Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

An assessment of the exposure of Americans to perfluorooctane sulfonate: A comparison of estimated intake with values inferred from NHANES data


To better understand human exposure to perfluorinated compounds (PFCs), a model that assesses exposure to perfluorooctane sulfonate (PFOS) and its precursors from both an intake and a body burden perspective and combines the two with a simple pharmacokinetic (PK) model is demonstrated. Exposure pathways were modeled under “typical” and “contaminated” scenarios, for young children and adults. A range of intakes was also estimated from serum concentrations of PFOS reported in the National Health and Nutrition Examination Survey (NHANES) using a first-order 1-compartment PK model. Total PFOS intakes (medians summed over all pathways) were estimated as: 160 and 2200 ng/day for adults and 50 and 640 ng/day for children under typical and contaminated scenarios, respectively. Food ingestion appears to be the primary route of exposure in the general population. For children, the contribution from dust ingestion is nearly as great as from food ingestion. Pathway-specific contributions span several orders of magnitude and exhibit considerable overlap. PK modeling suggests central tendency PFOS intakes for adults range between 1.6 and 24.2 ng/kg-bw/day, and the forward-based intake estimates are within this range. The favorable comparison reported between the forward-modeled and the back-calculated range of intake predictions lends validity to the proposed framework.


Although completely anthropogenic, perfluorinated compounds (PFCs) are found ubiquitously in marine and terrestrial animals (Giesy and Kannan, 2001). Animal toxicity studies have shown reproductive, developmental, and immune effects (Lau et al., 2007). The most widely known PFCs are the C8-chemicals: perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA), which have been found to persist in the environment and are not currently known to degrade by biotic or abiotic means (EPA, 2002a, 2002b). Because PFOA and PFOS themselves are relatively non-volatile, their occurrence in remote regions is believed to be the result of either translocation by oceanic transport and rain events or atmospheric transport of the relatively more volatile precursors (e.g., 8–2 fluorotelomer alcohol (FTOH) and N-ethyl perfluorooctane sulfonamidoethanol (EtFOSE)) that are later transformed (including metabolically by animals and microbes) into the terminal compounds (3M, 1999; Martin et al., 2002; Ellis et al., 2004; Stock et al., 2004; Nabb et al., 2007). Potential precursors of PFOA and other perfluorocarboxylates are commonly thought to include the FTOHs. Potential precursors of PFOS are commonly thought to include perfluoralkyl sulfonamides and sulfonamidoethanols (FOSAs/FOSEs) such as MeFOSA (N-methyl perfluorooctane sulfonamide), EtFOSA (N-ethyl perfluorooctane sulfonamide), MeFOSE (N-methyl perfluorooctane sulfonamidoethanol), and EtFOSE (Shoeib et al., 2005; Nabb et al., 2007).

With useful surfactant properties, PFCs are used extensively in the engineering and chemical, electronics, and medical industries, and production has increased substantially in the past few years (Lewandowski et al., 2006). Potential sources of exposure to PFOA and PFOS include direct industrial releases into air and water, intentional and unintentional releases of fire-fighting foam, long-range atmospheric transport and subsequent biotransformation and degradation of precursors, release from consumer products (including non-stick cookware, waterproof breathable textiles, electronics, and oil- and stain-protective coatings for carpets, apparel, and food containers), and release from degradation of telomer-based polymers (Giesy et al., 2006). PFOA is a surfactant used in the manufacturing of fluoropolymers and has been measured at low levels in finished consumer articles (Sinclair et al., 2007). Potential exposure media for PFOA, PFOS, and the precursors include air, dust, water, and food (possibly through migration from food packaging and cookware). Difficulties in measuring trace amounts of PFOS and other PFCs in the past have hindered the measurements in environmental media (Hansen et al., 2002). As a result, data on environmental concentrations in the United States, particularly in indoor environments, are sparse, and understanding the pathways of human exposure remains a challenge.

Geographically, proximity to industry appears to be an important determinant of exposure for some populations, with serum levels of PFOS and precursors higher in urban than in rural areas (Martin et al., 2002; Guruge et al., 2005), and serum levels of PFOS and PFOA higher in locations with known surface water and drinking water contamination (Harada et al., 2004; Emmett et al., 2006). Emmett et al. (2006) detected PFOA in public and private drinking water supplies near a Washington, WV fluoropolymer manufacturing facility and reported an average PFOA concentration of 3.55 μg/ml (range: 1.5–7.2 μg/ml) in the Little Hocking water distribution system from 2002 to 2005. This is about three orders of magnitude higher than what others have reported in drinking water (Sasaki et al., 2003; Hölzer et al., 2008). The authors reported mean serum PFOA levels of 775 ng/ml for those with known occupational exposures and 329 ng/ml for those in the community with no known occupational exposures. These levels were about 50 to 100 times higher than the mean serum PFOA levels (6 ng/ml) for a comparison group in Philadelphia. Those who reported drinking Little Hocking system water had mean levels about four times higher than in those who did not (320–374 vs 71–79 ng/ml, respectively), suggesting a clear drinking water ingestion exposure pathway. However, even those who did not report drinking the municipal water had a mean level more than 10 times higher than the mean of the reference population, suggesting other exposure pathways, including inhalation and dietary exposures (Emmett et al., 2006). Differences in patterns of PFC exposure between human and wildlife raise the possibility that consumption of fish and mammals may not be the major source of exposure in humans (Houde et al., 2006).

Few studies have attempted to quantify total human exposure for this class of compounds. The recent comprehensive evaluation of Trudel et al. (2008) sheds some light on the exposure of Europeans and North Americans to PFCs. They conducted an extensive exposure assessment considering pathways of exposures, exposure contact rates, and published exposure media concentrations for PFOS and PFOA to calculate intakes for Europeans and North Americans. They used a 1-compartment, first-order pharmacokinetic (PK) model for adult exposures only. They derived intake doses starting from measurements of PFOS and PFOA in the blood of adults, and compared those intakes to the exposure pathway-based intakes.

The most comprehensive description of current general population blood concentrations of PFCs in Americans is provided by Calafat et al. (2007). A total of 2094 serum samples from participants in the National Health and Nutrition Examination Survey (NHANES) 2003–2004 were measured for 11 PFCs. Four of these were detected in greater than 98% of the samples: PFOS, PFOA, perfluorohexane sulfonate, and perfluorononanoic acid (PFNA). Neither the 2003–2004 nor the 1999–2000 (Calafat et al., 2007) data suggest a difference among age groups. There appears, however, to be a clear sex difference for all compounds (males have higher concentrations than females), and a difference among ethnicities was also identified (Mexican Americans had the lowest concentrations of all PFCs in both NHANES examinations). Some other differences between non-Hispanic blacks and non-Hispanic whites were identified that were a function of age, but results between non-Hispanic blacks and non-Hispanic whites were fairly similar overall (Calafat et al., 2007).

The effort in our study mirrors that of Trudel et al. (2008), but it applies some different approaches for modeling intake and examines more of the available data specific to Americans. Also, this study carefully examines the areas of uncertainty in characterizing exposure of Americans to PFCs. The overall strategy for exposure evaluation used in this study, and also as studied by Trudel et al. (2008), is depicted in Figure 1. First, information on exposure is gathered on both ends of the study—on the occurrence of the contaminant in exposure media, and on the occurrence of the contaminant in human tissues. This information on media concentration levels is then combined with exposure contact information to estimate the exposure intake levels. Concurrently, toxicokinetic data are gathered to prepare for PK modeling. These lines of inquiry and study meet at the juncture of PK modeling. Intake quantities can be converted to body burdens with the use of such models in a “forward” modeling approach, and then predicted tissue concentrations can be compared with measured tissue concentrations. Alternatively, PK models are used to “back-calculate” intake quantities given tissue concentration levels, and these can then be compared with separately modeled intake quantities.

Figure 1

Generalized approach to assessing exposure that is demonstrated in this paper.

The U.S. Environmental Protection Agency (EPA) Office of Pollution Prevention and Toxics (OPPT) has been assessing PFCs since 1999, prompted by concerns over their potentially toxic properties and widespread presence in the environment. After negotiations between EPA and the only U.S. manufacturer of PFOS, the company voluntarily completed a phase out of PFOS chemistry in 2002, and EPA issued Significant New Use Rules in 2000 and 2002 to restrict the return of PFOS-related chemicals to the U.S. market (EPA, 2002a). Subsequent EPA efforts have focused on PFOA and fluorinated telomers, with EPA establishing an enforceable consent agreement negotiation process in 2003 for OPPT to obtain data on the sources of PFOA in the environment and the pathways leading to human and environmental exposures. The EPA's Office of Research and Development (ORD), in close collaboration with OPPT, has been conducting research in several areas, including telomer biodegradation, toxicology and PKs, analytical techniques development, and aged article analysis. The research presented in this article was initiated independently and used available data published in the peer-reviewed scientific literature. The authors recommend that as additional relevant data become publicly available, they be used to evaluate these research results.

In this study, we review exposure media and body burden data on PFCs, with an emphasis on published reports of exposures in the United States. Then, we demonstrate the approach of Figure 1 on one of the PFCs of interest, PFOS, as well as key precursors to PFOS, FOSEs/FOSAs. Initially, we describe pathways of exposure that would exist in three distinct scenarios characterized by differences in exposure media concentrations: (1) general background for PFOS, (2) an area highly impacted by PFOS as characterized by elevated concentrations in water to which individuals are exposed, and (3) general background for alkyl perfluorooctane sulfonyl precursors. We quantify estimates of exposures to an adult and a 2-year-old child in each scenario, and obtain pathway-specific as well as total intakes of PFOS and its precursors. Concurrently, we characterize adult background blood concentrations of PFOS, and using the PK model, we back-calculate a range of intake doses that might have resulted in this adult body burden. Finally, we compare the back-calculated range of intake predictions to a range of forward-modeled intakes for the adult in a background setting.

We examine the variability of possible intakes by incorporating the range of concentrations found in the intake estimates. We discuss the limitations and uncertainties of the exposure media data and discuss further the need to continue efforts to more fully quantify this variability. With regard to PK modeling, the issue is one more of uncertainty than variability, both in regard to the selection of the model used and the assignment of parameters for that model. We discuss issues associated with these uncertainties. Rather than answer all questions about exposure of Americans to PFCs, we hope to better frame the issue using the paradigm shown in Figure 1 and to identify gaps in our understanding of sources and pathways of exposure.


Compilation of Exposure and Biological Data

Available data on PFC concentrations in environmental exposure media and in human biological media were extracted from published, peer-reviewed journals and government agency reports, and to a very small extent from various public U.S. Federal Dockets related to PFCs. Although compiled, data from non-peer-reviewed sources were not included in the modeling. Although data were compiled and displayed for PFOS, PFOA, and other PFC compounds, discussions below focus only on PFOS.

PFOS Exposure Intake Estimates

Table 1 provides an overview of the methods used to calculate exposure including exposure factors and sources of assumed concentrations for PFOS and precursors MeFOSE, EtFOSE, MeFOSA, and EtFOSA. Aggregate exposures were estimated by a deterministic methodology consistent with EPA Guidelines for Exposure Assessment (EPA, 1992). Instead of using point estimates to represent potential exposures to all members of a population subgroup, distributions of media-specific concentrations culled from peer-reviewed scientific literature were used to more realistically represent the potential range of exposures along each pathway. Such analysis using distributions is more robust than that using central tendency alone because it allows examination of the entire range of environmental concentration data, comparison at various percentiles of interest (e.g., 95th percentile), and assessment of areas of overlap among pathways. Furthermore, it produces a more sophisticated assessment by allowing for the possibility that the “dominant pathway” may vary by individual within a sub-population.

Table 1 Data sources and assumptions used to estimate intakes.

The following procedure for generating distributions was used to characterize concentrations in dust, water, and air. Because environmental sampling data are typically lognormally distributed (Esmen and Hammad, 1977), only lognormal distributions of values were created. Mathematical relationships among the parameters of lognormal distributions, as summarized by Strom and Stansbury (2000), were applied to the available summary statistics reported in the scientific literature to estimate a geometric mean (GM) and a geometric standard deviation (GSD). When the only available parameters were a maximum value and either a median or mean, a distribution with a 99th percentile equal to the reported maximum was created. This extrapolation of distributions from summary statistics introduces modeling error of a small, but unknown, amount into the analysis. For concentrations of PFOS in dust, the actual measurement values reported in the Strynar and Lindstrom (2008) study were generously provided by those authors for use in this analysis. Point estimates of contact rate values were combined with the concentration distributions to model dust ingestion, dermal absorption of dust, water ingestion, and indoor and outdoor air inhalation.

However, a slightly different procedure was used for food ingestion. For this pathway, the final intakes of PFOS generated by Tittlemier et al. (2007) in their comprehensive assessment of the dietary exposure of Canadians to perfluorocarboxylates and PFOS was used. Specifically, a total of 54 composite food samples taken as part of the Canadian Total Dietary Survey were analyzed for PFCs. They combined concentrations with intake estimates of the foods to determine total dietary intakes. The GMs of PFOS dietary intakes for adults and children were taken from this reference, and then a distribution around each mean was created using GSDs in food concentration surveys as reported by Fromme et al. (2009) for German diets and by Mortimer et al. (2006) for English diets.

Route-specific PFOS and precursor intakes were estimated under three separate exposure scenarios. Note that both the 2-year-old child and the adult are modeled for each scenario. The three scenarios were as follows:

  1. 1)

    Intake of PFOS under typical exposure conditions: The pathways considered include ingestion of food, water, and dust; inhalation of indoor and outdoor air; and dermal absorption of surface residues. All available (as searched through the online PubMed database) published exposure media concentrations representative of background conditions in the United States were examined. Key data sources included Canadian food intakes from Tittlemier et al. (2007) and concentrations in dust from North Carolina and Ohio reported by Strynar and Lindstrom (2008). Outdoor air concentrations were based on values measured in Albany, NY and reported by Kim and Kannan (2007). Indoor air concentrations were assumed to be 20 times higher than outdoor; this assumption was based on data from Shoeib et al. (2005), who showed that average indoor FOSA concentrations were 13–24 times higher than average outdoor concentrations. Background drinking water concentrations were unavailable, but with evidence of poor water treatment removal efficiencies (Takagi et al., 2008), surface water concentrations were used as surrogates. Surface water concentrations from lakes and rivers in New York State (Sinclair et al., 2006) (Measurements from Lake Onondaga, a Superfund site impacted by several industries, were excluded.), Lakes Erie and Ontario (Boulanger et al., 2005), and rivers in the Cape Fear River Drainage Basin of North Carolina (Nakayama et al., 2007) were averaged. Recommended exposure contact rates for adults and for 1- to 2-year-old toddlers were obtained from EPA's Exposure Factors Handbook (EPA, 1997) and Child-Specific Exposure Factors Handbook (EPA, 2008), respectively.

  2. 2)

    Intake of PFOS in highly contaminated environments: This scenario was similar to the background scenario above with the exception that water concentrations were much higher than background. Water concentrations were derived from conditions observed in groundwater near a Minnesota perfluorochemical facility and a Michigan facility used for fire-fighting training (5 years after the cessation of activities). Specifically, ATSDR (2008) reported concentrations of PFOS from six wells in Lake Elmo and Oakdale, MN that averaged 465 ng/l, and Moody et al. (2003) reported a median concentration of 15 μg/l, nearly three orders of magnitude above the background water concentrations of 21 ng/l used in the typical scenario above, in well water in Oscoda Township, MI. Concentrations in this scenario are intended to represent only the most highly contaminated sites rather than all industrial sites.

  3. 3)

    Intake of common FOSA precursors to PFOS in a typical setting: This scenario used average contact rates as in the first scenario above, but used available information on the distribution of background concentrations of the PFOS precursors, principally FOSEs/FOSAs. Biotransformation of these compounds represents an indirect but potentially substantial source of exposure to PFOS (D'eon and Mabury, 2007). The intakes of these precursors in food came from Tittlemier et al. (2006); the water concentrations (including perfluorooctane sulfonamidoacetic acids) were from Boulanger et al. (2004); the dust concentrations from Shoeib et al. (2005); and the outdoor and indoor air concentrations were from Martin et al. (2002), Stock et al. (2004) (Measurements from the Griffin, GA sampling site were excluded because the area is known to be impacted by industrial contamination.), and Shoeib et al. (2005).

PFOS PK Modeling

A simple, single compartment, first-order PK model that predicts concentrations in blood serum as a function of dose, elimination rate, and volume of distribution is used in this study. This model was used for adults and for typical background exposures only. This model is given as:

where CP is the serum concentration (ng/ml), DP is the daily absorbed dose (ng/kg-bw/day), Vd is the volume of distribution (ml/kg), and kP is the first-order elimination rate in the body (1/day). Simplistically assuming steady state conditions exist, one can easily solve for intake dose as:

This steady state solution is used in this study. This implies that background exposures have been occurring for adults for a reasonably long period of time such that an assumption of steady state is reasonable. Besides this simplicity, discussions later in the paper focus on other issues associated with the toxicokinetics of PFCs and the appropriateness of the 1-compartment, first-order model. This approach has been used extensively to model PFCs in humans (Harada et al., 2003; Washburn et al., 2005; Sinclair et al., 2006; Trudel et al., 2008; Veronica et al., 2008; Lou et al., 2009), and its use here builds on the experience of these other researchers.

Fromme et al. (2007) used the model to predict intake dose given plasma concentrations of PFOS and PFOA for 31 study subjects. They used a median half-life of 1661 days for PFOS, derived from the single study that calculated human dissipation rates from an occupational cohort (Olsen et al., 2007a). Noting that there were no data on volume of distribution in humans, they used a value of 220 ml/kg for PFOS in accordance with use of this value by Andersen et al. (2006) for PK modeling in monkeys. Trudel et al. (2008) also used the 1-compartment, first-order PK model for adult exposures. Their volume of distribution term came from subchronic monkey studies (Griffith and Long, 1980; Noker, 2003) and was assigned low, intermediate, and high values of 1300, 3600, and 6000 ml/kg. These are about an order of magnitude higher than assumed by Fromme et al. (2007). The half-lives were taken from a European Hazard Assessment (OECD, 2002) and were 800, 3200, and 7800 days (low/intermediate/high) for PFOS. These inputs are similar to the ones used by Fromme et al. (2007). Harada et al. (2003) used subchronic monkey studies (Seacat et al., 2002) to arrive at a volume of distribution of 300 ml/kg, and used occupational data developed earlier by Olsen et al. (1999) to assume that the elimination half-life was between 1000 and 2000 days.

Parameters required to model PFOS in this paper include the first-order elimination rate, kP, and the volume of distribution for PFOS in serum, Vd. The first-order elimination rates for PFOS was assumed to be 0.00039 1/day, based on the reported median half-life of about 4.8 years from occupational exposures reported by Olsen et al. (2007a). With no available values for Vd based on human data, two bounding estimates were used, a “high” estimate of 3000 ml/kg and a “low” estimate of 200 ml/kg for PFOS, based on the review of studies above.


Compilation of Exposure Media Data

A summary of exposure media concentrations of PFCs from around the world is provided in Table 2. Most of the data available in the literature are focused on PFOS and PFOA. The FOSA and FOSE precursors to PFOS are included in the table, but the telomer alcohol precursors to PFOA are not included. A variety of exposure media are represented as follows: surface and ground water, indoor and outdoor air, house dust, and food. Discussions below focus on PFOS.

Table 2 Media-specific concentrations of PFOA, PFOS, and precursors.

The most abundant measurement data appear to be in surface waters, perhaps due to early interest in explaining transport to the arctic regions (Giesy and Kannan, 2001) and compelling evidence of the contribution of contaminated drinking water to elevated blood levels (Harada et al., 2003; Emmett et al., 2006; Skutlarek et al., 2006). Surprisingly, virtually no measurements of PFCs in tap water from areas not known to be contaminated are available (Hölzer et al., 2008). Surface water concentrations from Lakes Erie and Ontario and lakes and rivers in the states of New York and North Carolina (Boulanger et al., 2005; Sinclair et al., 2006; Nakayama et al., 2007) were compiled to derive a median PFOS concentration of 21.4 ng/l and a GSD of 2.6 for modeling typical water ingestion exposure among residents of the United States. As a point of comparison, the recently issued provisional health advisory value for PFOS was set at 200 ng/l (EPA, 2009). A median of 7.4 ng/l and a GSD of 3.1 were used to model exposure to precursors (principally 2-(perfluorooctane sulfonamido) acetic acid for this pathway alone).

While seemingly plentiful measurements of concentrations in air exist, typically only the volatile precursors are measured (Martin et al., 2002; Shoeib et al., 2004, 2005; Stock et al., 2004) or the measurements are from outside of North America (Sasaki et al., 2003; Harada et al., 2005; Barber et al., 2007; Jahnke et al., 2007a 2007b). As the meager data available for indoor PFOS are reported as below the limit of detection (Barber et al., 2007), indoor concentrations were assumed to be 20 times outdoor values (Kim and Kannan, 2007) based on the ratios for FOSAs and polybrominated diphenyl ethers reported by Shoeib et al. (2004). The values selected to model inhalation exposure were as follows: median of 2.2 pg/m3 and GSD of 1.9 for outdoor PFOS, median of 44 pg/m3 and GSD of 1.9 for indoor PFOS, median of 95 pg/m3 and GSD of 3.8 for outdoor fluorooctane precursors, and median of 2670 pg/m3 and GSD of 1.8 for indoor perfluorooctanylsulfonyl precursors.

PFOS measurements in house dust are available for both the United States (Strynar and Lindstrom, 2008) and Canada (Shoeib et al., 2005; Kubwabo et al., 2005), but these may not be representative of the general population because of the limited geographical locations from which environmental samples were collected. The values measured by Strynar and Lindstrom (2008), with a median of 201 ng/g and GSD of 5.4, were used for exposure modeling of PFOS. A median value of 460 ng/g and a GSD of 4.4 derived from Shoeib et al. (2005) were used for exposure modeling of fluorooctane precursors.

Concentrations in food are available from market basket surveys in Canada (Tittlemier et al., 2006, 2007) and Spain (Ericson et al., 2008) and a duplicate diet study performed in Germany (Fromme et al., 2007). The Canadian researchers focused on those foods expected to have measurable levels and reported an average dietary intake of PFOS two to three times higher than that estimated by the Europeans. Detectable levels of PFOS and various precursors were only measured in a small fraction of the food groups. The highest PFOS levels were measured in the beef steak (2.7 ng/g) and marine fish (2.6 ng/g) composites (Tittlemier et al., 2007). The highest total perfluorooctane sulfonamide levels (up to 27.3 ng/g) were measured in the fast food composites (Tittlemier et al., 2007) collected from 1992 to 1999. Median intake estimates (1.77 ng/kg/day PFOS and 1.15 ng/kg/day perfluorooctane solfonamides) published by Tittlemier et al. (2007, 2006) were directly used in the modeling.

Overall, measurement of PFCs in exposure media in North America was relatively sparse for all media compared with European data with the possible exception of dust. Food data are sparse for both continents. No measurements of PFOS in uncontaminated soil could be found. Although the data selected for the exposure models are not statistically representative of the general US population, they were deemed adequate for conducting a screening-level exposure intake assessment.

Compilation of Body Burden Data

Table 3 shows a summary of blood measurements taken of the PFCs in the United States, focusing first on five key PFCs and Table 4 presents values for some PFCs less regularly measured. As with the environmental media concentration, most of the data in the literature focuses of PFOS and PFOA, but discussions below are limited to PFOS.

Table 3 Summary of studies measuring general population blood concentrations of five key PFCs measured in the United States (all units=ng/ml; M=male; F=female).
Table 4 Summary of studies measuring general population blood concentrations of infrequently measured PFCs in the United States (all units=ng/ml).

Temporal evaluations of body burdens suggest a rise of PFC concentrations from the 1970s until the latter 1980s, with consistent levels throughout the 1990s, and then evidence of a drop in the early to mid 2000s. Olsen et al. (2003a) measured and reported GMs of 7 fluorinated compounds in 645 serum samples (332 males, 313 females) from 6 Red Cross blood banks from around the country in 2000–2001. PFOS was detected in all but one sample with a GM of 34.9 ng/ml, and a maximum of 1656 ng/ml. Olsen et al. (2007b) analyzed samples obtained from Red Cross blood banks in 2005 and compared concentrations found with samples collected from the same facility in Minneapolis, MN, in 2000. They found PFOS concentrations from 100 samples collected in 2000 at a GM of 33.1 ng/ml, whereas the 40 samples collected from this location in 2005 had a GM of only 15.1 ng/ml. It was noted that the 2005 samples were plasma, whereas the 2000 samples were serum, but Ehresman et al. (2007) found nearly identical results between serum and plasma in samples taken voluntarily from occupationally exposed individuals in a 3M plant.

Olsen et al. (2005) examined trends in earlier decades and reported higher concentrations in samples obtained in 1989 than in those obtained in 1974. The 1974 and 1989 median concentrations of PFOS were 29.5 ng/ml in 1974 and 34.7 ng/ml in 1989. The authors suggest that the increase could be due to exposure and/or bioaccumulation. They identify a fivefold increase in PFC production between 1975 and 1989. They also compared available data between 1989 and 2001, and show that concentrations of PFOS were virtually unchanged, suggesting that the similarity may be correlated to the relatively consistent production of PFOS by 3M between 1989 and 1999, the year before 3M's phaseout announcement. Calafat et al. (2006) confirm this trend of consistent concentrations throughout the 1990s. By comparing serum pools corresponding to years 1990, 1998, and 2002, they found PFOS detected in all US samples, with a median at 31.1 ng/ml, with little variation over time.

In the early 2000s, the serum PFC concentrations in blood samples from US volunteers were measured and recorded in NHANES. Recently, Calafat et al. (2007) compared NHANES 2003–2004 data with the NHANES 1999–2000 data and reported significantly lower concentrations for the later years in three of the four most commonly found PFCs, including PFOS: the GM of the PFOS concentrations dropped from 30.4 to 20.7 ng/ml. The authors attributed the reduction to discontinuation in 2002 of industrial production by electrochemical fluorination of PFOS and related perfluorooctanesulfonyl fluoride compounds.

In general, Table 3 shows PFOS measurements in the general population to range from 20 to 40 ng/ml with very high, near 100%, frequencies of detection. The lowest concentrations noted were in cord blood: the GMs for PFOS was 4.9 ng/ml (Apelberg et al., 2007).

PFOS Exposure Intake Estimates

Results of these route-specific intake estimates are presented in Table 5 and in box-and-whisker plots (boxplots) in Figures 2 and 3. The boxplots display estimates for the 5th and 95th percentiles (lower and upper whiskers, respectively) and the 25th percentile, median, and 75th percentile (bottom, middle, and top of box, respectively); extreme values (those beyond the 5th or 95th percentiles) are displayed as open circles. Dermal absorption, dietary and non-dietary ingestion, and inhalation are represented as separate boxes in each panel.

Table 5 Median estimates of exposure intake and percentage of total intake by each exposure pathway.
Figure 2

Estimated pathway-specific intakes for young children of (a) PFOS under a typical environmental exposure scenario, (b) PFOS under a highly contaminated environment scenario, and (c) perfluoralkyl sulfonamide compounds under a typical environmental exposure scenario. Pathways are ingestion of dust (ingest dust), food (ingest food), and water (ingest water), dermal absorption from contact with dust (dermal dust), and inhalation of indoor air (inhal indoor) and outdoor air (inhal outdoor).

Figure 3

Estimated pathway-specific intakes for adults of (a) PFOS under a typical environmental exposure scenario, (b) PFOS under a highly contaminated environment scenario, and (c) perfluoralkyl sulfonamide compounds under a typical environmental exposure scenario. Pathways are ingestion of dust (ingest dust), food (ingest food), and water (ingest water), dermal absorption from contact with dust (dermal dust), and inhalation of indoor air (inhal indoor) and outdoor air (inhal outdoor).

The median PFOS intake (i.e., the sum of the median route-specific intakes) for 2-year-old children under typical exposure conditions is estimated as 50 ng/day. Ingestion of food and of dust appear to be the primary routes of exposure (see Figure 2a), representing approximately 42% and 36% of the total (see Table 5). At the 95th percentile, however, intake from dust ingestion (220 ng/day) is roughly double that of intake from food ingestion (100 ng/day) due to greater observed variability in the dust concentrations. Intake from water ingestion is estimated to be the third most important source of PFOS intake at both the median (9.9 ng/day, or 20% of the total) and the 95th percentile (30 ng/day). Intake from dermal absorption, and inhalation of indoor air and of outdoor air all represent 2% or less of total intake.

Under the scenario of an environment with highly contaminated water, consumption of contaminated water appears to be the primary source of intake for 2-year-old children at about 600 ng/day (see Figure 2b) or 94% of total intake (see Table 5). With food concentrations assumed unchanged, the importance of dietary ingestion is much diminished (about 3% of total intake). Dermal absorption remains relevant under this scenario only for individuals at the highest percentiles of the distribution.

Returning to a typical exposure scenario, but extending the analysis to include intake of FOSE and FOSA precursors assigns even more importance to the incidental dust ingestion and dietary ingestion routes. Under a simplifying, and perhaps overly conservative, assumption that these precursors are fully metabolized to PFOS in the human body, intake of these precursors may add 41 ng/day (at the median) from ingestion of dust and 13 ng/day from dietary ingestion (see Figure 2c). Among the precursors, incidental dust ingestion, dermal absorption, and dietary ingestion represent the bulk of the total intake. As in the previous scenarios, ingestion of water remains a potentially important route (see Table 5).

Turning to adults for the typical exposure scenario, the aggregate median PFOS intake is estimated to be 160 ng/day, and dietary ingestion again appears to be the primary route of exposure (see Figure 3a). Incidental ingestion of dust is far less important among adults than among children. Median intake from dietary ingestion is estimated to be 110 ng/day, representing 72% of total intake, whereas median intake from dust (about 9 ng/day) represents only 6% of total intake (see Table 5). The relative contribution of ingestion of drinking water (22% of total intake) is similar among adults and children. In the “highly contaminated” scenario (Figure 3b), ingestion of contaminated drinking water, leading to 2100 ng/day of PFOS, would comprise 94% of the total intake (see Table 5), and dietary ingestion (maintained at 110 ng/day) would represent 5%. The contribution from the inhalation of indoor air and outdoor air is negligible. Examining the intake of the perfluorooactane precursors (Figure 3c), dietary ingestion is estimated to be the dominant route of exposure (see Table 5), leading to 59% of the precursor intake, and dust ingestion and inhalation of indoor air account for 16% and 12%, respectively, of total intake.

PFOS PK Modeling

As noted earlier, the best current characterization of American background body burden comes from the NHANES 2003/2004, with a GM of 20.7 ng/ml for PFOS in serum (Calafat et al., 2007). Given a first-order dissipation rate of 0.00039 1/day, two separate estimates of the serum volume of distribution (Vd) at 3000 and 200 ml/kg, the range in estimated intakes of PFOS is solved as 1.6–24.2 ng/kg-bw/day for PFOS. Because the NHANES data are from a nationally representative sample of participants aged 12 years and older, the pertinent scenario to compare this with is the typical exposure scenario for adults, the results of which are displayed in Table 5, and in Figures 3a and c for PFOS and FOSE/FOSA precursors, respectively. The total intake for PFOS (sum of the medians from each route) is 160 ng/day for adults, and assuming a 70 kg adult, this equals a body weight-based intake of 2.3 ng/kg-bw/day. The median intakes for the precursors total about the same, at 1.9 ng/kg-bw/day. If these precursors were to fully convert into PFOS in the body, then the total intake would be 4 ng/kg-bw/day, with nearly equal amounts from PFOS and from the precursors. The lower modeled intake value of PFOS, 1.6 ng/kg-bw/day, assumed the lower volume of distribution, at 200 ml/kg. While this might suggest that an appropriate volume of distribution might be closer to the lower end of the estimated range (200–3000 ml/kg), the substantial variability and uncertainty in both the intake estimate and the PK modeling are too great to draw this conclusion. The exposure pathway-based and PK-based intake estimates are compared in Figure 4.

Figure 4

Median intake of PFOS by adults based on an exposure pathway analysis compared with intakes predicted using the PK model, separately assuming volume of distribution (Vd) of 200 and 3000 ml/kg.


As has been noted frequently in this article, variabilities and uncertainties abound in both the forward- and back-calculated intake doses. There are uncertainties associated with use of the 1-compartment PK model and the input parameters assigned to the model. There are uncertainties with PFOS exposure media concentrations due to lack of measurements, particularly for food. Also, there are variabilities associated with exposure contacts not captured in the simple exercise, along with the variabilities within the measurements made for PFOS.

Uncertainties in Model Parameters

A large uncertainty in the backward PK modeling approach pertains to the use of the simple 1-compartment, first-order model and in particular, the steady state solution to this model. Andersen et al. (2006) describe why PFC pharmacokinetics are substantially more complicated than implied by use of this 1-compartment model: more rapid elimination in cynomolgus monkeys with increasing doses showed that capacity-limited, saturable processes must be involved in the kinetics of PFCs. Given the saturable reabsorption hypothesis and the dose dependency on elimination kinetics as described by Andersen et al. (2006), it is possible that higher doses lead to more rapid elimination; that a blood serum concentration of 20.7 ng/ml could result from a dose higher than the high estimate of 24.2 ng/kg-bw/day.

There is a lack of information on even the few parameters required for the 1-compartment approach for humans. No studies could be found which contain the data necessary to calibrate the volume of distribution in humans, and only one study could be found quantifying the half-lives of PFOS and PFOA in humans (Olsen et al., 2007a). This was a study on occupational exposures, and the half-lives gleaned from these high body burdens may not be appropriate for studying general population exposures. There were also no human data available for the route-specific absorption fractions for PFCs; those values remain a substantial uncertainty for this effort. They were assumed to be 0.9 for the ingestion pathways (food, water, dust), 0.50 for the inhalation pathway, and 0.03 for the dermal contact pathway. Absorptions near 1.00 for gastrointestinal ingestion are not uncommon for the modeling of persistent organic pollutants, such as dioxins (EPA, 2003) and PBDEs (McDonald, 2005; Lorber, 2008). Trudel et al. (2008) assumed absorption fractions at 0.66, 0.80, and 0.91 in the low, intermediate, and high scenarios for PFC modeling for all pathways.

In the comparison of predicted intakes using the PK model and intakes modeled based on an exposure pathway analysis, it was noted that perhaps the most appropriate volume of distribution might be closer to the lower value, 200 ml/kg, as compared with the higher value used, 3000 ml/kg. This was because use of the lower value led to a PK-modeled intake of 1.6 ng/kg-bw/day for PFOS, and the forward-modeled median intake of PFOS and precursors (assuming the precursors fully metabolized to result in an equal amount of PFOS in the blood) was 4 ng/kg-bw/day. Use of the higher volume of distribution, 3000 ml/kg, led to a much higher prediction of intake, 24.2 ng/kg-bw/day, so generally it might be surmised that the lower volume of distribution is more appropriate for use in this framework.

Finally, the suggestion that blood concentrations of PFOS could be due in relatively equal measure to intakes from PFOS itself and the sum of the precursors is likely to be overly conservative (despite the difficulties in identifying all likely precursors) and a major source of uncertainty. In their modeling of precursor compounds to PFOS, Vestergren et al. (2008) concluded that only 2–6% of the total dose of PFOS is due to precursors (where “dose” is internal dose, which considers transformation of precursors of PFOS to PFOS itself). However, this was the finding for the “median” set of results from their Monte Carlo exercise, in which they varied exposure contact rates, exposure media concentrations, and also PK modeling parameters. For modeling the internal transformation of precursors to PFOS, they assumed a “transformation factor” (fraction of precursor transforming to PFOS) of 0.01, 0.1, and 1.0 in low, intermediate, and high model runs. For the “high exposure” results, the precursors in fact dominated total exposures to PFOS, explaining between 60% and 80% of total exposures for the various receptors (infants, teens, adults, and so on). They cite the rate of transformation as a key uncertainty for characterizing the impact of PFOS precursors. They reported only one study on mammal internal transformation of PFOS precursors: Seacat et al. (2002) found that the yield of PFOS after oral dosing of N-EtPFOSE was 20% in vivo in rats.

Even with these uncertainties, several researchers have used the 1-compartment, first-order PK model to study exposures to PFCs (Harada et al., 2003; Washburn et al., 2005; Fromme et al., 2007; Trudel et al., 2008; Veronica et al., 2008). As the saturable reabsorption hypothesis (and the implied more complex modeling required) seems to be appropriate for higher doses, it is possible that the simple 1-compartment model and even the steady state solution may be quite acceptable for studying the long-term, low-dose exposures that are typical of general population exposures, or even for the level of exposure in contaminated settings also modeled in this study. Certainly, in earlier applications of this model to study PFOS and PFOA, this has been the assumption.

Total Human Exposure

Despite the sparseness of available data, others who have studied exposure to PFOS have arrived at typical background exposures that compare favorably with the range implied by the PK modeling (1.6–24.2 ng/kg-bw/day for adults) and by the forward-intake modeling (4 ng/kg-bw/day). Most recently, two studies provide detailed assessments of all relevant pathways of exposure from direct and indirect sources (Trudel et al., 2008; Fromme et al., 2009). In a comprehensive exposure assessment, Trudel et al. (2008) quantified water ingestion, inhalation, and dust-related exposures (dust ingestion, hand-to-mouth contact); included pathways involving clearly defined consumer product sources (e.g., transfer from contact with PFC-treated carpets); and performed an elaborate assessment of dietary exposure involving specific food types (e.g., dairy products, fish and shellfish, vegetables) and daily consumption of each food type by different age groups in North America and Europe. They incorporated data variability by deriving low, intermediate, and high exposure intakes for infants, toddlers, children, teens, and adults. Parameter uncertainty was incorporated into their modeling by combining models in which all parameters were set to their lowest values with models in which all were set to their highest values, resulting in PK-predicted intake dose estimates that spanned an order of magnitude.

The total intake doses of PFOS estimated by Trudel et al. (2008), considering all pathways, ranged from 3 to 220 ng/kg-bw/day across the age ranges modeled. Looking only at the intermediate values, their PK-predicted intake dose for adults is 20 ng/kg-bw/day for PFOS for North American exposures, compared with an exposure pathway estimate for adults at 15 ng/kg-bw/day. For PFOA, the PK-predicted value for adults was 10 ng/kg-bw/day, compared with an exposure pathway estimate for adults of 2 ng/kg-bw/day. While their intermediate estimate of 15 ng/kg-bw/day total PFOS intake falls well within the range implied by their PK-modeling approach, they did not consider precursor compounds. Excluding precursors, 15 ng/kg-bw/day is roughly seven times the median total adult PFOS intake estimated by forward modeling in this exercise (2.3 ng/day). Similarly, their estimate of total intake of 35 ng/kg-bw/day for toddlers exceeds the 3.8 ng/kg-bw/day estimated for 2-year olds in this exercise. The authors concluded that ingestion of food and drinking water are the primary contributors to PFOS intake for adults and that hand-to-mouth transfer of carpet-treatment residues represents the highest contribution for children. Their use of different food concentration data and their inclusion of a pathway accounting for contact with treated carpet/textiles are the main reasons for the large differences in estimated total intake between their assessment and this one.

The same research group continued their efforts by looking at precursors for PFOS and PFOA (Vestergren et al., 2008). Specifically, they looked at FTOHs, FOSAs, and FOSEs that are metabolized to form PFOA and PFOS in the body. Considering this metabolism, and also building on their previous effort, they recalculated a total dose of PFOS to range now from 3.9 to 520 ng/kg-bw/day (total intakes over all pathways and over all exposed populations modeled, including children). For their central tendency estimates, the precursors of PFOS accounted for between 2% and 6% of the total, but at one set of “high” assumptions for the precursors, they dominated the overall intakes.

Fromme et al. (2009) also assessed all major pathways for the general adult population. They estimated the average intake of PFOS and of perfluorooctylsulfonyl precursors as 1.6 ng/kg-bw/day, each. They estimated the median contribution from dietary PFOS exposure to be 1.5 ng/kg-bw/day (90 ng/day) and concluded that dietary exposure is the dominant intake pathway, comprising 91% of total intake. Their median adult dietary intake estimate, based on concentration data from a duplicate diet study in Germany (Fromme et al., 2007) that analyzed whole meals rather than composites of similar food items, is in accordance with the value estimated in this exercise (based on intake estimates published by Tittlemier et al. (2007). Nonetheless, their estimate of the fraction of total intake represented by dietary exposure is higher (91% compared with 72%) because of much lower estimates of intake from dust and drinking water ingestion.

Many other intake estimates have been derived by researchers who measured PFOS in exposure media data and combined their measurements with contact rates. Most of these efforts have focused on food exposures. Fromme et al. (2007) calculated median intakes for PFOS for German study subjects, based on duplicate dietary samples, and found it to be 1.4 ng/kg-bw/day. Tittlemier et al. (2007) conducted a market basket survey of a wide range of Canadian foods, and derived an adult intake estimate of 110 ng/day (1.6 ng/kg-bw/day assuming a body weight of 70 kg) for PFOS. This was the starting point for the generation of a distribution of dietary intakes that we used to characterize US exposures to PFOS in food. A similar total diet survey in the United Kingdom (Mortimer et al., 2006) resulted in a much broader range of adult dietary exposure estimates for PFOS (10–100 ng/kg-bw/day). However, unusually high findings of PFOS and PFOA in the potatoes food group were not explained. Ericson et al. (2008) measured 11 PFCs in 36 composite food samples (duplicates of 18 major food types) representing the diet of Spaniards living in Catalonia, Spain. They found frequent detections of PFOS only; positive concentrations of PFOA and PFHpA were found in milk, and there were non-detects (NDs) for all other PFCs at detection limits (DLs) generally near or less than 0.1 ng/g fresh weight. On the basis of the frequent occurrence of PFOS (11 of 18 foodstuffs), they calculated adult intakes of 62.5 ng/day (0.9 ng/kg-bw/day), setting ND values at 0, and 74.2 ng/day (1.1 ng/kg-bw/day), setting ND values at one-half the DL. Gulkowska et al. (2006) sampled 27 types of seafood from Chinese fish markets including finfish (croaker, mackerel), shrimp, crabs, and other species. They found 100% occurrence of PFOS, ranging from 0.3 to 13.9 ng/g. They also found positives for PFOA and PFUnDA, with infrequent positive occurrences for PFDA, PFNA, PFHpA, and PFHxA. By combining concentrations with high fish consumption rates (>100 g/day) found in dietary surveys for populations in Zhoushan and Guangzhou, they found PFOS intake rates totaling 9.3 ng/kg-bw/day for residents of Guangzhou, and 4.2 ng/kg-bw/day for Zhoushan residents.

Very few earlier studies focused on pathways other than food. Harada et al. (2003) calculated a water ingestion intake of PFOS of 1.4 ng/kg-bw/day, based on a water concentration of 50 ng/l found in measurements from a contaminated river. Shoeib et al. (2005) studied indoor dust and air exposures to FOSAs and FOSEs. Under most scenarios they devised for adult exposure, uptake through inhalation exceeded uptake through dust ingestion. For example, using median air and dust concentrations, total intake by inhalation and dust ingestion was estimated at 60 ng/day (0.86 ng/kg-bw/day), almost two-thirds of which is due to inhalation. For children, however, estimated intake by dust ingestion (44 ng/day) exceeded inhalation intake (27 ng/day). Tittlemier et al. (2006) compared their dietary intake results with the dust and air intake estimates of Shoeib et al. (2005) and found that the dietary intake estimates were of the same order of magnitude as the combined intake through indoor dust and air. Sasaki et al. (2003) reported GM PFOS levels of 0.6 pg/m3 in a rural area and 5.3 pg/m3 in a more urban setting in Japan, and estimated daily intakes for adults of 10 and 100 pg/day, respectively. They concluded that the contribution from outdoor air is about 1/1000th of the contribution from tap water, but their estimates of adult intake by inhalation are 2.5–25 times lower than our median estimate (about 0.25 ng/day).

Estimated Pathway-Specific Contributions

Forward-based calculations of intake doses, on the basis of exposure media concentrations, contact rates, and exposure factors, may be as uncertain as the back-calculated intakes based on body burdens and the 1-compartment PK model. While in typical exposure settings food contamination seems to dominate the intakes for adults, and food and dust contamination for children, as seen in Figures 2a, c, 3a and c, there is in reality a paucity of data on food concentrations of PFCs. Food seems to be an exceptionally difficult matrix in which to measure PFCs by conventional analytical techniques because of its highly variable composition. As a result, few researchers have reported concentrations in typical Western diets. In fact, we are unaware of any reports in the peer-reviewed literature of PFOS in food from the United States and we relied instead on a Canadian Total Diet Survey to estimate the intakes from this pathway. Although there is evidence that migration from fluorochemical-treated food packaging has been an important source of PFCs in food (Begley et al., 2005; D'eon and Mabury, 2007), there is also evidence that these concentrations, particularly in fast food wrappers, have decreased over the past 15 years (Tittlemier et al., 2006).

Water ingestion dominates the exposure in the “contaminated” scenario because of drinking water concentrations that are assumed to be similar to the high concentrations measured in the groundwater of affected areas, and because water treatment removal efficiencies are assumed to be negligible (Takagi et al., 2008). Although no data in the literature could be found linking elevated PFOS concentrations in water to elevations of PFOS in blood, there is evidence that water contaminated with PFOA leads to elevations of PFOA in blood. Elevations in both water and serum concentrations of PFOA were observed in a community located in West Virginia near a fluoropolymer manufacturing facility (Emmett et al., 2006; Steenland et al., 2009). Indeed, Veronica et al. (2008) investigated body burden impacts from consumption of contaminated water at that site using the same PK model as used in this evaluation.

Inhalation of indoor air only appears to be a major pathway, in terms of the percent contribution to total exposures (Table 5), under the scenario that considers only the FOSA precursors. Under the assumptions used, dermal absorption does not seem to be a meaningful route of exposure, and PFOS intake from dust ingestion is minimal for adults (both pathways explain less than 5% of total exposure). Dust ingestion is, however, a dominant pathway of exposure for children, explaining 37% of exposure to PFOS and 60% to the precursor compounds.

Although the modeling in this study suggests that dust ingestion together with food ingestion dominates the overall exposure for children, dust-related exposures represent a critical uncertainty for these compounds. The extent to which individuals actually ingest house dust is uncertain. The values of dust ingestion used in this assessment, 100 mg/day for children and 50 mg/day for adults, were the same values used by Lorber (2008) to characterize ingestion of house dust in his assessment of exposure to PBDEs. However, 100 and 50 mg/day are, in reality, recommended central tendency values for use in modeling soil ingestion in EPA's Exposure Factors Handbook (EPA, 1997). Although the influence of indoor dust is well recognized for PBDEs (Lorber, 2008; Stapleton et al., 2008), the precise way to model exposure to PBDEs in dust, and PFCs in dust for that matter, remains an uncertainty. Stapleton et al. (2008) focused on the hand-to-mouth pathway by measuring PBDEs on the surfaces of hands. As dust can be ingested when incidentally inhaled as well, we assume that hand-to-mouth and other mechanisms that transfer dust into the mouth are implicitly included in the soil ingestion rate used to model dust ingestion. It is possible that direct exposure to surfaces containing PFCs, which is not modeled in this study (dermal contact with impacted dust is modeled, but not direct dermal contact with treated surfaces), could contribute to the indoor exposures of these contaminants. For example, Trudel et al. (2008) suggest that hand-to-mouth transfer from contact with PFOS-treated carpets may be a major contributor to intake for children, and Jahnke et al. (2007a) note that surface-treated textiles could be a significant source of perfluorooctanesulfonyl precursors (FOSAs/FOSEs) in indoor air.

Most measurements used to estimate the pathway-specific intakes were performed before 2005. With the production of PFOS and related perfluorooctanesulfonyl alkyl compounds having ceased in the United States, the concentrations of PFOS and its precursors in exposure media are likely to decrease. As discussed earlier, a 32% decrease in body burden in the United States between 2003–2004 and 1999–2000 has already been documented (Calafat et al., 2007). Because of the persistence of PFOS in the environment, it is likely that intake through the food pathway would be most immediately affected, and may be most responsible for the apparent decline in body burden. Additional measurements of PFOS in food, particularly in the United States, are vitally needed. Also, as evidenced by the sparseness of data in Table 2, more measurements of concentrations in soil and air (particularly in areas affected by industrial emissions) are also desperately needed.

The approach used in this study relied on distributions of PFOS in exposure media to illustrate the potential variability and uncertainty (given the sparseness of data on PFOS in exposure media) in route-specific intakes, but relied on point estimates for contact rates and other exposure factors. Using distributions for contact rates is certainly possible, but requires adequate data, which may not exist for all pathways. By combining distributions for both concentrations and contact rates likely would have little effect on the estimates of central tendency, but would certainly have produced wider distributions of PFOS intake. Trudel et al. (2008) used a simplistic means of incorporating variability in contact factors in their exposure modeling and did indeed report a large range of possible exposures—over an order of magnitude between high and low estimates of exposure. There is no “correct” way to conduct such an analysis; we chose point estimates of the contact rates as a way of limiting the ranges of intakes that we model within a single pathway. A simplification with perhaps a larger effect on the uncertainty and variability in our aggregate exposure estimates is the practice of summing medians across route-specific intakes. As it is highly unlikely that a given individual within a sub-population will be at the same percentile with respect to intake along all exposure routes, a more accurate method incorporating the individual correlation among routes would be required to reduce the uncertainty within our estimate. However, the information on correlations among routes does not exist at this time. Again, only a true probabilistic assessment would provide an accurate assessment of population variability in aggregate intake. We present the route-specific distributions in Figures 2 and 3 to illustrate that it would be erroneous to conclude that there may be a dominant pathway for all individuals in a sub-population based on point estimates of intakes.


Assessing the most relevant routes of exposure to PFCs for humans remains a challenge and continues to be the topic of great debate. Given the pervasiveness of PFCs in our environment, it is surprising that so few published estimates of route-aggregated exposure levels exist. This document summarizes published existing human exposure estimates and provides a crude assessment of aggregate exposure to PFOS and its main precursors under scenarios of typical and contaminated environments, for 2-year-old children and for adults. The available measurement data and analysis in this paper point toward dietary ingestion as the major contributor to PFOS intake for adults, and dietary and dust ingestion as nearly equal contributors for young children, under typical exposure scenarios. The central estimates of forward-calculated intakes fell within the range of intakes back-calculated using the simple PK model. As this provides some level of cross-verification, there are critical uncertainties in the PK modeling along with the exposure media concentration data and contact rates. Crude estimates of route-specific intake are based on sparse data and are thus fraught with uncertainty. Estimates for intake from the food pathway, for example, are based on data from one of only four studies available with measured concentrations in foods found in the typical Western diet. Direct dermal exposure to treated articles, mouthing by children of treated fabrics, and inhalation associated with household use of contaminated water were not included in this analysis. Furthermore, assumptions regarding relevant exposure factors, in particular, the rate of dust ingestion, are used, often based on few data. As a result, assumptions and generalizations required for this analysis, together with the broad distributions of the route-specific estimates, make it difficult to draw definitive conclusions about the relative importance of the various exposure routes.

We believe that the elucidation of the framework for studying PFCs provided in this paper using PFOS as an example will allow for more meaningful future evaluations. The continued study of PFCs, particularly, measurements of concentrations in environmental and exposure media as well as toxicokinetic parameters, will allow for reduced uncertainties and more refined exposure assessments in the future.


  1. Andersen M.E., Clewell H.J., Tan Y-.M., Butenhoff J.L., and Olsen G.W. Pharmacokinetic modeling of saturable, renal resorption of perfluoroalkylacides in monkeys–probing the determinants of long plasma half-lives. Toxicology 2006: 227: 156–164.

    Article  CAS  Google Scholar 

  2. Apelberg B.J., Goldman L.R., Calafat A.M., Herbstram J.B., Kuklenyik Z., Heidler J., Needham L.L., Halden R.U., and Witter F.R. Determinants of fetal exposure to polyfluoroalkyl compounds in Baltimore, Maryland. Environ Sci Technol 2007: 41: 3891–3897.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. ATSDR. Public Health Assessment for Perfluorochemical Contamination in Lake Elmo and Oakdale, Washington County, Minnesota. Draft for Public Comment. Agency for Toxic Substances and Disease Registry, U.S. Department of Health and Human Services, Atlanta, GA, USA, 2008.

  4. Barber J.L., Berger U., Chaemfa C., Huber S., Jahnke A., Temme C., and Jones K.C. Analysis of per- and polyfluorinated alkyl substances in air samples from Northwest Europe. J Environ Monit 2007: 9 (6): 530–541.

    Article  CAS  Google Scholar 

  5. Begley T.H., White K., Honigfort P., Twaroski M.L., Neches R., and Walker R.A. Perfluorochemicals: potential sources of and migration from food packaging. Food Addit Contam 2005: 22 (10): 1023–1031.

    Article  CAS  Google Scholar 

  6. Boulanger B., Peck A.M., Schnoor J.L., and Hornbuckle K.C. Mass budget of perfluorooctane surfactants in Lake Ontario. Environ Sci Technol 2005: 39 (1): 74–79.

    Article  CAS  Google Scholar 

  7. Boulanger B., Vargo J., Schnoor J.L., and Hornbuckle K.C. Detection of perfluorooctane surfactants in Great Lakes water. Environ Sci Technol 2004: 38 (15): 4064–4070.

    Article  CAS  Google Scholar 

  8. Calafat A.M., Kuklenyik Z., Caudill S.P., Reidy J.A., and Needham L.L. Perfluorochemicals in pooled serum samples from United States residents in 2001 and 2002. Environ Sci Technol 2006: 40 (7): 2128–2134.

    Article  CAS  Google Scholar 

  9. Calafat A.M., Kuklenyik Z., Reidy J., Caudill S.P., Tully J.S., and Needham L.L. Serum concentrations of 11 polyfluoroalkyl compounds in the U.S. population: data from the National Health and Nutrition Examination Survey (NHANES) 1999–2000. Environ Sci Technol 2007: 41: 2237–2242.

    Article  CAS  Google Scholar 

  10. Calafat A.M., Needham L.L., Kuklenyik Z., Reidy J.A., Tully J.S., Aguilar-Villalobos M., and Naeher L.P. Perfluorinated chemicals in selected residents of the American continent. Chemosphere 2006: 63 (3): 490–496.

    Article  CAS  Google Scholar 

  11. Calafat A.M., Wong L.Y., Kuklenyik Z., Reidy J.A., and Needham L.L. Polyfluoroalkyl chemicals in the U.S. Population: Data from the National Health and Nutritional Examination Survey (NHANES) 2003-2004 and Comparisons to NHANES 1999–2000. Environ Health Perspect 2007: 115 (11): 1596–1602.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Cohen Hubal E.A., Egeghy P.P., Leovic K.W., and Akland G.G. Measuring potential dermal transfer of a pesticide to children in a child care center. Environ Health Perspect 2006: 114 (2): 264–269.

    Article  CAS  Google Scholar 

  13. CRG. 2004 APFO/PFOA monitoring summary report Washington Works Facility and Local, Letart and Dry Run Landfills Washington, WV. Corporate Remediation Group, An alliance between Dupont and URS Diamond, Barley Mill Plaza, Building 19, Wilmington, Delaware 1980. Project No: 7035-507532/7533 18984356. Available at;search term, EPA-HQ-OPPT-2003-0012 2005.

  14. D'eon J.C., and Mabury S.A. Production of perfluorinated carboxylic acids (PFCAs) from the biotransformation of polyfluoroalkyl phosphate surfactants (PAPS): exploring routes of human contamination. Environ Sci Technol 2007: 41 (13): 4799–4805.

    Article  CAS  Google Scholar 

  15. Davis K.L., Aucoin M.D., Larsen B.S., Kaiser M.A., and Hartten A.S. Transport of ammonium perfluorooctanoate in environmental media near a fluoropolymer manufacturing facility. Chemosphere 2007: 67 (10): 2011–2019.

    Article  CAS  Google Scholar 

  16. Ehresman D.J., Froehlichk J.W., Olsen G.W., Chang S.C., and Butenhoff J.L. Comparison of human whole blood, plasma, and serum matrices for the determination of perfluorooctanesulfonate (PFOS), perfluorooctanoate (PFOA), and other fluorochemicals. Environ Res 2007: 103 (2): 176–184.

    Article  CAS  Google Scholar 

  17. Ellis D.A., Martin J.W., De Silva A.O., Mabury S.A., Hurley M.D., Anderson M.P.S., and Wallington T.J. Degradation of fluorotelomer alcohols: a likely atmospheric source of perfluorinated carboxylic acids. Environ Sci Technol 2004: 38 (12): 3316–3321.

    Article  CAS  Google Scholar 

  18. Emmett E.A., Shofer F.S., Zhang H., Freeman D., Desai C., and Shaw L.M. Community exposure to perfluorooctanoate: relationships between serum levels and exposure sources. J Occup Environ Med 2006: 48 (8): 759–770.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. EPA. Provisional Health Advisories for Perfluorooctanoic Acid (PFOA) and Perfluorooctane Sulfonate (PFOS). United States Environmental Protection Agency, Office of Water, Washington, DC, USA, 2009.

  20. EPA. Guidelines for Exposure Assessment. Office of Health and Environmental Assessment, Washington, DC, USA (Publication no. EPA/6002-92/001 FR57:2288-22938). 1992.

  21. EPA. Exposure Factors Handbook. United States Environmental Protection Agency, Office of Research and Development, Washington, DC, USA, EPA/600/P-95/002B. 1997.

  22. EPA. Federal Register Vol. 67, No. 47. 40 CFR Part 721, Perfluoroalkyl Sulfonates; Significant New Use Rule (SNUR); Final Rule and Supplemental Proposed Rule. March 11, 2002a.

  23. EPA. Revised draft hazard assessment of perfluorooctanoic acid and its salts. U.S. Environmental Protection Agency, Office of Pollution Prevention and Toxics, Risk Assessment Division. 2002b.

  24. EPA. Exposure and Human Health Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. United States Environmental Protection Agency, Office of Research and Development, National Center for Environmental Assessment. NAS Review Draft. December, 2003. EPA/600/P-00/001C(a-f). Available at, 2003.

  25. EPA. Child-Specific Exposure Factors Handbook. United States Environmental Protection Agency, Office of Research and Development, Washington, DC, EPA/600/R/06/096F. 2008.

  26. Ericson I., Marti-Cid R., Nadal M., Bavel B.V., Lindstrom G., and Domingo J.L. Human exposure to perfluorinated chemicals through the diet: intake of perfluorinated compounds in foods from the Catalan (Spain) market. J Agr Food Chem 2008: 56: 1787–1794.

    Article  CAS  Google Scholar 

  27. Esmen N.A., and Hammad Y.Y. Log-normality of environmental sampling data. J Environ Sci Health, Part A 1977: 12 (1): 29–41.

    Article  Google Scholar 

  28. Fasano W.J., Kennedy G.L., Szostek B., Farrar D.G., Ward R.J., Haroun L., and Hinderliter P.M. Penetration of ammonium perfluorooctanoate through rat and human skin in vitro. Drug Chem Toxicol 2005: 28 (1): 79–90.

    Article  CAS  Google Scholar 

  29. Fromme H., Schlummer M., Moller A., Gruber L., Wolz G., Ungewiss I., Bohmer S., Dekant W., Mayer R., Liebl B., and Twardella D. Exposure of an adult population perfluorinated substances using duplicate diet portions and biomonitoring data. Environ Sci Technol 2007: 41: 7928–7933.

    Article  CAS  Google Scholar 

  30. Fromme H., Tittlemier S.A., Völkel W., Wilhelm M., and Twardella D. Perfluorinated compounds – exposure assessment for the general population in western countries. Int J Hyg Environ Health 2009: 212: 239–270.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. FSA. Fluorinated Chemicals: UK Dietary Intakes (FSIS 11/06). UK Food Standards Agency. Available at,, 2006.

  32. Giesy J.P., and Kannan K. Global distribution of perfluorooctane sulfonate in wildlife. Environ Sci Technol 2001: 35 (7): 1339–1342.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Giesy J.P., Mabury S.A., Martin J.W., Kannan K., Jones P.D., Newsted J.L., and Coady K.S. Perfluorinated compounds in the Great Lakes. Hdb Env Chem 2006: 5 (Part N): 391–438.

    Google Scholar 

  34. Griffith F.D., and Long J.E. Animal toxicity studies with ammonium perfluorooctanoate. Am Ind Hyg Ass J 1980: 41 (8): 576–583.

    Article  CAS  Google Scholar 

  35. Gulkowska A., Hang Q., So M.K., Taniyasu S., Lam P.K.S., and Yamashita N. Persistent perfluorinated acids in seafood collected from two cities in China. Environ Sci Technol 2006: 50: 3736–3741.

    Article  CAS  Google Scholar 

  36. Guruge K.S., Taniyasu S., Yamashita N., Wijeratna S., Mohotti K.M., Seneviratne H.R., Kannan K., Yamanaka N., and Miyazaki S. Perfluorinated organic compounds in human blood serum and seminal plasma: a study of urban and rural tea worker populations in Sri Lanka. J Environ Monit 2005: 7 (4): 371–377.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Hansen K.J., Johnson H.O., Eldridge J.S., Butenhoff J.L., and Dick L.A. Quantitative characterization of trace levels of PFOS and PFOA in the Tennessee River. Environ Sci Technol 2002: 36 (8): 1681–1685.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Hansen K.J, Clemen L.A., Ellefson M.E., and Johnson J.O. Compound-specific, quantitative characterization of organic fluorochemicals in biological matrices. Environ Sci Technol 2001: 35: 766–770.

    Article  CAS  Google Scholar 

  39. Harada K., Nakanishi S., Saito N., Tsutsui T., and Koizumi A. Airborne perfluorooctanoate may be a substantial source contamination in Kyoto area, Japan. Bull Environ Contam Toxicol 2005: 74: 64–69.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Harada K., Saito N., Inoue K., Yoshinaga T., Watanabe T., Sasaki S., Kamiyama S., and Koizumi A. The influence of time, sex and geographic factors on levels of perfluorooctane sulfonate and perfluorooctanoate in human serum over the last 25 years. J Occup Health 2004: 46 (2): 141–147.

    Article  CAS  Google Scholar 

  41. Harada K., Saito N., Sasaki K., Inoue K., and Koizumi A. Perfluorooctane sulfonate contamination of drinking water in the Tama River, Japan: estimated effects on resident serum levels. Bull Environ Contam Toxicol 2003: 71 (1): 31–36.

    Article  CAS  Google Scholar 

  42. Hölzer J., Midasch O., Rauchfuss K., Kraft M., Reupert R., Angerer J., Kleeschulte P., Marschall N., and Wilhelm M. Biomonitoring of perfluorinated compounds in children and adults exposed to perfluorooctanoate-contaminated drinking water. Environ Health Perspect 2008: 116 (5): 651–657.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Houde M., Martin J.W., Letcher R.J., Solomon K.R., and Muir D.C. Biological monitoring of polyfluoroalkyl substances: a review. Environ Sci Technol 2006: 40 (11): 3463–3473.

    Article  CAS  Google Scholar 

  44. Jahnke A., Ahrens L., Ebinghaus R., and Temme C. Urban versus remote air concentrations of fluorotelomer alcohols and other polyfluorinated alkyl substances in Germany. Environ Sci Technol 2007b: 41 (3): 745–752.

    Article  CAS  Google Scholar 

  45. Jahnke A., Huber S., Temme C., Kylin H., and Berger U. Development and application of a simplified sampling method for volatile polyfluorinated alkyl substances in indoor and environmental air. J Chromatogr A 2007a: 1164 (1-2): 1–9.

    Article  CAS  Google Scholar 

  46. Kallenborn R., Berger U., and Järnberg U. Perfluorinated alkylated substances (PFAS) in the Nordic environment. Nordic Council of Ministers, Copenhagen DK. TemaNord, 2004, 552.

    Google Scholar 

  47. Kannan K., Corsolini S., Falandysz J., Fillmann G., Kumar K.S., Loganathan B.G., Mohd M.A., Olivero J., Van Wouwe N., Yang J.H., and Aldous K.M. Perfluorooctanesulfonate and related fluorochemicals in human blood from several countries. Environ Sci Technol 2004: 38: 4489–4495.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Kannan K., Tao L., Sinclair E., Pastva S.D., Jude D.J., and Giesy J.P. Perfluorinated compounds in aquatic organisms at various trophic levels in a Great Lakes Food Chain. Arch Environ Contam Toxicol 2005: 48: 559–566.

    Article  CAS  Google Scholar 

  49. Kim S.K., and Kannan K. Perfluorinated acids in air, rain, snow, surface runoff, and lakes: relative importance of pathways to contamination of urban lakes. Environ Sci Technol 2007: 41 (24): 8328–8334.

    Article  CAS  Google Scholar 

  50. Kubwabo C., Stewart B., Zhu J., and Marro L. Occurrence of perfluorosulfonates and other perfluorochemicals in dust from selected homes in the city of Ottawa, Canada. J Environ Monit 2005: 7 (11): 1074–1078.

    Article  CAS  Google Scholar 

  51. Lau C., Anitole K., Hodes C., Lai D., Pfahles-Hutchens A., and Seed J. Perfluoroalkyl acids: a review of monitoring and toxicological findings. Toxicol Sci 2007: 99 (2): 366–394.

    Article  CAS  Google Scholar 

  52. Lewandowski G., Meissner E., and Milchert E. Special applications of fluorinated organic compounds. J Hazard Mater 2006: 136 (3): 385–391.

    Article  CAS  Google Scholar 

  53. Loewen M., Halldorson T., Wang F., and Tomy G. Fluorotelomer carboxylic acids and PFOS in rainwater from an urban center in Canada. Environ Sci Technol 2005: 39 (9): 2944–2951.

    Article  CAS  Google Scholar 

  54. Lorber M. Exposure of Americans to polybrominated diphenyl ethers. J Expo Sci Environ Epidemiol 2008: 18: 2–19.

    Article  CAS  Google Scholar 

  55. Lou I., Wambaugh J.F., Lau C., Hanson R.G., Lindstrom A.B., Strynar M.J., Zehr R.D., Setzer R.W., and Barton H.A. Modeling single and repeated dose pharmacokinetics of PFOA in mice. Toxicol Sci 2009: 107 (2): 331–341.

    Article  CAS  Google Scholar 

  56. 3M. Fluorochemical Use, Distribution and Release Overview. St. Paul, MN. U.S. Public Docket EPA-HQ-OPPT-2002-0051-0003. 1999.

  57. 3M. Analysis of PFOS, FOSA and PFOA From Various Food Matrices Using HPLC Electrospray/Mass Spectrometry. 3M Environmental Technology and Safety Services, Building 2-3E-09, PO Box 33331, St. Paul, MN 55133-3331. Centre Protocol Number: 00P-023-057; Centre Study Number: 023-057. June 21, 2001.

  58. 3M. Addendum Number 1 to the second semiannual progress report. Fluorochemical (FC) site related environmental assessment program 3M Cottage Grove, MN Facility. Prepared for 3M Company by Weston Solutions, Inc., West Chester, Pennsylvania 19380. W.O. No. 02181.002.010.00. Available at term, EPA-HQ-OPPT-2003-0012 2006.

  59. Martin J.W., Muir D.C., Moody C.A., Ellis D.A., Kwan W.C., Solomon K.R., and Mabury S.A. Collection of airborne fluorinated organics and analysis by gas chromatography/chemical ionization mass spectrometry. Anal Chem 2002: 74 (3): 584–590.

    Article  CAS  Google Scholar 

  60. McDonald T.A. Polybrominated diphenyl ether levels among United States residents: daily intake and risk of harm to the developing brain and reproductive organs. Int Env Ass Mgmt 2005: 1 (4): 343–354.

    Article  CAS  Google Scholar 

  61. Moody C.A., Hebert G.N., Strauss S.H., and Field J.A. Occurrence and persistence of perfluorooctanesulfonate and other perfluorinated surfactants in groundwater at a fire-training area at Wurtsmith Air Force Base, Michigan, USA. J Environ Monit 2003: 5 (2): 341–345.

    Article  CAS  Google Scholar 

  62. Moriwaki H., Takata Y., and Arakawa R. Concentrations of perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA) in vacuum cleaner dust collected in Japanese homes. J Environ Monit 2003: 5 (5): 753–757.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  63. Mortimer D.N., Clarke D.B., Gem M., and Rose M. Perfluorinated compounds in the UK 2004 Total Diet. Organohalogen Comp 2006: 68: 371–374.

    CAS  Google Scholar 

  64. Nabb D.L., Szostek B., Himmelstein M.W., Mawn M.P., Gargas M.L., Sweeney L.M., Stadler J.C., Buck R.C., and Fasano W.J. In vitro metabolism of 8-2 fluorotelomer alcohol: interspecies comparisons and metabolic pathway refinement. Toxicol Sci 2007: 100 (2): 333–344.

    Article  CAS  Google Scholar 

  65. Nakayama S., Strynar M.J., Helfant L., Egeghy P., Ye X., and Lindstrom A.B. Perfluorinated compounds in the Cape Fear Drainage Basin in North Carolina. Environ Sci Technol 2007: 41 (15): 5271–5276.

    Article  CAS  Google Scholar 

  66. Noker P A pharmacokinetic study of potassium perfluorooctanoate in the cynaomologus monkey. Southern Research Institute Study ID: 99214. 2003.

  67. OECD. Co-operation on existing chemicals – hazard assessment of perfluorooctane sulfonate (PFOS) and its salts. Organisation for Economic Co-operation and Development 2002.

  68. Olsen G.W., Burris J.M., Ehresman D.J., Froehlich J.W., Seacat A.M., Butenhoff J.L., and Zobel L.R. Half-life of serum elimination of perfluorooctanesulfonate, perfluorohexanesulfonate, and perfluorooctanoate in retired fluorochemical production workers. Env Health Persp 2007a: 115 (9): 1298–1305.

    Article  CAS  Google Scholar 

  69. Olsen G.W., Burris J.M., Mandel J.H., and Zobel L.R. Serum perfluorooctane sulfonate and hepatic and lipid clinical chemistry tests in fluorochemical production employees. J Occup Environ Med 1999: 41: 799–806.

    Article  CAS  Google Scholar 

  70. Olsen G.W., Church T.R., Larson E.B., van Belle G., Lundbert J.K., Hansen K.J., Burrsi J.M., Mandel H.J., and Zobel L.R. Serum concentrations of perfluorooctanesulftonate and other fluorochemicals in an elderly population from Seattle, Washingon. Chemosphere 2004: 54: 1599–1611.

    Article  CAS  Google Scholar 

  71. Olsen G.W., Church T.R., Miller J.P., Burris J.M., Hansen K.J., Lundberg J.K., Armitage J.B., Herron R.M., Medhdizadehkashi Z., Nobiletti J.B., O'Neill E.M., Mandel J.H., and Zobel L.R. Perfluorooctanesulfonate and other fluorochemicals in the serum of American Red Cross adult blood donors. Environ Health Perspect 2003a: 111 (16): 1892–1901. Erratum in: Environ Health Perspect. 111(16):1900.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  72. Olsen G.W., Hansen K.J., Stevenson L.A., Burris J.M., and Mandel J.H. Human donor liver and serum concentrations of perfluorooctanesulfonate and other perfluorochemicals. Environ Sci Technol 1 2003b: 37 (5): 888–891.

    Article  CAS  Google Scholar 

  73. Olsen G.W., Huang H.Y., Helzlsouer K.J., Hansen K.J., Butenhoff J.L., and Mandel J.H. Historical comparison of perfluorooctanesulfonate, perfluorooctanoate, and other fluorochemicals in human blood. Environ Health Perspect 2005: 113 (5): 539–545.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  74. Olsen G.W., Mair D.C., Reagen W.K., Ellefson M.E., Ehresman D.J., Butenhoff J.L., and Zobel L.R. Preliminary evidence of a decline in perfluorooctanesultonate (PFOS) and perfluorooctaneoate (PFOA) concentrations in American Red Cross blood donors. Chemosphere 2007b: 68: 105–111.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  75. Pang Y., MacIntosh D.L., Camann D.E., and Ryan P.B. Analysis of aggregate exposure to chlorpyrifos in the NHEXAS-Maryland investigation. Environ Health Perspect 2002: 110 (3): 235–240.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  76. Saito N., Sasaki K., Nakatome K., Harada K., Yoshinaga T., and Koizumi A. Perfluorooctane sulfonate concentrations in surface water in Japan. Arch Environ Contam Toxicol 2003: 45 (2): 149–158.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  77. Sasaki K., Harada K., Saito N., Tsutsui T., Nakanishi S., Tsuzuki H., and Koizumi A. Impact of Airborne Perfluorooctane Sulfonate on the Human Body Burden and the Ecological System. Bull Environ Contam Toxicol 2003: 71: 408–413.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  78. Schultz M.M., Barofsky D.F., and Field J.A. Quantitative determination of fluorotelomer sulfonates in groundwater by LC MS/MS. Environ Sci Technol 2004: 38 (6): 1828–1835.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  79. Seacat A.M., Thomford P.J., Hansen KJ., Olsen G.W., Case M.T., and Butterhoff J.L. Subchronic toxicity studies on perfluorooctane sulfonate potassium salt in Cynomolgus Monkeys. Toxicol Sci 2002: 68: 249–264.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  80. Shoeib M., Harner T., Ikonomou M., and Kannan K. Indoor and outdoor air concentrations and phase partitioning of perfluoroalkyl sulfonamides and polybrominated diphenyl ethers. Environ Sci Technol 2004: 38 (5): 1313–1320.

    Article  CAS  Google Scholar 

  81. Shoeib M., Harner T., Wilford B.H., Jones K.C., and Zhu J. Perfluorinated sulfonamides in indoor and outdoor air and indoor dust: occurrence, partitioning, and human exposure. Environ Sci Technol 2005: 39 (17): 6599–6606.

    Article  CAS  Google Scholar 

  82. Sinclair E., Kim S.K., Akinleye H.B., and Kannan K. Quantitation of gas-phase perfluoroalkyl surfactants and fluorotelomer alcohols released from nonstick cookware and microwave popcorn bags. Environ Sci Technol 2007: 41 (4): 1180–1185.

    Article  CAS  Google Scholar 

  83. Sinclair E., Mayack D.T., Roblee K., Yamashita N., and Kannan K. Occurrence of perfluoroalkyl surfactants in water, fish, and birds from New York State. Arch Environ Contam Toxicol 2006: 50 (3): 398–410.

    Article  CAS  Google Scholar 

  84. Skutlarek D., Exner M., and Farber H. Perfluorinated surfactants in surface and drinking waters. Environ Sci Pollut Res Int 2006: 13 (5): 299–307.

    Article  CAS  Google Scholar 

  85. So M.K., Yamashita N., Taniyasu S., Jiang Q., Giesy J.P., Chen K., and Lam P.K.S. Health risks to infants associated with exposure to perfluorinated compounds in human breast milk from Zhoushan, China. Env Sci Tech 2006: 40: 2924–2929.

    Article  CAS  Google Scholar 

  86. Stapleton H.M., Kelly S.M., Allen J.G., McClean M.D., and Webster T.F. Measurement of polybrominated diphenyl ethers on hand wipes: estimating exposure from hand-to-mouth contact. Env Sci Tech 2008: 42: 3329–3334.

    Article  CAS  Google Scholar 

  87. Steenland K., Jin C.F., MacNeil J., Lally C., Ducatman A., Vieira V., and Fletcher T. Predictors of PFOA levels in a community surrounding a chemical plant. Environ Health Perspect 2009: 117: 1083–1088.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  88. Stock N.L., Lau F.K., Ellis D.A., Martin J.W., Muir D.C., and Mabury S.A. Polyfluorinated telomer alcohols and sulfonamides in the North American troposphere. Environ Sci Technol 2004: 38 (4): 991–996.

    Article  CAS  Google Scholar 

  89. Strom D.J., and Stansbury P.S. Determining parameters of lognormal distributions from minimal information. AIHAJ 2000: 61 (6): 877–880.

    Article  CAS  Google Scholar 

  90. Strynar M.J, and Lindstrom A.B Perfluorinated compounds in house dust from Ohio and North Carolina, USA. Environ Sci Technol 2008: 42 (10): 3751–3756.

    Article  CAS  Google Scholar 

  91. Takagi S., Adachi F., Miyano K., Koizumi Y., Tanaka H., Mimura M., Watanabe I., Tanabe S., and Kannan K. Perfluorooctanesulfonate and perfluorooctanoate in raw and treated tap water from Osaka, Japan. Chemosphere 2008: 72 (10): 1409–1412.

    Article  CAS  Google Scholar 

  92. Taniyasu S., Kanna K., Horii Y., Hanari N., and Yamashita N. A survey of perfluorooctane sulfonate and related perfluorinated organic compounds in water, fish, birds, and humans from Japan. Environ Sci Technol 2003: 37: 2634–2639.

    Article  CAS  Google Scholar 

  93. Tao L., Kannan K., Wong C.M., Arcaro K.F., and Butenhoff J.L. Perfluorinated compounds in human milk from Massachusetts. USA Environ Sci Technol 2008: 42 (8): 3096–3101.

    Article  CAS  Google Scholar 

  94. Tittlemier S., Edwards L., and Pepper K. Concentrations and temporal trends of two perfluorooctyl sulfonamides in fast food composites collected during the Canadian Total Diet Study. Organohalogen Comp 2003: 62: 315–318.

    CAS  Google Scholar 

  95. Tittlemier S.A., Pepper K., and Edwards L. Concentrations of perfluorooctanesulfonamides in Canadian Total Diet Study composite food samples collected between 1992 and 2004. J Agric Food Chem 2006: 54 (21): 8385–8389.

    Article  CAS  Google Scholar 

  96. Tittlemier S.A., Pepper K., Seymour C., Moisey J., Bronson R., Cao X.L., and Dabeka R.W. Dietary exposure of Canadians to perfluorinated carboxylates and perfluorooctane sulfonate via consumption of meat, fish, fast foods, and food items prepared in their packaging. J Agric Food Chem 2007: 55 (8): 3203–3210.

    Article  CAS  Google Scholar 

  97. Trudel D., Horowitz L., Wormuth M., Scheringer M., Cousins I.T., and Hungerbuhler K. Estimating consumer exposure to PFOS and PFOA. Risk Analysis 2008: 28: 251–269.

    Article  Google Scholar 

  98. Veronica V., Webster T., Bartell S., Steenland K., Savitz D., and Fletcher T. PFOA community health studies: exposure via drinking water contaminated by a teflon manufacturing facility. Organohalogen Comp 2008: 70: 730–732.

    Google Scholar 

  99. Vestergren R., Cousins I.T., Trudel D., Wormuth M., and Sheringer M. Considering the role of precursor compounds in consumer exposure to PFOS and PFOA. Organohalogen Comp 2008: 70: 1442–1466.

    CAS  Google Scholar 

  100. Washburn S.T., Bingham T.S., Braithwaite S.K., Buck R.C., Buxton L.W., Clewell H.J., Haroun L.A., Kester J.E., Rickard R.W., and Shipp A.M. Exposure assessment and risk characterization for perfluorooctanoate in selected consumer articles. Environ Sci Technol 2005: 39: 3904–3910.

    Article  CAS  Google Scholar 

  101. Yuan W., Yin Y.L.W., Nobuyoshi Y., Sachi T., Ka S.M., Murphy M.B., and Sing L.P.K. Perfluorooctane sulfonate (PFOS) and related fluorochemicals in chicken egg in China. Chinese Science Bulletin 53 2008.

Download references


We thank our colleagues in the U.S. Environmental Protection Agency Office of Research and Development (ORD) and the Office of Pollution Prevention and Toxics for their discussions, reviews, and comments. We particularly thank Linda Sheldon, Andrew Lindstrom, and Mark Strynar for their expert advice and discussions about the published data. We are also indebted to our reviewers, Nicolle Tulve and Daniel Chang of the ORD.

Author information



Corresponding author

Correspondence to Peter P Egeghy.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information


The U.S. Environmental Protection Agency through its Office of Research and Development funded and managed the research described here. Although it has been subjected to Agency's administrative review and approved for publication, it does not necessarily reflect the views of the Agency and no official endorsement should be inferred.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Egeghy, P., Lorber, M. An assessment of the exposure of Americans to perfluorooctane sulfonate: A comparison of estimated intake with values inferred from NHANES data. J Expo Sci Environ Epidemiol 21, 150–168 (2011).

Download citation


  • human exposure assessment, perfluorinated compounds, pharmacokinetic modeling

Further reading


Quick links