The European Food Safety Authority's 2020 Tolerable Weekly Intake for PFAS is 4.4 nanograms per kilogram of body weight per week, defined as the sum of PFOA + PFOS + PFNA + PFHxS, with the critical effect being reduced antibody response to childhood vaccination at age one. The US EPA does not publish a directly comparable intake figure, but its April 2024 Maximum Contaminant Levels of 4.0 ng/L for PFOA and PFOS in drinking water are the enforceable backstop in the United States, and pregnancy is the life stage where the gap between these two frameworks matters most.
If you are pregnant, nursing, or planning a pregnancy and you live on a public water system in the United States, the most defensible reference value is EFSA's 4.4 ng/kg bw/week TWI for the sum of four PFAS, and the practical action it implies is a point-of-use filter certified to NSF/ANSI 53 with the P473 protocol or NSF/ANSI 58 reverse osmosis if your finished water shows any quantifiable PFAS. This article walks through how that number was derived, why the EPA's framework looks different, what the C8 Science Panel and Project Viva cohorts found in pregnancy specifically, and a triage approach for households on a hotspot utility. It is educational; if you are pregnant or breastfeeding, take any specific concerns to your obstetrician, midwife, or pediatrician, and to your state's PFAS hotline if your utility is on a known impact list.
PFAS are not metabolised in the human body, the long-chain compounds are eliminated only by slow renal clearance and, in lactating women, by transfer into breast milk. The Agency for Toxic Substances and Disease Registry's 2021 Toxicological Profile for Perfluoroalkyls puts the geometric mean serum half-life at roughly 3.8 years for PFOA and 5.4 years for PFOS in general adult populations. That timescale is the central fact behind the pregnancy concern: a body burden accumulated over a decade of low-dose exposure is the burden that will be present during gestation and lactation, irrespective of what the woman drinks during the nine months of pregnancy itself.
Three exposure routes specific to the perinatal window matter. First, placental transfer is well documented, cord blood PFOA and PFOS concentrations in observational cohorts run at roughly half to three-quarters of paired maternal serum levels, and the shorter-chain PFHxS and PFNA cross at similar or higher ratios. Second, lactational transfer concentrates several PFAS into breast milk; ATSDR cites breast-milk-to-serum ratios in the 0.01–0.05 range, which is small as a fraction but is the dominant exposure route for an exclusively breastfed infant whose only "diet" is that milk. Third, fetal and infant kidneys have lower glomerular filtration than the adult kidney, so the same dose produces a longer internal residence time. The net effect is that the infant exposure window begins at conception and extends through weaning.
None of this means breastfeeding should be discouraged, the consensus across ATSDR, WHO, and every paediatric authority is that the benefits of breastfeeding outweigh the PFAS exposure for the vast majority of women. It does mean the question "how much PFAS is in the household water" matters more for a pregnant or lactating woman than for an otherwise comparable adult. The background toxicology for the parent compounds themselves is in PFOA and PFOS toxicology, half-lives, endpoints, dose-response.
The European Food Safety Authority's CONTAM Panel published its current PFAS opinion in September 2020 ("Risk to human health related to the presence of perfluoroalkyl substances in food," EFSA Journal 2020;18(9):6223). The headline output is a single Tolerable Weekly Intake of 4.4 nanograms per kilogram of body weight per week, expressed as the sum of four substances: PFOA, PFOS, PFNA, and PFHxS. EFSA chose a sum-of-four because the four compounds share mechanism and cumulate in serum, and because expressing the TWI as a single grouped value sidesteps the otherwise impossible problem of compound-by-compound dose attribution from a mixed-exposure diet.
The critical endpoint EFSA selected was reduced antibody response to childhood vaccination, observed in one-year-old children in the Faroe Islands birth cohort followed by Grandjean and colleagues. The Faroese population is highly exposed to PFAS via traditional pilot-whale consumption, and the cohort has been followed since the late 1990s. EFSA used a benchmark dose modelling approach on the maternal-serum / child-antibody-titre data to derive a BMDL10 (the lower 95% confidence bound on a dose producing a 10% response) and worked back through a PBPK model to the weekly intake that would produce a steady-state serum concentration at that benchmark.
The arithmetic for a household is easier than the derivation. For a 70 kg adult, 4.4 ng/kg/week × 70 kg = 308 ng/week, or about 44 ng of summed PFAS per day from all sources combined, food, drinking water, indoor dust, consumer-product contact. Drinking water at the new US MCL of 4.0 ng/L for PFOA alone would, at the standard 2 L/day intake assumption, contribute 8 ng/day from PFOA, roughly a fifth of the weekly tolerable intake budget for the whole sum-of-four, from a single compound in a single exposure route. That is the back-of-envelope arithmetic behind why European regulators and several US state programs treat 4.0 ng/L as a ceiling rather than a target.
EFSA's TWI sits two orders of magnitude below the older 2008 EFSA TDIs (1500 ng/kg/day for PFOS and 150 ng/kg/day for PFOA) that it superseded. The order-of-magnitude shift came from new exposure data and from the choice of immune effect over the previously-used liver effect as the critical endpoint, vaccine-antibody suppression manifests at substantially lower serum concentrations than the liver and developmental endpoints that drove the earlier values.
The US framework is structured differently. The EPA does not publish a Tolerable Weekly Intake in the EFSA sense; it publishes Maximum Contaminant Levels (MCLs) for drinking water and, for substances that lack an MCL, non-enforceable Health Advisories. In June 2022 the EPA issued interim updated lifetime Health Advisories of 0.004 ng/L for PFOA and 0.02 ng/L for PFOS, replacing the 2016 70 ng/L combined advisory. The 2022 numbers were strikingly low, below the analytical method detection limit for EPA Method 537.1, meaning no commercial laboratory could actually verify a sample met the advisory. They were derived from updated epidemiology and from incorporating the EFSA immune endpoint into the EPA's own reference-dose calculation.
Because the 2022 values sat below the analytical floor, they were not workable as enforcement thresholds. On April 10 2024 the EPA signed the first federally-enforceable PFAS National Primary Drinking Water Regulation, published April 26 2024 at 89 Fed. Reg. 32532 and codified at 40 CFR §141.61(c). The final MCLs are 4.0 ng/L for PFOA and 4.0 ng/L for PFOS, with 10 ng/L each for PFHxS, PFNA, and HFPO-DA, and a unitless Hazard Index of 1 for the four-compound mixture. The 2024 MCL Goals (the aspirational targets, non-enforceable) for PFOA and PFOS remain at zero, reflecting the EPA's classification of both as likely human carcinogens with no demonstrated safe threshold.
The 2022 interim Health Advisories were withdrawn and superseded for enforcement purposes when the 2024 NPDWR took effect. They are no longer the operative US guidance; the 4.0 ng/L MCLs are. Compliance monitoring is required from 2027 and the MCLs are enforceable from 2029. The full mechanics of the rule are covered in EPA's April 2024 PFAS MCL, the 6 regulated compounds, and the UCMR 5 dataset that underpinned it is in UCMR 5 explained, what every utility had to report.
The C8 Science Panel was the independent epidemiological body convened under the 2005 class-action settlement with DuPont in the Mid-Ohio Valley, with access to serum PFOA measurements and health records for roughly 69 000 residents exposed to drinking water contaminated by the Washington Works plant. Between 2008 and 2012 the Panel issued a series of "Probable Link" reports defining the conditions for which they considered a causal relationship between PFOA exposure and the disease more likely than not in this exposed population.
Six probable links were identified: kidney cancer, testicular cancer, ulcerative colitis, thyroid disease, hypercholesterolaemia, and pregnancy-induced hypertension (the category that includes preeclampsia). The pregnancy-induced hypertension finding was based on a nested analysis of women in the C8 cohort with documented pregnancies and serum PFOA measurements, and is the most directly pregnancy-specific finding in the entire PFAS epidemiological record. The Panel's full methodology and the underlying analyses are summarised in the C8 Science Panel, what 70 000 exposed residents revealed.
The C8 finding does not establish that any individual exposure causes preeclampsia, nor does it quantify a dose-response that can be applied directly to a household on a different utility. It does establish biological plausibility for a vascular / endothelial effect during pregnancy, and it is one of the standing reasons that obstetric and environmental-health authorities treat known PFAS exposure as a flag worth recording in the prenatal history.
The other pregnancy-relevant outcome in the PFAS literature is reduced birth weight. The most-cited US analysis is from Project Viva, the longitudinal birth cohort run out of Harvard Pilgrim Health Care Institute. Sagiv et al. (2018, American Journal of Epidemiology) reported that higher maternal plasma PFOA and PFOS concentrations measured in early pregnancy were associated with modest reductions in birth weight, on the order of 50–100 grams per inter-quartile-range increase in exposure, with the effect more pronounced in girls.
The effect sizes are small at the individual level, well below the 500-gram threshold typically used to define "low birth weight" in clinical practice, and the cohort exposures were predominantly background, not hotspot-utility, exposures. The signal is consistent across multiple cohorts (Danish National Birth Cohort, MIREC in Canada, HOME Study in Cincinnati), which is why the EPA's hazard assessments now treat reduced fetal growth as a co-critical endpoint alongside the immune effect, and why several state programs include developmental outcomes in their reference-value derivations.
For a household, the take-away is that the literature does not suggest large individual-level risk from background exposures, but it does suggest that reducing avoidable exposure during pregnancy is sensible, and drinking water is the most controllable single route once a household has tested or located the utility's UCMR 5 result.
This table is educational, not medical advice. The categories are descriptive based on UCMR 5 / state monitoring results and the April 2024 MCLs; specific health decisions belong with your obstetrician, midwife, or pediatrician, and your state's PFAS hotline if your utility is on a known impact list.
| Finished-water status | Risk framing | Filter posture | Clinical conversation |
|---|---|---|---|
| Non-detect across UCMR 5 and state sampling | Background risk only | Optional; carbon-block POU certified to NSF/ANSI 53 P473 is reasonable | No PFAS-specific action; standard prenatal care |
| Detectable but below all individual MCLs and HI < 1 | Compliant but quantifiable | POU recommended during pregnancy and lactation; NSF 53 P473 or NSF 58 RO | Mention exposure in prenatal history; no specific intervention |
| At or above any MCL, or HI > 1 | Above the enforceable federal limit (effective 2029; non-compliant today is still above the rule) | NSF/ANSI 58 reverse osmosis at the kitchen tap, used for all drinking and cooking water | Discuss with obstetrician and pediatrician; consider serum PFAS testing per ATSDR clinician guide; document in chart |
| Known hotspot utility (Cape Fear / Decatur / Merrimack / etc.) | Documented elevated exposure on file | NSF/ANSI 58 RO from before conception if possible; verify filter cartridge change schedule | Active obstetric and pediatric awareness; state PFAS hotline; serum testing per ATSDR guide |
The filter selection logic, why NSF/ANSI 53 with the P473 protocol or NSF/ANSI 58 reverse osmosis, and not generic NSF 42, is covered in reverse osmosis vs a pitcher filter for PFAS. For confirming what is in the tap before relying on the utility's own sampling, see home PFAS testing labs that use EPA Method 537.1.
The Agency for Toxic Substances and Disease Registry maintains a Clinician's Guide on per- and polyfluoroalkyl substances aimed at primary-care, obstetric, and paediatric practitioners. The guide is not a treatment protocol, there is no clinical intervention that lowers serum PFAS appreciably faster than the natural elimination half-life, and routine serum PFAS testing is not recommended for general populations. What ATSDR does recommend is a structured exposure history when the patient is from a community with documented PFAS-contaminated drinking water, occupational exposure (firefighters and AFFF-handling personnel, fluorochemical manufacturing workers), or known consumer-product exposure (heavy use of stain-resistant treatments, food packaging, certain personal-care products).
For pregnant and lactating women on a hotspot utility, the guide's principal recommendations are: take an exposure history including drinking-water source and duration of residence; document the exposure in the chart so it is available to the paediatrician; counsel on reducing further exposure through a certified point-of-use filter; and continue to recommend breastfeeding given the established benefits. Serum PFAS testing is suggested when the result will change clinical management, typically meaning the patient is in a documented exposed cohort, where the result enters the medical record and can inform paediatric monitoring (for example, watching for sub-optimal vaccine response or thyroid function). The guide explicitly counsels against routine population-level testing.
The National Academies of Sciences, Engineering, and Medicine 2022 consensus report on PFAS clinical guidance is the companion document and tracks the same conclusions: above a summed-serum threshold of 2 ng/mL, clinicians should counsel on exposure reduction; above 20 ng/mL, additional clinical follow-up for the C8-Panel-listed outcomes is appropriate. Neither document positions PFAS as something to test routinely in a non-exposed pregnant patient, the focus is on the documented-exposure cohort and on reducing forward exposure.
One ZIP, one PDF brief. UCMR 5 results for your state, hotspot utility match if applicable, and the filter that actually removes PFAS from your tap. Tick the pregnancy / infant box on the order form and the report adds an EFSA TWI + ATSDR clinician-guide callout calibrated to your utility's finished-water result.
Pull my report, $15