How does PFAS affect the body? Assessments by different bodies, and what a blood test can tell you
IARC classified PFOA as “carcinogenic to humans,” while the FSCJ concluded that “the evidence is not certain enough to clearly determine carcinogenicity in humans.” Using official documents and papers, we set out why assessments of the same substance split, the trend in blood levels in Japanese people, and what a blood test can and can’t tell you.
Short answer
Assessments differ by body. IARC (2023) classified PFOA as Group 1 (carcinogenic to humans) and PFOS as Group 2B. The FSCJ (2024) rated only the effects on the offspring of animals as having “strong evidence,” said carcinogenicity “may be due to a mechanism specific to rodents,” and set the TDI at 20 ng/kg body weight/day each.
Median blood levels in Japanese people in Ministry of the Environment surveys were PFOS 11 → 2.9 ng/mL and PFOA 3.9 → 2.0 ng/mL (FY2008–FY2010 → FY2024). However, the survey method changed from FY2020, and the report says simple comparison is difficult.
The Ministry of the Environment says “it is not clear what health effects occur in individuals at what blood levels.” Japan has no criteria for judging individual health effects from blood test values. It also says no cases have been confirmed in Japan of individual health damage mainly attributable to PFOS or PFOA.
Assessments by different bodies: why they differ for the same substance
| Body, year | Assessment | Benchmark value |
|---|---|---|
| IARC (International Agency for Research on Cancer), November 2023 | PFOA is Group 1 (carcinogenic to humans). Evidence “sufficient” in animals, “strong” mechanistic evidence in exposed humans, “limited” for cancer in humans (renal cell carcinoma and testicular cancer). PFOS is Group 2B (possibly carcinogenic) | None (the classification reflects the strength of the evidence and does not indicate the potency of carcinogenicity or the amount involved) |
| FSCJ, June 2024 | Liver, lipids, thyroid, immune system, nervous system, genotoxicity and carcinogenicity: evidence “insufficient / limited.” Only effects on the offspring of animals had “strong certainty of evidence.” On cancer: “possibly due to a mechanism specific to rodents, and difficult to apply to humans” | TDI 20 ng/kg body weight/day each (none set for PFHxS) |
| EFSA (European Food Safety Authority), 2020 | The most critical effect is on the immune system. Calculated from reduced vaccine antibodies in 1-year-olds. Some people in Europe exceed this, which is a concern | 4.4 ng/kg body weight per week for the sum of PFOA, PFNA, PFHxS and PFOS (about 0.63 per day) |
| U.S. EPA, 2024 | PFOA and PFOS are both “likely to be carcinogenic to humans.” Priority non-cancer effects: immune, cardiovascular, liver and developmental | Reference dose for PFOA: 0.03 ng/kg body weight/day (antibodies in children) |
| U.S. National Academies (NASEM), 2022 | Evidence of an association “sufficient”: reduced antibody response, dyslipidemia, reduced fetal and infant growth, kidney cancer. “Limited/suggestive”: breast cancer, liver enzymes, pregnancy-induced hypertension, testicular cancer, thyroid disease, ulcerative colitis | Clinical guidance based on blood levels (section below) |
- Japan’s TDI (20) and the U.S. EPA’s reference dose for PFOA (0.03) differ by about 670 times. The FSCJ’s assessment report notes that benchmark values from overseas bodies vary by about 600 times for PFOS and about 5,300 times for PFOA.
- A big reason for the difference is which effect is chosen as the basis. Japan chose effects on the offspring of animals; EFSA and the U.S. EPA chose reduced vaccine antibodies in human children. On reduced antibodies, the FSCJ says “the possibility cannot be ruled out, but there are issues with the quality and sufficiency of the evidence.”
- The FSCJ also notes that blood levels in animal experiments are more than 1,000 times higher than in humans.
The C8 Health Project and recent research (mind the funding and positions)
The C8 Health Project (2005–2013), which came out of the settlement of a class action over PFOA contamination of tap water around a plant in West Virginia and Ohio, found a “probable link” (a link is more likely than not) for six conditions: pregnancy-induced hypertension, testicular cancer, kidney cancer, thyroid disease, ulcerative colitis and high cholesterol. A 2020 review by former panel members said the evidence for kidney and testicular cancer was “supportive but not definitive,” and that the association with cholesterol was consistent but there was no association with heart disease.
| Study | Type | Results | Funding and positions |
|---|---|---|---|
| Li S et al. 2025 | Ecological study (U.S. counties) | PFAS in tap water associated with incidence of digestive, endocrine and other cancers (incidence rate ratios 1.02–1.33) | One co-author is an expert for plaintiffs. Individual exposure not measured |
| Chang CJ et al. 2024 | Meta-analysis | No support for an association with breast cancer overall (PFOA RR 0.95, PFOS 0.98) | No conflicts of interest |
| Spyrakis F et al. 2025 | Meta-analysis (3 studies) | No significant association between serum PFOA and kidney cancer (mRR 1.05) | Authors received research funding from and served as advisers to fluorochemical manufacturers |
| Hussey MR et al. 2025 | Systematic review and meta-analysis | Pooled, PFOA and PFOS were positively associated with total cholesterol and LDL. When limited to longitudinal studies, the association tends to disappear | Two authors disclosed being retained by companies as PFAS toxicity experts |
| Antoniou EE and Dekant W 2024 | Meta-analysis (immunity in children) | Infections “suggestive”; reduced vaccine antibodies “moderate to none” | Authors received payment from 3M |
- Studies on both sides, those strongly showing associations and those denying them, include some with biases in funding or in their position in litigation. Urayomi states the type of study (ecological, cross-sectional, longitudinal) together with its funding.
- In a Japanese birth cohort (the Japan Environment and Children’s Study, 23,058 children), higher PFAS in mothers’ plasma was reported to be associated with a lower height growth trajectory up to age 4 (Yamamoto et al. 2025).
How many years it takes to halve in the body
| Source | PFOS | PFOA | PFHxS |
|---|---|---|---|
| Data compiled by EFSA (Ministry of the Environment Q10) | Average 5.7 years | Average 3.2 years | – |
| Ronneby, Sweden (after supply of contaminated water stopped, 106 people) | 3.4 years | 2.7 years | 5.3 years |
| Tama, Tokyo (after the 2019 water source switch, 17 women; estimate after subtracting Japan’s average level) | 2.7 years | 5.1 years | 5.6 years |
- The Tama values are estimates after subtracting Japan’s average blood level (background). Without subtracting it, they were 3.9 years for PFOS, 8.0 years for PFOA and 5.7 years for PFHxS.
- PFOS and PFOA stay in the body for a long time. Even when water values fall, blood levels take years to come down.
- In the U.S., serum PFOS fell by more than 85% and PFOA by more than 70% over about 20 years from 1999 (Ministry of the Environment Q10).
Trend in blood levels in Japanese people
Estimated plasma values (median, linear isomers only, ng/mL) from the Ministry of the Environment’s “Monitoring Survey of Human Exposure to Chemical Substances.”
| Survey fiscal year | People | PFOS | PFOA | PFHxS |
|---|---|---|---|---|
| 2008–2010 | 609 | 11 | 3.9 | – |
| 2011, 2013–2016 | 406 | 6.4 | 3.3 | 0.66 |
| 2020 | 80 | 2.2 | 1.4 | 0.38 |
| 2021 | 119 | 3.4 | 2.0 | 0.76 |
| 2022 | 89 | 3.2 | 1.9 | 0.51 |
| 2023 | 108 | 3.3 | 1.8 | 0.44 |
| 2024 | 117 | 2.9 | 2.0 | 0.72 |
- From FY2020 on, these are pilot surveys, and how participants were selected and their ages are not consistent. The report says “comparison with results from previous years or with other surveys is difficult.” The prefectures where the surveys were done are not disclosed.
- Some areas where contamination became an issue have high values. Among 17 women in the Tama area of Tokyo (2023), medians were PFOS 8.4, PFHxS 14.7 and PFOA 4.1 ng/mL (Lyu et al. 2026). The participants and methods differ, so these can’t simply be compared with the table above.
- ng/mL in blood is a unit 1,000 times larger than ng/L in water. Be careful not to confuse them.
What a blood test can and can’t tell you
The U.S. National Academies (NASEM, 2022) issued guidance for clinicians based on the sum of seven PFAS in serum.
| Sum of 7 | NASEM guidance |
|---|---|
| Below 2 ng/mL | No adverse effects expected |
| 2–20 ng/mL | Potential for effects, especially in sensitive people. Advice on reducing exposure, screening for dyslipidemia, etc. |
| 20 ng/mL or more | Increased risk. Thyroid function testing, examination for kidney and testicular cancer, etc. |
- These are guides for additional testing and advice, not a line where “above 20 means you will get sick.”
- The Ministry of the Environment says “it is not clear what health effects occur in individuals at what blood levels,” and explains that foreign blood-level benchmarks “do not mean they cause health problems in each individual, but are mainly for understanding the situation as a population.” Japan has no criteria for judging individual health effects from blood test values.
- What a blood test shows is “a rough idea of how much has entered your body so far.” Because it takes years to fall, it does not match the current water value.
Health damage in Japan, and everyday life
- The Ministry of the Environment says “no cases have been confirmed in Japan of individual health damage in which intake of PFOS or PFOA is considered the main cause.” Okinawa Prefecture, Kibichuo Town (吉備中央町) and Settsu City (摂津市) compared local health indicators using existing statistics, and reportedly found no differences.
- The FSCJ says the levels of PFOS and PFOA taken in through a normal diet (including drinking water) are not causing significant health effects, and that changing your diet out of excessive worry carries other risks, such as nutritional excesses or deficiencies.
- What happened in areas with high values is covered in the guide to areas with high values, and your local value in the guide to checking.
Sources
- Food Safety Commission of Japan, “Q&A on the assessment report on organic fluorine compounds (PFAS)”
- Ministry of the Environment, FAQ on PFAS (Q9–Q11)
- IARC Monographs Volume 135 Q&A (PFOA and PFOS, 2023)
- EFSA 2020 Risk to human health related to the presence of PFAS in food（PMID 32994824）
- NASEM 2022 Guidance on PFAS Exposure, Testing, and Clinical Follow-Up
- Ministry of the Environment, Monitoring Survey of Human Exposure to Chemical Substances (FY2018–FY2024)
- Steenland K et al. 2020, review of the C8 Health Project and subsequent research (Environ Int)（PMID 32950793）
- Lyu et al. 2026, serum PFAS in residents of the Tama area, Tokyo (Environ Health Prev Med)（PMID 41548900）
Last checked 2026.10.04 (against the original government documents and papers, collected and cross-checked in Gemini Notebook). Utility values are as in the national table; Urayomi sorted and summarized them.