Evaluation of Drinking Water Disinfectant Byproducts Compliance Data as an Indirect Measure for Short-Term Exposure in Humans

Abstract

In the absence of shorter term disinfectant byproducts (DBPs) data on regulated Trihalomethanes (THMs) and Haloacetic acids (HAAs), epidemiologists and risk assessors have used long-term annual compliance (LRAA) or quarterly (QA) data to evaluate the association between DBP exposure and adverse birth outcomes, which resulted in inconclusive findings. Therefore, we evaluated the reliability of using long-term LRAA and QA data as an indirect measure for short-term exposure. Short-term residential tap water samples were collected in peak DBP months (May–August) in a community water system with five separate treatment stations and were sourced from surface or groundwater. Samples were analyzed for THMs and HAAs per the EPA (U.S. Environmental Protection Agency) standard methods (524.2 and 552.2). The measured levels of total THMs and HAAs were compared temporally and spatially with LRAA and QA data, which showed significant differences (p < 0.05). Most samples from surface water stations showed higher levels than LRAA or QA. Significant numbers of samples in surface water stations exceeded regulatory permissible limits: 27% had excessive THMs and 35% had excessive HAAs. Trichloromethane, trichloroacetic acid, and dichloroacetic acid were the major drivers of variability. This study suggests that LRAA and QA data are not good proxies of short-term exposure. Further investigation is needed to determine if other drinking water systems show consistent findings for improved regulation.

Exposure Characterization of Haloacetic Acids in Humans for Exposure and Risk Assessment Applications: An Exploratory Study

“…Site A contained higher amounts of chlorinated HAAs than Site B, which contained slightly higher levels of TOBr and brominated HAA species (e.g., BCAA). The difference in TOX and TOCl levels at the two sites is likely due to differences in amount of natural organic matter and water treatment conditions at the two sites, such as chlorine dose, reaction time, and residual time [6,28,34,54]. Contrary to TOX and TOCl, the measured levels of individual HAAs (except DCAA) in drinking water showed no significant site-specific differences.…”

“…Most population-based epidemiologic studies use aggregated compliance data on five-HAAs for assessing their association with adverse reproductive outcomes. However, only a few studies reported the use of individual HAAs, especially urinary measures and reproductive health risk, which may provide more conclusive evidence of health risk [28]. For example, a retrospective cohort study from Central Arizona reported that the water concentrations of DCAA > 18 µg/L and TCAA > 17.8 µg/L increase the risk of small for gestational age (SGA), and the consumption of DBAA (>5 µg/L) and DCAA (>8 µg/L) later in pregnancy (weeks 33–40) elevate the risk of low birthweight and intrauterine growth restriction [29].…”

“…However, the composition of individual HAAs changes with water distribution systems, disinfectant type, and nature of NOM, and the sum total of the five regulated HAAs for compliance use does not account for differences in individual HAAs. It also does not account for toxicokinetic mechanisms or intra- and interindividual variability due to behavior differences in water use activities and water source characteristics, which could potentially change HAA exposure [1,6,28,33,34].…”

“…Parvez et al’s (2017) study on short-term (weekly and biweekly) exposure assessments of regulated HAAs in drinking water showed high exposure variability. Thirty-five per cent of water specimens had HAA levels above the regulatory limit, demonstrating that compliance data does not capture short-term exposure variability, which is essential to determine their correlation with reproductive health risks [28]. Moreover, the compliance-based drinking water data do not account for differences in exposure routes, multiple water sources (e.g., bottle water, office, home, and gym), or variability in HAA composition between urinary and drinking water metrics.…”

Maternal Exposure to Disinfection By-Products and Risk of Hypospadias in the National Birth Defects Prevention Study (2000–2005)

“…Given the temporal and spatial variability of DBPs throughout a distribution system [ 37 , 38 ], we cannot rule out that such sparse water sampling requirements contribute to exposure misclassification in our study. This is an important consideration due to the relatively short window of exposure relevant to birth defects and that LRAAs and quarterly averages can be problematic when used for brief exposure periods [ 39 ]. In addition to this Berkson-type measurement error, we must also consider that the DBP concentration estimates utilized in this study are susceptible to classical measurement error [ 40 ].…”