Health Effects and Water Quality Following Low Pressure Events in Drinking Water Distribution Systems in the United States

Abstract

Background: Low pressure events (LPEs), defined as a water service disruption that presumably lowers system water pressure, can cause drinking water contamination, resulting in increased illness risk to consumers.

Objectives: This study sought to examine whether LPEs increase the risk for highly credible acute gastrointestinal illness (HCGI) and acute respiratory illness (ARI) and to compare water quality in exposed and unexposed areas in the United States.

Methods: A matched cohort study was conducted during the period 2015-2019. For each LPE, household survey exposed areas were matched 1:2 with unexposed areas based on water-main size and material at the point of repair, as well as the housing type and demographic characteristics of the local population from the most recent census tract data. Water samples were collected to monitor physicochemical and microbiological water quality parameters. Households (HHs) were surveyed about water use and illness during the 2 wk after the LPE. Multivariable log-binomial models clustered on utility and LPE number were used to investigate associations between LPE exposure and HCGI and ARI.

Results: Five water utilities reported 58 LPEs, including planned maintenance (76%) and emergency (24%) events. Controlling for livestock near home, private well presence, number of people in HH, and travel away from home, exposed HHs were at higher risk of HCGI in comparison with unexposed HHs ($\text{risk ratio}=1.20$; 95% confidence interval: 1.05, 1.37). No associations between LPE and ARI were detected. Certain LPEs characteristics like pipe material, size, and depth were associated with an increased HH HCGI risk. HHs experiencing LPEs where low disinfectant residual, high adenosine triphosphate, or general Bacteroidales were detected in water following an LPE repair were also at a higher risk for HCGI.

Conclusions: LPEs were associated with 20% higher risk of HCGI in HHs. Planned improvements to water distribution system infrastructure, adherence to industry standard distribution repair practices, and water monitoring following pipe repairs could supplement community alert systems to reduce illnesses from LPEs. https://doi.org/10.1289/EHP15564.

Figure 1.

Response rates of household and individual surveys through mail and web link — United States, 2015–2019. Note: HCGI, highly credible gastrointestinal illness (any vomiting, diarrhea ($>2$ loose stools in 24 h), or abdominal pain with nausea); HH, household.

Figure 2.

Unadjusted and adjusted RRs of health outcomes (HCGI or ARI) by LPE exposure—United States, 2015–2019. $p$-Values represent the probability of a significant difference between the presence of the health outcome and LPE exposure. HCGI HH aRR adjusted for presence of private well, livestock near home, household size, and any nights spent away from home. HCGI individual aRR adjusted for sex, drinking from untreated surface water, and chronic conditions. ARI HH aRR adjusted for HH size, and any nights spent away from home. ARI individual aRR adjusted for swimming in untreated surface water, exposure to childcare, any chronic abdominal or respiratory illness, and being Black/African American. Note: ARI, acute respiratory illness (defined as cough, sore throat, running or stuffy nose, shortness of breath, fever); aRR, adjusted risk ratio; HCGI, highly credible gastrointestinal illness (defined as any vomiting, diarrhea ($>2$ loose stools in 24 h), or abdominal pain with nausea); LPE, low pressure event; RR, risk ratio.