PFAS / 💧 The Poison | 👨‍👩‍👧 The Affected | 🩸✚ The Plan | 🔬 Research Notes
OPEN RESEARCH QUESTIONS

Research Notes

These are not published findings. They are documented hypotheses — coherent arguments constructed from peer-reviewed evidence, recorded here as a timestamped public record before the studies exist to confirm or refute them.

A Note on Methodology

Science advances through hypothesis formation. The studies on this page have not been conducted. The evidence cited in support of each hypothesis is peer-reviewed and documented — the logical connection proposed here is the original contribution. Date-stamping is intentional: if any of these hypotheses are later confirmed by formal research, the public record will show when the pattern was first identified and where.

Nothing on this page should be used as medical or clinical guidance. These are research proposals, not conclusions.

Hypothesis #1 Recorded: June 13, 2026 Awaiting Study

The Uncontrolled Variable in Fetal Alcohol Syndrome Research

Could PFAS co-exposure in contaminated beer explain why FAS outcomes don't track linearly with alcohol consumption?

The Documented Puzzle

Fetal Alcohol Syndrome is one of the most well-studied prenatal exposure conditions in medicine — and one of the most inconsistent in its dose-response relationship. The same reported alcohol consumption produces radically different outcomes across individuals and communities. Standard explanations include timing within pregnancy, maternal genetics affecting alcohol metabolism, binge patterns, and nutritional status. These are real factors. They do not fully explain the variance.

Some communities have FAS rates 3–5× higher than comparable communities with similar alcohol consumption rates. Some women who drank heavily have children with minimal FAS markers. Some women who drank moderately have children with significant effects. The "per drink" dose calculation has always been a poor predictor of outcome. The question is why.

The Evidence Base for This Hypothesis

PFAS is documented in beer — at levels that exceed EPAEPA — Environmental Protection AgencyThe U.S. Environmental Protection Agency (EPA) is the federal agency responsible for protecting human health and the environment by regulating pollutants — including the drinking-w… drinking water limits

A peer-reviewed study from RTI International (Environmental Science & Technology, April 24, 2025) tested 23 beers from U.S. and international breweries using EPA-approved methods. Over half contained measurable PFAS. Some contained PFOS or PFOA at levels exceeding EPA safe drinking water standards. Approximately 18% of U.S. breweries operate in ZIP codes served by PFAS-contaminated municipal water. Craft breweries are disproportionately affected — most lack the reverse osmosis or activated carbon filtration needed to remove long-chain PFAS.

Mullin AJ et al. "Hold My Beer: The Linkage between Municipal Water and Brewing Location on PFAS in Popular Beverages." Environmental Science & Technology, 2025. DOI: 10.1021/acs.est.4c11265.

Alcohol and PFAS share at least one fetal brain development pathway

Alcohol harms the developing brain through several mechanisms — including cell death, oxidative stress, and disruption of thyroid hormone (T4) signaling during critical windows of cortical development. PFAS independently disrupts thyroid hormone function through a documented and peer-reviewed mechanism (reviewed in Current Environmental Health Reports). Thyroid disruption is one pathway both share, with fetal neurodevelopment as the downstream harm. A woman drinking beer that also delivers PFAS is subjecting the fetus to an added load on that shared pathway that neither the alcohol nor the PFAS research has separately accounted for.

The existing FAS research never controlled for this variable

Decades of FAS research have measured alcohol consumption — number of drinks, timing, binge patterns. None of that research controlled for the PFAS content of the alcohol consumed, because PFAS-in-beer was not documented until 2025. A woman drinking three local craft beers from a brewery downstream of a military base receives a fundamentally different combined exposure than a woman drinking three imported beers brewed under stricter water standards. The studies were measuring the same variable — "drinks" — while the actual exposure was different each time.

The Formal Hypothesis

The poor dose-response correlation in Fetal Alcohol Syndrome research is partly attributable to an uncontrolled co-exposure variable: PFAS content in the alcoholic beverages consumed during pregnancy. Because both ethanol and PFAS independently disrupt fetal thyroid hormone signaling — and because PFAS levels in beer vary dramatically based on local water contamination — two pregnancies with the same number of drinks may represent substantially different combined neurodevelopmental toxin loads. The "per drink" variable was never measuring what it appeared to measure.

What Would Confirm or Refute It

What the test would look like

A retrospective study comparing FAS case severity against maternal PFAS serum levels at time of pregnancy, controlling for total alcohol consumption. Alternatively: a geographic correlation study comparing FAS rates in PFAS-positive vs. PFAS-negative drinking water regions, controlling for regional alcohol consumption rates. The EPA UCMR5 data now provides ZIP-code-level PFAS contamination data. FAS rates are tracked. The necessary data exists.

What would need to be true

PFAS and alcohol must show synergistic — not merely additive — disruption of fetal thyroid function (mechanistically plausible, not yet directly tested in co-exposure models). FAS outcome severity would need to correlate with regional PFAS contamination at rates exceeding what alcohol consumption alone would predict. Unexplained FAS cases in "low-consumption" profiles would cluster in PFAS-positive regions.

Why this is not a published finding

The co-exposure studies have not been conducted. PFAS-in-beer was not confirmed by EPA-method testing until April 2025. FAS research has not looked for this confounding variable because the data on PFAS in alcohol did not exist when most FAS dose-response studies were designed. The hypothesis is ahead of the evidence base — which is why it is documented here as a hypothesis, not presented as a conclusion.

Supporting Evidence (Peer-Reviewed)

  • Mullin AJ et al. "Hold My Beer: The Linkage between Municipal Water and Brewing Location on PFAS in Popular Beverages." Environmental Science & Technology, 2025. DOI: 10.1021/acs.est.4c11265.
  • Brewers Association staff review of the RTI study findings, May 13, 2025. brewersassociation.org/association-news/pfas-compounds-found-in-beer/
  • EWG. "Toxic substances in your suds? Study finds PFAS in local water used to brew beer." May 12, 2025. ewg.org.
  • PFAS thyroid disruption mechanism: documented in peer-reviewed literature reviewed in Current Environmental Health Reports and cited across the AbilityForge PFAS Affected documentation.
  • FAS dose-response variability: established in the FASD literature; documented reviews available through NIAAA and SAMHSA.
Hypothesis #2 First documented: June 13, 2026 Researcher Contact Made

A Contained Ovalbumin Insert to Remove PFAS from Pumped Breast Milk

Could a sealed mesh device — functioning like a teabag — use ovalbumin's documented PFAS-binding chemistry to strip contamination from breast milk before infant feeding?

The Problem This Would Solve

Current guidance (ATSDR) tells mothers in PFAS-exposed communities to keep breastfeeding — the benefits outweigh the risks. That is the right call with the tools we have, but it leaves families with two imperfect options: nurse and pass part of the mother's accumulated PFAS body burden to the baby, or give up the most nutritionally and immunologically irreplaceable food of early infancy. There is no clinical tool that treats the milk, and no screening step between exposure and infant feeding.

Two paths could change that. One is a way to make the milk itself safer — the device proposed below. The other is an expanded national milk-bank system that screens donated milk for PFAS, giving exposed families a tested alternative instead of a forced choice.

The documented outcome: lactation functions as a PFAS offloading mechanism for the mother and a concentration event for the infant. The body burden accumulated over an adult lifetime is delivered to an organism that may weigh seven pounds, with an immature blood-brain barrier and organs still developing. The hypothesis here is that the chemistry needed to interrupt this exists — and has now been confirmed at the molecular level.

The Published Mechanism

Ovalbumin binds PFAS — confirmed at the atomic level

Research from North Dakota State University and Iowa State University (Arshad et al., with corresponding authors Bezbaruah and Xia, Cell Reports Physical Science, 2026) found that ovalbumin — the primary protein in egg white — binds PFAS spontaneously, with arginine and lysine residues critical to binding stability. Hydrophobic interaction is the primary driver. In lab tests it removed 87% of PFOS and 80% of PFOA from water, with weaker binding for short-chain PFAS. The authors propose it as a platform for PFAS sensing and remediation.

Arshad A, Mazumdar S, Goswami J, Quadir M, Nadagouda M, Bezbaruah AN, Xia W. "Ovalbumin as a PFAS carrier protein in aquatic environments." Cell Reports Physical Science 7, 103089 (2026). DOI: 10.1016/j.xcrp.2025.103089

The same binding chemistry operates in blood — which is why PFAS persists

PFAS binds to serum albumin in blood — the primary endogenous carrier protein — with a half-life of 3–8 years because the enterohepatic loop continuously recirculates it before hepatic clearance. In breast milk, PFAS binds similarly to β-lactoglobulin and fatty-acid binding proteins. This is not a new discovery: it is the documented mechanism of PFAS persistence. The novel question is whether an exogenous ovalbumin source could out-compete the milk proteins currently holding PFAS in a milk matrix under practical conditions. Ovalbumin's binding has been measured in water; how it compares with milk proteins has not — that comparison is part of what the experiment below would test.

The Device Concept

A sealed mesh insert — functioning on the principle of a teabag — would contain immobilized ovalbumin in a food-safe porous membrane. Placed into a bottle of pumped breast milk and agitated for a defined contact period, the ovalbumin would compete with β-lactoglobulin for PFAS binding. If ovalbumin's affinity proves higher, PFAS would migrate toward it. The insert would be removed before feeding, taking the bound PFAS with it.

Binding Chemistry

Documented. Ovalbumin binds PFOA and PFOS at arginine/lysine docking sites (Bezbaruah/Xia, 2026). The mechanism is confirmed.

Research Gap

Whether binding efficiency holds in a complex milk matrix — with competing proteins, lipids, and lactose — vs. water. This is the untested variable.

Food Safety Baseline

Ovalbumin is GRAS. Egg protein is present in infant-safe foods. If fully contained, no transfer to milk occurs. If trace transfer occurs — see allergen note below.

The Allergen Objection — and the LEAP Counter

The immediate objection to ovalbumin in proximity to infant feeding is egg white allergy — one of the most common pediatric allergens. This objection was well-founded under the prevailing immunological model through approximately 2015. It has since been substantially revised.

LEAP and the oral tolerance principle

The Learning Early About Peanut Allergy (LEAP) trial (NEJM, 2015 — NIH-funded) demonstrated that early oral exposure to food proteins does not cause allergy — it confers tolerance. Children introduced to peanut protein in early infancy had dramatically lower peanut allergy rates than those who avoided it. Follow-up published 2024 confirmed this protection persisted into adolescence. The NIH, AAP, and ACAAI have subsequently revised guidance for multiple food proteins, including egg, to recommend early introduction rather than avoidance.

Du Toit G et al. "Randomized Trial of Peanut Consumption in Infants at Risk for Peanut Allergy." NEJM 372:803-813 (2015). Follow-up: Randomized Trial of Peanut Consumption — Adolescence cohort, NEJM 2024. NIH: nih.gov/news-events/news-releases/introducing-peanut-infancy-prevents-peanut-allergy-into-adolescence.

Applied here: if the membrane is truly impermeable to protein, no ovalbumin enters the milk and no allergen exposure occurs. If trace ovalbumin transfers — following LEAP's oral tolerance logic — early low-dose exposure during infancy may confer tolerance rather than sensitization. The allergen objection, under current immunological understanding, operates in the opposite direction from the pre-2015 assumption. The concern does not disappear entirely — infants with known sensitization or family history of severe egg allergy would warrant different consideration — but the blanket objection no longer holds as an absolute barrier to the concept.

Researcher Contact — May 26, 2026

This hypothesis was communicated directly to Dr. A. Bezbaruah (North Dakota State University), a corresponding author on the ovalbumin PFAS binding study, on May 26, 2026 — eighteen days before this page was published. The email proposed the milk bank style treatment application, identified the PFAS-binding mechanism as the foundation, noted the food safety baseline for ovalbumin, and framed the milk matrix performance question as the unknown worth testing. The AbilityForge site was cited as a reference.

Subject: "Breast Milk Application of Ovalbumin PFAS Binding — Research Gap Worth Exploring." Sent May 26, 2026, 12:39 PM.

What Would Confirm or Refute It

The experiment that hasn't been run

Spiked breast milk (PFAS-dosed in controlled quantity) + immobilized ovalbumin insert + defined agitation and contact time → measure PFAS concentration before and after. Compare removal efficiency to plain water. Secondary: assess milk protein integrity, IgA concentration, and fat content before and after to verify the milk remains nutritionally intact. This is a straightforward benchtop experiment requiring no novel technology.

What would need to be true

Ovalbumin's PFAS binding affinity must exceed that of β-lactoglobulin in milk matrix conditions (temperature ~37°C, pH ~7.0, high protein/lipid environment). The milk proteins, immunoglobulins, and fatty acids must be preserved — the device must be selective for PFAS without stripping beneficial components. The removal rate must be clinically meaningful given PFAS levels documented in breast milk from exposed communities.

Supporting Evidence

  • Arshad A, Mazumdar S, Goswami J, Quadir M, Nadagouda M, Bezbaruah AN, Xia W. "Ovalbumin as a PFAS carrier protein in aquatic environments." Cell Reports Physical Science 7, 103089 (2026). DOI: 10.1016/j.xcrp.2025.103089.
  • PFAS in breast milk / β-lactoglobulin binding: documented in systematic review in Toxicological Sciences, cited in AbilityForge PFAS Affected documentation (Step 3).
  • Du Toit G et al. "Randomized Trial of Peanut Consumption in Infants at Risk for Peanut Allergy." NEJM 372:803–813 (2015). DOI: 10.1056/NEJMoa1414850.
  • LEAP Adolescence Follow-Up. NEJM, 2024. NIH announcement: nih.gov/news-events/news-releases/introducing-peanut-infancy-prevents-peanut-allergy-into-adolescence.
  • Srinivasan R et al. "Fenugreek and Okra Polymers as Treatment Agents for the Removal of Microplastics from Water Sources." ACS Omega 10(15):14640 (2025). DOI: 10.1021/acsomega.4c07476. [Referenced for biopolymer-contaminant capture concept.]
Hypothesis #3 Recorded: June 13, 2026 Connecting Argument

The Wrong Attribution Pattern

What if PFAS creates the condition, the condition produces a visible variable, and the visible variable gets blamed? This connecting argument looks at three debates in the neurodevelopmental literature where that question was never asked — and at the mothers who most urgently need testing.

The Structure of the Misattribution

Epidemiologists identify associations between variables and outcomes. When a signal appears — a temporal relationship, a dose correlation, a cluster — the most recent or most visible intervention tends to attract the causal attribution. What this framework misses is the chronic background exposure that may have set the biological trajectory long before the attributed cause arrived. PFAS is exactly this kind of background exposure: invisible, persistent, accumulated over years, operating through pathways that produce outcomes indistinguishable from those attributed to acute exposures.

Below are three debates in the literature on childhood neurodevelopmental harm. Each centered on a visible, recent exposure. In none of them was the PFAS body burden measured. The hypothesis is that, in exposed communities, PFAS may be part of the background that shaped the outcomes — a question that can be tested.

Instance 1 — The Vaccine Claim

What's been claimed: Some parents report their child changed after a round of vaccines, and HHS Secretary Robert F. Kennedy Jr. has long promoted the idea that vaccines cause autism.

What the evidence shows: Vaccines have been studied for this more than almost anything else. A Danish study of more than 650,000 children (Hviid et al., Annals of Internal Medicine, 2019) found no link between the MMR vaccine and autism. Thimerosal was studied and cleared.

Why parents see it: The first signs of autism often become noticeable between 12 and 24 months — the same age the childhood vaccine schedule is busiest. The timing parents notice is real. The cause isn't the vaccine.

What PFAS actually does to vaccines: Children with higher PFAS levels make fewer antibodies after their routine shots (Grandjean et al., JAMA, 2012). Contamination may be making vaccines work less well — a documented harm in its own right.

The open question: Families looking for an environmental explanation for rising autism have been pointed at one of the most-studied exposures there is. PFAS — associated with ASD in the peer-reviewed literature — has been studied far less.

Instance 2 — Prenatal Acetaminophen and the Pharmacology Problem

The signal: Multiple epidemiological studies have identified an association between prenatal acetaminophen (Tylenol) use and neurodevelopmental outcomes in offspring, including ASD and ADHD presentations. A 2021 consensus statement from researchers across multiple institutions called for precautionary guidance. The signal is real and has been replicated.

The pharmacology problem: Pregnant women take acetaminophen for pain and fever — headaches, musculoskeletal discomfort, infections. They cannot take ibuprofen. NSAIDs (ibuprofen, naproxen) are contraindicated in pregnancy — particularly after 20 weeks — due to risk of premature closure of the ductus arteriosus and other fetal vascular effects. Acetaminophen is the main option left.

What acetaminophen cannot do: Treat inflammation. Acetaminophen is a pain and fever reducer, not an anti-inflammatory — it works mainly in the central nervous system and does little for inflammation in the body. The drugs that do treat inflammation are off the table in pregnancy. If inflammation is part of what's driving the symptoms, the mother gets the drug that eases her pain while leaving the underlying process running. The PFAS continues to cross the placenta. The fetal neurological disruption continues. The child is born with neurodevelopmental differences. The researcher records: prenatal Tylenol use.

What was never measured: Why the mother needed the Tylenol. PFAS exposure is associated with preeclampsia (Swedish SELMA, OR 2.68) and with inflammatory and endocrine disruption. If PFAS contributed to the conditions that drove the prescription, then Tylenol use may be a marker of that underlying exposure — not the cause of the outcome. The studies on prenatal Tylenol do not control for maternal PFAS serum levels. The uncontrolled variable appears again.

Confounding by indication: This is the established epidemiological term for exactly this problem. Women who take Tylenol during pregnancy take it because something is already wrong. In PFAS-exposed communities, "something wrong" may include PFAS-related illness. The indication confounds the outcome — which is exactly what the sibling study below found.

The sibling study: direct refutation — 2,480,797 children, association disappears

A 2024 nationwide cohort study (Ahlqvist, Sjöqvist, Dalman et al., Karolinska Institutet / Drexel University — JAMA, April 9, 2024) analyzed every child born in Sweden from 1995 to 2019. 185,909 children were exposed to acetaminophen during pregnancy. In population-level models without sibling controls, marginal associations appeared — consistent with prior literature. Then the researchers applied sibling control analysis: comparing outcomes between full siblings within the same family, controlling for shared genetics and shared household environment. The signal disappeared completely.

HR 0.98

Autism (95% CI 0.93–1.04)

HR 0.98

ADHD (95% CI 0.94–1.02)

No dose-response pattern in sibling analysis. High-dose acetaminophen: autism HR 0.88. Authors' conclusion: "Acetaminophen use during pregnancy was not associated with children's risk of autism, ADHD, or intellectual disabilityDisabilityPermanent or long-term functional impairment resulting from an untreated or inadequately treated medical condition. in sibling control analyses. This suggests that associations observed in other models may have been attributable to familial confounding."

The PFAS reading of this result: The sibling design controls for what siblings share — genetics, household, neighborhood. It does not control for what siblings don't share. Each child's gestational PFAS exposure depends on when they were conceived relative to contamination events, how much PFAS the mother's body had accumulated and shed by then, placental transfer efficiency, and breastfeeding duration. Maternal PFAS body burden decreases with each pregnancy and lactation cycle — a documented offloading mechanism. A first child gestated during a contamination peak, breastfed for 12 months, receives a meaningfully higher cumulative dose than a sibling born three years later after two rounds of maternal PFAS offloading. The sibling analysis assumed equal PFAS exposure within families. That assumption is almost certainly wrong — and the attenuation it found is exactly what PFAS confounding would predict once the genetics variable is removed.

Ahlqvist VH, Sjöqvist H, Dalman C, et al. "Acetaminophen Use During Pregnancy and Children's Risk of Autism, ADHD, and Intellectual DisabilityDisabilityPermanent or long-term functional impairment resulting from an untreated or inadequately treated medical condition.." JAMA. 2024;331(14):1205–1214. DOI: 10.1001/jama.2024.3172.

Instance 3 — Fetal Alcohol Syndrome Dose Variance (see Hypothesis #1)

The same structure: PFAS in contaminated beer co-delivers thyroid-disrupting chemicals alongside ethanol. The study measures drinks. The outcome varies unexpectedly. The FAS researchers look harder at alcohol dose, timing, and maternal genetics — without measuring the PFAS content of what the mother was drinking. The uncontrolled variable is the same one. See Hypothesis #1 for the full argument.

The Downstream Indictment

Each of these debates focused on a visible, recent exposure — a shot, a pill, a drink. None of them measured the invisible background one. The companies that contaminated drinking water near military installations, airports, and manufacturing sites left a variable in place that the research never looked for.

The mothers at the center of these debates — blamed for a prescription, a drink, or a decision — deserve the same question the research skipped: what was already in their blood? In contaminated communities, they may be among the most exposed. They are the patients who most urgently need PFAS serum panels. They have been treated as variables instead of as victims.

The mortality anchor: 191% increased first-year infant deathDeaththe act of dying; the end of life; the total and permanent cessation of all the vital functions of an organism. rate

Baluja, Guo, Howden, Langer, and Lemoine (University of Arizona / National Bureau of Economic Research / Cornerstone Research) published in PNAS (December 2025) an analysis of New Hampshire birth records from 2010–2019. Mothers receiving water that had flowed beneath a PFAS-contaminated site — as opposed to comparable mothers receiving water that had flowed toward a contaminated site — had 191% higher first-year infant mortality (95% CI: 83–298%), 168% more births before 28 weeks of gestational age, and 180% more births below 1,000g birthweight. Extrapolated to the contiguous United States: approximately $8 billion in annual social costs from infant mortality and adverse birth outcomes alone.

This is the quantified outcome at the end of the chain this hypothesis describes. The wrong attributions in the literature identify the proximate variables. This number documents the endpoint. The babies dying in contaminated communities are not dying because their mothers took Tylenol. They are dying because the water was contaminated before anyone thought to test it.

Baluja R, Guo B, Howden W, Langer A, Lemoine D. "PFAS-contaminated drinking water harms infants." Proc Natl Acad Sci U S A. 2025 Dec 16;122(50):e2509801122. DOI: 10.1073/pnas.2509801122. PMID: 41359852.

Note on structure

Unlike Hypotheses #1 and #2, this is not a testable benchtop experiment. It is a connecting argument across documented facts — each element independently established in peer-reviewed literature, the connection proposed here. The test would be epidemiological: re-analyze the existing acetaminophen-autism cohort studies controlling for maternal PFAS serum levels, and study vaccine antibody response by PFAS level, and re-run sibling analyses controlling for each sibling's individual gestational PFAS exposure rather than assuming equivalent within-family exposure. The data exists. The analysis has not been done.

Key Supporting Evidence (Peer-Reviewed)

  • Ahlqvist VH, Sjöqvist H, Dalman C, et al. "Acetaminophen Use During Pregnancy and Children's Risk of Autism, ADHD, and Intellectual Disability." JAMA 2024;331(14):1205–1214. DOI: 10.1001/jama.2024.3172. [Sibling analysis: autism HR 0.98, ADHD HR 0.98, intellectual disability HR 1.01 — association absent when family-level confounders controlled]
  • Hviid A, Hansen JV, Frisch M, Melbye M. "Measles, Mumps, Rubella Vaccination and Autism: A Nationwide Cohort Study." Annals of Internal Medicine 170(8):513–520 (2019). DOI: 10.7326/M18-2101. [657,461 Danish children — no association between MMR and autism]
  • Grandjean P et al. "Serum Vaccine Antibody Concentrations in Children Exposed to Perfluorinated Compounds." JAMA 308(2):185–192 (2012). DOI: 10.1001/jama.2012.7598. [PFAS reduces vaccine antibody titers — documented immune disruption]
  • Baluja R, Guo B, Howden W, Langer A, Lemoine D. "PFAS-contaminated drinking water harms infants." Proc Natl Acad Sci U S A. 2025 Dec 16;122(50):e2509801122. DOI: 10.1073/pnas.2509801122. PMID: 41359852. [191% increased first-year infant mortality (95% CI: 83–298%), NH births 2010–2019]
  • Swedish SELMA cohort (Wikström et al., 2019): PFOS OR 2.68 for preeclampsia — the condition driving Tylenol prescriptions in pregnancy.
  • PFAS thyroid/neuroinflammation mechanism and BH4 pathway disruption: documented across AbilityForge PFAS Affected, Steps 5–6.

Additional Hypotheses

This page will receive additional entries as patterns emerge from the evidence. Each will be date-stamped at first documentation.