Occurrence and Risks of Per- and Polyfluoroalkyl Substances in Shellfish

Abstract

Purpose of review:

Per- and polyfluoroalkyl substances (PFAS) are a diverse class of persistent, fluorinated surfactants used widely in industrial and commercial applications with known adverse health effects. Seafood consumption is thought to be an underappreciated source of PFAS exposure in the general population. This review synthesizes the current understanding of PFAS occurrence in shellfish, a term used to describe animals such as mollusk bivalves, certain gastropods (snails), cephalopods (e.g., octopuses and squid) and crustaceans, and highlights scientific gaps relative to bioaccumulation and the protection of shellfish consumers.

Recent findings:

A range of sampling methodologies are used across studies and the suite of PFAS surveyed across studies is highly variable. Concentrations of PFAS observed in shellfish vary by geographic location, shellfish species, habitat, and across PFAS compounds, and studies informing estimates of bioaccumulation of PFAS in shellfish are extremely limited at this time.

Summary:

This review identifies several important opportunities for researchers to standardize PFAS sampling techniques, sample preparation, and analytical methodologies to allow for better comparison of PFAS analytes both within and across future studies. Increasing the range of geographic locations where samples are collected is also a critical priority to support a greater knowledge of worldwide PFAS contamination. When put into the context of risk to consumer, concentrations of PFAS, especially PFOS, found in shellfish collected from sites containing aqueous film-forming foam (AFFF) and industrial contamination may present risks to frequent consumers. Further research is needed to protect shellfish consumers and to inform shellfish advisories and health protective policies.

Introduction

Per- and polyfluoroalkyl substances (PFAS) are a diverse class of fluorinated surfactants used in industrial and commercial applications and have resulted in contamination of surface water, groundwater, soil, and sediments. Examples of their applications include oil and stain repellants, food packaging, corrosion inhibition, and aqueous film-forming foam (AFFF) used against fuel fires [1–3]. These chemicals are prized for their unique properties and durability under extreme physicochemical conditions [4,5]. However, these same properties make PFAS highly persistent and mobile in the environment. Such traits have made PFAS contamination a global issue since it is detected in groundwater supplies [6,7], freshwater and marine surface water [8–11], and food products [12,13]. These chemicals are also detected in the blood, placenta and breastmilk of people from nearly every continent, which is most concerning [14–17].

Current evidence indicates that the primary route of exposure to PFAS is ingestion, especially contaminated drinking water [18,19] and seafood consumption [20,21]. In a cohort of Swedish mothers, Shu et al. [22] reported that maternal serum concentrations of certain PFAS were strongly correlated to seafood consumption rates, along with maternal age, parity and smoking history. The European Food Safety Authority (ESFA) CONTAM Panel has estimated that upwards of 86% of dietary PFAS exposure may be attributed to seafood consumption [12,13]. While these studies point to seafood consumption as a significant source of PFAS exposure, they do not differentiate the contribution of distinct seafood categories (e.g., fish, shellfish, roe).

Shellfish is a subcategory of seafood that is frequently scrutinized for its safety relative to other environmental hazards (e.g. microbial toxins and heavy metals) [23], but has received less attention when it comes to PFAS or other emerging contaminants. In the context of international trade or aquaculture, the term “shellfish” is restricted to only include invertebrates that both have an external shell and belong to either the phylum Mollusca or subphylum Crustacea [24]. However, for the purposes of this review, the colloquial definition of “shellfish” was adopted to include animals such as mollusk bivalves, certain gastropods (snails), cephalopods (e.g. octopuses and squid) and crustaceans. Most shellfish do not sit at the top of their respective food-chains and therefore are not considered demonstrative of the full biomagnification potential seen with PFAS in polar bears, seabirds, marine mammals, certain finfish and humans (reviewed by [25]). However, shellfish are a part of exposure pathways leading to the accumulation and magnification of PFAS to their consumers, including both larger aquatic organisms and humans residing at the top of the food chain (Figure 1). Yet the specific contribution of shellfish remains poorly understood [26]. Thus, efforts to reduce potentially harmful PFAS exposures from seafood requires an understanding of chemical occurrence, bioaccumulation, and dietary exposure in shellfish.

Figure 1.

Conceptual model for PFAS exposure to humans via the consumption of shellfish. Boxes represent compartments where PFAS may accumulate. Arrows represent processes like bioaccumulation (A), bioconcentration (B), and biomagnification (C). Red text denotes taxa that meet the definition of shellfish in this review.

This review paper will evaluate the current understanding of PFAS occurrence in mollusks (e.g., bivalves, gastropods, and cephalopods) and crustaceans to identify knowledge gaps relative to bioaccumulation and the protection of consumers. The purpose of this is to provide researchers and regulators with a narrative summary on the current state of PFAS science related to this class of seafood. To address this goal, we first provide an overview of studies to evaluate global PFAS occurrence in freshwater, brackish, and marine invertebrates, and discuss methods for sampling and detection. The subsequent text summarizes the limited literature around the process of PFAS bioaccumulation into shellfish, and the implication for shellfish harvesters, and consumers. To capture taxonomic variability in the uptake of PFAS, this paper takes a broad definition of shellfish. This includes several non-dietary aquatic invertebrates, which are either part of the food chain for dietary species or may help to inform our understanding of PFAS uptake among dietary species for which data are not yet available (Supplemental Material, Table S1). We refer to these non-dietary and dietary aquatic invertebrates as shellfish throughout this manuscript. This summary allows for comparison of study methodologies, an understanding of temporal trends in the number of PFAS analytes being analyzed, and current challenges of studying PFAS that should be addressed in future studies.

Identification of Studies Investigating PFAS Concentrations in Shellfish

A total of thirty-one studies were identified from database searches relating to the occurrence of PFAS in shellfish worldwide. These studies were screened to determine if: 1) the study is peer-reviewed and published prior to December 2021, 2) the study analyzes raw, soft tissues, from a known origin and 3) the study reports PFAS concentrations by individual species. If any of the above criteria were not met, the study was excluded from this review. Of the original thirty-one studies, eight were excluded from this review based on the criteria and partial data from two studies were also excluded (See Supplemental Methods M1).

Twenty-three studies, representing a range of geographic locations (Figure 2, Supplemental Table S2, Supplemental Figure S1) met the review criteria. Authors were contacted for additional information as needed. These studies were published between 2005 and 2021 and provided measurements of perfluoroalkyl carboxylic acids (PFCA), perfluoroalkyl sulfonic acids (PFSA), and some precursor PFAS (i.e., perfluoroalkyl sulfonamides and fluorotelomer alcohols). When the species from these reports were divided into broad taxonomic classes, Bivalvia were studied most often (21 studies), followed by Crustacea (12 studies), Gastropoda (4 studies) and Cephalopoda (3 studies). At the time of writing, there is a growing body of grey literature that reveals significant regional efforts to evaluate shellfish through various monitoring programs (reviewed by [25,27]).

Figure 2.

Locations of PFAS occurrence studies described in this review (n = 23).

Methodologies for Shellfish Sampling

Although there are several valid approaches for assessing PFAS concentrations in seafood, aspects of the methodology for sampling shellfish influence the interpretation of study findings. These include the geographic location of the study, whether shellfish were sourced from fish markets or directly harvested, and use of individual or composite tissue samples.

Geographic location and background PFAS contamination

PFAS are globally distributed contaminants, but a study’s location can impact the reported concentrations found in shellfish. Studies sampling from areas without specific sources represent the majority (71%) of PFAS occurrence studies in shellfish worldwide (see Supplemental Material, Methods M1). These sites are representative of the widespread occurrence of PFAS and include bivalve monitoring program sites [28,29], commercial fishing zones, such as the coastal waters of China [30–33] and the Mediterranean Sea [34–37], and communal waterways that are not linked to specific industrial or waste facilities [38–41]. These types of studies are critical to understanding the potential for background occurrence of PFAS in biota as these persistent chemicals are subject to long-term circulation in ecosystems [8,21,25]. However, the tissue burdens in shellfish are expected to be lower than those of animals near sites with industrial releases or uncontrolled spills.

A smaller number of studies (n = 6) deliberately sampled around sites with known PFAS contamination, such as AFFF use near airports or military installations [42–45], waste-water treatment facilities [46,47], and industrial manufacturing areas [48]. Military sites and airports that currently or previously used AFFF are likely hotspots for PFAS contamination, especially compounds like perfluorooctane sulfonic acid (PFOS), perfluorooctanoic acid (PFOA), perfluorohexane sulfonic acid (PFHxS) and certain long-chain PFAS [45,49]. Given that the set of PFAS analytes differs depending on the application, it is important to consider the source of samples when evaluating shellfish data from different studies.

Shellfish sources and collection

The primary methods for collecting shellfish samples for PFAS analysis included buying shellfish at local fish markets (17% of studies) and direct sampling from the site(s) of interest (83% of studies) (Supplemental Figure S1). There were 3 studies that used samples from local markets [30,37,50]. None of these studies discuss any verification of shellfish sources. However, two of the studies [30,37] purchased shellfish from fish markets in large, coastal cities connected to globally significant fisheries. These fisheries are more than likely the sources of the shellfish analyzed.

Individual versus composite sampling

Some studies analyzed PFAS concentrations of individual animals from a given species (25% of studies), though the majority (75% of studies) created homogenous composites from multiple individuals of a single species. Using composites of many individuals creates enough tissue for analysis, especially in the case of species with very little soft tissue and reduces inter-individual variability of PFAS occurrence. Furthermore, homogenous composites are more representative of human dietary exposure to PFAS and therefore, applicable to dietary risk assessment, which is consistent with regulatory approaches taken for other chemical contaminants. However, non-composite sampling allows for the estimation of intra- and interspecies variability in occurrence which has value for toxicokinetic modeling of uptake and refined estimates of bioaccumulation.

Limitations and future directions

In some geographic regions, knowledge of PFAS in environmental media is limited but insights about local industrial and municipal activities (e.g., wastewater treatment plants) remain a critical component of contextualizing PFAS concentrations observed in shellfish. Additional concerns arise from cross contamination from food packaging [51], sampling equipment and a sampler’s gear or personal care products (reviewed by ITRC [2]) which have plagued studies of various environmental media.

Future studies would benefit from considering potential sources of shellfish tissue, as market seafood samples are of special interest for assessing consumer risks. Direct sampling from the target site(s) is favored when the study goals include analysis of abiotic factors from the source environment (i.e., water and sediment) for quantifying bioaccumulation. This also reduces the chance of post-collection sample contamination by fisher people and vendors due to the aforementioned widespread use of PFAS in commercial products. Choice of individual or composite sampling must match the overarching needs of the study with individual sampling typically being more useful for toxicokinetic modeling and composite sampling being favored to inform regulatory guidance.

Detection of PFAS in Shellfish: Differences Across Habitats and Taxonomic Groups

Below we summarize the analytic chemistry methods and PFAS analytes observed in shellfish. Further, we assessed patterns and variations in analytes detected and their concentrations according to shellfish habitat (brackish, freshwater, saltwater) and taxonomic groupings (bivalves, cephalopods, crustaceans, and gastropods).

Analytic chemistry methods for PFAS assessment

The two primary platforms for measuring PFAS in biological matrices are liquid chromatography with tandem mass spectrometry (LC-MS/MS) and high-performance liquid chromatography (HPLC) (Supplemental Figure S1). Due to the evolving nature of PFAS analyses in various biological matrices, these analytical methods were not consistently applied across all studies creating large variation in detection limits and data quality. This lack of consistency is partially attributed to inter-lab and inter-instrument variability and the lack of a standardized detection method for a growing list of PFAS analytes found within different biological samples. Most studies report their data relative to their limit of detection (LOD) (62% of studies), while others use limit of reporting (LOR) or limit of quantification (LOQ) to report their findings. Either way, the ability to detect lower concentrations and a growing list of analytes has improved over time (Supplemental Figure S2). In one early study, Bossi, et al. [46] used LC-MS/MS to measure PFOS and PFOA with detection limits of 0.2 ng/g and 1.2 ng/g, respectively, which was comparable to other studies at that same time [35]. More recently, detection limits with LC-MS/MS have decreased by several orders of magnitude to 0.006 ng/g and 0.01 ng/g for PFOS and PFOA, respectively [31,52]. Similar trends are reported for other PFAS analytes (Supplemental Table S3).

Detected PFAS and their concentrations

Across all studies, the number of PFAS targeted ranged from one to 25 analytes, averaging nine PFAS analytes targeted in each study. A full table of these compounds and corresponding acronyms is provided in Supplemental Table S3. Of those analytes, PFOS was the most targeted analyte (n = 24 studies), followed by PFOA (n = 23 studies), perfluorodecanoic acid (PFDA) (n = 18 studies), and perfluoroundecanoic acid (PFUnDA) (n = 18 studies). The dominance of PFOS and PFOA in most studies is a result of research and regulatory interests in the early 2000s, but this has been followed by growing awareness of novel PFAS replacement products after the phase out of PFOS and PFOA in North America and Europe in the subsequent decade [2,21]. Advances in detection limits have been matched by a growing list of PFAS analytes resulting in a significant increase in reported PFAS profiles (Supplemental Figure S2). The analytical profile tends to be biased towards PFCAs over PFSAs, fluorotelomers and precursor sulfonamides. This growing list of analytes creates a moving target for both academic researchers and regulators seeking to prioritize PFAS monitoring efforts.

PFOS was the predominant PFSA reported in all studies with the highest average and maximum concentrations of any PFAS (Figure 3). The average PFOS concentrations ranged from 0.001 ng/g (DL) to 72.0 ng/g across all taxonomic groups, [28,30,32–34,36,38,39,41,42,45,47,48], and a maximum reported concentration of 125.9 ng/g in Mediterranean mussels (Mytilus galloprovincialis) off the coast of Portugal [38]. In the reviewed studies, the average concentration of PFOS in cephalopods (1.08 ± 1.11 ng/g_ww) (mean ± SEM)) tended to be lower than other shellfish, where bivalves (1.82 ± 10.4 ng/g_ww) and crustaceans (3.94 ± 8.40 ng/g_ww) had higher average tissue burdens of PFOS (Supplemental Figure S3). Bivalves are generally primary consumers (i.e., herbivores and detritivores) that feed lower in the food web whereas cephalopods and crustaceans can be secondary consumers (i.e., predators and omnivores). This pattern across these shellfish studies is consistent with reports from aquatic predators, where PFOS is the primary PFAS found to biomagnify through aquatic food webs [25], although some discrepancy has been reported across the limited number of studies evaluating cephalopods.

Figure 3.

Concentrations of perfluoroalkyl carboxylic acids (PFCAs) for bivalves (A), cephalopods (C), crustaceans (E) and gastropods (G) along with perfluoroalkyl sulfonic acids (PFSAs) bivalves (B), cephalopods (D), crustaceans (F) and gastropods (H). All concentrations are reported as mean ± standard error of the mean (SEM) of averages reported in occurrence studies

PFOA was also frequently detected in studies (n = 22) with overall average concentrations similar to PFOS. Unlike PFOS, the average tissue concentration of PFOA in bivalves (1.51 ± 3.44 ng/g_ww) tended to be higher than those observed in cephalopods (0.610 ± 0.959 ng/g_ww), crustaceans (0.356 ± 0.440 ng/g_ww) and gastropods (0.122 ± 0.114 ng/g_ww). A longer history of monitoring for PFOA has likely contributed to an overall higher average concentration of PFOA in bivalves, with some of the earliest studies predating the PFOA/PFOS Phase Out in several Western countries. At least 4 studies found PFOA concentrations exceeding those of all other PFAS [32–34,48] while some only report the occurrence of PFOA [31,35,36,39,47].

The remaining compounds detected in these studies can be broadly divided into long-chain (> 6 carbons) or short chain (≤ 6 carbons) PFAS. Commonly measured long-chain PFAS, outside of PFOS and PFOA, include perfluorononanoic acid (PFNA), perfluorodecanoic acid (PFDA), PFUnDA, and perfluorodecane sulfonic acid (PFDS). With the exception of PFDS, these other long-chain PFAS have been detected in bivalves (i.e. oysters and mussels) and some crustaceans [28,37,41,53]. Vassiliadou, et al. [37] attempted to measure a wide array of long-chain PFCAs in cooked and uncooked shellfish, including analytes like perflurotridecanoic acid (PFTrDA), perfluorotetradecanoic acid (PFTeDA), and perfluoropentadecanoic acid (PFPeDA). While the authors reported detections for some of these compounds, the LOD (0.18 ng/g_ww – 2.65 ng/g_ww) were likely too high to accurately quantify these substances [37]. More recently, Guo et al. [31] utilized lower LODs (0.003–0.005 ng/g) to screen for long-chain PFAS in bivalves (e.g. clams, scallops, oysters and mussels) from Bohai Bay, China, and did not detect PFAS other than PFOA.

Like long-chain PFAS, shorter-chain compounds have been understudied and therefore infrequently reported in the current literature. Compounds like perfluorobutane sulfonic acid (PFBS) and perfluorobutanoic acid (PFBA) are known replacement chemistries for older PFAS [2,20] and potentially the breakdown products of precursor PFAS, so they are expected to co-occur with other PFAS in the environment. However, the studies that have tried to measure PFBS or PFBA have failed to detect either analyte above existing LODs [31,33,37,40,42,53]. Perfluoropetnaoic acid (PFPeA) is present in crustaceans, including deep water nose shrimp (Parapenaeus longirostris [37]), school prawn (Metapenaeus macleaya), and mud crab (Scylla serrata) [44], but has not been detected in mussels or clams [31,37]. As discussed later, the lack of detection despite their environmental presence is likely due to an interaction of current detection limits and differences in bioaccumulation for certain PFAS.

Habitat type and major taxonomic groups represented

The limited available studies suggest that PFAS concentrations vary across different habitat types and major taxonomic groups within those habitats (Figure 3). Across all habitats, PFOS and perfluorooctane sulfonamide (PFOSA) had the highest concentrations of any analyte. The highest average concentration of PFOS (7.10 ng/g) was in brackish water and the highest average concentration of PFOSA (1.37 ng/g) was in freshwater organisms. This variation may be related to the number of species within each group of shellfish, differences in their feeding strategies or location of study. For example, longer chained PFAS are associated with water column and sediment solids while short-chain PFAS are frequently dissolved in water [54]. Differences between bivalve taxa have been observed in which concentrations of PFAS in clams>mussels>scallops>oysters, but this trend depends on the analyte [31]. The authors suggest that the higher concentrations of PFOA in clams is related to their feeding habits and lifestyle habits in the sediments where PFOA concentrations may be higher than in the water column, where mussels feed [31,55]. Thus, benthic feeders might contain higher concentrations of long-chain PFAS while filter feeders might preferentially bioaccumulate short-chain compounds.

For cephalopods and crustaceans, each representative of different phyla, the highest concentrations of PFCAs and PFSAs were found in saltwater environments over brackish and there appeared to be increased concentration of long-chained PFCAs (>6 carbons) with salinity. A similar pattern was seen in experiments with blackrock fish and Pacific oysters where uptake and elimination constants increased with salinity [56]. In this review, while the differences in PFAS concentrations by salinity for the cephalopods and crustaceans may be due to the effects of salinity directly on uptake, these differences could also be due to the individual species within each group found in the different habitats (Figure 3). For crustaceans, some species of shrimp feed on particles in the water column while lobsters are predatory and benthic feeders. Within cephalopods, there are highly predatory species of octopus that consume large benthic prey and others like cuttlefish that feed in the water column on small invertebrates and fish. Future studies could compare taxa representing different feeding strategies and sources of food as well as PFAS occurrence in individual species across different habitat types and/or salinities. Given the limited data on PFAS, trends within each habitat of each taxonomic group cannot be easily identified and require further research.

Limitations and future directions

A key limitation to these studies is the suite of PFAS analyzed over time, with a distinct bias towards PFOA and PFOS that may influence the averages described above. The growing list of analytes creates a moving target for both academic researchers and regulators seeking to prioritize PFAS monitoring efforts. While this allows for the identification of additional PFAS that may present risks to human and ecological health, this also confounds some comparisons as detection limits and the presentation of data varies greatly across studies. A critical need for future studies is providing as much data as possible for comparison across different analytical platforms (e.g., LC-MS/MS and HPLC).

The narrow geographic scope of these studies creates another challenge to understanding PFAS occurrence in shellfish, where limited data from understudied regions (i.e., Africa, South America, and Australia) creates uncertainty for the broader application of findings from heavily studied regions such as Europe and Asia. When considering potential impacts of feeding strategies or food sources of marine invertebrates, more studies are needed to determine if there are unique PFAS bioaccumulation processes related to organism life histories that result in differences in PFAS concentrations and compound profiles.

To improve the literature in this field, future monitoring efforts should identify a standard suite of PFAS or analytes for “finger-printing” the sources of PFAS. This could include the use of extractable organic fluorine or total organic fluorine analyses to better understand what is possibly being missed in studies that target a limited number of analytes across a diverse array of shellfish [57]. There is also a need for more studies that conduct individual sampling for comparisons within and between species to inform the role of ecological behaviors and physiological differences in PFAS uptake.

Bioaccumulation in Shellfish

Overall, the process of PFAS bioconcentration and bioaccumulation in shellfish is poorly understood. Broadly, bioconcentration factors (BCFs) quantify the process of chemical uptake into an organism via respiration (or ventilation) and dermal contact, while bioaccumulation factors (BAFs) include these same routes along with ingestion[2,58,59]. This is complex for shellfish as the variety in species is matched by variety in ecological roles and habitats. By understanding and quantifying bioaccumulation, it is possible to predict where environmental concentrations of PFAS may cause tissue burdens of concern for shellfish consumers.

Bioconcentration and bioaccumulation Factors

A recent review by Burkhard [27] provides a comprehensive summary and evaluation of available BCF and BAF studies in aquatic organisms for a universe of nearly 1,500 PFAS as defined by the Organization for Economic Cooperation and Development (OECD). This included BCFs/BAFs from a large body of literature for fish, as well as bioaccumulation data for 6 PFCAs (PFHxA, PFHpA, PFOA, PFNA, PFDA, PFUnDA) and 4 PFSAs (PFBS, PFHxS, PFOS, FOSA) from the smaller number of studies on bivalves, gastropods, and crustaceans. A significant finding from this review is that ~60% of PFAS BAFs are greater than the BCF for the same organisms, which indicates that uptake of PFAS includes ingestion and not water exposure alone [27]. Burkhard [27] also found that the BAFs for several of the shellfish taxa included in this current review displayed orders-of-magnitude variation in BAFs within and across taxa. Taken together, this points to differences in PFAS bioaccumulation that need to be assessed in order to evaluate risks to shellfish consumers. For readers interested in more technical details about BAF/BCF derivation, we recommend the review of Burkhard [27].

Functional groups and carbon-chain length

There are suspected differences in bioaccumulation based on chemical structure of specific PFAS; namely the functional groups and the carbon-chain length. Prior research indicates that BAFs increase with carbon chain length in other aquatic organisms [28]. Based on currently available BAFs for PFAS in shellfish, there are no definitive trends between the PFAS structure and BAFs (Figure 4) although individual studies suggest that BAFs may be related to both chain length and functional group [60–64]. BAFs for PFCAs in a sampling of oysters and crustaceans increased linearly with carbon chain length [28]. Liu et al. [62] similarly reported higher BAFs for long-chain PFAS compared to short chain PFAS in mussels, with a significantly higher BAF for PFSAs. Hong et al. [61] elaborated on this finding regarding functional group where PFSAs had higher BAFs than PFCAs of the same chain length in bivalves, gastropods, and crustaceans. Current variability in BAFs may be due to the limited number of studies on PFAS bioaccumulation in shellfish, highlighting the need for additional bioaccumulation studies.

Figure 4:

Log transformed bioaccumulation factors (BAFs) for PFAS identified in this review. Y-axis represents log values for BAF while x-axis is studied PFCA and PFSA ordered by chain length.

Limitations and future directions

There are two clear gaps in the literature related to PFAS bioaccumulation in shellfish. The first is the small number of studies in a very limited number of species. Only nine studies derived BAF/BCFs for 12 PFAS across 5 bivalve species[28,40,56,61,62], 7 crustacean species [28,40,60,61,64], and 3 gastropod species [40,61,64]. The second gap is a poor understanding behind mechanisms of PFAS uptake in shellfish. In laboratory studies using Daphnia magna, a non-dietary species, there was no association between consumption of food and BAFs, indicating that dietary uptake was not as important as interaction with dissolved PFAS [65]. Other studies found increasing concentrations of proteins, either bovine albumin or soy peptone [66], or humic substances [67] resulted in reduced bioaccumulation of PFAS in D. magna. Understanding of these mechanistic processes for uptake is critical to predicting toxicokinetics of these and similar compounds in dietary species as the family of PFAS analytes continues to grow.

Potential Risk to Consumers

Epidemiological studies support associations of PFAS with a wide range of adverse health effects, including diminished immune response, thyroid disorders, liver and kidney dysfunction, hyperlipidemia, low birth weight, and some cancers [68]. For a detailed review of the human health effects of PFAS, we recommend the review by Fenton et al. 2021[68]. Estimating the risk of PFAS exposure for shellfish consumers is a challenging task as risk is quantified by the amount of a given chemical in shellfish, frequency of consumption, and the known toxicity of a substance. Based on the previously discussed studies, we provide an example of how risk might be considered using quantitative risk assessment methods combined with observed ranges of PFAS concentrations. Specifically, this approach uses a ratio of the ingested dose of chemical consumed divided by a threshold dose for risk to the consumer to calculate a hazard quotient (See Supplemental Methods M2). Our approach estimates risks for 5 PFAS using typical exposure assumptions [67,69] (See Supplemental Methods M2). These are not intended to serve as recommended regulatory thresholds, rather an example for how risk might be estimated based on necessary assumptions given available data.

Our modeled exposure scenario to estimate risks uses concentrations from 5 PFAS reported in bivalves using a simple approach used for site-specific risk assessments (Supplemental Equation 1; [70]. In these scenarios, we considered routine exposure to PFAS in shellfish consumers from three specific segments of the population: adults (≥21 years of age), women of childbearing age (13–49 years of age) and children (ages 3–6 years of age) representing typical U.S. residents [69,71]. The average and high-end (90^th percentile) consumption rates were applied to adults (3.1 and 15.6 g/d), women of childbearing age (2.0 and 11.4 g/day) and children (0.5 and 3.5 g/day); central body weight assumptions were applied to adults (80 kg), women of child bearing age (74 kg) and children (18.6) to be conservative with exposure estimates [69,71].

We compared calculated daily exposure doses (mg/kg/day) against a Reference Dose (RfD) or Minimal Risk Level (MRL) to create a hazard quotient as follows: $\frac{\left(\frac{\left[\mathrm{T}\mathrm{i}\mathrm{s}\mathrm{s}\mathrm{u}\mathrm{e}\mathrm{P}\mathrm{F}\mathrm{A}\mathrm{S}\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{c}.\left(\mathrm{m}\mathrm{g}/\mathrm{g}\right)\right]\times \left[\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{s}\mathrm{u}\mathrm{m}\mathrm{p}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{R}\mathrm{a}\mathrm{t}\mathrm{e}\right(\mathrm{g}/\mathrm{d}\mathrm{a}\mathrm{y}\left)\right]}{\left[\mathrm{B}\mathrm{o}\mathrm{d}\mathrm{y}\mathrm{W}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\right(\mathrm{k}\mathrm{g}\left)\right]}\right)}{\left[\mathrm{R}\mathrm{f}\mathrm{D}\mathrm{o}\mathrm{r}\mathrm{M}\mathrm{R}\mathrm{L}\right(\mathrm{m}\mathrm{g}/\mathrm{k}\mathrm{g}/\mathrm{d}\left)\right]}$ (Supplemental Material, Supplemental Methods M2). To estimate the toxicity of 5 PFAS, we applied the Agency for Toxic Substances and Disease Registry’s (ATSDR) sub-chronic oral MRLs for PFOA (3×10⁻⁶ mg/kg/d), PFNA (3×10⁻⁶ mg/kg/d), PFHxS (2×10⁻⁵ mg/kg/d)and PFOS (2×10⁻⁶ mg/kg/d) [76] as these are currently suggested to be applicable for screening tools by the U.S. FDA [72]; but for PFBS, which lacks an ATSDR MRL, we used the recently published EPA chronic oral Reference Dose RfD (3×10⁻⁴ mg/kg/d)[73]. Although alternative toxicity values have been proposed by various U.S. states (reviewed by Post [19]) and European agencies [12,13], we deferred to these ATSDR and USEPA values for a simple example for how toxicity might be defined.

Tissues concentrations of the 5 PFAS were used to evaluate risks for each of the major 4 taxa discussed throughout this review (Supplemental Methods M2: Tissue Concentration Table). To create ranges of hazard quotients, we applied the maximum, minimum and median reported concentrations for each compound. The median value was generated from across the averages reported in the various identified studies, and where minimums were not reported, we applied ½ the lowest detection limit for that specific compound. For each taxa group and concentrations range (min, max and median), we assumed that concentrations neither increased nor declined over the course of long-term exposure (consumption) by the three population segments evaluated in this exercise.

Bivalves are used in our main illustration of risk (Figure 5), as the range of reported PFAS concentrations in bivalves generally presented higher hazard quotients than other taxonomic groups. However, similar trends in estimated exposures were seen in crustaceans, cephalopods, and gastropods, with variability across species and geographic locations of sampling (Supplemental Figures S4–S6). Among all average and high-end consumers (i.e., adults, women of reproductive age, and children), the highest reported concentrations of PFOS in bivalves have the potential to result in PFOS exposure greater than the sub-chronic MRL of 2×10⁻⁶ mg/kg/d. This is largely attributable to a maximum reported concentration of PFOS in Mediterranean mussels (125.9 ng/g) from a region of Portugal affected by paper, textile, and leather production plants [38]. PFOA had the second highest exposure ratios for both consumer groups, with the highest tissue concentrations ranging from 32.9–53.0 ng/g in various bivalves from Bohai Bay, China [31]. For PFNA, high estimates of exposure ratios were driven by brown mussels from Brazil (24.6 ng/g; [41]), and excess exposure was apparent among all high-end consumers. PFBS and PFHxS in bivalves were not observed to result in excess exposure in either consumer group.

Figure 5.

Hazard quotients for various consumers of bivalves based on concentrations reported from reviewed studies. The grey line at “1” indicates a consumed dosage equal to the threshold dose for risk described by either the Agency for Toxic Substances and Disease Registry (ATSDR) or U.S. EPA for respective PFAS. Each symbol represents the hazard quotient based on a calculation using the median observed PFAS concentration from the reviewed studies, with the bars extending to show the range of hazard quotients based on the minimum (left bar) and maximum (right bar) observed PFAS concentrations. Bars crossing the gray line indicate the potential for excessive exposure risk relative to threshold doses for respective PFAS.

Limitations and future directions

The assumptions that underly these estimates highlight the need for additional studies on PFAS in shellfish and the overall toxicity of PFAS. The highest ranges of PFOS, PFOA, and other PFAS were frequently found in shellfish collected from sites with either AFFF use or industrial releases, emphasizing the need for additional sampling both near these sources and at sites without known PFAS contamination. Additionally, the evolving thresholds for PFAS toxicity [13,72–74] create different interpretations of risk [31,75] at concentrations that would exceed more recent estimates of PFAS toxicity applied in the present example.

Finally, a key issue is that seafood is not the sole source of PFAS exposure. While our simple exposure estimates (Supplemental Equation 2) might describe exposure to individual PFAS from shellfish consumption, other sources of PFAS [1,3,12,13,16] would contribute to the total exposure, potentially increasing the exposure ratios by an order of magnitude. Additionally, multiple PFAS are likely to contaminate the same organisms and the influence of co-exposure to multiple PFAS on health has not been well delineated [74,76].

Conclusions

These results highlight the diverse range of methodologies and concentrations of PFAS in shellfish around the world and the ways in which PFAS researchers can improve upon and standardize PFAS research in the future. There is a need for the standardization of PFAS sampling techniques, sample preparation, and analytical methodologies to allow for better comparison of different PFAS analytes both within and across studies. There is also a need for more PFAS shellfish studies that are conducted in diverse geographic locations. This would allow for a better understanding of PFAS contamination globally and how it affects different communities around the world, as well as help keep global shellfish research, shellfishing and shellfish consumption advisories and policies up to date.