Exposure to human source-associated fecal indicators and self-reported illness among swimmers

Background: Current fecal indicator bacteria used to assess illness risks in recreational waters (E. coli, enterococci) cannot discriminate among sources of contamination. To address this limitation, human-associated Bacteroides and chemical markers have been proposed, but the risk of illness associated with human fecal indicators is unclear. We estimated associations between microbial and chemical markers of human fecal pollution and self-reported illness among body immersion swimmers at U.S. beaches during 2003 – 2007. Methods: Participants were surveyed about beach activities and water exposure on the day of their beach visit and followed up 10 to 12 days later to document illness experienced since the beach visit. At 6 beaches, water was analyzed for the presence of human-associated Bacteroides markers: HF183, BsteriF1, BuniF2, HumM2. At 5 beaches, water was analyzed for 56 anthropomorphic chemicals. Adjusted standardized risk differences (RD) and 95% confidence intervals (CI) for the indicator-illness associations were estimated using model-based standardization. Human associated markers were assessed as modifiers of the association between general Enterococcus and illness using interaction contrast. Results: Overall we observed little evidence of association between Bacteroides markers and illness, and between chemical markers and illness among body immersion swimmers. There was a pattern of increased risks of GI illness (RD=1.9%; 0.1%, 3.7%), diarrhea (RD=1.3%; -0.2%, 2.7%), and respiratory illness (RD=1.1%; -0.2%, 2.5%) associated with the BsteriF1 marker. There was no evidence that Bacteroides markers acted as modifiers of general Enterococcus and illness. Several chemicals also showed a pattern of increased risks, including bisphenol A-GI illness, cholesterol-GI illness, household wastewater products-respiratory illness, and tributyl phosphate-respiratory illness. Phenol exposure increased the magnitude of association between general Enterococcus dichotomized at policy-relevant cut-points and GI illness, eye ailments, and respiratory illness by 3-5%. Conclusions: Human-associated Bacteroides and chemical markers were not consistently associated with swimming-associated illness, though patterns suggest possible increased risks. It is not clear that these findings are generalizable to beach sites impacted predominantly by animal sources, runoff, or sporadic sources of contamination. Additional research is needed to support the use of human-associated indicators in predicting illness risks from human fecal pollution of recreational water.Doctor of Philosoph