Perfluoroalkylated substances (PFAS) is the collective name for a vast group of fluorinated compounds, including oligomers and polymers, which consist of neutral and anionic surface active compounds with high thermal, chemical and biological inertness. Perfluorinated compounds are generally hydrophobic but also lipophobic and will therefore not accumulate in fatty tissues as is usually the case with other persistent halogenated compounds. An important subset is the (per)fluorinated organic surfactants, to which perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA) belong.
The analytical detection method of choice for PFOS and PFOA is currently liquid chromatography-mass spectrometry/mass spectrometry (LC-MS/MS), whereas both LC-MS/MS and gas chromatography-mass spectrometry (GC-MS) can be used for the determination of precursors of PFOS and PFOA. There are few reports of analysis of food items using these methods. Due to the substantial lack of suitable analytical data, many assumptions have been made in order to derive exposure estimates. Therefore, figures on levels in food and exposure provided in this opinion should be taken as indicative.
PFOS, PFOA and other perfluorinated organic compounds have been widely used in industrial and consumer applications including stain- and water-resistant coatings for fabrics and carpets, oil-resistant coatings for paper products approved for food contact, fire-fighting foams, mining and oil well surfactants, floor polishes, and insecticide formulations. A number of different perfluorinated organic compounds have been widely found in the environment.
PFOS has been analysed in a limited number of European environmental and food samples (mainly fish). The PFOS concentrations are almost invariably higher than PFOA concentrations and the PFOS concentrations in fish liver are consistently higher than those in fillet. PFOS has been shown to bioaccumulate in fish and a kinetic bioconcentration factor has been estimated to be in the range 1000 – 4000. The time to reach 50% clearance in fish has been estimated to be around 100 days.
Fish seems to be an important source of human exposure to PFOS, although the data might be influenced by results of studies in relatively polluted areas, which is likely to over-estimate exposure from commonly consumed fish. There are very few data, especially for Europe, that can serve as reliable indicators of the relative importance of most other kinds of food. Drinking water is estimated to contribute less than 0.5% of the indicative exposure. The importance of fish is, however, not supported by all studies, indicating other important sources of human exposure might exist which have not yet been identified. It is possible that additional exposure to PFOS could result from precursors and other sources.
Such possible sources could be related to food (e.g. via packaging material or cookware) or be a result of more direct exposure from the technosphere (e.g. household dust). Based primarily on the available data for fish and fishery products, indicative estimates of dietary exposure to PFOS were 60 ng/kg body weight (b.w.) per day for average consumers, and 200 ng/kg b.w. per day for high consumers of fish. In contrast, recent studies have indicated much lower exposures, demonstrating the uncertainty in the assessments. The importance of possible pathways of non-food human exposure to PFOS has been estimated to decrease when moving from childhood into adulthood. The total contribution from non-food articles was estimated to be less than 2% compared to the average total PFOS exposure. In individuals with high fish consumption, the percentage contribution from non-food exposure is expected to be lower.
Following absorption, PFOS is slowly eliminated and therefore accumulates in the body. PFOS shows moderate acute toxicity. In subacute and chronic studies the liver was the major target organ and also developmental toxicity was seen. Other sensitive effects were changes in thyroid hormones and high density lipoprotein (HDL) levels in rats and Cynomolgus monkeys. PFOS induced liver tumours in rats, which appears to be due to a non-genotoxic mode of action.
Epidemiological studies in PFOS exposed workers have not shown convincing evidence of increased cancer risk. An increase in serum T3 and triglyceride levels was observed, which is the opposite direction to the findings in rodents and monkeys. The very few epidemiological data available for the general population do not indicate a risk of reduced birth weight or gestational age.
From a subchronic study in Cynomolgus monkeys, the Scientific Panel on Contaminants in the Food Chain (CONTAM) identified 0.03 mg/kg b.w. per day as the lowest no-observed-adverse-effect level (NOAEL) and considered this a suitable basis for deriving a Tolerable Daily Intake (TDI). The CONTAM Panel established a TDI for PFOS of 150 ng/kg b.w. per day by applying an overall uncertainty factor (UF) of 200 to the NOAEL. An UF of 100 was used for inter and intra-species differences and an additional UF of 2 to compensate for uncertainties in connection to the relatively short duration of the key study and the internal dose kinetics.
The CONTAM Panel noted that the indicative dietary exposure of 60 ng/kg b.w. per day is below the TDI of 150 ng/kg b.w. but that the highest exposed people within the general population might slightly exceed this TDI.
The CONTAM Panel recognised that a significant part of the body burden could result from exposure to other sources and also from precursors that could be transformed into PFOS in the body. However, there was no reliable information on body burdens in humans, and therefore the Panel decided to compare blood levels in humans and animals recognising the uncertainties in attainment of steady-state conditions. The margin between serum levels in the monkeys at the NOAEL and the serum levels in the general population was between 200 and 3,000. Given this margin, the Panel considered it unlikely that adverse effects of PFOS are occurring in the general population.
PFOA has been analysed in a limited number of European environmental and food samples (mainly fish) and concentrations are almost invariably lower than PFOS concentrations. PFOA has been shown to bioaccumulate in fish but probably less than PFOS. The importance of possible pathways of non-food human exposure to PFOA has been estimated to decrease when moving from childhood into adulthood. For PFOA, the total contribution from the non-food sources, mainly indoor exposure, could be as high as 50% compared to the estimated average dietary exposure to PFOA.
Fish seems to be an important source of human exposure to PFOA, although the data might be influenced by results of studies in relatively polluted areas, which is likely to over-estimate exposure from commonly consumed fish. There are very few data, especially for Europe, that can serve as reliable indicators of the relative importance of most other kinds of food. Drinking water is estimated to contribute less than 16% to the indicative exposure. Based on the limited data, the CONTAM Panel identified the indicative average and high level dietary exposures of 2 and 6 ng/kg b.w. per day, respectively. Persons with higher fish consumption do not always show higher levels of PFOA in blood compared to persons with “normal” fish consumption. It is possible that additional exposure to PFOA could result from non food sources and precursors.
PFOA is readily absorbed. Elimination is dependent on active transport mechanisms which vary between different species, and between sexes in some species. PFOA shows moderate acute toxicity. In sub acute and chronic studies, PFOA affected primarily the liver and can cause developmental and reproductive toxic effects at relatively low dose levels in experimental animals. It increased the tumour incidence in rats, mainly in the liver. Based on the weight of evidence at present, the carcinogenic effects in rats appear to be due to indirect/non-genotoxic modes of action.
Epidemiological studies in PFOA-exposed workers do not indicate an increased cancer risk. Some have shown associations with elevated cholesterol and triglycerides, or with changes in thyroid hormones, but overall there is no consistent pattern of changes. In two recent studies, PFOA exposure of pregnant women, measured by maternal and/or cord serum levels was associated with reduced birth weight. The Panel noted that these observations could be due to chance, or to factors other than PFOA.
The lowest NOAEL identified of 0.06 mg/kg per day, originated from a subchronic study in male rats, whereas results from long-term studies indicated higher NOAELs for effects on the liver. The Panel noted that the 95% lower confidence limit of the benchmark dose for a 10% increase in effects on the liver (BMDL10) values from a number of studies in mice and male rats were in the region of 0.3 - 0.7 mg/kg b.w. per day. Therefore, the CONTAM Panel concluded that the lowest BMDL10 of 0.3 mg/kg b.w. per day was an appropriate point of departure for deriving a TDI. The CONTAM Panel established a TDI for PFOA of 1.5 µg/kg b.w. per day by applying an overall UF of 200 to the BMDL10. An UF of 100 was used for inter- and intra-species differences and an additional UF of 2 to compensate for uncertainties relating to the internal dose kinetics.
The CONTAM Panel noted that the indicative human average and high level dietary exposure for PFOA of 2 and 6 ng/kg b.w. per day, respectively, are well below the TDI of 1.5 µg/kg b.w. per day.
The serum levels in rats at the BMDL10 are expected to be in the region of three orders of magnitude higher than in serum levels of PFOA from European citizens who do not have occupational exposure. Given this margin, the CONTAM Panel considered it unlikely that adverse effects of PFOA are occurring in the general population, but noted uncertainties with regards to developmental effects.
Finally the CONTAM Panel recommended that further data on PFAS levels in food and in humans would be desirable, particularly with respect to monitoring trends in exposure.