. The findings support safe water system use to reduce diarrhea risk among HIV-exposed infants particularly in the early months of life. Non-waterborne factors may be responsible for increased peri-weaning diarrhea rates.