PFAS Policies and Reductions in Drinking Water Contamination Across the United States: Incorporated Policy Perspectives from Los Angeles Residents

Abstract

Background

Many U.S. states have adopted per- and polyfluoroalkyl substances (PFAS)-related policies, but evidence on whether these actions reduce detections in drinking water is limited.

Methods

We conducted a nationwide analysis of the EPA Unregulated Contaminant Monitoring Rule (UCMR) and a community study in Los Angeles (LA). State-wide PFAS policies correlated with PFAS detections in public water systems (n = 56,856) using UCMR3 (2013–2015) and UCMR5 (2023–2025) for six PFAS (PFBS, PFHpA, PFHxS, PFNA, PFOA, PFOS). We examined associations between detections or changes in detections and two metrics: (1) total PFAS policy counts and (2) the proportion of water-related PFAS policies. In LA, we conducted focus groups/interviews (n = 29) and tested a 55-PFAS panel in household tap water (n = 23; Pasadena/Torrance n = 9; Central/Southeast LA n = 14).

Results

Total policy counts were not associated with lower PFAS detection in UCMR5, except for a borderline reduction in PFBS (β = −0.0003, p = 0.06). A higher proportion of water-focused policies was associated with increased odds of PFAS detection, with significant associations for PFBS (OR = 3.79, 95% CI: 2.83–5.10), PFHpA (2.66, 1.64–4.33), PFHxS (3.10, 2.10–4.62), and PFOS (3.37, 2.41–4.71). Similar results were observed for changes in detection. In LA, residents reported distrust in tap water, low PFAS awareness, and emphasized independent testing, subsidized filtration, and clearer communication.

Conclusions

Greater concentrations of water-focused policies were linked to increased PFAS detections, likely reflecting enhanced surveillance or responses in high-burden states. Community perspectives highlight that policy also requires trust, transparency, and resources.

Graphical abstract

1. Introduction

Access to safe drinking water is a basic human right and a major goal of the California Human Right to Water bill (A.B. 685).(1) It was estimated that 2.2 million people lack access to running water, and over 46 million people do not have safe drinking water worldwide.(2) Under the United Nations Sustainable Development Goals, Target 6.1 aims to “achieve universal and equitable access to safe and affordable drinking water for all” by 2030, tracking progress via the proportion of the population using safely managed drinking water services.(3) Although 9 out of 10 people in the United States obtain their tap water from a public water system,(4) aging infrastructure and chemical contamination continue to threaten safe access.(2) The Safe Drinking Water Act (SDWA) authorizes the U.S. Environmental Protection Agency (EPA) to set and enforce health-based standards for contaminants in public water systems, requiring routine monitoring, reporting, and actions to ensure safety. The Clean Water Act (CWA) governs discharges of pollutants into waters by requiring facilities to obtain permits and adhere to effluent limits and monitoring schedules. Yet, regulation of emerging contaminants often lags behind, allowing harm to occur before protective measures are in place. PFAS exemplify this regulatory gap, as exposures became widespread decades before enforceable limits were established.

Per- and polyfluoroalkyl substances (PFAS) are a large class of synthetic chemicals that have been widely used since the 1940s in products such as nonstick cookware, waterproof textiles, firefighting foams, and food packaging.(5) PFAS are highly persistent in the environment and in humans.(6–8) In April 2024, the EPA finalized its first National Primary Drinking Water Regulation for six PFAS chemicals (subsequently narrowed to two), requiring public water systems to comply by 2029.(9) Additionally, perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS) have been designated as hazardous substances under the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), enabling EPA to compel cleanup and pursue responsible parties at contaminated sites.(10) At the same time, many U.S. states have implemented their own PFAS-related legislation over the past decade, ranging from drinking water standards, bans on consumer products, industrial discharge limits, to requirements for water system monitoring. These state-level policies are designed both to reduce upstream emissions and to expand surveillance of PFAS in public water systems. However, no studies have assessed whether these policies are associated with changes in PFAS detection at the drinking water system level.

Los Angeles (LA), as the nation’s second-largest metropolitan area, presents a uniquely complex and instructive context for examining drinking water perceptions and PFAS exposure. LA’s water supply is drawn from multiple sources, including the Owens Valley, Sacramento-San Joaquin Delta, Colorado River, and numerous interconnected groundwater basins.(11) Each of these sources has distinct chemical contaminant profiles and treatment histories, resulting in uneven contaminant burdens across neighborhoods. Reliance on contaminated groundwater, in particular, contributes to PFAS detections in certain areas. For example, Southeast LA is heavily burdened with industry and environmental hazards, leading to disproportionate exposure to a wide range of water contaminants, including PFAS.(12) Understanding how residents in such contexts perceive water safety and respond to PFAS concerns provides important localized context that may be coupled with other communities’ perspectives to inform national and state-level policy and translation into meaningful local action.

Although widespread PFAS contamination is well documented in the scientific literature, there remains a notable lack of qualitative research examining how community members perceive the quality and safety of their drinking water, the pathways through which they obtain information, and the types of policy interventions they find most acceptable. Community-engaged approaches, such as those pioneered in North Carolina (13) and Louisville,(14) demonstrate the value of participatory testing and dissemination. To build on this work, the present study had two aims. First, we analyzed national monitoring data from the EPA’s Unregulated Contaminant Monitoring Rule (UCMR3 and UCMR5) to evaluate whether state-level PFAS policy activity is associated with changes in PFAS detection at the drinking water system level across the US. Second, we conducted semi-structured focus groups and interviews with residents of Torrance, Pasadena, Southeast LA, and Central LA, coupled with PFAS testing in tap water, to identify how perceptions of water quality are formed, what media and social networks inform opinion, and what community-driven policy recommendations emerge. These approaches link national policy with local community experiences, providing a more comprehensive understanding of potential solutions to address PFAS contamination in drinking water.

2. Method

National Analysis of State-wide PFAS Policy Activity and Detection Trends

2.1. Policy Data

PFAS policy data from 2016 to 2023 were abstracted from a national database of state legislation and coded by issue area from The Nelson A. Rockefeller Institute of Government.(15) Policy actions were categorized by issue area including agriculture, air quality or emissions, animals, appropriations, biosolids, children’s products, composting, consumer products, disposal or incineration, drinking water, environmental justice, federal government (or federal site), firefighting foam or protective equipment, food, food packaging, groundwater, hazardous substance designation, hazardous waste, health monitoring or testing, industrial products or production, insurance, landfills, liability, monitoring & detection, oil & gas, packaging, pesticides, plastics, private wells, recycling, remediation, soil, solid waste, statute of limitations, study or recommendation(s), surface water, task force, and wastewater. All policy actions were counted equally in the policy totals. We did not weight policies by scope or enforceability, which is a limitation in our current method. “Water-focused policy actions” were defined as those falling within water, drinking water, groundwater, surface water, wastewater, stormwater, or PFAS in water. Full descriptions of the policy data are provided in Supplementary Table S1. For each state, we calculated two measures: (1) the total number of PFAS-related policies introduced or enacted from 2016 through 2023 (“all policy actions”), and (2) the subset specifically focused on water systems (“water-focused policy actions”). Policies were linked to water systems by state.

2.2. UCMR Detection Data

We analyzed data from the U.S. Environmental Protection Agency’s Unregulated Contaminant Monitoring Rule (UCMR), focusing on UCMR3 (2013–2015) and UCMR5 (2023–2025).(16,17) UCMR is a Safe Drinking Water Act monitoring program that requires specified public water systems to sample for selected unregulated contaminants to support national occurrence estimates. UCMR results are reported at the public water system level based on compliance monitoring samples collected at designated locations (e.g., entry points to the distribution system), rather than household taps. To ensure comparability across cycles, the analysis was restricted to six PFAS measured in both: perfluorobutanesulfonic acid (PFBS), perfluoroheptanoic acid (PFHpA), perfluorohexane sulfonic acid (PFHxS), perfluorononanoic acid (PFNA), perfluorooctanoic acid (PFOA), and perfluorooctanesulfonic acid (PFOS). For each analyte and public water system (PWSID), we calculated two variables: (1) a binary indicator of whether any sample exceeded the minimum reporting level (MRL), and (2) the proportion of samples exceeding the MRL. The EPA lowered reporting thresholds substantially between UCMR3 and UCMR5 (for example, PFOA and PFOS from 20 ng/L to ~4 ng/L, PFBS from 90 ng/L to ~3 ng/L, and PFHxS from 30 ng/L to ~3 ng/L). Because analytical methods and minimum reporting levels differed across UCMR cycles, cross-cycle changes in detection may reflect both true changes in occurrence and improved detectability. Accordingly, we interpreted UCMR3-to-UCMR5 “change” outcomes cautiously and prioritized detection-based comparability over cross-cycle concentration comparisons.

We did not use cross-cycle concentration-based outcomes (e.g., mean concentration or exceedance) as primary endpoints because differences in reporting limits and left-censoring across cycles complicate direct concentration comparisons without additional harmonization assumptions. Concentration-based outcomes (e.g., mean concentrations) remained important for future work, particularly within a single monitoring cycle with consistent reporting limits.

2.3. Covariates

To reduce potential confounding, we adjusted for both system-level and state-level factors. System-level covariates included population served. State-level covariates included socioeconomic variables from the 2022 American Community Survey: median household income, poverty rate, educational attainment (proportion of adults with at least a bachelor’s degree), and population size.

2.4. Statistical Analysis for Policy Analysis

For each PFAS, we conducted two sets of models at the water system levels. First, logistic regression models estimated the associations between PFAS policy and any detection in UCMR5 while adjusting for detection in UCMR3. Second, linear regression models estimated the associations between PFAS policy and the change in detection rate between UCMR3 and UCMR5. Two sets of exposure variables were included: 1) the total number of PFAS policies and 2) total number of PFAS policies and proportion of water-related PFAS policies, and cluster-robust standard errors were calculated at the state level to account for correlation among water systems within the same state. All covariates mentioned above were included in all models.

We conducted sensitivity analyses by calculating implementation duration as the number of years since the first enacted water-focused PFAS policy in each state (as of 2023) and examining whether associations with UCMR3–UCMR5 changes in detection rates differed between short- and long-duration implementation.

Furthermore, we conducted lagged-exposure sensitivity analyses by recalculating policy counts and water-policy share excluding policies enacted in 2022–2023 and re-estimating associations with UCMR5 detection and UCMR3–UCMR5 changes.

To better explain what contributes to the associations between PFAS policies and detection of PFAS, we further tested whether the number of PFAS tests increased from UCMR3 to UCMR5, whether PFAS policies were associated with the number of PFAS tests, and whether the number of PFAS tests was associated with higher PFAS detections. In addition, we also tested whether associations between PFAS policies and PFAS detections would be attenuated by additionally controlling for the number of PFAS tests. Furthermore, we conducted sensitivity analyses testing whether baseline state-level PFAS burden in UCMR3 was associated with PFAS policy later.

Community-Engaged Focus Groups and PFAS Testing in Los Angeles

2.5. PFAS Testing in Tap Water

Participants provided household tap-water samples for PFAS testing (n = 23 total; Pasadena/Torrance n = 9; Central/Southeast LA n = 14). All tap-water samples were collected using the Cyclopure Water Test Kit Pro, which comprises a pre-conditioned collection cup. Two hundred and fifty mL of tap water was poured directly into the collection cup and allowed to drain completely through the DEXSORB^® disc by gravity over approximately 10 to 15 minutes, trapping PFAS molecules via host-guest complexation within the media’s uniform 0.78 nm hydrophobic cavities and eliminating the need to ship bulk water samples. Tap-water samples were collected as flushed grab samples from the household tap after running the cold water tap for 30 seconds. Collection procedures followed the study protocol and manufacturer instructions. Upon arrival at Cyclopure’s analytical laboratory, each spent extraction disc was transferred to a solid-phase extraction workflow: PFAS were quantitatively eluted with methanol (or 0.1% ammonium hydroxide in methanol) following standard EPA SPE protocols, and the eluates were analyzed by isotope-dilution HPLC-MS/MS in negative electrospray mode. Analyses adhered to EPA Methods 533, 537, and 1633 for internal standardization, calibration, and quality control. The reporting limits were 1 ng/L for 55 target PFAS chemicals (full analyte list provided in Table S2). Quality assurance/quality control procedures included instrument calibration and isotope-dilution internal standardization as described above. Field QA/QC samples were not collected, which represented a limitation of our community-based household sampling design.(18) To protect the privacy of participants, we only linked participants’ zip code information with the PFAS testing results for analysis.

Household tap-water sampling in Los Angeles was distinct from UCMR monitoring, which is conducted at public water system compliance sampling locations rather than at household taps. The LA community study was conducted as a complementary, context-generating component and was not used to parameterize or statistically power the national policy models.

2.6. Focus Group and Interview

Before each focus group or interview, participants were asked to provide basic demographic information, including age, gender, race/ethnicity, education, as well as perceptions of drinking water quality (See Supplemental Text for the questionnaires implemented). To protect the privacy of our participants, we did not link participants’ individual information to the focus group or interview, but were separated by two major neighborhoods (Torrance or Pasadena, and Southeast and Central LA, which included including Bell Gardens, Boyle Heights, East LA, Koreatown, University Park and Exposition Park, Glassell Park, Whittier, and Vermont). The water systems included Golden State Water Company (GSWC), Los Angeles Department of Water and Power, California Water Service Company - East Los Angeles, California Water Service Company – Dominguez, Whittier Water Department, and Pasadena Water and Power.

Neighborhood groupings were selected purposively to capture variation in community context and water perceptions and to leverage feasible community partnerships. Recruitment occurred through community meetings, individual meetings, and community health fairs, and sessions were conducted in English or Spanish. These recruitment channels may have influenced participant demographics and the perspectives captured. Therefore, qualitative findings were intended to provide contextual insight rather than population-representative estimates.

We did not ask questions specific to PFAS in the perception of drinking water quality questionnaire to avoid leading the participants to answer questions specific to PFAS during the focus group/interview.

We grouped participants from Torrance and Pasadena together due to the small sample size in the Pasadena group and similar demographic factors.

We conducted six semi-structured focus groups and seven interviews with residents across Los Angeles. The assignment of a participant to an interview or focus group was based on participants availability. We created a focus group/interview guide to ask questions around observations on the community-level concerns, personal experience, information and awareness, and policy recommendations for drinking water quality. It’s organized purely around these four practical domains and the questions were not based on particular theory. Interviews or focus groups were conducted in either English or Spanish (led by SL and RS). All focus groups/interviews for Torrance and Pasadena participants were conducted in English, while the others were in Spanish. Sessions were recorded and transcribed. The English and Spanish sessions were translated and transcribed by a bilingual speaker (led by TG).

Questionnaire and interview data were used to characterize perceptions and behaviors and to contextualize household testing. Individual survey responses were not linked to individual PFAS results to protect privacy, and analyses were conducted at the grouped neighborhood/ZIP-area level.

2.7. Community Engagement

We have engaged local residents and community organizations by presenting at community meetings, individual meetings, and attending community health fairs to raise awareness of PFAS contamination, the health effects of PFAS, and potential solutions to address PFAS contamination in LA.

To provide results back to participants, we returned both individual-level results through email or text and invited participants back for a community-level report back on community-level findings and information on potential sources of PFAS and prevention strategies.

2.8. Statistical Analysis for Focus Group

For questionnaire data, we compared the difference between the two groups using either Fisher’s exact test for categorical variables or Welch’s t-test for continuous variables.

For qualitative data analysis, we began with an inductive, line-by-line open-coding process, an approach derived from grounded theory,(19) to ensure that our codes emerged directly from participants’ own language. These initial codes were then organized through thematic analysis, grouping similar concepts into four higher-order themes based on the focus group/interview guide (Community Observations; Personal Experiences; Information & Awareness; Policy). We quantified the prevalence of codes and themes and examined word-use patterns to provide objective counts and visual summaries of the data.

Coding was conducted by two coders [SL and RS]. Coders independently coded an initial subset of transcripts to develop a shared codebook. Discrepancies were resolved through discussion and consensus, with periodic adjudication during full coding.

3. Results

3.1. Water System Characteristics

Shown in Table 1, among the 56,856 public water systems included in the analysis, the majority were classified as large systems, which serve over 10,000 people per system, (83.8%), with only 16.2% considered small, which serve 10,000 or fewer people. At baseline (UCMR3), PFAS detections were rare, with only 0.7% of systems showing any detectable levels and a mean detection rate of 0.002 (SD = 0.03). By the UCMR5 cycle, detections had increased markedly: 9.0% of systems reported PFAS detections, with a mean detection rate of 0.04 (SD = 0.17), showing an average increase of 0.04 from baseline.

Table 1.

Characteristics of public water systems included in PFAS policy analysis (N = 56,856)

Characteristic	Statistics	N = 56,856
State policy variables
Total PFAS policies	Mean (SD)	36.73 (46.24)
Water-focused PFAS policies	Mean (SD)	21.31 (33.29)
PFAS monitoring (UCMR3 baseline)
Detection rate (UCMR3)	Mean (SD)	0.002 (0.03)
PFAS monitoring (UCMR5 follow-up)
Detection rate (UCMR5)	Mean (SD)	0.04 (0.17)
Change in detection rate (UCMR5–UCMR3)	Mean (SD)	0.04 (0.17)
Socioeconomic indicators
Median household income (USD)	Mean (SD)	73,823 (12,768)
Poverty rate (%)	Mean (SD)	0.13 (0.04)
Bachelor’s degree or higher (%)	Mean (SD)	0.33 (0.05)
State population	Mean (SD)	12,570,183 (10,952,834)
Detection status
Any PFAS detection in UCMR3	N (%)	375 (0.7%)
Any PFAS detection in UCMR5	N (%)	5,132 (9.0%)
Water system size (UCMR3)
Large	N (%)	24,726 (83.8%)
Small	N (%)	4,794 (16.2%)

3.2. Policy Analysis

Shown in Figure 1 and Table S3, when examining the total number of PFAS-related policies proposed/enacted by a state, no consistent associations were observed with PFAS detection. We found mostly null findings across PFAS, with odds ratios approximating 1 (e.g., PFBS OR = 1.00, 95% CI: 0.99–1.00; PFOS OR = 1.00, 95% CI: 1.00–1.00). Similarly, results on the detection rate change showed small and largely non-significant effects for total policy counts except for a borderline statistically significant association between total policy counts and change in PFBS (β = −0.0003, p = 0.06).

Figure 1.

A) Odds of PFAS detection in UCMR5 associated with state PFAS policy actions, and B) Change in PFAS detection rate from UCMR3 to UCMR5 associated with state PFAS policy actions.

In contrast, the proportion of policies specifically focused on drinking water showed consistent and statistically significant associations with PFAS outcomes. Systems in states with a higher share of water-focused PFAS policies had higher odds of PFAS detection in UCMR5 for PFBS (OR = 3.79, 95% CI: 2.83–5.10, p < 0.001), PFHpA (OR = 2.66, 95% CI: 1.64–4.33, p = 0.04), PFHxS (OR = 3.10, 95% CI: 2.10–4.62, p < 0.001), and PFOS (OR = 3.37, 95% CI: 2.41–4.71, p = 0.03). Associations were borderline for PFOA (OR = 3.20, p = 0.07) and not significant for PFNA (OR = 2.58, p = 0.44). In linear models of change in detection rate between UCMR3 and UCMR5, the proportion of water-focused policies was positively associated with changes in PFBS (β = 0.1061, p < 0.001), PFHpA (β = 0.0281, p = 0.02), PFHxS (β = 0.0575, p < 0.001), and PFOS (β = 0.0864, p = 0.03). Associations were not significant for PFNA (β = 0.0018, p = 0.21) or PFOA (β = 0.0767, p = 0.09).

In sensitivity analyses incorporating policy implementation duration, longer duration since the first enacted water-focused policy was associated with lower changes in detection rates for PFBS, PFHpA, PFHxS, PFOA, and PFOS (β range: −0.010 to −0.036; all p≤0.043), while PFNA showed no clear association (p=0.10). However, the association between the proportion of water-focused policies and changes in detection became stronger with increasing duration for PFBS, PFHpA, PFHxS, and PFOS (positive interaction terms; p≤0.035), with a similar borderline interaction for PFOA (p=0.058) and no interaction for PFNA (p=0.28). Consistently, stratified models indicated that the water-policy share-change association was concentrated in the long-term group (significant interactions for PFBS, PFHpA, PFHxS, PFOA, and PFOS; p≤0.021) and short-term interactions were null across PFAS (all p≥0.39). Full summary statistics are in Tables S4 and S5.

In lagged-exposure sensitivity analyses that excluded policies enacted after 2020 or 2021 to reduce temporal mismatch, lagged total PFAS policy counts remained null for UCMR5 detections and UCMR3–UCMR5 changes across all six PFAS (all p>0.13). In contrast, the lagged proportion of water-focused policies remained positively associated with PFAS outcomes. Using the 2021 cutoff, higher lagged water-policy share was associated with higher odds of detection for PFBS (OR=2.76, 95% CI: 1.94–3.91; p=0.015), PFHpA (OR=4.56, 2.72–7.57; p<0.001), PFHxS (OR=6.04, 4.00–9.07; p<0.001), PFOA (OR=5.03, 3.47–7.27; p<0.001), and PFOS (OR=5.16, 3.60–7.37; p<0.001), with no clear association for PFNA (p=0.15). Lagged water-policy share was also positively associated with increases in detection-rate change for PFBS, PFHpA, PFHxS, PFOA, and PFOS (all p≤0.003), while PFNA remained non-significant. Full summary statistics are in Table S6.

In Table S7, overall, the number of PFAS tests increased from UCMR3 to UCMR5 across almost all states (paired t-test p-values: p<0.01).

Shown in Table S8, PFAS policies were not associated with a greater number of PFAS tests, and a greater number of PFAS tests was associated with greater PFAS detection for each PFAS. Furthermore, controlling for the number of PFAS tests in UCMR5 did not attenuate the effect estimates compared to the main analysis (Table S9). Baseline PFAS detection was associated with the number of PFAS policies (Table S10).

3.3. Demographics

Shown in Table 2, of the 29 participants, nine (31%) lived in Pasadena or Torrance, and 20 (69%) lived in Southeast or Central Los Angeles. In the Pasadena/Torrance group, residents were older (mean age 61 years), predominantly female (89%), and ethnically diverse (44% Caucasian/White, 33% Asian, 22% Latino/Hispanic). 89% had completed college or higher. The Southeast/Central LA participants were younger (mean age 48 years), all female (100%), and all Latino/Hispanic (100%). 55% had completed college or above.

Table 2.

Demographics of study participants.

		Pasadena/Torrence	Central/Southeast LA
	Statistics	N = 9	N = 20	p value
Female	n (%)	8 (89%)	20 (100%)	1
Age (yrs)	mean (sd)	61 (17)	48 (11)	0.06
Ethnicity	n (%)
White		4 (44%)	-
Hispanic		2 (22%)	20 (100%)	<0.01
Asian		3 (33%)	-
Education: College or above	n (%)	8 (89%)	11 (55%)	0.09

^1.Fisher’s exact test was used for categorical variable and Welch’s t-test for continuous variable.

3.4. Perception of Drinking Water Quality and Safety

Shown in Table 3, in the pre-focus-group survey, Pasadena/Torrance participants (n=9) most often combined bottled (40%) and filtered tap water (56%) as their primary drinking sources. Meanwhile, Southeast/Central LA respondents (n=20) more frequently supplemented bottled water (70%) with filtered tap water (55%).

Table 3.

Source of drinking water, perceived drinking water safety, source of information.

		Pasadena/Torrence	Central/Southeast LA
	Statistics	N = 9	N = 20	p value
Source of Drinking Water	n (%)
Unfiltered Tap Water		2 (20%)	6 (30%)	1
Filtered Tap Water		5 (56%)	11 (55%)	1
Bottled Water		4 (40%)	14 (70%)	0.24
Other		1 (11%)	5 (25%)	0.63
Safety of Water (1: very safe to 4: very unsafe)	mean (sd)	2 (0.6)	2.8 (0.9)	0.01
Safety of Water Stayed the Same	n (%)	4 (44%)	10 (50%)	1
Know Your Water Provider	n (%)	5 (60%)	8 (40%)	0.69
Satisfied with your water provider (1: very satisfied to 4: very dissatisfied)	mean (sd)	2.3 (1.3)	2.3 (0.5)	1
Information about Water Safety	n (%)
Water Company		6 (67%)	16 (80%)	0.64
Internet		3 (33%)	7 (35%)	1
Local government website		2 (22%)	11 (55%)	0.13
State government website		1 (11%)	4 (20%)	1
Local news		0 (0%)	5 (25%)	0.15
Friends, family or neighbors		2 (22%)	2 (10%)	0.57
Social media		2 (22%)	4 (20%)	1
Federal website		1 (11%)	4 (20%)	1
Nonprofit website		1 (11%)	6 (30%)	0.38

^1.Fisher’s exact test was used for categorical variable and Welch’s t-test for continuous variable.

Pasadena/Torrance participants rated their tap water as somewhat or very safe (mean [SD]: 2.0 [0.6]) and reported that safety had largely stayed the same over the past five years (44%). Over half of them (60%) knew who their water provider was, and most expressed neutral satisfaction with that provider (mean [SD]: 2.3 [1.3]). Southeast/Central LA participants were more divided on tap-water safety (mean [SD]: 2.8 [0.9]), and half of the participants felt the water quality stayed the same. Far fewer could identify their water provider (40%), and satisfaction with their water providers was more neutral or mixed (mean [SD]: 2.3 [0.5]).

When asked where they would go for more information, the Pasadena/Torrance group relied on the water company’s website (67%) whereas Southeast/Central LA participants also turned to water company’s website (80%) and internet (33%) along with local government website (55%) and internet (35%).

We did not see statistically different answers between the two groups except for perceived safety of drinking water (p=0.01).

3.5. Detection of PFAS

Overall, PFAS were not detected in any tap water sample collected in Pasadena or Torrance, but were frequently detected in Central or Southeast LA tap water. At least one PFAS compound was detected in 50% of samples and the average concentration was 9.64 ng/L (See Table 4). As illustrated in Figure 2, within Central or Southeast LA, PFAS were detected in only two of the eight ZIP-code areas sampled (East Los Angeles and Bell Gardens). All other areas returned non-detects across all PFAS chemicals tested. In East Los Angeles, only PFBA was detected (mean: 0.275 ng/L; SD: 0.55 ng/L; range 01.1 ng/L). Table 4 reports Central/Southeast LA as a combined group and PFBA ranged from 0 to 1.6 ng/L across the full Central/Southeast LA sample. In Bell Gardens, total PFAS concentrations were on average 22.3 ng/L (SD 17.4 ng/L; range 8.7–51.3 ng/L). The highest individual chemical concentration observed was PFOS at 22.4 ng/L, and the highest total PFAS concentration measured in any single sample was 51.3 ng/L, both in Bell Gardens. A broad range of PFAS chemicals was detected in Bell Gardens, including perfluorobutanoic acid (PFBA), perfluoropentanoic acid (PFPeA), perfluorohexanoic acid (PFHxA), perfluoroheptanoic acid (PFHpA), PFOA, perfluorononanoic acid (PFNA), perfluorobutanesulfonic acid (PFBS), perfluorohexanesulfonic acid (PFHxS), perfluorooctanesulfonic acid (PFOS), and perfluoro-2-ethoxyethanesulfonic acid (PFECHS).

Table 4.

Detection and concentrations of PFAS in tap water.

	Pasadena/Torrence (N = 9)		Central/Southeast Los Angeles (N = 14)
PFAS (ng/L)	Detection (n [%])	Concentrations (mean [SD])	Detection (n [%])	Range (min to max)	Concentrations (mean [SD])
PFBA	0 (0)	--	3 (21%)	0–1.6	0.28 (0.56)
PFPeA	0 (0)	--	4 (29%)	0–2.9	0.61 (1.07)
PFHxA	0 (0)	--	5 (36%)	0–4.8	1.2 (1.78)
PFHpA	0 (0)	--	1 (7%)	0–1.4	0.17 (0.44)
PFOA	0 (0)	--	3 (21%)	0–8.4	1.19 (2.64)
PFNA	0 (0)	--	1 (7%)	0–1	0.07 (0.27)
PFBS	0 (0)	--	4 (29%)	0–3.7	0.8 (1.39)
PFHxS	0 (0)	--	3 (21%)	0–5.2	0.84 (1.79)
PFOS	0 (0)	--	3 (21%)	0–22.4	3.63 (7.8)
PFECHS	0 (0)	--	3 (21%)	0–6.1	0.84 (1.88)
Total PFAS	0 (0)	--	7 (50%)	0–51.3	9.64 (15.69)

Figure 2.

Detection of PFAS in tap water in Southeast Los Angeles.

3.6. Focus Group/Interview: Community Observations

We conducted one focus group and four semi-structured interviews among residents of Pasadena and Torrance. All reported using municipal tap water for drinking and cooking, and most supplemented with household filtration systems or bottled water. We conducted five Spanish-language focus groups and three semi-structured interviews with residents of Southeast and Central Los Angeles. Most supplemented their drinking or cooking water with bottled supplies, bulk-refill stations, or home filtration systems.

In the first group (Pasadena/Torrence), participants overwhelmingly mentioned invisible chemical pollutants, especially PFAS, and concerns about plastics in bottled and tap water over traditional microbiological or heavy-metal risks. As one interviewee explained, “I am very concerned about PFAS and all the other persistent organic pollutants… I don’t know whether my water system actually cares about those,” and during the focus group another resident added, “I’m concerned about the plastic that they say that it’s in the water, especially with the water bottles.” Sensory cues, most often intermittent sewer-like odors, served as potent triggers of distrust: “The water smells like sewer… It’s on some days and not others.”

In the second group (Central/Southeast LA), across all sessions, deep mistrust of tap water centered on chemical and infrastructural concerns. Many had “heard that it’s contaminated, that it has lead, that the pipes are old, and sometimes it has a lot of fluoride,” with some concerned that fluoride may be linked with dental issues in children. Chlorine taste and odor were nearly universal worries: “Si tomas directamente del grifo, sale con mucho cloro” (“If you drink straight from the tap, it comes out with a lot of chlorine”). Visual cues, black flecks, yellowish scale, or white residue after boiling, served as further red flags: “Cuando hiervo agua para el café… siempre queda eso blanco alrededor de la olla” (“When I boil water for coffee… there’s always that white residue around the kettle”).

3.7. Focus Group/Interview: Personal Experiences

In the first group, to protect themselves, nearly all participants reported relying on household treatment or bottled water, though confidence in these measures varied. Many used basic carbon-block pitchers: “Brita filter, that barely filters anything…but it makes me feel a little bit better,” yet admitted that cartridge replacement was a “guessing game.” High-performance units (e.g., Berkey) were widely viewed as too expensive: “I’d love to have a Berkey, but you know what they cost.”

In the second group, nearly all participants avoided drinking tap water without treatment, often boiling it first or purchasing from refill stations: “Mis vecinos van y llenan sus galones de agua en otro lugar” (“My neighbors go and fill their water jugs somewhere else”) after seeing black or yellow particles. Household filters were common, but people showed mixed confidence, and many worried about unknown filter efficacy. Boiling was seen as cost-effective.

3.8. Focus Group/Interview: Information and Awareness

In the first group, when seeking updates on water quality, participants most often turned to their municipal utility’s website and mainstream news outlets. Convenience drove initial consultations of the news, but many felt that official reports lacked depth: “If they do publish anything about chemical contaminants on their website, I will find that more trustworthy than the media.” A few also cross-checked nonprofit or county health sites, but none felt fully satisfied by existing sources.

In the second group, residents primarily relied on informal channels, including word of mouth, local news, and occasional Los Angeles Department of Water and Power (LADWP) bulletins, for water-quality information. Awareness of emerging contaminants like PFAS was limited. Those familiar typically learned about them through media reports on firefighting foams or jet-fuel chemicals.

3.9. Focus Group/Interview: Policy

In the first group, residents discussed around four actionable policies including:

1. Independent Testing for water quality

   One participant suggested that “They (the water companies) shouldn’t be testing their own water… let a third-party lab handle it.” One other participant suggested that the water companies should set up “clinic” for companies to take samples from residents’ home and get them tested.

2. “Polluter Pays”

   Several participants called for industries (e.g., chrome platers, microchip fabrication facility) to fund routine community monitoring and face heavier fines for violations, with revenues reinvested into local water-quality programs.

3. Real-Estate Disclosure Requirements

   One participant mentioned that “I wish my realtor had warned me about the old industrial site 100 yards away.” It was suggested that mandatory notification of nearby contamination sites or Superfund listings during property transfers is essential.

4. Subsidized Home Filtration

   To overcome cost barriers, participants recommended voucher programs or tax credits for certified filter systems, especially for renters, who often cannot install permanent units.

Residents in Southeast/Central LA proposed a bottom-up suite of solutions tailored to local needs:

1. Culturally Tailored Workshops & Outreach

   Participants suggested that it is beneficial for water companies or local government to host in-person sessions in multiple languages. One participant said that these water-related worskshops could be “like diabetes classes, but for water,” with practical tips (e.g., “run the tap for five seconds to flush lead”) and plain-language overviews of treatment processes.

2. School Field Trips to Treatment Plants

   One participant suggested that water companies or schools should consider organizing visits to the water treatment plants so that families see firsthand how water is sourced, treated, and tested. This would be an essential step for building transparency and trust.

3. Transparent, Community-Focused Reporting

   To foster better water quality reporting, participants mentioned that regular neighborhood bulletins detailing reservoir cleanings, boil-water advisories, and pipe inspections are useful. One participant said that “we need to know when they’re flushing lines or repairing tanks.” Other participants also mentioned more governmental oversight and regular reporting from the government to verify the quality of the water.

3.10. Focus Group/Interview: Quantitative assessment

The relative frequencies of excerpts by bigger themes and codes are illustrated in Figure 3. Personal Experiences dominate the thematic landscape in both groups, but are more prevalent in SE/Central LA than in Torrance/Pasadena, while Policy-related content is also relatively higher in SE/Central LA. Further assessing the specific codes, Figure 3 shows that mitigation behaviors (boiling water, bottled/refill water) and health-risk perceptions drive the SE/Central LA profile, whereas Torrance/Pasadena showed a greater share of sensory observations and (dis)trust in municipal reports excerpts.

Figure 3.

Distribution of Overarching Themes and Specific Codes in Drinking-Water Perceptions Across SE/Central LA and Torrance/Pasadena Participants.

Discussion

In the present study, we conducted the first study to examine whether PFAS policies are associated with PFAS detections in more than 56,000 public water systems across the US. We found that the total number of PFAS policies was not associated with reductions of PFAS in drinking water except for PFBS. A higher proportion of water-related PFAS policies was associated with an increase in PFAS detection. In addition, we evaluated the community concerns, personal experience, source of information, and suggestions for policy for safer access to drinking water in LA among residents from Torrance, Pasadena, Southeast, and Central LA. Residents across the communities shared the experience of drinking water that sometimes smelled, looked, or tasted bad, yet the threat of PFAS contamination is invisible. Several policies were suggested by residents, including additional testing of tap water, “Polluter Pays,” transparent information, and educational materials on drinking water safety. Based on PFAS testing results, we identified neighborhoods in Bell Gardens and East LA that had detectable amounts of PFAS, including both legacy PFAS (PFOS) and other emerging PFAS.

The finding that water-focused policies are associated with higher detection should not be interpreted as evidence that such policies worsen PFAS contamination. Rather, this association may reflect policy responsiveness in states with greater pre-existing PFAS burden and was not explained by the number of tests performed in UCMR5. Moreover, many state policies may not yet have had sufficient time to drive remediation or infrastructure upgrades capable of lowering PFAS concentrations in drinking water. Previous studies have shown state-level differences and challenges in setting maximum contaminant levels or requiring treatment interventions,(20) but few have empirically linked policy adoption to national monitoring data. One study specifically assessed California’s legislation in reducing PFOA and PFOS and found that these policies can reduce PFOA and PFOS levels.(21) California has been at the forefront of PFAS regulation, often adopting stricter standards and implementing policies more quickly than other states.(22) This variation underscores the importance of considering state-level heterogeneity when evaluating changes in PFAS contamination associated with state-level policies. In our analysis, we found that a greater proportion of water-focused PFAS policies was associated with higher odds of detecting PFAS, reflecting the reality that stronger policies may first lead to increased monitoring and identification of contamination before ultimately contributing to reductions in exposure. At the same time, our findings suggest that the overall presence of PFAS policies across states has the potential to uncover hidden contamination and, in the long term, reduce PFAS levels in drinking water nationwide.

We found a borderline association between total PFAS policies and reduction of PFBS detections. PFBS is a short-chain PFAS that has been used as a replacement for PFOS and has been included in drinking-water advisory actions, which may increase the likelihood that state water-focused policy activity corresponds to monitoring and mitigation actions that affect PFBS.(23) PFBS also differs from longer-chain PFAS in toxicokinetics (e.g., shorter human half-life compared with PFOS and PFHxS), and detection dynamics can vary depending on baseline occurrence, reporting limits, and treatment changes.(23) Given the evaluation of multiple analytes and the potential for sparse detections, the PFBS finding should be interpreted cautiously and viewed as hypothesis-generating.(24)

The national and LA findings suggested a coherent surveillance–trust pathway. Nationally, a higher proportion of water-focused policy actions was associated with higher odds of PFAS detection in UCMR5, a pattern most consistent with targeted surveillance and disclosure in states that recognized or suspected PFAS risk rather than policies increasing contamination. This interpretation aligned with community recommendations: Los Angeles residents repeatedly emphasized the need for independent testing, transparent reporting, and clear, actionable communication about what PFAS results mean for household decisions. The combined evidence therefore implied that policy approaches centered only on monitoring may be incomplete. Policies that expand surveillance should be coupled with public-facing transparency (e.g., easy-to-access results dashboards, plain-language summaries) and verification pathways (support for independent testing or confirmatory sampling) to better align implementation with public expectations and strengthen trust in tap water.

In our study, we observed high distrust of the unfiltered tap water with or without specific sensory cues (bad smell, taste, or color) in the water. It has been documented before that LA residents have the nation’s second-highest distrust in tap water.(25) Our survey results further helped explain how residents formed trust (or distrust) in tap water. Southeast/Central LA participants reported lower perceived tap-water safety, greater reliance on bottled water, and less ability to identify their water provider, while also using a broader mix of information sources (utility, local government, internet). However, we emphasized that these survey findings were not intended to represent objective contamination levels or be population-representative, but rather to contextualize household testing and identify actionable barriers to policy implementation.

One previous qualitative study of Latina mothers in LA also shared similar concerns about tap water, including yellow colored tap water, and reported using filters or purchasing water from stores.(26) Therefore, these water-drinking behaviors add to the economic burden of a family, especially for those from lower-income families.(27) One study attributed the source of “mistrust” to contamination from premise plumbing issues or the distribution network rather than a “real” unsafe source of water after treatment.(25) However, the reasoning does not apply to the case of PFAS contamination, given that contamination of PFAS happened partly due to contamination of groundwater due to industrial activities, use of aqueous fire-fighting foams, or discharge of waste streams containing PFAS. Therefore, the distrust of tap water was, to some level, justifiable. In addition, it is difficult for residents to understand the responsibilities of drinking water systems, such as distinguishing the cause of contamination by either public infrastructure or private service line, where landlords are responsible. In our present study, the level of distrust in unfiltered tap water helped the residents to avoid contaminated tap water with PFAS in Southeast LA. For some participants, the total PFAS concentration in tap water can be as high as 51 ng/L, with PFOS at the level of 22 ng/L, which are well above the current proposed EPA MCL (4ng/L).(9) One previous study showed that using a water filter may reduce about 90% of exposure to PFAS.(28) Community-based or cultural practice of water-drinking behavior may be beneficial to prevent exposure to PFAS.

Although few studies focused specifically on LA, one recent study found that PFAS contaminations were heavily concentrated in Southeast LA, which aligned well with our current findings.(12) Our study identified both long-chain and short-chain PFAS in tap water, including legacy PFAS such as PFOA, PFNA, PFHxS, and PFOS and emerging PFAS such as PFBA, PFPeA, PFHxA, PFHpA, PFBS, and PFECHS. Substantial research exists on the potential health risks of exposure to legacy PFAS,(29) but emerging PFAS remain understudied. However, experimental studies suggest these emerging PFAS, such as PFHpA, PFHxA, and PFPeA, are hepatotoxic and are just as toxic or slightly less toxic than legacy PFAS, which raises questions on the suitability of using these PFAS as replacements for legacy PFAS.(30,31) It is important for regulatory agencies to consider the potential health risk of exposure to these emerging PFAS and preemptively set regulations for all PFAS chemicals as a class rather than chemical-by-chemical, when human studies would take years to complete.

Although this study focused on drinking-water monitoring, drinking water represents only one component of overall PFAS exposure. PFAS exposure can also occur through dietary intake and food-related pathways (including food contact materials), consumer products, and other environmental media, and the relative contribution of each pathway can vary by community and consumption patterns.(32–34) Policies that improve drinking-water quality remain critical, but PFAS exposure in general should be interpreted within the broader context of cumulative exposure, consumption patterns, and co-exposure to multiple PFAS beyond drinking water.(32)

In addition, our study shows that LA residents in general are not fully aware of what PFAS are, and since PFAS contamination is invisible, it is challenging to know the risk of PFAS exposure through tap water. Many study participants suggested that workshops on drinking water safety and quality, transparent reporting, and free or subsidized testing or filters are needed. However, different neighborhoods may require tailored policies to address the community’s concern for drinking water safety, which has been proposed by one previous case study in other PFAS-contaminated communities.(35) For example, in a neighborhood with greater distrust of the regulatory agencies, like Torrance/Pasadena, without any PFAS detected, independent testing may be needed to supplement government testing, and repairing the trust may be important to effectively communicate the safety and quality of tap water. In other neighborhoods like Southeast LA, the majority of the participants still rely on information from water companies for water quality and safety, and therefore, providing digestible and accessible information is needed to help residents understand the potential risk of exposure.

Previous research has shown that there is inconsistent awareness of PFAS sources, health risks, and prevention strategies.(36) In addition, one previous qualitative study assessed the source of PFAS contamination, government actions on PFAS contamination, community-level activists in Pease Tradeport in Portsmouth, NH, and Hyannis, MA, which were one of the first communities identified with major PFAS contamination.(37) The case study pointed out that some small initial steps from local residents to inform the public about PFAS contamination in these areas can spark greater governmental actions and collaboration of scientific communities, but greater barriers still existed for further community action.(37) Previous qualitative research also suggests that there is a great need for collaboration between scientific communities and the grassroots movement, and one can influence the other to help regulate PFAS and prevent harm.(38,39) Our study is uniquely focused on a large metropolitan area and presents its unique challenges in health risk communication for PFAS in the absence of an obvious single source of contamination events. Therefore, our study can serve as a new model for engaging communities within large metropolitan areas like LA.

One of the key strengths of this study is our integration of national-level policy analysis with community-engaged qualitative research, allowing for both a comprehensive assessment of PFAS policies and a more nuanced understanding of both actual PFAS exposure and residents’ perceptions, concerns, and policy preferences. We were able to contextualize PFAS testing data within community experiences and highlight community-driven solutions. Furthermore, we provide PFAS testing results at both the individual and community level, which is a critical step towards fostering transparency and increasing residents’ knowledge to make informed decisions about drinking water. However, our study also has limitations. First, PFAS policies often required multi-year timelines for enactment, implementation, and infrastructure change. Therefore, policies enacted later in the 2016–2023 window may not have had time to influence UCMR5 outcomes. In addition, enforcement intensity and implementation effectiveness likely varied within and between states, and these dimensions were not directly measured in our policy metrics. In addition, even if these policies were enacted, the enforcement of these policies may vary by water systems within the same state and between states. Future work should consider including proxy variables such as policy age, enacted-only measures and enforcement proxies to account for the confounding.

We had a relatively small sample size for testing PFAS in tap water, which may not fully capture the geographic distribution or temporal variability of PFAS contamination across Los Angeles. Our water testing was not intended to represent the entire LA population. Additionally, while qualitative findings provide rich insights into community perspectives, they may not be generalizable beyond the specific neighborhoods included. Findings should be interpreted as exploratory and hypothesis-generating. Future studies should include larger, probability-based household sampling across more ZIP codes and repeated sampling over time to support population inference and stronger quantitative-qualitative linkage

These results should be interpreted as associations rather than causal estimates of policy effectiveness. PFAS policies are often enacted in response to prior detections, suspected contamination, and public concern, and may expand monitoring and disclosure. Therefore, the observed positive association between water-focused policy share and UCMR5 detections potentially reflects policy response in higher-burden states, rather than policies increasing contamination. PFAS occurrence is highly localized and shaped by historical and ongoing sources (e.g., industrial releases and AFFF use), which are not fully captured by state-level covariates. Residual confounding is therefore possible.

Conclusion

In conclusion, our study revealed important differences in perceptions of water safety, access to information, and PFAS exposure across Los Angeles neighborhoods, while also providing national evidence on the role of state-level PFAS policies. At the national level, we found that states with stronger PFAS regulatory actions, particularly those focused on water, had increased detection rates of PFAS such as PFBS, PFHpA, PFHxS, PFOA, and PFOS in public water systems. However, our Los Angeles case study demonstrated that even when broad state-level policies are in place, local disparities persist: detectable PFAS was observed in Southeast and Central LA, where communities also reported limited awareness and inconsistent access to trustworthy information. Community-level behaviors, such as avoiding unfiltered tap water, may have inadvertently mitigated exposure in some households, but such adaptations are not a substitute for structural interventions. Together, these findings underscore that while state policies are essential for reducing PFAS contamination nationally, policy responses must be complemented by locally tailored strategies that address community concerns, ensure access to water quality data, and strengthen public trust in drinking water safety.

Highlights.

• Total PFAS policies at the state level may borderline reduce PFBS in drinking water.

• A higher proportion of water-related PFAS policies may lead to greater detection of PFAS, reflecting enhanced surveillance or responses.

• Residents reported low PFAS awareness and high distrust of tap water.

• Policies identified: independent testing, subsidized filters, clear reporting.

Availability of data and materials

The coded deidentified transcript and PFAS water testing data are available upon request.