The peer-reviewed evidence base linking PFAS forever chemicals to cancer, immune suppression, thyroid disease, and developmental harm. Every finding on this page is drawn from government agencies, independent scientific bodies, and peer-reviewed literature. Sources are cited at the end of each section.
The 2022 National Academies of Sciences, Engineering, and Medicine (NASEM) report — commissioned by the U.S. Department of Health and Human Services — synthesized the PFAS health literature into four tiers. This is the most authoritative current synthesis of the evidence.
| Tier | What it means | PFAS health outcomes in this tier |
|---|---|---|
| Sufficient evidence | Positive association observed where chance, bias, and confounding can be ruled out with reasonable confidence across multiple independent studies | Decreased vaccine antibody response; elevated cholesterol; kidney cancer; decreased infant/fetal birth weight |
| Limited / suggestive | Evidence suggests association but cannot fully rule out chance, bias, or confounding | Testicular cancer; breast cancer; liver enzyme elevation; pregnancy-induced hypertension; thyroid disease; ulcerative colitis; chronic kidney disease |
| Inadequate / insufficient | Evidence exists but cannot yet be classified | Most cardiovascular outcomes; most developmental outcomes beyond birth weight |
| Limited evidence of no association | Studies suggest absence of association | Very few endpoints reach this threshold |
Why this matters for Central New York: Lake Ontario already has detectable PFAS at or near the EPA 4 ppt drinking water limit — before Micron discharges a single drop. Semiconductor fab wastewater averages 840 ppt — 210 times the EPA limit. The health effects below describe what happens when people are exposed to PFAS at levels already present in our water system.
The four health outcomes below represent the strongest PFAS evidence base. These are the endpoints where the National Academies concluded that chance, bias, and confounding can be ruled out with reasonable confidence.
This is the most alarming finding in the PFAS literature — because it affects every vaccinated person and is measurable in children at background exposure levels, not just in contaminated communities.
Reduction in diphtheria antibody levels in children at age 5 per doubling of maternal PFOS — enough to fall below clinically protective thresholds
Overall antibody concentration reduction per 2× increase in combined PFAS in child serum (Grandjean et al. 2012, Faroe Islands cohort, n=656)
Per doubling of PFOA across all vaccine antibody types in adults — consistent across diphtheria, tetanus, rubella, measles (meta-analysis, ~4,830 participants)
Diphtheria antibody reduction per doubling of PFOA at age 13 in the same cohort — immune suppression persists into adolescence
What this means in plain terms: Children living near PFAS contamination may receive their scheduled vaccines but still not develop adequate immune protection. A child who has been vaccinated against diphtheria or tetanus may test below clinically protective antibody levels because PFAS have blunted their immune response. This is not a theoretical risk — it has been measured in children at PFAS exposure levels lower than those present in many U.S. communities.
A 2025 in-vitro study using human blood cells directly confirmed the mechanism: PFOA, PFOS, PFNA, and PFHxS suppress T-cell-dependent antibody production — the same pathway that makes vaccines work.
The Lake Ontario connection: OCWA's Executive Director Jeff Brown publicly stated in July 2025 that Lake Ontario already contains 2 ppt PFAS in the raw water source. Children in the OCWA service area are already being exposed to PFAS through drinking water — before Micron begins operations. Adding an uncontrolled industrial PFAS load to that system is not precautionary. It is compounding a known risk.
Some adult studies (COVID-19 vaccine, influenza) found no significant association. Evidence is strongest and most consistent in children and for diphtheria/tetanus vaccines. NASEM still rates overall evidence as sufficient.
The most replicated finding in PFAS research. Confirmed across occupational cohorts, community studies, and general population surveys spanning four decades.
Core finding: A 2025 systematic review and meta-analysis of 74 epidemiological studies confirmed positive associations between PFOA and PFOS exposure and increased total cholesterol and LDL cholesterol in adults. Studies near contamination sources showed the strongest effects. The association holds in general population NHANES data — meaning it applies to people without unusually high PFAS exposure, not just contaminated communities.
C8 cohort: In the 69,000-person C8 Health Project near DuPont's Parkersburg, WV plant, elevated PFOA was associated with significantly higher total cholesterol across all age groups in a dose-dependent pattern.
Mechanism: PFAS activate peroxisome proliferator-activated receptors (PPARs) in the liver, disrupting fatty acid metabolism and increasing LDL synthesis — a direct, well-characterized biochemical pathway, not just a statistical correlation.
The absolute cholesterol increase is modest (typically 5–15 mg/dL in epidemiological studies). Industry will argue this is clinically insignificant. The public health argument is that even modest cholesterol elevation at population scale produces meaningful cardiovascular risk — especially compounding other PFAS effects.
The best-documented PFAS cancer link. Established through the largest PFAS health study ever conducted and confirmed in multiple independent cohorts. The legal and scientific record here is unusually strong — industry has already lost in court on this finding.
Relative risk of kidney cancer for people with elevated PFOA exposure vs. unexposed population — C8 Science Panel 2012 (funded independently as part of DuPont class action settlement)
Settlement paid by DuPont and Chemours in 2017, based on ~3,500 personal injury cases rooted in the C8 "probable link" kidney cancer finding
Increase in U.S. kidney cancer incidence from 1975 to 2022 — from 7.4 to 16.1 per 100,000. No PSA-type screening artifact explains this rise (NCI SEER)
IARC classification for PFOA as of 2023 — upgraded from "possible" to definite human carcinogen. PFOS newly classified Group 2A (probable)
Why the C8 finding matters: The C8 Science Panel was not an advocacy study. It was funded by DuPont as part of a class action settlement and staffed by independent scientists. The "probable link" finding for kidney cancer required seven years of analysis of 69,000 people. It is not contested by any major scientific or regulatory body. When DuPont tried to contest it in court, they settled for over a billion dollars.
IARC 2023 upgrade: The International Agency for Research on Cancer upgraded PFOA from Group 2B (possible carcinogen) to Group 1 — meaning the evidence is sufficient to conclude PFOA causes cancer in humans. This is the same classification as tobacco smoke and asbestos. PFOS was simultaneously classified for the first time as Group 2A (probable human carcinogen).
Mechanism: Kidneys are the primary organ for PFAS excretion; PFAS accumulate preferentially in kidney tubular cells. Mechanisms include oxidative stress, PPAR activation promoting cell proliferation, and suppression of the VHL tumor suppressor gene pathway — the same pathway implicated in inherited kidney cancer syndromes.
Consistent findings across multiple birth cohort studies on multiple continents — U.S., Denmark, Norway, Japan, and Taiwan. This is the developmental endpoint with the strongest and most replicated evidence base.
Core finding: Prenatal PFOA and PFOS exposure is associated with statistically significant reductions in birth weight across multiple independent cohorts. A per-interquartile-range increase in PFOA is associated with approximately 20–50 gram reductions in birth weight. While modest per individual, at population scale this represents a meaningful shift in the birth weight distribution.
Mechanism: PFAS cross the placenta. Prenatal exposure affects fetal growth via disruption of growth hormone signaling and thyroid hormone availability during critical developmental windows — periods when exposure to endocrine-disrupting chemicals can have disproportionate effects relative to adult exposure.
Why it matters here: Pregnant women in the OCWA service area — the 500,000 people whose drinking water comes from the Lake Ontario intake 1.5 miles downstream of the planned discharge — are directly in the exposure pathway. Lake Ontario PFAS is already detectable in that intake before Micron starts operations.
The following outcomes have evidence suggesting a PFAS association but where chance, bias, or confounding cannot yet be fully excluded. Several are likely to move to the "sufficient" tier as evidence accumulates — the trajectory of the science is toward stronger associations, not weaker ones.
| Health outcome | Key quantitative finding | Primary PFAS | Status |
|---|---|---|---|
| Testicular cancer | +97% NCI SEER incidence rise since 1975; consistent cohort associations with no PSA-equivalent confounder; strongest PFAS cancer signal without the C8 contested-science issue | PFOS, PFOA | Limited / suggestive |
| Breast cancer (postmenopausal) | French cohort (n=194 cases) found positive association with pre-diagnostic PFOS for hormone receptor-positive tumors; NCI PLCO follow-up ongoing | PFOS, PFOA | Limited / suggestive |
| Liver enzyme elevation (ALT/AST) | Multiple cross-sectional studies show elevated liver enzymes with higher PFAS serum levels; consistent in occupational and general population cohorts; direct hepatotoxicity confirmed in animal models | PFOA, PFOS, PFHxS | Limited / suggestive |
| Pregnancy-induced hypertension / preeclampsia | Multiple prospective cohort studies; dose-response pattern; mechanism involves endothelial disruption; particularly concerning given fetal growth effects | PFOA, PFOS | Limited / suggestive |
| Thyroid disease and dysfunction | PFAS compete with thyroid hormones for serum binding proteins; most consistent findings in occupational cohorts at higher exposures; mixed results in general population | PFOA, PFOS, PFHxS | Limited / suggestive |
| Chronic kidney disease / reduced GFR | Multiple NHANES cross-sectional studies; propensity score analysis (Jain & Ducatman 2019) supports causal direction; kidneys are primary PFAS excretion organ | PFOA, PFOS, PFNA | Limited / suggestive |
| Ulcerative colitis | C8 cohort probable link finding; independent replications limited; immune dysregulation pathway biologically plausible | PFOA | Limited / suggestive |
The health associations above were established at specific exposure levels. Understanding those levels — and how they compare to what Micron's discharge would add — is essential context for regulators and legislators.
| Reference point | Concentration | Significance |
|---|---|---|
| EPA Maximum Contaminant Level for PFOA in drinking water | 4 ppt (ng/L) | Legally enforceable limit effective 2029. EPA Federal Register, April 26, 2024 |
| EPA MCL for PFOS in drinking water | 4 ppt | Same standard, same source |
| Lake Ontario total PFAS — current, before Micron | 11 ppt total; PFOS 1.3 ng/L; PFOA 1.4 ng/L | Already at or above EPA MCL for individual compounds. PMC11137863, EPA-funded IADN monitoring 2021–2023 |
| Average PFAS in semiconductor fab wastewater — SIA 2025 survey | 840 ppt total | Industry's own data. 210× the EPA 4 ppt drinking water limit. SIA Semiconductor PFAS Consortium, January 2025 |
| PFAS compounds found in fab wastewater — targeted methods (EPA 533/537.1) | ~7 compounds | Standard regulatory testing — what the signed permit requires |
| PFAS compounds found in fab wastewater — non-targeted analysis (Cornell) | 133 compounds | 19× to 126× more than targeted methods detect. Jacob et al., ES&T 2021. "Dark PFAS" invisible to standard monitoring |
| Maternal PFOS at time of Grandjean vaccine suppression findings | ~27 ng/mL serum | Level at which −39% diphtheria antibody reduction was measured in children. Grandjean et al. 2012 |
| NASEM threshold for clinical screening | 7–19.9 ng/mL serum | Standard monitoring recommended at this exposure level. NASEM 2022 |
The central problem: Lake Ontario already has detectable PFAS at or near the EPA drinking water limit — from background and industrial sources — before Micron discharges anything. Semiconductor fab wastewater averages 210 times that limit. And standard monitoring (EPA Method 533/537.1) would detect only 7 of the 133 PFAS compounds Cornell found in fab wastewater — leaving the rest invisible. This is why the campaign demands EPA Method 1633A plus the Total Oxidizable Precursor (TOP) Assay, not just standard methods.
The epidemiological associations are supported by well-characterized biochemical mechanisms. These are not statistical artifacts — they reflect specific molecular interactions that explain why PFAS produce the health outcomes described above.
PFAS disrupt T-cell activation and B-cell antibody production. 2025 in-vitro studies using human peripheral blood mononuclear cells directly confirmed that PFOA, PFOS, PFNA, and PFHxS suppress both T-cell-dependent and T-cell-independent antibody production at concentrations relevant to environmental exposure — not just occupational or accidental exposures.
PFAS activate peroxisome proliferator-activated receptors (PPARs) in the liver — transcription factors that regulate fatty acid oxidation and lipid synthesis. PPAR activation by PFAS dysregulates LDL cholesterol production, explaining the consistent cholesterol elevations seen across populations.
PFAS compete with thyroid hormones — specifically thyroxine (T4) — for binding sites on serum transport proteins, particularly transthyretin. By displacing active thyroid hormone from circulation, PFAS effectively reduce the amount of thyroid hormone available to tissues, disrupting metabolism, development, and fetal growth.
Kidneys are the primary organ responsible for filtering and excreting PFAS from the body. PFAS accumulate preferentially in kidney tubular cells. Proposed mechanisms for kidney cancer include oxidative stress, PPAR-mediated cell proliferation, epigenetic modification, and suppression of the VHL tumor suppressor pathway — the same pathway implicated in hereditary kidney cancer syndromes.
PFAS cross the placenta and concentrate in breast milk. During critical developmental windows, endocrine disruption by PFAS — particularly thyroid hormone displacement and growth factor interference — can have disproportionately large effects compared to equivalent adult exposure. This explains why prenatal PFAS exposure produces birth weight reductions and developmental delays at exposure levels that produce subtler effects in adults.
Multiple converging mechanisms: PPAR activation promoting cell proliferation, oxidative DNA damage, epigenetic modifications silencing tumor suppressor genes, and immune suppression reducing cancer immune surveillance. The convergence of epidemiological evidence and mechanistic data was sufficient for IARC to upgrade PFOA to Group 1 — definite human carcinogen — in 2023.
Credible scientific communication requires acknowledging uncertainty. These are the genuine limits of what the current evidence base shows.
Legacy vs. semiconductor PFAS: The vast majority of health research focuses on PFOA and PFOS — legacy long-chain compounds largely phased out by 2002–2013. The short-chain and ultrashort-chain PFAS that Micron will actually use in Clay have almost no human health data. The absence of data is not evidence of safety.
Mixture toxicity: Almost all health studies assess individual PFAS compounds. Humans are exposed to mixtures of dozens to hundreds of compounds simultaneously. How PFAS compounds interact with each other — and with other semiconductor chemicals — is essentially unknown.
Causation vs. association: Epidemiological studies establish association; randomized controlled trials to prove causation are impossible for obvious reasons. Industry will exploit this distinction. The IARC Group 1 classification for PFOA means causation is considered established — but this applies only to PFOA, not the broader class.
Safe thresholds: It is unknown whether there is a safe threshold below which PFAS exposure produces no harm, or whether effects occur at any detectable level. EPA's 4 ppt MCL reflects the assumption that there is no completely safe level for carcinogens.
Why uncertainty is itself the argument: The short-chain and semiconductor-specific PFAS that Micron will discharge are not the compounds for which we have health data. The 126 of 133 compounds that standard monitoring methods miss — the "dark PFAS" identified by Cornell — have essentially no toxicological characterization. You cannot set a safe discharge limit for compounds you haven't tested. You cannot design a treatment system for compounds whose behavior you don't understand.
This is not precautionary overreach. This is basic engineering and basic public health. Limits first, design second, contract third.
"Even in the best of cases there will not be standards or guidance values for the preponderance of PFAS that Micron will actually be using. Yet current science demonstrates that all PFAS are toxic as well as persistent. Any release of PFAS into the environment adds to the already unacceptable load." — Lenny Siegel, Center for Public Environmental Oversight, comments on Micron DEIS, 2025
All findings cited on this page are drawn from government agencies, independent scientific bodies, or peer-reviewed literature. None are advocacy publications.
PFAS cause measurable harm at concentrations already present in Lake Ontario. Semiconductor fab wastewater averages 210 times the EPA drinking water limit. The signed SPDES permit contains zero enforceable PFAS discharge limits. That has to change in Micron’s Industrial Wastewater Discharge Permit — before the County Legislature authorizes the bond for the Industrial Treatment Plant.