Private, public, and bottled drinking water: Shared contaminant-mixture exposures and effects challenge

Abstract

Background: Humans are primary drivers of environmental-contaminant exposures worldwide, including in drinking-water (DW). In the United States, point-of-use DW (POU-DW) is supplied via private tapwater (TW), public-supply TW, and bottled water (BW). Differences in management, monitoring, and messaging and lack of directly-intercomparable exposure data influence the actual and perceived quality and safety of different DW supplies and directly impact consumer decision-making.

Objectives: The purpose of this paper is to provide a meta-analysis (quantitative synthesis) of POU-DW contaminant-mixture exposures and corresponding potential human-health effects of private-TW, public-TW, and BW by aggregating exposure results and harmonizing apical-health-benchmark-weighted and bioactivity-weighted effects predictions across previous studies by this research group.

Discussion: Simultaneous exposures to multiple inorganic and organic contaminants of known or suspected human-health concern are common across all three DW supplies, with substantial variability observed in each and no systematic difference in predicted cumulative risk between supplies. Differences in contaminant or contaminant-class exposures, with important implications for DW-quality improvements, were observed and attributed to corresponding differences in regulation and compliance monitoring.

Conclusion: The results indicate that human-health risks from contaminant exposures are common to and comparable in all three DW-supplies, including BW. Importantly, this study's target analytical coverage, which exceeds that currently feasible for water purveyors or homeowners, nevertheless is a substantial underestimation of the breadth of contaminant mixtures in the environment and potentially present in DW. Thus, the results emphasize the need for improved understanding of the adverse human-health implications of long-term exposures to low-level inorganic-/organic-contaminant mixtures across all three distribution pipelines and do not support commercial messaging of BW as a systematically safer alternative to public-TW. Regardless of the supply, increased public engagement in source-water protection and drinking-water treatment is necessary to reduce risks associated with long-term DW-contaminant exposures, especially in vulnerable populations, and to reduce environmental waste and plastics contamination.

Keywords: Bottled water; Contaminant mixtures; Drinking water; Inorganics; Organics; Point of use; Private supply; Public supply.

Fig. 1.

Group comparison of concentrations of select inorganics detected in private–tapwater (Private; cyan), public–tapwater (Public; white), and bottled–water (Bottled; purple) samples. Solid red lines indicate Environmental Protection Agency (EPA) maximum contaminant level (MCL: enforceable for public tapwater; reference only for private tapwater) and Food and Drug Administration (FDA) standard of quality (SOQ: enforceable for bottled water). Maximum contaminant level goals (MCLG) for As, U, and Pb are zero. For NO₃-N, MCLG and MCL are the same. Circles (●) are data for individual samples. Boxes, centerlines, and whiskers indicate interquartile range, median, and 5th and 95th percentiles, respectively. Numbers at top center of plots indicate the permuted probability that the centroids and dispersions are the same (PERMANOVA; 9999 permutations) across all drinking-water types. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2.

Detected concentrations (circles, μg/L) of organics in private–tapwater (Private; left), public–tapwater (Public; center), and bottled–water (Bottled; right), arranged top to bottom in order of decreasing overall detection frequency. Circles (●) are data for individual samples. Boxes, centerlines, and whiskers indicate interquartile range, median, and 5th and 95th percentiles, respectively. Box colors identify disinfection byproduct (DBP), pesticide, per/polyfluoroalkyl substances (PFAS), pharmaceutical, and volatile organic chemical (VOC) classes, as shown in legend. Numbers on right Y-axis indicate compound-/supply-specific detection frequencies (percent of samples).

Fig. 3.

Group comparisons of cumulative concentrations (circles, μg/L) of all organics (upper left) and of select organic classes detected in private–tapwater (Private; cyan fill), public–tapwater (Public; white), and bottled–water (Bottled; purple) samples. DBP, VOC, and PFAS are disinfection byproducts, volatile organic chemicals, and per/polyfluoroalkyl substances, respectively. For DBP, the solid red line indicates the Environmental Protection Agency (EPA) maximum contaminant level (MCL: enforceable for public tapwater; reference only for private tapwater) and Food and Drug Administration (FDA) standard of quality (SOQ: enforceable for bottled water) for trihalomethanes (80 μg/L). For PFAS, the solid red line indicates the Environmental Protection Agency (EPA) maximum contaminant level (MCL: enforceable for public tapwater; reference only for private tapwater) and presumptive Food and Drug Administration (FDA) standard of quality (SOQ: enforceable for bottled water) for PFOS or PFOA (0.004 μg/L). Circles (●) are data for individual samples. Boxes, centerlines, and whiskers indicate interquartile range, median, and 5th and 95th percentiles, respectively. Numbers at top center of plots indicate the permuted probability that the centroids and dispersions are the same (PERMANOVA; 9999 permutations) across all drinking-water types. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4.

Left Plot – Group comparisons of heterotrophic plate count (HPC) results, a general indicator of concentrations (colony forming units [CFU] per 100 mL) of viable heterotrophic microorganisms in private–tapwater (Private; cyan fill), public–tapwater (Public; white), and bottled–water (Bottled; purple) samples. Circles (●) are data for individual samples. Boxes, centerlines, and whiskers indicate interquartile range, median, and 5th and 95th percentiles, respectively. Right Plot – Scatterplot of HPC versus cumulative concentration of disinfection byproducts (DBP, μg/L) in private–tapwater (Private; cyan fill), public–tapwater (Public; white), and bottled–water (Bottled; purple) samples, illustrating the public-health trade-off of chlorine disinfection of drinking water. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5.

Cumulative TQ (${\sum }_{\text{TQ}}$, left) and EAR (${\sum }_{\text{EAR}}$, right) for analytes detected in private–tapwater (Private; cyan fill), public–tapwater (Public; white), and bottled–water (Bottled; purple) samples. Circles (●) are data for individual samples. For ${\sum }_{\text{TQ}}$ (left), solid red and dashed orange lines, respectively, indicate benchmark equivalent concentrations and risk–screening level (TQ = 0.1) below which little to no risk expected. For ${\sum }_{\text{EAR}}$ (right), solid red and dashed orange lines indicate concentrations shown to modulate effects in vitro and effects–screening–level (EAR = 0.001), respectively. Boxes, centerlines, and whiskers indicate interquartile range, median, and 5th and 95th percentiles, respectively. For each plot, number at the top indicates the permuted probability that the centroids and dispersions are the same across all three groups (PERMANOVA; 9999 permutations). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)