Per- and polyfluoroalkyl substances (PFAS) are ubiquitous chemicals whose use began in the 1940′s [1]. They are found in numerous items of daily use, including food packaging, cleaning products, carpeting, and firefighting foam [1]. There are over 12,000 PFAS chemicals, and some of these chemicals are pervasive and persistent environmental contaminants; termed “forever chemicals”, these PFAS do not degrade easily and maintain legacy status, meaning they remain persistent in the environment even after they are no longer used in production. Some PFAS analytes, such as perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS), are thought to persist in the environment for up to 1000 years [2], and some PFAS analytes, including PFOA and PFOS, have a half-life of 3–8 years within the human body [3,4]. PFAS analytes are also known to bioaccumulate in human tissue, act as endocrine disrupting chemicals, and/or promote carcinogenesis [5,6]. PFAS are now ubiquitous in the human population; U.S. data from NHANES in 2011–2012 reported detectable PFAS in 97 % of 1887 human serum samples evaluated [7].

PFAS exposure has been linked to various health outcomes and diseases, including pregnancy related outcomes (e.g., preeclampsia and lower birth weight), higher cholesterol levels, and kidney and testicular cancer [8–10]. More importantly, PFAS exposure have been linked specifically to immune-related endpoints, including altered vaccine responses and immune suppression through decreasing antibody production, hindering immunoglobulin production, and decreasing the amount of circulating immune cells [11–14]. To date, there are limited epidemiology-based studies that have investigated the association between PFAS and cytokine levels circulating in human serum, the latter serving as a surrogate measure for immune function and response [14–20]. Gaps in the current literature include 1) the limited number of PFAS exposures and immune markers evaluated and 2) the conflicting results that have been reported thus far [14]. Because immune dysregulation and chronic inflammation are recognized as key biological mechanisms for disease development and progression [21], we sought to specifically evaluate the link between PFAS and immune dysregulation.

Here, we measured 9 PFAS analytes and 16 immune markers and cytokines in 722 women cross-sectionally from the California Teachers Study. Specifically, we aimed to understand the association between levels of PFAS chemicals detected in the blood to circulating levels of cytokine and immune markers in blood at the same point in time, thereby providing critical clues on whether exposures to PFAS are linked to inflammation or immune alterations. Additionally, this study of largely post-menopausal women is particularly significant as a study of a potentially susceptibility population where immune response is known to be depressed and the occurrence of autoimmune conditions is higher. Moreover, the older age group of our population further embodies a higher cumulative exposure, particularly for this population which has been exposed to PFAS virtually their entire lifetime. Our investigation of PFAS and altered immune regulation in this population is thus timely and of biological significance in understanding potential downstream health effects [22,23].

The PFAS analytes included in our study (specifically, PFOA, PFOS, PFNA, and PFHxS [perfluorohexane sulfonate]) are considered legacy analytes, thus still prevalent in our environment despite being phased out in the early 2000′s by the EPA (Environmental Protection Agency) [24]. As recently as 2022, the Centers for Disease Control and Prevention reported that legacy PFAS analytes continue to be detectable in human serum samples. Specifically, 10 communities identified across multiple states within the United States, including Massachusetts, West Virginia, Delaware, New York, Washington, Texas, Colorado, and Arkansas, had legacy PFAS analytes present in their water [25], and, in these same 10 communities, the populations had higher age-adjusted blood levels of PFHxS, PFOS, PFOA and PFNA than national levels [25]. A 2016–2021 study further reported legacy PFAS contamination in 45 % of United States drinking water samples (of 716 tap water samples) [26]. Given the ubiquity of PFAS, the growing number of health outcomes associated with PFAS exposure and the dearth of understanding of what biological mechanisms by which PFAS leads to these outcomes, our study is timely and fills current gaps in the literature. We hypothesized that individuals with a higher burden of detected PFAS exposure would be associated with detectable differences in immune markers.

In our evaluation of 722 women from the California Teachers Study whose sera, collected from 2013 to 2016, were evaluated for both PFAS and cytokine/immune markers at a single time point, we found: 1) positive associations between increasing levels of PFHxS and elevated levels of BAFF, TNFRII, and IL2Rα, 2) positive associations between increasing levels of PFOA and PFOS with elevated BAFF, and 3) inverse associations between PFNA and sCD14. Pathway-based analyses did not appear to add additional information beyond the individual cytokine/immune marker associations. These results add important data to the literature by providing robust data linking PFAS analytes with specific immune markers, specifically BAFF, TNFRII and IL2Rα, which, to our knowledge, have not previously been reported. Our data adds to the growing evidence from prior studies that have reported associations between PFAS analytes and increased risk for long-term adverse immune-related health outcomes, such as persistent infections, rheumatoid arthritis, vaccine responses, and ulcerative colitis [11,42,43].

Consistent with the current literature, older age was significantly associated with higher detectable levels for each of the PFAS analytes [37,44]. Our results are consistent with Kato et al.’s (2011) prevalence estimates for PFOA, PFOS, PFNA, and PFHxS at an earlier timepoint (1999–2008) in the United States among both men and women whereby those in the oldest age group (60 + years) had the highest prevalences and levels of PFOA, PFOS and PFNA [44]. Similar results were reported by Stein et al. (2016) in a more recent but smaller subset of younger U.S. men and women. [37]. We also previously reported elevated levels of PFAS in older women compared to younger women in a different subset of 1,257 women in the California Teachers Study [45].

The descriptive statistics for PFOA and PFNA analytes are relatively consistent with other studies that utilized serum and reported median values for each analyte [16,18,41,46]. We reported similar medians (1.28 ng/mL) for PFHxS [18,41,46], except for Barton et al. (2022) who reported a much higher median (16.6 ng/mL) and whose results were based on both men and women [16]. Notably, Zota et al. (2018) reported a similar PFOS median (2.83 ng/mL) to our median (4.63 ng/mL), which could be attributed to study population similarities (all female cohorts) [18]; otherwise, higher PFOS medians were reported in general though these studies also included both men and women (8.2 ng/mL [16] to 45.86 ng/mL [41]). Geographic differences and differences in laboratory methods may also contribute to quantitative differences. It is noteworthy that the median levels of at least two of our PFAS analytes measured in the present study exceeded 2 ng/mL, which is the threshold whereby advanced screenings and health assessments are recommended [47].

Each PFAS analyte displayed different associations with the immune markers assessed in our study. With respect to PFOA exposure, there was a positive association with higher PFOA and elevated levels of BAFF. Of note, BAFF has been shown to play a role in controlling B cell maturation and acting as a survival factor for B cells [48]. Increased levels of BAFF due to a genetic deletion have been linked to an increased risk of developing autoimmune diseases, such as multiple sclerosis (MS) [49]. Similar to PFOA, higher levels of PFOS were also associated with elevated BAFF. PFOA and PFOS are among the most well documented PFAS analytes, but, to our knowledge, they have not been previously linked to BAFF. The robust and novel associations found here thus warrant further investigation as a potentially important biological pathway for understanding downstream health effects.

Our data yielded an association between increased PFNA exposure with decreased levels of sCD14. CD14 is expressed by monocytes, such as macrophages, and plays a role in innate immunity [50]. CD14 has also been shown to contribute to cancer-related inflammation and alternatively promote an immunosuppressive environment that helps drive tumor development [50]. According to Zhang et al. (2022), PFNA exposure and CD14 have been minimally studied via various study types, and the results are conflicting [14]. In an in vivo mouse study, PFNA exposure decreased the number of overall CD14 + cells [51]. In a follow-up in vivo mouse study, Rockwell et al. (2017) described that PFNA exposure caused an initial decease in CD14 + cells followed by an increase in CD14 + cells over time [52]. It is unclear how our results, which found increased PFNA levels associated with decreased levels of sCD14, fit into the larger published data, but further investigations in population studies appear warranted.

Similar to both PFOA and PFOS, elevated levels of PFHxS exposure were robustly associated with higher levels of BAFF. The positive associations of all three PFAS analytes with BAFF are notable and warrant further investigation. PFHxS exposure also yielded a positive association with elevated levels of TNFRII. TNFRII has been linked to susceptibility to immune conditions, such as multiple sclerosis, due to its role in suppressing the activation of regulatory T cells (Tregs) [53]. As well-documented in current literature, the number of Tregs increases as an individual ages, serving as a mechanism to prevent excessive inflammation or possible autoimmune conditions; however, when the production and activation of Tregs is altered, this protective function is decreased (e.g., such as in the case of PFAS exposure) [23,54]. Aging women represent a particularly susceptibility population – as immune response is depressed among post-menopausal women and as autoimmune condition risk is higher – understanding the role that PFAS plays in perturbing the immune response is thus of particular importance for understanding downstream health effects in this susceptible population [22,54]. Lastly, TNFRII has also been linked to the activation of the NFkB signaling pathway, which subsequently increases the production of TNF-α, a ligand for TNFRII [55]. PFHxS exposure was positively associated with increased IL2Rα. As a receptor for IL-2, IL2Rα and has been shown to assist with Treg activation and development [56]. PFAS exposure has also been found to increase the levels of Th2 cytokines, including IL2Rα, that can increase the risk for immune conditions, such as rheumatoid arthritis [43,57]. Further investigations to confirm the associations between PFHxS with TNFRII and IL2Rα in population-based studies are thus warranted to illuminate its potential role in Treg and/or NFkB activation that impacts long-term health outcomes.

When assessing PFAS exposure in relation to our constructed immune pathways, there were no notable associations other than those already driven by the specific cytokines. For example, the PFHxS association with the B-cell activation pathway is likely attributed to the specific association between PFHxS and BAFF association which was in fact more pronounced than that of the overall pathway, likely diluted by the null associations from the other markers in the pathway. Overall, the pathway associations did not appear to indicate a cumulative effect of all markers in a given pathway, but rather appeared largely driven by one specific marker.

Our cross-sectional study strengths include the use of a relatively large study cohort, the number of immune markers assessed, and involved standardized specimen collection and minimized variability due to the processing and storage by a single laboratory. Due to the nature of a cross-sectional study, our study does not permit us to measure the long-term implications of PFAS exposure on immune dysregulation. Study limitations also include the evaluation among females only, thereby limiting the generalizability of the study overall; however, we note the strength of evaluating these effects among aging women. Studies have demonstrated that PFAS exposure have been associated with altered female reproductive function, including later age of menarche, irregular and longer menstrual cycles, earlier age of menopause, and altered levels of hormones [58,59]. These effects are posited to result in downstream health outcomes; to date, higher levels of PFOS have been positively associated with ER+, PR+and ER+/PR+breast cancer tumors, and PFOA have been positively associated with ER−, PR−, ET+/PR− and ER−/PR− breast cancer tumors [60]. Investigating the role of PFAS exposure within an aging female population is thus a particularly important aspect of understanding the effects of the environment on immunity and related health outcomes, particularly in an immunologically susceptible and highly exposed population. Moreover, based on studies to date, a recent review emphasized that future epidemiology studies should include the evaluation of immunotoxicity among PFAS-associated disease endpoints [11]. Despite the overall consistency of our study results to the current literature, we acknowledge that cannot exclude the possibility that the associations present in our results were due to chance, particularly given the number of PFAS analytes and immune markers evaluated.

In summary, our study results assessed the associations between various PFAS analyte levels and circulating inflammatory markers at a single timepoint, providing important insight regarding the biological underpinnings that PFAS exposures may have in downstream development of disease endpoints. Specifically, our data suggests that PFAS exposure may be associated with the elevation of markers in the TNF family or superfamily, and its effects on immune function dysregulation may be associated with Treg production and activation. Importantly, each PFAS analyte displayed different associations with measured cytokines, thus warranting future studies to investigate the mechanistic differences between the PFAS analytes to identify the individual effects of these exposures on immune dysregulation and the array of disease endpoints that may result.

All data associated with this publication are available for research use. The California Teachers Study welcomes all inquiries (https://www.calteachersstudy.org/for-researchers). Additionally, the PFAS analyte dataset presented in this study can be found in the online Human Health Exposure Analysis Resource data repository (https://hhearprogram.org/data-services).

Data associated with this publication are also publicly available through the Human Health Exposure Analysis Resource (HHEAR) Data Center (https://hheardatacenter.mssm.edu/).

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.cyto.2024.156753.