Community Water Trihalomethanes and Chronic Kidney Disease

Abstract

Importance: Over 90% of the US population relies on community water supplies (CWS), which generally use chlorine for disinfection. Trihalomethanes are regulated disinfection byproducts associated with bladder cancer and adverse birth outcomes. Animal studies report trihalomethanes, especially brominated compounds, may damage kidney function, but epidemiologic research is limited.

Objective: To evaluate long-term exposure to trihalomethanes in residential CWS and its association with chronic kidney disease (CKD) risk.

Design, setting, and participants: The California Teachers Study (CTS) is an ongoing prospective cohort of female teachers and administrators enrolled between 1995 and 1996 with data linked to mortality and health care records. This cohort study analyzed CTS data from January 1, 2005, once CKD diagnostic coding was adopted, through December 31, 2018. Statistical analysis was conducted from July 2023 to December 2024.

Exposures: Residence time-weighted mean concentrations of 4 trihalomethanes, including 3 brominated trihalomethanes and chloroform, were calculated using annual measurements from CWS serving participants' homes from 1995 to 2005. Uranium and arsenic (potentially nephrotoxic metals, previously evaluated in the cohort) from CWS were included as part of a g-computation mixture analysis.

Main outcomes and measures: Cases of moderate (stage 3) to end-stage CKD were identified with diagnostic codes or dialysis-related procedures. Mixed-effects multivariable-adjusted Cox proportional hazards regression models were used to estimate hazard ratios (HRs) and 95% CIs for CKD by exposure to trihalomethane levels (US maximum contaminant level of 80 μg/L).

Results: The study sample included 89 320 female participants (median age, 50 years [IQR, 43-61 years]) with 6242 CKD cases. Median concentrations were 5.5 μg/L (IQR, 0.5-24.1 μg/L; 95th percentile, 57.8 μg/L) for total trihalomethanes and 2.7 μg/L (IQR, 0.7-11.3 μg/L; 95th percentile, 30.0 μg/L) for brominated trihalomethanes. In flexible spline-based models, a clear exposure-response association was observed between trihalomethanes and CKD risk, with the highest risk for brominated trihalomethanes. The HRs for CKD risk associated with brominated trihalomethanes at the highest 2 exposure categories (75th percentile and at or above the 95th percentile) were 1.23 (95% CI, 1.13-1.33) and 1.43 (95% CI, 1.23-1.66), respectively (P < .001 for trend). Brominated trihalomethanes were the largest contributor (52.9%) to the association of the overall mixture with CKD risk, followed by uranium (35.4%), arsenic (6.2%), and chloroform (5.5%).

Conclusions and relevance: In this prospective cohort study of California female teachers, exposure to trihalomethane concentrations less than 80 μg/L (US current standard) increased CKD risk, particularly brominated trihalomethanes, which are not separately regulated in community water. The findings may have public health implications given the widespread use of water chlorination and growing burden of CKD.

Figure 1.. Map of California Community Water Supplies and Average Total Trihalomethane Concentrations for the Exposure Period (1995-2005)

Includes 2933 community water supplies. Gray shading indicates supplies that had missing data or were not linked to California Teacher Study participants. The 10 census regions of California are outlined and labeled. ^aTwo community water supplies had concentrations of 80 μg/L or higher, the US Environmental Protection Agency maximum contaminant level.

Figure 2.. Exposure-Response Associations Between Trihalomethane (THM) Concentrations in Residential Community Water Supplies and Moderate to End-Stage Chronic Kidney Disease (1995-2005)

Solid lines indicate hazard ratios (HRs) for chronic kidney disease and shading, 95% CIs, modeled as restricted cubic splines in the California Teachers Study (N = 89 320; analytic follow-up, 2005-2018). HRs were adjusted for body mass index as a continuous variable, smoking status, race and ethnicity, neighborhood socioeconomic status quartiles, US Census region as a random effect, and age as the time scale. Concentrations were log-transformed with knots placed at the 25th, 75th, and 95th percentiles. Vertical dashed lines indicate the 95th percentiles and the lowest concentration in each graph indicates the 25th percentile. A histogram of the population density is shown for each spline. To better visualize trends at the lower tails, plots with x-axes on a log scale are provided in eFigure 3 in Supplement 1.

Figure 3.. Relative Contributions of Water Contaminants to Chronic Kidney Disease (CKD) Risk

Mixture analysis used g-computation to estimate the relative contributions of mean brominated trihalomethane (THM), chloroform (nonbrominated), uranium, and arsenic concentrations from residential community water supplies (1995-2005) to the overall association with moderate to end-stage chronic kidney disease (CKD) risk in the California Teachers Study (analytic follow-up, 2005-2018). Analysis included 88 169 participants with complete uranium and arsenic data (all but 1151 of the total 89 320 participants). The overall outcome (hazard ratio and 95% CI) was the risk of CKD associated with a joint IQR increase of all water contaminants. Findings were adjusted for body mass index as a continuous variable, smoking status, race and ethnicity, neighborhood socioeconomic status quartiles, census region as a stratification term, and age as the time scale.