Introduction
Per- and polyfluoroalkyl substances (PFAS) are a class of nearly 15,000 synthetic chemicals that are exclusively anthropogenic and not naturally occurring. They exhibit high thermal and chemical stability because their hydrogen atoms are partially or completely replaced by fluorine atoms, making them resistant to biological and abiotic degradation (Brunn et al., 2023; Maruzzo et al., 2025). First manufactured in the late 1940s, PFAS are now found in food packaging, cookware, cosmetics, and firefighting foams (Brunn et al., 2023; International Agency for Research on Cancer, 2023, 2025). They have been ubiquitously detected in the environment and are particularly common in drinking-water supplies near industrial sites that extensively use or produce them (IARC, 2023). A model by Tokranov et al. (2024) estimates that over 71 million people in the United States rely on PFAS-contaminated groundwater. This body of literature indicates that PFAS exposure is no longer an isolated hazard but a nationwide feature of the U.S. drinking-water supply.
The scientific evidence on PFAS-associated health risks has also recently reached a critical juncture. In 2023, the IARC classified the two most widely studied PFAS compounds, perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS), as a Group 1 human carcinogen and a Group 2B possible human carcinogen, respectively (IARC, 2023, 2025). This classification places PFOA in the same carcinogenicity tier as asbestos and benzene (IARC, 2023, 2025).
In response, the United States Environmental Protection Agency (EPA) established the first-ever legally enforceable national Maximum Contaminant Levels (MCLs) for six PFAS compounds in April 2024 under the National Primary Drinking Water Regulation (NPDWR) (Schnettler et al., 2024; U.S. Environmental Protection Agency, 2024, 2026a). This regulatory trajectory was, however, subject to reversal. In 2025, the EPA began rolling back both the NPDWR’s compound coverage and the Justice40 Initiative that had directed remediation resources toward the most affected populations, a reversal detailed further below (2025; U.S. EPA, 2026a; U.S. Government Accountability Office, 2025).
These regulatory shifts occur alongside a growing, but disjointed, body of evidence on the disproportionate distribution of PFAS exposure. Studies have found that PFAS and other unregulated drinking water contaminants are unevenly concentrated in public water systems serving counties with predominantly Hispanic, Black, and low-income residents (Maruzzo et al., 2025). However, no existing synthesis has connected these exposure disparities to the 2025-2026 regulatory changes, nor predicted their equity implications.
This review bridges that gap by addressing a core question: how do the regulatory rollbacks of 2025-2026 interact with structural determinants of PFAS exposure disparity, and what do competing explanatory frameworks reveal about the adequacy of monitoring and remediation infrastructure? To answer this question, the review is organized into four sections: (1) peer-reviewed evidence on the health effects of PFAS; (2) structural determinants of PFAS exposure disparities by race and socioeconomic status; (3) the federal regulatory landscape and its 2025-2026 changes; and (4) directions for future equitable drinking-water policies.
Methods
A structured narrative literature review was conducted to synthesize peer-reviewed evidence and federal regulatory documents focused on PFAS exposure disparities and drinking water policy in the United States. Literature was identified through targeted searches of PubMed and Google Scholar using terms including “PFAS,” “per- and polyfluoroalkyl substances,” “drinking water,” “environmental justice,” “exposure disparities,” and “maximum contaminant levels”, supplemented by citation tracing. Sources were limited to English-language publications addressing U.S. populations, prioritizing recent nationwide or multi-state analyses. Regulatory literature was retrieved directly from agency websites, including the Government Accountability Office (GAO), the Agency for Toxic Substances and Disease Registry (ATSDR), the IARC, and the National Academies. Building on this evidence base, the review applies a comparative analysis between two regulatory states: the 2024 regime, defined by the PFAS NPDWR and the Justice40 Initiative, and the 2025-2026 regime, marked by the NPDWR’s proposed partial rescission, Justice40’s elimination, and proposed exemptions to the Toxic Substances Control Act (TSCA) reporting requirements. This isolates policy variables that interact with established sociodemographic exposure disparities, yielding an original analytical typology of three regulatory gaps: the science-policy gap, the equity monitoring gap, and the compliance equity gap.
Results
PFAS Health Effects
Understanding the health effects associated with PFAS exposure provides the scientific foundation for evaluating whether current regulatory decisions adequately protect vulnerable populations. As previously indicated, IARC classified PFOA as a Group 1 carcinogen in November 2023. This rested on “sufficient” evidence for cancer in experimental animals, “strong” mechanistic evidence in exposed humans, and “limited” but consistent evidence for renal cell carcinoma and testicular cancer in human populations (IARC, 2023, 2025). In a Good Laboratory Practice study, there was an increased incidence of benign and malignant neoplasms in both sexes of a single species, thereby meeting the threshold of “sufficiency” in experimental animal evidence (IARC, 2025). The mechanistic evidence was classified as “strong” in exposed humans because PFOA induced persistent epigenetic alterations and immunosuppression across multiple independent populations of children and adults; it also generated oxidative stress and modulated nuclear receptor pathways (IARC, 2025). PFOS received a Group 2B classification based on strong mechanistic evidence but only limited animal and inadequate human cancer evidence (IARC, 2025). This collectively demonstrates that the toxicological case against PFOA and PFOS had already reached regulatory maturity before the 2025-2026 rollback began.
The evidence base has since matured from laboratory to clinical guidance with the release of the National Academies of Sciences, Engineering, and Medicine (NASEM) 2022 review and the ATSDR’s 2024 clinical evaluation and management report (U.S. ATSDR, 2024; NASEM, 2022). Sufficient evidence was found of an association between PFAS exposure and decreased antibody response, dyslipidemia, and decreased infant and fetal growth (NASEM, 2022). However, the two reports diverge on screening thresholds: NASEM established cutoffs at 2 and 20 ng/mL, while ATSDR has not adopted these, leaving clinical practice without a unified standard (U.S. ATSDR, 2024; NASEM, 2022). This divergence foreshadows what the review later terms the science-policy gap, which is discussed further below.
Beyond carcinogenicity, PFAS exposure has immunotoxic effects with particular consequences for children. As mentioned previously, reduced antibody production in response to vaccination is one of the most consistently recorded health effects associated with PFAS exposure, specifically in children receiving tetanus and diphtheria vaccines (Bline et al., 2024). This is significant because immunotoxic exposures during a developmental stage produce more severe and long-lasting effects (Bline et al., 2024). This establishes a direct pathway that carries public health consequences: a reduction in individuals who can effectively respond to vaccines due to contaminated drinking water lowers the threshold required for community immunity, compounding infectious disease risk in certain populations (Bline et al., 2024).
Emerging evidence links PFAS exposure to adverse mental health outcomes, including anxiety, depression, and elevated markers of chronic physiological stress. These effects are hypothesized to result from PFAS-induced neuroinflammation and disruption of dopaminergic signaling via increased blood-brain barrier permeability (Sukhram et al., 2025). Mental health burdens caused by PFAS are not equally distributed: the same communities of color that face inordinate exposure also experience heightened psychological distress (Schnettler et al., 2024; Sukhram et al., 2025). This anxiety stems from a lack of control and eroded trust in government institutions, patterns linked to historical and institutional trauma (Sukhram et al., 2025; Vázquez et al., 2024). The combined effect of environmental and psychosocial stressors creates mental health outcomes that exceed what either would produce independently (Sukhram et al., 2025).
Structural Drivers of Exposure Inequity
The spatial disparity in PFAS exposure is corroborated by multiple peer-reviewed studies. Collectively, nationwide and multi-state analyses demonstrate that PFAS contamination is not randomly distributed but concentrated in historically industrialized and urban regions, including the Northeast, Mid-Atlantic, Rust Belt, and Gulf Coast (Garrett et al., 2025). Contamination is similarly concentrated in water systems serving high proportions of Black and Hispanic residents, with each one-percentage-point increase in either group associated with 3-6% higher odds of detection above state MCLs (Liddie et al., 2023).
These sociodemographic patterns are not incidental. Rather, historical segregation and discriminatory land-use practices have shaped the locations of industrial pollution sources (Schnettler et al., 2024). Marginalized groups have occupied the “path of least resistance” for polluting infrastructure, given their constrained decision-making power and limited access to environmental information. As a result, the largest PFAS-emitting sources—airports, fire-training areas, and manufacturing sites—are systematically concentrated in historically disadvantaged and racially segregated regions, aligning with the legacy of redlining and inequitable zoning (Liddie et al., 2023; Schnettler et al., 2024; Shkembi & Neitzel, 2025; Sukhram et al., 2025). Interestingly, one study found that poverty’s relationship with PFAS contamination was inconsistent, with a negative association nationally but a positive association in rural areas. By contrast, associations with race/ethnicity were consistently positive, independent of poverty (Liddie et al., 2023).
The same racial disparity extends beyond urban and industrial settings into agricultural contexts, where regions with larger Latinx populations exhibit significantly higher PFAS application density due to pesticide use, producing a 27% increase in exposure risk in areas where Latinx residents exceed 10% of the population (Libenson et al., 2024; Sukhram et al., 2025). Similarly, in New Jersey, 93% of people of color relied on water systems with detectable PFAS contamination, compared with 76% of non-Hispanic white residents (Mueller et al., 2024; Sukhram et al., 2025).
Unequal access to water infrastructure, monitoring capacity, and healthcare resources further compounds PFAS-related risks for marginalized communities. The Drinking Water Disparities Framework identifies the technical capacity of water treatment facilities as a structural determinant of exposure, independent of proximity to contamination sources (Liddie et al., 2023; Schnettler et al., 2024). Because state-level PFAS testing programs vary substantially in scope and rigor, documented contamination levels often reflect the quality of testing infrastructure rather than the actual contamination burden, masking exposure in under-monitored areas (Garrett et al., 2025). For example, the EPA’s toxic release inventory captured only 76 out of an estimated 1,582 textile facilities nationwide, demonstrating how contamination sources can be substantially undercounted (Garrett et al., 2025). Rural communities that rely on domestic wells and small water systems are subject to less rigorous federal oversight than larger urban systems, and they are frequently excluded from drinking water testing and reclamation efforts altogether (Garrett et al., 2025; Morales-Pate, 2026; Sukhram et al., 2025). The same inconsistencies apply to regulations addressing contamination on tribal lands (Sukhram et al., 2025). These regulatory and infrastructural gaps intersect with healthcare inequity: fenceline communities already carry disproportionate disease burdens from non-PFAS causes, and a historical legacy of medical mistreatment may stigmatize or discourage underserved populations from seeking care at health services, even when PFAS-related health effects emerge (Sukhram et al., 2025).
Federal Regulatory Changes and Environmental Justice
The Biden administration EPA promulgated the first-ever legally enforceable MCLs for PFAS in 2024, establishing limits of 4.0 parts per trillion (ppt) for both PFOA and PFOS, 10 ppt each for PFHxS, PFNA, and HFPO-DA, and a Hazard Index MCL of 1 for mixtures containing two or more of PFHxS, PFNA, HFPO-DA, and PFBS (U.S. EPA, 2025; Schnettler et al., 2024). Public water systems were required to complete monitoring by 2027 and achieve compliance by 2029 (U.S. EPA, 2024, 2026a). The EPA projected it would prevent approximately 100 million people from being exposed to PFAS, while averting thousands of deaths and tens of thousands of PFAS-attributable illnesses (Schnettler et al., 2024; U.S. EPA, 2024, 2026b). To offset the financial burden this would place on utilities, the EPA paired the rule with the Bipartisan Infrastructure Act, which allocated $9 billion for PFAS-related water quality testing, infrastructure, and treatment (Schnettler et al., 2024; U.S. EPA, 2023, 2024). It specifically invested $5 billion over five years to help small or disadvantaged communities that are most heavily impacted by contamination (Schnettler et al., 2024; U.S. EPA, 2023, 2024).
This progress was not permanent. In 2025, the EPA announced its intent to rescind MCLs for PFHxS, PFNA, HFPO-DA, and PFBS, and in May 2026, it published proposed rules that are currently open for public comment (Eagle & Firstenberg, 2026; Nathanson & Venus, 2025; U.S. EPA, 2024, 2025a). Citing Loper Bright Enterprises v. Raimondo, this was characterized as a valid procedural correction. The agency argued that the Safe Drinking Water Act (SDWA) required a final determination before rulemaking, and because Biden’s EPA finalized both simultaneously for those four compounds, the rule was therefore unlawful, exempting it from the Act’s anti-backsliding provision (Eagle & Firstenberg, 2026; Nathanson & Venus, 2025). Furthermore, the EPA extended the compliance deadlines for PFOA and PFOS by two years to 2031, citing water systems’ need for additional time to secure funding and install treatment infrastructure (Eagle & Firstenberg, 2026; Nathanson & Venus, 2025; U.S. EPA, 2024, 2025a, 2026a). Finally, the agency proposed exemptions to the TSCA’s PFAS data reporting requirements as well, including imported articles, de minimis concentrations below 0.1%, byproducts, impurities, non-isolated intermediates, and research and development activities, changes projected to reduce reporting obligations for an estimated 127,469 small article importers and create significant surveillance gaps (U.S. EPA, 2025b). Taken together, these revisions substantially narrow the scope of federal PFAS regulation.
The same year brought a second rollback, this time targeting the equity infrastructure meant to direct relief to the communities most affected by contamination. On January 20th, 2025, President Trump signed EO 14148 (“Initial Rescissions of Harmful Executive Orders and Actions”), which revoked former President Biden’s EO 14008 and EO 14096, thus terminating the Justice40 Initiative (EELP, 2025). Justice40 directed 40 percent of overall benefits derived from federal climate programs to “disadvantaged communities”, encompassing over 518 programs across 19 agencies by late 2024 (EELP, 2025; McCall, 2025). It allocated nearly $30 billion in funding for a large variety of services, from the cleanup of legacy pollution to workplace development to the provision of affordable and low-emissions transport (McCall, 2025). With the rescission of EO 14008, the Climate and Economic Justice Screening Tool (CEJST) was also removed from operation on federal government websites (EELP, 2026). The CEJST used data across eight categories—climate change, energy, health, housing, legacy pollution, transportation, water and wastewater, and workforce development—to identify “disadvantaged communities” and help achieve the goals set by Justice40 (Council on Environmental Quality, n.d.; EELP, 2026).
Discussion
A Three-Gap Analytical Framework
The science-policy gap refers to the systematic failure to translate robust health evidence into commensurate regulatory action. In the context of PFAS contamination via drinking water, this gap is caused by regulatory failure rather than scientific uncertainty, and manifests at both the clinical and regulatory levels. Clinically, the divergence between NASEM’s screening thresholds and ATSDR’s refusal to adopt them leaves physicians in contaminated communities without a unified standard for when to test or refer patients. At the regulatory level, no formal mechanism in U.S. law requires MCL review following an updated IARC classification (Samet et al., 2020). PFOA’s 2023 designation as a Group 1 carcinogen carried no legal weight under the SDWA (IARC, 2023; Samet et al., 2020). The 2025 rescissions deepen this gap, eliminating mandatory monitoring and public reporting for four compounds with the documented negative health effects discussed above. By framing the decision as a corrective to procedural non-compliance, the EPA avoids engagement with scientific evidence and with questions about the racial and socioeconomic composition of exposed populations, exemplifying the displacement of science by administrative process in regulatory decision-making.
The equity monitoring gap is the failure of existing surveillance systems to generate data stratified by race, ethnicity, and income that would enable documentation of disparate PFAS exposure. This gap originates from the monitoring frameworks established under the Unregulated Contaminant Monitoring Rule (UCMR), which were designed to characterize aggregate contamination patterns rather than identify equity dimensions, and are not consistently cross-referenced with the community demographic data (Liddie et al., 2023). Now, the removal of Justice40’s CEJST eliminates the federal infrastructure that had just begun to address this issue (EELP, 2025; U.S. GAO, 2025). Equity-stratified monitoring becomes practically impossible at the federal level without it.
The compliance equity gap describes the differential capacities of water utility systems to achieve and maintain MCL compliance due to their varying financial and technical resources. Compliance requires a costly process: many small systems lack grant writers or licensed engineers, and the administrative complexity of federal funding programs only adds to this access barrier (Morales-Pate, 2026). Without Justice40, funding allocation will likely revert to administrative patterns that historically favored systems with greater grant-writing capacity; this leaves communities most exposed to contamination without a federal mechanism to ensure relief reaches them (EELP, 2025; U.S. GAO, 2025).
Table 1 summarizes how the 2024 regulatory baseline and the 2025–2026 rollback map onto these three gaps, as well as the research gap each dimension leaves unresolved.
The primary contribution of this review is that the original framework established is not limited to PFAS. The science-policy gap, equity monitoring gap, and compliance equity gap function as a diagnostic model that other researchers and advocates could apply to varying rollbacks of unregulated contaminant standards.
This analysis also investigates a point of tension in the literature base: whether socioeconomic status or race is the primary driver of exposure disparities. Because Liddie et al. (2023) found race/ethnicity associations with PFAS contamination to be consistently positive independent of poverty, these findings raise questions about whether policy responses built primarily around income-based targeting would adequately serve communities where race, rather than income, is the stronger predictor of exposure.
Why Regulatory Rollback Disproportionately Burdens Vulnerable Communities
The compliance equity gap, detailed above, is not unique to PFAS regulation. Evidence from other environmental rollbacks and rulemaking contexts shows the same pattern recurring through two mechanisms that compound the disparities already discussed. First, populations with fewer civic resources are less able to generate the public comments or mount the legal challenges that often determine whether a proposed rollback proceeds to a final rule (Eagle & Firstenberg, 2026; Freudenberg et al., 2011; Nathanson & Venus, 2025). Second, analysis of the 2001 Revised Arsenic Rule found that community water systems with higher home-ownership rates—a proxy for community wealth—had significantly lower contaminant levels, posing another example of how the communities least able to absorb compliance costs face the greatest contamination risk (Balazs et al., 2012). This asymmetry likely holds in reverse: a system with more resources can keep monitoring voluntarily even after the federal requirement disappears, while one without them cannot.
Another precedent is the Flint water crisis, which shows how not only contamination, but institutional response, can be unevenly distributed. Block-level lead exposure was most consistently predicted by the spatial clustering of contamination and the concentration of single-father Black families, and official health advisories were issued only in English for nearly two years, leaving Flint’s Spanish-speaking residents without the information needed to protect their children (Liévanos et al., 2021). The cleanup of the Superfund sites offers a lesson about durability. After Executive Order 12898 made environmental justice federal policy in 1994, cleanup delays in predominantly Black, urban neighborhoods appear to have shortened noticeably. But this protection was uneven: more recent research found that communities with larger Asian populations are still significantly less likely to have their sites cleaned up, a gap earlier studies missed because they never analyzed that population separately (Burda & Harding, 2014; Topaz et al., 2024). These findings show how dedicated infrastructure can measurably narrow disparities, but only where monitoring is disaggregated finely enough to detect them.
Broader Implications and Directions for Future Research and Policy
Beyond substantive gaps, the current rollback is further enabled by a broader shift in administrative law. Specifically, the 2024 overturn of the Chevron doctrine has opened the same door for agencies more broadly to walk back environmental rules once protected by judicial deference (Sabharwal, 2025). Taken with the precedents above, these dynamics point toward one overarching lesson: equity-oriented protections are most durable when embedded in statute rather than executive action, since statutory protections require affirmative congressional repeal rather than unilateral rescission. These cases show that future rulemaking should invest in multilingual, community-embedded risk communication and should sequence health-effects determinations and enforceable standards as clearly separable actions to reduce vulnerability to procedural challenge.
We identify three priority areas for future research and policy. First, equity-stratified monitoring should become a regulatory default rather than an exception initiated by independent research. Second, future frameworks should require all PFAS monitoring data collected under the SDWA to be cross-referenced with community demographic data from the outset, building on the methodology of Liddie et al. (2023) but embedded in regulatory design. Third, cumulative impact assessment should be formally integrated into PFAS MCL decision-making, paired with statutory remediation-funding requirements resilient to administrative turnover and a cost-benefit framework weighing compliance costs against the health burden of continued exposure, since both fall on the same communities (EELP, 2025; McCall, 2025; U.S. GAO, 2025).
Limitations
Several limitations apply. First, this review relies on monitoring data subject to the same equity gaps described here, meaning the true scope of exposure is likely underestimated. Second, studies are cross-sectional and cannot establish direct causality between regulatory changes and health outcomes. Third, the 2025-2026 changes are too recent for outcome data to exist. This means that the framework’s predictions cannot yet be empirically validated.
Conclusion
This mini review connects literature on PFAS exposure disparities in drinking water to specific 2025-2026 regulatory changes, an intersection no prior synthesis has examined through an environmental justice lens. The resulting three-gap model explains how the 2025-2026 rollbacks compound pre-existing PFAS exposure disparities and offers a transferable diagnostic tool for evaluating future regulatory retrenchment in other areas of environmental health (Garrett et al., 2025; Liddie et al., 2023; Nathanson & Venus, 2025). Ultimately, this review demonstrates that the environmental justice challenges associated with PFAS are no longer solely scientific questions but governance questions. Unless future oversight integrates equity into monitoring, funding, and implementation from the outset, existing disparities are likely to widen, regardless of how far toxicological knowledge advances.