Association of Community Water Lead Levels and Erythropoietin Stimulating Agent Use among End-Stage Kidney Disease Patients

Introduction Although U.S. Environmental Protection Agency (U.S. EPA) regulations mandate action when the 90th percentile of tested community water samples exceeds 15 lg=L lead, the great majority of water systems in the United States fail to reach the U.S. EPA’s maximum contaminant goal of zero (https://www.epa. gov/laws-regulations/summary-safe-drinking-water-act). Patients with chronic kidney disease (CKD) have heightened susceptibility to heavy-metal exposure due to a combination of increased proportions absorbed across the gastrointestinal tract and reduced urinary excretion.1,2 We have recently demonstrated that lead levels found widely in United States water systems associate with lower hemoglobin concentrations among individuals with advanced CKD.3 Herein, focusing on the End-Stage Kidney Disease (ESKD) population, we describe whether exposure to drinking water lead is associated with higher use of medications used to treat anemia during the first 12 months of maintenance dialysis therapy.


Introduction
Although U.S. Environmental Protection Agency (U.S. EPA) regulations mandate action when the 90th percentile of tested community water samples exceeds 15 lg=L lead, the great majority of water systems in the United States fail to reach the U.S. EPA's maximum contaminant goal of zero (https://www.epa. gov/laws-regulations/summary-safe-drinking-water-act). Patients with chronic kidney disease (CKD) have heightened susceptibility to heavy-metal exposure due to a combination of increased proportions absorbed across the gastrointestinal tract and reduced urinary excretion. 1,2 We have recently demonstrated that lead levels found widely in United States water systems associate with lower hemoglobin concentrations among individuals with advanced CKD. 3 Herein, focusing on the End-Stage Kidney Disease (ESKD) population, we describe whether exposure to drinking water lead is associated with higher use of medications used to treat anemia during the first 12 months of maintenance dialysis therapy.

Methods
Using the U.S. Renal Data System (U.S. RDS), a national data system cataloging patients with incident ESKD, we identified all patients who began dialysis between 1 January 2011 and 31 December 2017. We identified the primary city of residence, defined as a steady city residence spanning 6 months or more prior to and 15 months after dialysis initiation. We used "Crown-Web," a data-management system from the Centers for Medicare & Medicaid Services (CMS) available from 2011 that allows Medicare-certified dialysis facilities to communicate prevalent ESKD data, to characterize monthly Erythropoietin Stimulating Agent (ESA) use during the first 12 months of maintenance dialysis therapy [Epoetin alfa (n = 1,022,349 months), beta (n = 1,506 months) and Darbopoetin (n = 165,317 months)] among a primary cohort of 191,693 ESKD patients with 1,619,204 months of follow up data. Data from the first 90 d after dialysis initiation was excluded given the heightened clinical instability during this time period. Package insert dosing guidelines were used to calculate an ESA equivalent for Darbopoetin users.
We used the standard 2728 form, used to certify ESKD status and typically completed by the nephrologist of record, to characterize pre-ESKD patient characteristics, including age; gender; race (White, Black, Asian, other/unknown); diagnosis of diabetes mellitus, congestive heart failure, or hypertension; current tobacco use; ESKD geographic network; access type (arteriovenous fistula, arteriovenous graft, dialysis catheter, unknown); and calendar year of dialysis initiation. Given nonuniformity in data collection and adjudication, baseline characteristics are vulnerable to misclassification. Monthly measures of dialysis adequacy (kt/v) and nutrition (albumin) were included. Missing measures of adequacy (n = 309,647 months) and albumin (n = 122,279 months) were replaced with the average patient value over the study period, or the cohort mean for those missing all measurements. The average number of missing monthly measures per patient was 2.2, 2.3, 2.3, 2.1, and 2.0 across increasing categories of city water lead, respectively. We also included measures of median household income and unemployment rate, ascertained at the patient's residential ZIP code level, using data from the United States Census American Community Survey 5-y estimates (2011-15).
We used 90th percentile community water lead levels, as reported in the Safe Drinking Water Information System (SDWIS) Federal Data Warehouse (https://www.epa.gov/ enviro/sdwis-search), to characterize drinking water lead exposure. We averaged the maximum reported lead level per year of all water systems serving a city, weighted for the number of individuals served by each water system, to derive an annual citywide lead level. For each patient, we calculated the average citywide level during the 5 y preceding dialysis initiation. Research was deemed exempt by the institutional review board (IRB) of Beth Israel Deaconess Medical Center (IRB Protocol No. 2020P000200).
We used robust Poisson regression in generalized estimating equations (GEE) with a log link to examine the adjusted association of citywide lead levels with monthly mean ESA dosing. Those categorized as not receiving monthly ESA were categorized as zero units of exposure. In addition, we examined ESA resistance, defined as a weekly Epoetin dose >300 units=kg or Darbepoetin dose >1:5 mcg=kg, according to best practice guidelines. 4 For the 321,154 months missing documentation of predialysis weights, the average weights over the study period were used. To examine whether the effects of lead exposure were higher among patients with concomitant iron deficiency, among whom gastrointestinal absorption of ingested lead is thought to be more avid, we explored multiplicative interaction terms between iron deficiency (transferrin percent saturation <20%) and lead levels, defined continuously.

Results
Individuals with ESKD (n = 157,183, 82%) ( The authors have no competing financial interests. Note to readers with disabilities: EHP strives to ensure that all journal content is accessible to all readers. However, some figures and Supplemental Material published in EHP articles may not conform to 508 standards due to the complexity of the information being presented. If you need assistance accessing journal content, please contact ehpsubmissions@niehs.nih.gov. Our staff will work with you to assess and meet your accessibility needs within 3 working days. confidence interval (CI): 1,012, 1,865] units higher monthly ESA dosing and 1.04 (95% CI: 1.00, 1.07) higher risk of ESA resistance (Table 2). These associations were observed at exposure levels below the threshold currently allowed by the U.S. EPA and were particularly apparent when associated with iron deficiency (multiplicative interaction p-values between lead and transferrin saturation <20% were <0:001 and 0.01 for monthly ESA dose and risk of ESA resistance, respectively).

Discussion
Levels of lead found widely in U.S. water systems and well below the current U.S. EPA threshold that mandates regulatory action are associated with higher monthly ESA dose and risk of ESA resistance during the first year of maintenance dialysis therapy, particularly when accompanied by iron deficiency.
Further extending our previous work highlighting the myelosuppressive effects of lead, our findings underscore the heightened susceptibility to environmental toxicants for patients with CKD. Although epidemiological studies have suggested that lead toxicity may cause renal damage, 5,6 impaired renal function may simultaneously be a risk factor for lead toxicity. That levels of lead found widely in drinking water associate with significantly higher ESA use among maintenance dialysis patients suggest an accumulative hazard from repeated exposure. In addition to the financial burden and known cardiovascular risks of higher ESA use, two questions that remain to be studied are whether lead exposure meaningfully contributes to other diseases highly prevalent in ESKD (including depression, dementia, and other neurocognitive disorders 7,8 ) and whether disparate lead exposure among minoritized communities might explain their greater prevalence of anemia, iron deficiency, and ESA resistance. 9,10 The possibility that inexpensive and readily available water filters may mitigate toxicity from environmental hazards for those with CKD adds urgency to these public health questions. Given the inherent limitations of our analysis, further welldesigned studies with individual levels of heavy-metal exposure and tissue accumulation are warranted.

Acknowledgments
This study is unfunded. The data reported here have been supplied by the U.S. RDS. The interpretation and reporting of these data are the responsibility of the author(s) and in no way should be seen as an official policy or interpretation of the U.S. government.