Skip to main content
NCBI home page
Elsevier Sponsored Documents logoLink to Elsevier Sponsored Documents
. 2021 Nov;156:106599. doi: 10.1016/j.envint.2021.106599

Perfluoroalkyl substances and immune cell counts in adults from the Mid-Ohio Valley (USA)

Maria-Jose Lopez-Espinosa a,b,c,⁎, Christian Carrizosa a, Michael I Luster d, Joseph B Margolick e, Olga Costa a, Giovanni S Leonardi f,g, Tony Fletcher f,g
PMCID: PMC8381762  PMID: 33993002

Highlights

  • •

    The literature on PFASs and circulating levels of immune cells is scarce.

  • •

    Study of the association between PFASs and white blood cells at two different time points.

  • •

    Cross-sectional study on the association between PFASs and lymphocyte subsets.

  • •

    Positive association between PFHxS and total lymphocytes, especially for cell count.

  • •

    Positive, but less consistent, association between PFOA and PFOS and total lymphocytes.

Keywords: PFASs, Immune system, White blood cells, Lymphocytes

Abstract

Background

Although perfluoroalkyl substances (PFASs) may be immunotoxic, evidence for this in humans is scarce. We studied the association between 4 PFASs (perfluorohexane sulfonate [PFHxS], perfluorooctanoic acid [PFOA], perfluorooctane sulfonate [PFOS] and perfluorononanoic acid [PFNA]) and circulating levels of several types of immune cells.

Methods

Serum PFASs and white blood cell types were measured in 42,782 (2005–2006) and 526 (2010) adults from an area with PFOA drinking water contamination in the Mid-Ohio Valley (USA). Additionally, the major lymphocyte subsets were measured in 2010. Ln(cell counts) and percentages of cell counts were regressed on serum PFAS concentrations (ln or percentiles). Adjusted results were expressed as the percentage difference (95% CI) per interquartile range (IQR) increment of each PFAS concentration.

Results

Generally positive monotonic associations between total lymphocytes and PFHxS, PFOA, and PFOS were found in both surveys (difference range: 1.12–7.33% for count and 0.36–1.77 for percentage, per PFAS IQR increment), and were stronger for PFHxS. These associations were reflected in lymphocyte subset counts but not percentages, with PFHxS positively and monotonically associated with T, B, and natural killer (NK) cell counts (range: 5.51–8.62%), PFOA and PFOS with some T-cell phenotypes, and PFOS with NK cells (range: 3.12–12.21%), the associations being monotonic in some cases. Neutrophils, particularly percentage (range: −1.74 to −0.36), showed decreasing trends associated with PFASs. Findings were less consistent for monocytes and eosinophils.

Conclusion

These results suggest an association between PFHxS and, less consistently, for PFOA and PFOS, and total lymphocytes (although the magnitudes of the differences were small). The increase in absolute lymphocyte count appeared to be evenly distributed across lymphocyte subsets since associations with their percentages were not significant.

1. Introduction

Perfluoroalkyl substances (PFASs) are a diverse group of chemicals used in industrial applications and consumer products since the 1950s. They persist in the environment, resist degradation, bioaccumulate in humans, and have relatively long half-lives in the body of humans (approximately 4–6 years) (Agency for Toxic Substances and Disease Registry [ATSDR], 2018). Whereas mean serum concentrations of PFASs have decreased in recent years in Western countries (Sunderland et al., 2019), probably because their production has been reduced or has been banned (ATSDR, 2018), the stability of these compounds and recent increases in serum levels observed in Asia and the Middle East (International Pollutants Elimination Network, 2019) mean that PFASs, and their possible health effects, remain significant public health concerns.

To date, most of the animal and human studies on the health effects of some PFASs have focused on perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS), and to a lesser extent on perfluorohexane sulfonate (PFHxS) and perfluorononanoic acid (PFNA). The results of those studies suggest that they may affect metabolic, endocrine, and immune processes (Ballesteros et al., 2017, EFSA [European Food Safety Authority], 2018, Lau et al., 2007, Matilla-Santander et al., 2017, United States Department of Health and Human Services-National Toxicology Program [USDHHS-NTP], 2016). Regarding the immune system, an extensive review of the literature by the United States National Toxicology Program (USDHHS-NTP, 2016) concluded that both PFOA and PFOS are immunotoxic. This conclusion was based primarily upon a high level of evidence from animal studies showing that at high concentrations they suppress the antibody response, and a moderate level of evidence from human studies indicating that they affect vaccine responses. This conclusion is also supported by a more recent review from the European Food Safety Authority (EFSA), which reported an adverse effect on serum antibody response after vaccination due to PFOS (and possibly PFOA) exposure, with children considered to be the most vulnerable group (EFSA, 2018). Regarding PFHxS and PFNA exposure, there is still insufficient epidemiological data on their possible immunotoxicity (Chen et al., 2018, Dong et al., 2013, Grandjean et al., 2017) to reach a conclusion.

Most observational studies in humans have investigated associations between PFASs and immune outcomes such as asthma, allergies, propensity to infections, and serum antibody response to vaccination (EFSA, 2018). However, immunotoxicity can also be studied by measuring peripheral counts of white blood cells (WBCs: neutrophils, lymphocytes, monocytes, eosinophils, and basophils) and subsets of peripheral blood lymphocytes delineated by cell-surface markers (CD3+T cells, CD3+CD4+T helper cells, CD3+CD8+T-cytotoxic cells, CD3+CD4+ CD8+ double positive [DP] T cells, CD3− CD16+ CD56+ natural killer [NK] cells, CD3− CD19+B cells, and CD4+/CD8+ ratio). WBC count and peripheral blood lymphocyte measurements are powerful diagnostic tools used primarily in the identification and progression of diseases such as acquired immunodeficiency syndrome (AIDS), and leukemias and lymphomas (Castilho et al., 2016, Davis et al., 2014). They have also been used occasionally in clinical studies for testing both the efficacy and the toxicity of therapeutics (e.g., Aziz et al., 2013), and to help identify immunotoxicity following environmental chemical exposure (Haase et al., 2016, Leonardi et al., 2000, Tryphonas, 2001). Nevertheless, data obtained from human epidemiologic studies on PFAS exposure and these immune markers is still limited, with some data derived from monitoring humans exposed to PFASs in the workplace (Costa et al., 2009, Olsen et al., 2003) or from cross-sectional (Abraham et al., 2020, Dong et al., 2013, Emmett et al., 2006, Gaylord et al., 2019, Knudsen et al., 2018) or longitudinal (Oulhote et al., 2017) non-occupational human studies. Furthermore, only a limited number of immunological end points have been studied in most of the previous studies (predominantly, total WBCs, and/or eosinophils). Given these limitations, further studies are warranted.

Previous work by our group showed that PFOA may be associated with a decrease in influenza vaccine efficacy in adults who had had an A/H3N2 seasonal flu vaccination (Looker et al., 2014). To further address immunotoxicity, we aimed to evaluate the association between PFAS concentrations and WBC counts (measured in both 2005–2006 and 2010), and the major lymphocyte subsets (measured in 2010) amongst adults who had been highly exposed to PFOA due to contaminated water supplies, but with similar exposure to other PFASs compared to other populations from North America (Frisbee et al., 2009).

2. Methods

2.1. Study design and subjects

The C8 Health Project arose from the settlement of a class action lawsuit filed in 2002 by residents living near the DuPont Washington Works plant in West Virginia regarding the release of PFOA from the facility that had contaminated drinking water supplies along the Mid-Ohio River Valley. The project collected data on 69,030 subjects (estimated participation rate for residents at the time: 80%) between August 2005 and August 2006. Individuals were eligible for the study if they had consumed water for at least one year between 1950 and 2004 from any of the six contaminated water districts or from private wells in the proximity of the chemical plant. Details of the study population have been described elsewhere (Frisbee et al., 2009).

Between March and July 2010, participants were recalled and invited to participate in a second survey, as described elsewhere (Fitz-Simon et al., 2013). Briefly, the second survey was restricted to residents of the three Ohio and five West Virginia ZIP code areas in the region most affected by PFOA contamination. The follow-up population was restricted to those aged 20–59 at the time of the initial survey (2005–2006), and individuals were chosen randomly from age–gender strata to give balanced numbers of men and women in each ten-year age group. Altogether 973 people were contacted, and after excluding people who reported a past or present diagnosis of cancer and those unwilling to provide blood samples, 748 participated, with ages between 24 and 64 years at the time of the second survey. From the first survey containing participants of all ages, we focused on the subset of adults aged 24–64 years at baseline (2005–2006) so as to be in the same age range as those in the follow-up study (2010).

Information on infections was not available in the first survey, but this information was available in 2010. Participants were asked if, at the time of the survey, they had any of the following infections or conditions associated with inflammation: cold, influenza, sore throat, bronchitis, pneumonia, gastroenteritis, cold sore, shingles, and/or sinus, ear, tooth, and/or mouth infections. A total of 213 out of 748 participants reported at least one of the above-mentioned conditions, with sinus and skin infections being the most common. Analyses for the follow-up study focused on those who reported not having any infections or conditions associated with inflammation plus PFAS analyses. We chose to exclude those with infections as they could have an effect on immune cell counts (George-Gay and Parker, 2003). Altogether, 42,782 and 526 adults in the first and second surveys, respectively, gave a blood sample, had analyses of both PFASs and immune markers (2005–2006: types of WBCs, and 2010: types of WBCs and subtypes of lymphocytes), as well as data on sociodemographic and anthropometric factors, and were included in the present study. The London School of Hygiene and Tropical Medicine Ethics Committee approved this study and informed consent was obtained from all participants in each period.

2.2. Serum PFAS concentrations

Laboratory analyses of serum PFASs were conducted by Exygen, State College, PA, USA (a commercial laboratory) in 2005–2006 and by the Centers for Disease Control and Prevention (CDC) in Atlanta, GA, USA in 2010. The analytical (Flaherty et al., 2005, Frisbee et al., 2009, Kato et al., 2011) and quality control (Van Leeuwen et al., 2006) procedures employed by both laboratories have been described elsewhere. Briefly, laboratory analyses of PFASs used online solid phase extraction coupled with reversed-phase high-performance liquid chromatography separation and detection by isotope-dilution tandem mass spectrometry. The limits of detection (LODs) in 2005–2006 were 0.5 ng/mL for all PFASs. In 2010, LODs were 0.1 ng/mL for PFNA and PFHxS, 0.2 ng/mL for PFOS, and 0.5 ng/mL for PFOA.

2.3. Immune cell analysis

Whole blood collected in EDTA vacutainer tubes was shipped to laboratories for analysis of immune cells. In LabCorp Inc., Burlington, NC, USA (an accredited clinical diagnostic laboratory), analyses of types of WBCs (neutrophils, monocytes, eosinophils, lymphocytes, and basophils) were performed using an automated hematology analyzer (LH Series, Beckman Coulter). In the flow cytometry core laboratory at Johns Hopkins Bloomberg School of Public Health in Baltimore, immunophenotypic analysis was performed on peripheral blood lymphocytes (CD3+ T cells, CD3+ CD4+ T helper cells, CD3+ CD8+ T-cytotoxic cells, CD3+ CD4+ CD8+ DP T cells, CD3− CD16+ CD56+ NK cells, CD3− CD19+ B cells) using standard flow cytometry methods, as previously described (Giorgi et al., 1990, Schenker et al., 1993). The CD4+/CD8+ ratio was also calculated. The laboratory was certified by two proficiency testing programs, i.e., the Immunology Quality Assurance (IQA) administered by the US National Institutes of Health, and the College of American Pathologists (CAP) program. The laboratory was also certified under the provisions of the Clinical Laboratory Improvement Act (CLIA).

2.4. Covariates

We obtained information on sociodemographic and anthropometric characteristics, as well as on lifestyle variables, through questionnaires that were administered by trained interviewers in person (2005–2006) and by telephone (2010). We considered the following variables: gender, race/ethnicity, age, education, body mass index (BMI, only available in the 2005–2006 survey), tobacco consumption, alcohol intake, present or past diagnosis of immune diseases and/or cancer, and anti-inflammatory medication (only available for the 2010 survey). The participants were asked if they were taking or had taken in the last four years paracetamol, aspirin, or other anti-inflammatory medications on a regular basis (i.e., at least once a week for six months or more).

Serum creatinine levels were determined using the Kinetic Jaffe method (LabCorp, Burlington, NC). We estimated glomerular filtration rate (eGFR) using the Modification of Diet in Renal Disease (MDRD) Study 4-variable equation (Levey et al., 2006).

2.5. Statistical analysis

We performed statistical analyses for all immune cell types except basophils (due to the low proportion of samples with these immune cells present: 25.6% in 2005–2006 and 17.9% in 2010). For PFASs, the % ≤LOD was very low (<0.1% for PFOA up to 2.6% for PFHxS) and they were set to half the LOD for the analyses. Pearson’s correlations between the four ln(PFASs), and these contaminants and eGFR were performed, and a heatmap was also generated to display them. We used ordinal mixed models to assess the baseline to follow-up differences in sociodemographic characteristics. Linear mixed models adjusted for covariates were also employed to study the changes in exposure levels of ln(PFASs) over time.

We used multivariate linear regression analysis to assess the relation between ln(counts) and percentages of immune cells and ln(PFASs) at the baseline (n = 42,782) and follow-up (n = 526). All models were adjusted for gender, age in 10-year categories, smoking, month of sampling (because of a decreasing trend in serum PFASs during the collection period in this population), alcohol intake, and educational level (hereafter called ‘main analysis’).

We also conducted several sensitivity analyses including: BMI (since excess adiposity has been related to both impaired immune function (Martí et al., 2001) and PFASs (Braun, 2017, Cardenas et al., 2018), we preferred not to include this variable in the main analysis); the four PFASs simultaneously; eGFR (to evaluate whether a common disease process may be contributing to altered PFAS excretion from the kidney and increasing immune cell counts) (Dhingra et al., 2017, Watkins et al., 2013); and use of anti-inflammatory drugs (only available in the 2010 survey). We also excluded from the models those participants with self-reported autoimmune diseases (n = 1,606 in 2005–2006, n = 28 in 2010) to account for medical conditions that could affect immunological measurements (specifically, systemic lupus erythematosus, scleroderma, multiple sclerosis, Sjögren’s syndrome, Addison's disease, and rheumatoid arthritis) and those who reported cancer (n = 2,614 in the 2005–2006 survey, since in the follow-up study participants with this condition were not included in the study). Although the main analyses in 2010 were restricted to participants who reported no infection at the time of the interview (n = 526), a sensitivity analysis was performed on the full population (n = 736) repeating the main analyses. Finally, we also investigated possible differences by gender by including the interaction term with the contaminants in the models.

We expressed results as percentage difference (with a 95% confidence interval – CI) in the outcomes associated with an interquartile range (IQR) increment in PFAS concentrations. The percentage difference was calculated as the complement of the exponentiated regression coefficient [(exp(β × IQR) – 1) × 100] for cell counts and just the regression coefficient (β × IQR) for cell percentages. We used the 2005–2006 IQRs of ln(PFASs) for both populations to make results comparable. For easier clinical interpretation, results for count data were also expressed in terms of absolute differences, taking the median values as the reference [median(cells) * (exp(β × IQR) – 1)]. Differences in percentages of cells were evaluated to determine whether there were differential effects of PFASs on this relative scale as opposed to the effects on absolute counts (e.g., an increase in total lymphocyte count might not occur evenly across the different subsets).

We also re-ran the main analysis with deciles (2005–2006 population) and tertiles (2010 population) of PFASs to explore the shape of the exposure–response curve. To be consistent with the regression on the ln(PFASs) scale, tests for trend were conducted across percentiles including the ln-transformed geometric mean of each decile or tertile added to the models as a numerical regressor. Results of categorical analyses are presented graphically as the adjusted value of each outcome, by exposure group, with 95% CI.

To be conservative, we used heteroscedasticity-consistent standard errors in all the analyses to account for potential deviations from the linear model assumptions (Long and Ervin, 2000). We conducted statistical analyses using R.3.6.2 (R Core Team, 2020).

3. Results

3.1. Study population, immune outcomes, and PFAS concentrations

The baseline population (2005–2006) was comprised of similar numbers of women (53%) and men (47%), 29% of whom were smokers, and 72% were overweight or obese. Significant differences between the 2005–2006 and 2010 surveys were found in some characteristics, with higher educational levels, fewer smokers and alcohol drinkers, and slightly older participants in 2010 (n = 526, mean age: 46 years) compared to 2005–2006 (n = 42,782, mean age: 44 years) (Table 1).

Table 1.

Study population and PFAS concentrations, Mid-Ohio Valley, USA (2005–2010).

Variable 2005–2006 2010
(n = 42,782) (n = 526) pa
Sex 0.292
Female 22,542 (52.7) 265 (50.4)
Male 20,240 (47.3) 261 (49.6)
Race/ethnicity 0.071
Non-Hispanic white 41,735 (97.6) 519 (98.7)
Others 1,047 (2.45) 7 (1.33)
Age <0.001
≤30 years 7,745 (18.1) 62 (11.8)
31–40 years 10,444 (24.4) 122 (23.2)
41–50 years 11,804 (27.6) 132 (25.1)
51–60 years 9,808 (22.9) 138 (26.2)
>60 years 2,981 (6.97) 72 (13.7)
Education <0.001
<12 years 3,902 (9.12) 16 (3.04)
HSD or GED 17,461 (40.8) 158 (30.0)
Some college 14,881 (34.8) 242 (46.0)
≥Bachelor’s degree 6,538 (15.3) 110 (20.9)
BMI 0.436
Underweight 486 (1.14) 6 (1.14)
Normal weight 11,607 (27.1) 139 (26.4)
Overweight 14,826 (34.7) 206 (39.2)
Obese I 9,198 (21.5) 96 (18.3)
Obese II and III 6,665 (15.6) 79 (15.0)
Smoking <0.001
Never smoked 20,017 (46.8) 331 (62.9)
Ex-smoker 10,509 (24.6) 111 (21.1)
Current smoker 12,256 (28.6) 84 (16.0)
Alcohol intake <0.001
Lifetime non-drinker 7,898 (18.5) 303 (57.6)
Ex-drinker 12,207 (28.5) 52 (9.89)
Drinker 22,677 (53.0) 171 (32.5)
PFAS (ng/mL)
PFHxS 2.90(1.80, 4.60) 2.10(1.20, 3.20) <0.001
PFOA 26.9(13.2, 69.2) 35.7(15.0, 93.7) <0.001
PFOS 19.7(13.3, 28.4) 9.60(6.10, 14.9) <0.001
PFNA 1.40(1.10, 1.80) 1.40(1.10, 1.80) <0.001
Open in a new tab

Numbers (%) or median (interquartile range) are presented. BMI, Body mass index (only available for the 2005–2006 survey); GED, General education development; HSD, High school diploma. PFAS, Perfluoroalkyl substance. PFHxS, Perfluorohexane sulfonate; PFNA, Perfluorononanoic acid; PFOA, Perfluorooctanoic acid; PFOS, Perfluorooctane sulfonate.

a

Likelihood ratio test p-value from unadjusted ordinal mixed models (for sociodemographic covariates), or linear mixed models adjusted for gender, age, smoking, month of sampling, alcohol intake, and educational level (for ln(PFAS)).

PFHxS, PFOA, PFOS, and PFNA were detectable (above the LOD) in 97.4, 99.9, 99.5, and 97.6% of people in 2005–2006. Corresponding data for 2010 were: 99.4, 99.8, 99.4, and 99.2% (data not shown in table). The median serum PFASs in the follow-up compared to baseline was higher for PFOA (due to the study population being located in the six most contaminated water districts in 2010), lower for PFOS, similar for PFHxS, and the same for PFNA (Table 1). In 2005–2006, the IQR contrast for PFOA was bigger (56.0 ng/ml), PFOS was intermediate (15.1 ng/ml), and both PFHxS and PFNA were smaller (2.8 and 0.7 ng/ml, respectively). IQR contrasts for 2010 were: 69.8, 9.0, 2.0, and 0.7 ng/mL for PFOA, PFOS, PFHxS, and PFNA, respectively (data not shown in table). Furthermore, positive correlations (p < 0.001 in all cases) were found between all the PFASs analyzed (r range: 0.32–0.57 in 2005–2006 and 0.44–0.77 in 2010). We found weak inverse correlations between PFASs and eGFR (r range from −0.13 to −0.03, with p < 0.05 except for PFOS and PFNA in 2010) (Figure S1).

Most median cell counts were slightly lower in the later survey sample, though this was not the case for eosinophils (Table 2). Median percentages of the different immune cells were mostly very similar between the two surveys (Table 3).

Table 2.

Counts of WBCs and lymphocyte subsets, and association between PFASs and relative (percent) differences in these counts amongst the population in 2005–2006 (n = 42,782) and in 2010 (n = 526), Mid-Ohio Valley, USA.

Cell count PFHxS PFOA PFOS PFNA
Outcome Median (IQR) % diff. (95%CI) % diff. (95%CI) % diff. (95%CI) % diff. (95%CI)
Total WBCs 2005–2006 7,100 (5,900, 8,600) 0.55 (0.25, 0.86) −0.27 (-0.62, 0.08) −0.55 (-0.84, −0.26) −0.21 (-0.47, 0.05)
Total WBCs 2010 6,300 (5,300, 7,600) 0.90 (-1.47, 3.33) 0.84 (-2.20, 3.97) 0.55 (-1.35, 2.49) −0.08 (-1.95, 1.82)
Neutrophils 2005–2006 4,400 (3,500, 5,500) −0.04 (-0.44, 0.35) −0.92 (-1.37, −0.47) −1.56 (-1.93, −1.19) −0.77 (-1.11, −0.44)
Neutrophils 2010 4,000 (3,200, 5,000) −1.99 (-4.97, 1.08) −1.44 (-5.12, 2.39) −0.86 (-3.38, 1.72) −1.14 (-3.68, 1.46)
Monocytes 2005–2006 400 (300, 500) 1.16 (0.71, 1.63) 0.61 (0.08, 1.15) 0.38 (-0.06, 0.82) 0.15 (-0.24, 0.54)
Monocytes 2010 300 (300, 400) 1.85 (-2.71, 6.63) 4.29 (-1.05, 9.92) 1.41 (-2.38, 5.35) 0.55 (-2.81, 4.02)
Eosinophils 2005–2006 100 (100, 200) −0.96 (-1.76, −0.15) −0.06 (-0.98, 0.87) −0.29 (-1.08, 0.50) −0.76 (-1.44, −0.07)
Eosinophils 2010 100 (100, 200) 1.15 (-6.64, 9.59) 4.77 (-4.11, 14.46) 4.22 (-2.33, 11.21) 6.37 (0.68, 12.38)
Lymphocytes 2005–2006 2,000 (1,700, 2,500) 2.00 (1.63, 2.37) 1.12 (0.70, 1.55) 1.95 (1.57, 2.33) 1.47 (1.15, 1.79)
Lymphocytes 2010 1,800 (1,500, 2,200) 7.33 (4.12, 10.64) 5.50 (1.52, 9.63) 3.39 (0.70, 6.15) 1.29 (-1.04, 3.68)
CD3+ T cells 1,356 (1,085, 1,645) 7.43 (3.80, 11.19) 5.89 (1.46, 10.51) 3.12 (0.20, 6.13) 1.30 (-1.25, 3.92)
CD3+CD4+T-helper cells 904 (714, 1105) 8.28 (4.28, 12.44) 7.42 (2.78, 12.26) 3.94 (0.60, 7.40) 2.04 (-0.84, 5.01)
CD3+CD8+T-cytotoxic cells 401 (293, 534) 5.51 (0.52, 10.75) 3.99 (-2.40, 10.80) 1.62 (-2.31, 5.71) −0.88 (-3.97, 2.32)
CD3+CD4+CD8+DP T cells 16 (10, 25) 6.03 (-1.21, 13.79) 12.21 (2.47, 22.89) −0.23 (-5.68, 5.55) 1.06 (-4.01, 6.40)
CD3− CD16+CD56+NK cells 184 (129, 260) 8.62 (3.45, 14.04) 4.94 (-0.75, 10.96) 5.40 (1.35, 9.60) 0.31 (-3.34, 4.10)
CD3− CD19+B cells 238 (170, 318) 7.26 (2.08, 12.70) 3.32 (-2.82, 9.86) 3.89 (-0.64, 8.63) 2.26 (-1.79, 6.47)
CD4+/CD8+ratio 2.35 (1.71, 3.02) 2.63 (-2.19, 7.68) 3.29 (-2.53, 9.46) 2.28 (-1.78, 6.52) 2.94 (-0.31, 6.30)
Open in a new tab

Immune cell count > 0 in 100% of samples except for monocytes (99.9% and 99.6% in 2005–2006 and 2010, respectively) and eosinophils (91.9% and 93.2%). PFASs and counts of immune cells were log transformed (ln) for analyses. Likelihood ratio test p-value < 0.001 for the comparison between the two time points for all outcomes except eosinophils (linear mixed models). Cell counts are expressed as cells/μL. Results are expressed as the difference in the outcomes associated with IQR increments in PFAS levels in 2005–2006. All models were adjusted for gender, age, smoking, month of sampling, alcohol intake, and educational level.

% diff., % difference; CI, Confidence interval; DP, Double positive; IQR, Interquartile range; NK, Natural killer; PFASs, Perfluoroalkyl substances; PFHxS, Perfluorohexane sulfonate; PFNA, Perfluorononanoic acid; PFOA, Perfluorooctanoic acid; PFOS, Perfluorooctane sulfonate; WBCs, White blood cells.

Table 3.

Percentages of WBCs and lymphocyte subsets, and association between PFASs and percentages of these cells amongst the population in 2005–2006 (n = 42,782) and in 2010 (n = 526), Mid-Ohio Valley, USA.

Cell percentage PFHxS PFOA PFOS PFNA
Outcome Median (IQR) diff. (95%CI) diff. (95%CI) diff. (95%CI) diff. (95%CI)
Neutrophils 2005–2006 62 (57, 67)a −0.36 (-0.46, −0.26) −0.41 (-0.52, −0.30) −0.65 (-0.74, −0.55) −0.37 (-0.45, −0.29)
Neutrophils 2010 63 (57, 68)a −1.74 (-2.54, −0.94) −1.41 (-2.38, −0.44) −0.85 (-1.63, −0.07) −0.58 (-1.31, 0.15)
Monocytes 2005–2006 6.0 (5.0, 7.0)a 0.02 (0.00, 0.05) 0.04 (0.02, 0.07) 0.04 (0.02, 0.06) 0.00 (-0.02, 0.02)
Monocytes 2010 5.0 (4.0, 7.0)a 0.02 (-0.2, 0.24) 0.16 (-0.08, 0.39) 0.03 (-0.15, 0.20) 0.04 (-0.11, 0.20)
Eosinophils 2005–2006 2.0 (1.0, 3.0)a −0.04 (-0.06, −0.02) 0.00 (-0.02, 0.02) −0.01 (-0.03, 0.01) −0.03 (-0.04, −0.01)
Eosinophils 2010 2.0 (1.0, 3.0)a −0.04 (-0.23, 0.15) 0.05 (-0.21, 0.32) 0.06 (-0.11, 0.22) 0.13 (0.01, 0.26)
Lymphocytes 2005–2006 29 (24, 34)a 0.37 (0.28, 0.46) 0.36 (0.26, 0.45) 0.61 (0.53, 0.69) 0.39 (0.32, 0.46)
Lymphocytes 2010 29 (24, 34)a 1.77 (1.05, 2.50) 1.24 (0.36, 2.12) 0.77 (0.13, 1.42) 0.43 (-0.23, 1.08)
CD3+T cells 75 (70, 79)b 0.02 (-0.70, 0.74) 0.27 (-0.54, 1.07) −0.20 (-0.79, 0.39) 0.00 (-0.52, 0.52)
CD3+CD4+T-helper cells 50 (45, 55)b 0.45 (-0.38, 1.28) 0.88 (-0.05, 1.82) 0.24 (-0.45, 0.93) 0.40 (-0.15, 0.95)
CD3+CD8+T-cytotoxic cells 22 (17, 27)b −0.44 (-1.29, 0.41) −0.36 (-1.40, 0.68) −0.51 (-1.26, 0.24) −0.42 (-1.01, 0.17)
CD3+CD4+CD8+DP T cells 0.9 (0.6, 1.3)b −0.03 (-0.12, 0.07) 0.15 (-0.02, 0.33) −0.05 (-0.13, 0.03) 0.02 (-0.07, 0.11)
CD3− CD16+CD56+NK cells 10 (7.3, 14)b 0.16 (-0.37, 0.70) 0.04 (-0.62, 0.71) 0.27 (-0.12, 0.66) −0.04 (-0.44, 0.37)
CD3− CD19+B cells 13 (10, 17)b −0.18 (-0.77, 0.42) −0.33 (-0.93, 0.28) −0.04 (-0.55, 0.46) 0.06 (-0.39, 0.51)
Open in a new tab

Immune cell count > 0 in 100% of samples except for monocytes (99.9% and 99.6% in 2005–2006 and 2010, respectively) and eosinophils (91.9% and 93.2%). PFASs were log transformed (ln) for analyses. Likelihood ratio test p-value < 0.001 for the comparison between both time points for all outcomes except lymphocytes and eosinophils (linear mixed models). Results are expressed as the difference in the outcomes associated with IQR increments in PFAS levels in 2005–2006. All models were adjusted for gender, age, smoking, month of sampling, alcohol intake, and educational level.

CI, Confidence interval; diff., difference; DP, Double positive; IQR, Interquartile range; NK, Natural killer; PFASs, Perfluoroalkyl substances; PFHxS, Perfluorohexane sulfonate; PFNA, Perfluorononanoic acid; PFOA, Perfluorooctanoic acid; PFOS, Perfluorooctane sulfonate.

a

Percentage of total WBCs.

b

Percentage of total lymphocytes.

3.2. Serum PFASs and WBC count

For total WBC count in the 2005–2006 survey, positive (PFHxS) and inverse (rest of the contaminants) associations were found, reaching statistical significance for PFHxS and PFOS (difference: 0.55% and −0.55% per PFHxS and PFOS IQR increment, corresponding to 39 and −39 cells/μl when referring to the median population count). WBC associations showed wide confidence intervals in the smaller follow-up population (Table 2, Table 3 and S1 and Fig. 1 and S2).

Fig. 1.

Fig. 1

Open in a new tab

Adjusted WBC counts (mean and 95% confidence interval) by PFAS deciles (2005–2006, n = 42,782) or tertiles (2010, n = 526), Mid-Ohio Valley, USA. Statistically significant trends at p < 0.05 and p < 0.01 are denoted by one and two asterisks, respectively. All models were adjusted for gender, age, smoking, month of sampling, alcohol intake, and educational level. PFAS, Perfluoroalkyl substance; PFHxS, Perfluorohexane sulfonate; PFNA, Perfluorononanoic acid; PFOA, Perfluorooctanoic acid; PFOS, Perfluorooctane sulfonate, WBCs, White blood cells.

3.3. Serum PFASs and myeloid lineage

There was some evidence of an association between PFOA, PFOS, PFNA, and neutrophil count (difference range: −1.56% to −0.77% per PFAS IQR increment [–69 to –34 cells/μl]) and between the four PFASs and neutrophil percentage (−0.65 to −0.36 per PFAS IQR increment) in the first survey. The effect was notably larger for the percentage of neutrophils in relation to PFHxS, PFOA, and PFOS in the follow-up population (difference: −1.74, −1.41, and −0.85, per PFAS IQR increment, respectively). We also found some evidence of a positive association between PFHxS, PFOA, and monocyte count (difference: 1.16% and 0.61% per PFAS IQR increment, respectively) and between PFHxS, PFOA, PFOS, and monocyte percentage (difference range: 0.02 to 0.04 per PFAS IQR increment, respectively) in the 2005–2006 but not in the 2010 survey. Finally, eosinophils were associated with PFHxS in the first survey (difference: −0.96% and −0.04 for count and percentage, per PFHxS IQR increment, respectively) and PFNA in both surveys, but the direction of the latter association changed between the two time points (difference: −0.76% and 6.37% for count and −0.03 and 0.13 for percentage, per PFNA IQR increment) (Table 2, Table 3, and Fig. 1 and S2). Given the smaller contribution of monocytes and eosinophils to the WBC count, such changes were almost negligible (range: 1–6 cells/μl) when expressed in absolute terms (Table S1).

3.4. Serum PFASs and lymphoid lineage

For total lymphocytes, positive associations were clearly evident for PFHxS, PFOA, and PFOS. These associations were stronger for absolute numbers compared to percentages and with a larger effect size in the 2010 subset compared to the 2005–2006 population. Specifically, the relative (percent) differences in the counts of lymphocytes for an IQR contrast in PFHxS was 2.00% in the large population and 7.33% in the follow-up, 1.12% and 5.50% for PFOA, and 1.95% and 3.39% for PFOS (Table 2). Concerning absolute counts, these estimates represented changes ranging from 22 to 132 cells/μl (Table S1). Respective data for the difference increase in lymphocyte percentage were 0.37 and 1.77 (PFHxS), 0.36 and 1.24 (PFOA), and 0.61 and 0.77 (PFOS) (Table 3). Fig. 1 and S2 show these graphically across PFAS deciles (population in 2005–2006) and tertiles (population in 2010) with, in general, clear monotonic increases for these three contaminants and lymphocytes. A monotonic association was also found between PFNA and lymphocyte cells (difference for count and percentage: 1.47% [29 cells/μl] and 0.39 per PFNA IQR increment) in the 2005–2006 population but not in that of 2010 (Table 2, Table 3 and Fig. 1 and S2).

The above-mentioned statistically significant changes in absolute lymphocyte count led to changes in the numbers of lymphocyte subsets. Specifically, we found associations between most of the types of numbers of lymphocyte subsets and PFHxS concentrations, with differences ranging from 5.51% in CD3+ CD8+ T-cytotoxic cells to 8.62% in CD3− CD16+ CD56+ NK cells per PFHxS IQR increment (Table 2). Fig. 2 shows clear monotonic increases across PFHxS tertiles for all B, NK, and T cell numbers except the CD3+ CD4+ CD8+ DP subset. For other PFASs the associations were more limited. For PFOA, associations with the numbers of the above-mentioned lymphocytes were also evident for CD3+ T cells (5.89%) and CD3+ CD4+ T-helper cells (7.42%), and more strongly for CD3+ CD4+ CD8+ DP T cells (12.21%). For PFOS, again there were associations with both CD3+ and CD3+ CD4+ T cells, and CD3− CD16+ CD56+ NK cells (3.12%, 3.94%, and 5.40%, respectively) (Table 2). The magnitudes of the associations increased more clearly monotonically across tertiles for both CD3+ T cells and CD3+ CD4+ T cells in relation to PFOA, and CD3− CD16+ CD56+ NK cells in relation to PFOS (Fig. 2). PFNA was not associated with any subtype of lymphocyte (Table 2 and Fig. 2).

Fig. 2.

Fig. 2

Open in a new tab

Adjusted lymphocyte subtype counts (mean and 95% confidence interval) by PFAS tertiles (n = 526), Mid-Ohio Valley, USA (2010). Statistically significant trends at p < 0.05 and p < 0.01 are denoted by one and two asterisks, respectively. All models were adjusted for gender, age, smoking, month of sampling, alcohol intake, and educational level. DP, Double positive; NK, Natural killer; PFAS, Perfluoroalkyl substance; PFHxS, Perfluorohexane sulfonate; PFNA, Perfluorononanoic acid; PFOA, Perfluorooctanoic acid; PFOS, Perfluorooctane sulfonate.

Regarding percentage data, associations were mostly null in all cases, meaning that changes in actual cell numbers strongly resembled those seen for total lymphocytes, but now distributed accordingly across the different phenotypes (Table 3 and Figure S3).

3.5. Sensitivity analyses

Various models including other variables which might have confounded the results are set out in detail in the Supplementary material (Figures S4–S9). Firstly, modest rather than major changes in the associations resulted after adding BMI, eGFR, or anti-inflammatory drugs, or excluding self-reported cancer or autoimmune diseases. Secondly, in models containing all four PFASs, the regression coefficients fell or were even close to null in some cases and their CIs were wider. Specifically, for the myeloid lineage bigger changes were shown in the 2005–2006 survey compared to that of 2010. Regarding lymphocytes, the association was somewhat smaller for PFHxS versus count and null for percentage in the 2005–2006 survey. For PFOA, the associations were smaller, reaching the null in 2005-2005 for count and losing statistical significance for count and percentage in 2010. For PFOS and both count and percentage the associations were smaller in 2005–2006 and disappeared in 2010. For numbers of lymphocyte subtypes including the four PFASs simultaneously, the associations were, in general, a little smaller for PFHxS and PFOA but completely disappeared for PFOS. Thirdly, multivariate analysis amongst the full follow-up population, i.e., population reporting (or not) infections in 2010 (n = 736) also suggested associations between PFHxS, PFOA, and PFOS in relation to immune cell counts, but with generally lower % differences and some with CIs including the null. Finally, although some p-values of the interaction with sex were < 0.05, especially for the WBC types, no clear evidence of a differential effect between males and females was found on considering the two surveys.

4. Discussion

We determined peripheral WBC counts in two surveys (2005–2006 and 2010) and lymphocyte subtypes in 2010 to characterize potential health effects related to the immune system associated with PFAS concentrations. Our community-based sample had a wide range of PFOA serum concentrations (IQR: 13.2–69.2 ng/mL; median: 26.9 ng/mL in 2005–2006), and background levels of PFHxS, PFOS, and PFNA (medians: 2.9, 19.7, and 1.4 ng/mL in 2005–2006, respectively). The trends with PFASs for WBCs were slight, with small percentage differences across the IQRs. For most PFASs there were negative slopes for neutrophils, with larger positive slopes for lymphocytes. Our results suggest a consistent positive association between serum PFHxS and absolute lymphocyte count and weaker evidence of an association for percentage of total lymphocytes. For PFHxS and absolute counts of lymphocyte subtypes, significant and positive associations were found for T, B, and NK cell counts, which likely reflect the changes in the absolute lymphocyte count. A less consistent association was also shown for both PFOA and PFOS and total and subtypes of lymphocytes.

No clear association was found between total WBC count and PFAS levels, statistical significance only being reached for PFHxS and PFOS in the 2005–2006 survey. The effects were minimal, and the direction of the association was not consistent with PFHxS, showing a positive association and the others showing negative trends. Two studies measuring serum PFOA and PFOS in workers at 3M (n = 53) and DuPont (n = 518) facilities located in Italy, Alabama (USA), and Belgium did not report any significant associations between these contaminants and total WBC count (range of PFOA and PFOS means were much higher than in the present study: 0.07–19.7 and 0.13–1.40 μg/mL, respectively) (Costa et al., 2009, Olsen et al., 2003). Diverging results have been reported in non-occupational studies. Specifically, in a study on a population (n = 371) with longstanding environmental PFOA exposure (median: 354 ng/mL) in the same area as the present study, marginally significant positive associations with WBC count were found (Emmett et al., 2006). A cross-sectional study in pregnant women (n = 189) in Greenland (Knudsen et al., 2018) reported a significant negative association between the sum of 17 PFASs (ΣPFAS median: 18.5 ng/mL) and WBC count. However, levels of 4 PFASs in maternal and child samples at 18 months and 5 years of age (ΣPFAS geometric mean range: 0.47–0.73 ng/mL) were not associated with WBC count at 5 years of age in a cohort (n = 56) of Faroese children (Oulhote et al., 2017).

Examination of cell differentials resulted in mixed associations for the myeloid lineage. Neutrophils showed slightly decreased trends associated with PFASs in the first survey and less clearly in the second survey, where the association was only with percentages of this cell type. In line with our results, negative associations were reported between the sum of 17 PFASs and numbers of neutrophils in the above-mentioned pregnant women from Greenland (Knudsen et al., 2018). Conversely, other studies did not find such an association (Dong et al., 2013, Emmett et al., 2006, Oulhote et al., 2017). Regarding other myeloid cells, we found positive associations between PFHxS, PFOA, and PFOS and monocytes in the first survey, while no consistent pattern of an association was found for monocytes in the second survey or for eosinophils. In the above-mentioned study on the Mid-Ohio Valley, significant positive associations with monocyte count but not percentage were found (Emmett et al., 2006). Interestingly, none of the above-mentioned immune cell types measured in children aged 5 years were associated with prenatal, 18-month or 5-year concentrations of PFASs in Faroese children (Oulhote et al., 2017). Cross-sectional studies reported conflicting results for eosinophil counts, with one study finding a positive association with PFOA (median in cases [n = 231] and controls [n = 225]: 1.2 and 0.5 ng/mL, respectively) and PFOS (33.9 and 28.9 ng/mL) among asthmatic compared to nonasthmatic Taiwanese teenagers (Dong et al., 2013). However, another study reported null results for 6 PFAS congeners (median ΣPFASs for cases [n = 118] and controls [n = 169]: 7.3 and 5.7 ng/mL) among New Yorkers aged 16–17 years exposed to high levels of PFASs, among other contaminants, in the World Trade Center disaster who were matched to a control group (Gaylord et al., 2019).

Results were clearer for total lymphocytes, showing statistically significant positive and generally monotonic associations with all PFAS levels except PFNA in the two surveys. Specifically, PFHxS presented the strongest effect for a small absolute contrast in levels (i.e., for an IQR contrast of only 2.8 ng/mL it showed the biggest increase with both count and percentage of lymphocytes). For PFOA and PFOS, there was also an association, but less clear. With regard to the previous epidemiological studies measuring lymphocytes, none have reported PFAS exposure as being associated with alterations in this particular immune cell type during childhood (Abraham et al., 2020, Oulhote et al., 2017) or adulthood (Emmett et al., 2006, Knudsen et al., 2018).

To better characterize the lymphocytosis and help determine whether there was any evidence of altered immunoregulation, immunophenotyping of peripheral blood was conducted in the 2010 survey to enumerate the major lymphocyte populations. For lymphocyte subtypes, PFHxS concentrations were positively and monotonically associated with absolute counts of T, B, and NK cells. In addition, higher PFOA serum levels were positively associated with some T cell subsets and PFOS with some T cell subsets and NK cells, with monotonicity shown in some cases. Since counts of lymphocyte subtypes are not independent of the absolute lymphocyte count and there were only significant changes in counts but not in percentages of the lymphocyte subsets, the observed changes likely reflect changes in absolute lymphocyte counts rather than any specific effect on lymphocyte subsets themselves. Something similar has been reported in smoker’s leukocytosis, a well-documented effect of active smoking (Pedersen et al., 2019, Sopori, 2002) where the effect of smoking raises WBCs without greatly changing subset percentages of the different cell types (Park et al., 1992). With regard to PFASs, only two epidemiological studies, both focused on child populations, have been conducted on peripheral blood lymphocytes and PFASs, and no significant associations were reported (Abraham et al., 2020, Oulhote et al., 2017). On a different note, animal experiments have reported altered lymphocyte populations following exposure to PFASs, although effects seem species- or even strain-dependent (DeWitt et al., 2009). Increased CD3+ CD4+ T-helper cells and decreased CD3+ CD8+ T-cytotoxic cells in spleens and increased CD3+ CD8+ T–cytotoxic cells in thymus were reported in male ICR mice exposed to PFOA for 21 days (Son et al., 2009). Reduced B and T cells were also reported, particularly CD4+ CD8+ thymocytes in male C57BL/6 mice treated with PFOA for 7 or 10 days (Yang et al., 2001). In contrast, our finding of a positive association between PFAS exposure and CD3− CD16+ CD56+ NK cells might be more consistent with the increase in NK cell activity reported in B6C3F1 male mice exposed to PFOS (Peden-Adams et al., 2008).

In the present study, despite the evidence of an association with increased total lymphocyte cells, the absolute magnitudes of difference in cell numbers were small (difference range: 1.12–7.33% per PFAS IQR increment, which when compared to the population median represented at most 132 [95% CI: 74, 191] cells/μL for PFHxS in 2010) and the percentages of total WBCs that these cells represented changed very little (difference range: 0.36–1.77 per PFAS IQR increment). Therefore, lymphocyte changes in the present study cannot be considered clinically meaningful. The same is true for the decreases observed in absolute neutrophil count: though statistically significant for some PFASs, especially for the 2005–2006 data, these were quite small and would not be clinically important in themselves. Further studies would be needed to determine whether these effects could be meaningful in specific clinical contexts, such as a serious bacterial infection.

Overall, these findings, along with previous findings of suppressed antibody response (Abraham et al., 2020, Grandjean et al., 2012, Kielsen et al., 2016), suggest that the effect of PFASs on immune function, if there is one, is not mediated solely through altered lymphocyte numbers but by changes in cell signaling or function (as measured by antibody response).

Previous studies in the population considered here have suggested that impaired kidney function may lead to decreased net PFOA excretion via the kidneys and, accordingly, increased serum levels of PFOA (Watkins et al., 2013). Thus, confounding by kidney function may plausibly occur due to the fact that inflammation or ill health, which can be associated with renal impairment, may also be reflected in the elevated lymphocyte count. Since eGFR reflects the filtering capacity of the kidney (Dhingra et al., 2017, Watkins et al., 2013) and was available for the study population, we repeated the regression analysis including eGFR in the models, with no evidence of eGFR confounding the association between PFASs and immune cell outcomes. Estimates in the main analysis did not differ markedly when adding other possible confounders such as 2005–2006 BMI or self-reported anti-inflammatory drug therapy (only available for the follow-up study). The main differences were found for the inclusion of the four PFASs at the same time since CIs were wider, and the estimates fell when including other PFASs or were even close to null in some cases. The exclusion of people who reported autoimmune disease (both surveys) or cancer (first survey) hardly modified the estimated associations. Finally, gender-specific differences in the development of reproductive organs and physiology may result in gender-specific responses to a given toxicant (Bach et al., 2015). For PFASs, two studies have indicated a higher vulnerability among asthmatic boys compared to girls (Qin et al., 2017, Zhu et al., 2016). Mixed results were reported in another study with girls or boys at higher risk depending on the outcome (asthma, atopic dermatitis, and rhinitis) and the class of PFAS studied (Kvalem et al., 2020). However, although we found some significant gender interactions indicating that higher PFAS exposure was associated with higher counts of types of WBCs in females, and lower counts in males, such a pattern was not consistent.

Our study has several limitations. First, information on residential histories, infections, and other potential confounders was self-reported, and participants were only asked about the presence or absence (not severity) of specific infections. Nevertheless, restricting our analysis to those who reported no infections in the 2010 survey should have reduced potential bias due to infection reports. Second, information on infections was only available for the follow-up population and BMI only for the first survey. Nevertheless, the results for participants who did not report infection in 2010 (n = 526) are somewhat stronger than those for the full population in 2010 (n = 736) or for the 2005–2006 cohort. In addition, sensitivity analyses including 2005–2006 BMI did not substantially change the results in either of the two surveys, thus ruling out any role of BMI as a confounder. Furthermore, in this population PFOA exposure was largely driven by residential water district and consumption, and less correlated with individual self-reported characteristics and behaviors (Steenland et al., 2009). Third, given the high-exposure nature of our population for PFOA, the results for this contaminant may not be generalizable to other populations. Fourth, the immune system has many components but we studied only the absolute counts and percentages of the main types of immune cells, with no assessment of the function of these cells. However, it is pertinent to highlight that, in a previous study of part of the follow-up population of the present study, increased PFOA concentrations were associated with a lower antibody response to A/H3N2 influenza vaccine (Looker et al., 2014). Fifth, due to the high number of statistical analyses, there is the possibility of potential false positive statistically significant associations (i.e., type I errors). The estimates for the coefficients and their confidence intervals should be taken as a global picture of the pattern of the relations between the variables involved in the study. Finally, humans are potentially exposed to a large variety of other types of contaminants, and further study of the relation between exposures to multiple chemicals with immunotoxic capability is warranted.

The major strengths of this study are the large sample size, assessments of PFAS levels and some immune cells at two time points with an interval of approximately five years, and the inclusion of individual potential confounders. In addition, it is believed that these participants are representative, at least for the first survey, given the high participation rates in the C8 Health Project (~85% in the range of age of the present study) (Frisbee et al., 2009) of those adults who drank contaminated water in the Mid-Ohio Valley, and this diminishes concern about potential selection biases. Thirdly, despite an increasing body of literature on the possible immunotoxic effects of PFASs in humans and specific animal models demonstrating specific lymphocyte changes, to our knowledge studies involving human immunophenotypes are scarce and there is therefore a gap in the literature on this topic.

In summary, the strongest association was observed between absolute lymphocyte count and PFHxS. If causal, our results suggest that PFHxS may be more potent compared to the other PFASs, since a contrast of only 2.8 ng/ml was associated with as much as a 7% increase in absolute lymphocyte count in the 2010 survey. Weaker evidence of an association was also shown for PFHxS and percentage of lymphocytes. For PFOA and PFOS, the association was less clear. For all PFASs, the magnitudes of the differences in absolute lymphocyte count were small. A similar pattern was found for lymphocyte cell subtypes with, again, stronger evidence for PFHxS. Nevertheless, the changes observed in absolute counts of lymphocyte subsets reflected the change in total lymphocyte count rather than any specific effect on the subsets themselves, since no association was reported for percentage of the subtypes.

CRediT authorship contribution statement

Maria-Jose Lopez-Espinosa: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing - original draft, Writing - review & editing. Christian Carrizosa: Data curation, Formal analysis, Investigation, Methodology, Software, Visualization, Writing - original draft, Writing - review & editing. Michael I. Luster: Conceptualization, Investigation, Methodology, Resources, Validation, Validation, Writing - review & editing. Joseph B. Margolick: Investigation, Methodology, Resources, Visualization, Writing - original draft, Writing - review & editing. Olga Costa: Data curation, Formal analysis, Investigation, Methodology, Software, Writing - review & editing. Giovanni S. Leonardi: Conceptualization, Funding acquisition, Investigation, Methodology, Validation, Writing - review & editing. Tony Fletcher: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing - original draft, Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

Acknowledgments

The authors wish to thank the participants for their contributions to this study. We are also grateful to the Centers for Disease Control and Prevention (CDC) laboratory for analyzing the PFASs in serum samples.

This project was funded by the C8 Class Action Settlement Agreement (Circuit Court of Wood County, WV) between DuPont and plaintiffs, which resulted from releases of perfluorooctanoate (PFOA, or C8) into drinking water. It was one of the C8 Science Panel Studies undertaken by the Court-approved C8 Science Panel, of which Dr Tony Fletcher was a member, established under the same Settlement Agreement. Dr Fletcher is a consortium partner of the European Union’s Horizon 2020 research and innovation program under grant agreement 733032 HBM4EU (www.HBM4EU.eu); part-funded by the National Institute for Health Research (NIHR) Health Protection Research Unit in Environmental Exposures and Health, a partnership between Public Health England, the Health and Safety Executive, and the University of Leicester; and part funded with a PFAS research grant from CORIS/REGIONE VENETO (Italy). The views expressed are those of the author(s) and not necessarily those of the NIHR, Public Health England, the Health and Safety Executive or the Department of Health and Social Care. MJ Lopez-Espinosa holds grants from the Spanish Carlos III Health Institute (Miguel Servet-FSE: MSII16/00051 and FIS-FEDER: PI14/00891 and PI17/00663), Alicia Koplowitz Foundation 2017, and Ministry of Education, Culture and Sports (Jose Castillejo Grant: CAS17/00052).

Handling Editor: Shoji F. Nakayama

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.envint.2021.106599.

Appendix A. Supplementary material

The following are the Supplementary data to this article:

Supplementary data 1
mmc1.pdf (2.3MB, pdf)

References

  1. Abraham K., Mielke H., Fromme H., Völkel W., Menzel J., Peiser M., Zepp F., Willich S.N., Weikert C. Internal exposure to perfluoroalkyl substances (PFASs) and biological markers in 101 healthy 1-year-old children: associations between levels of perfluorooctanoic acid (PFOA) and vaccine response. Arch. Toxicol. 2020;94:2131–2147. doi: 10.1007/s00204-020-02715-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Agency for Toxic Substances and Disease Registry [ATSDR], 2018. Toxicological profile for perfluoroalkyls (Draft for public comment). Atlanta, GA: US Department of Health and Human Services, Public Health Service.
  3. Aziz N., Margolick J.B., Detels R., Rinaldo C.R., Phair J., Jamieson B.D., Butch A.W. Value of a quality assessment program in optimizing cryopreservation of peripheral blood mononuclear cells in a multicenter study. Clin. Vaccine Immunol. 2013;20:590–595. doi: 10.1128/CVI.00693-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bach C.C., Bech B.H., Brix N., Nohr E.A., Bonde J.P.E., Henriksen T.B. Perfluoroalkyl and polyfluoroalkyl substances and human fetal growth: A systematic review. Crit. Rev. Toxicol. 2015;45:53–67. doi: 10.3109/10408444.2014.952400. [DOI] [PubMed] [Google Scholar]
  5. Ballesteros V., Costa O., Iñiguez C., Fletcher T., Ballester F., Lopez-Espinosa M.J. Exposure to perfluoroalkyl substances and thyroid function in pregnant women and children: A systematic review of epidemiologic studies. Environ. Int. 2017;99:15–28. doi: 10.1016/j.envint.2016.10.015. [DOI] [PubMed] [Google Scholar]
  6. Braun J.M. Early-life exposure to EDCs: Role in childhood obesity and neurodevelopment. Nat. Rev. Endocrinol. 2017;13:161–173. doi: 10.1038/nrendo.2016.186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Cardenas A., Hauser R., Gold D.R., Kleinman K.P., Hivert M.F., Fleisch A.F., Lin P.I.D., Calafat A.M., Webster T.F., Horton E.S., Oken E. Association of Perfluoroalkyl and Polyfluoroalkyl Substances With Adiposity. JAMA Netw. open. 2018;1 doi: 10.1001/jamanetworkopen.2018.1493. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Castilho J.L., Shepherd B.E., Koethe J., Turner M., Bebawy S., Logan J., Rogers W.B., Raffanti S., Sterling T.R. CD4+/CD8+ratio, age, and risk of serious noncommunicable diseases in HIV-infected adults on antiretroviral therapy. AIDS. 2016;30:899–908. doi: 10.1097/QAD.0000000000001005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chen Q., Huang R., Hua L., Guo Y., Huang L., Zhao Y., Wang X., Zhang J. Prenatal exposure to perfluoroalkyl and polyfluoroalkyl substances and childhood atopic dermatitis: A prospective birth cohort study. Environ. Heal. 2018;17:8. doi: 10.1186/s12940-018-0352-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Costa G., Sartori S., Consonni D. Thirty years of medical surveillance in perfluooctanoic acid production workers. J. Occup. Environ. Med. 2009;51:364–372. doi: 10.1097/JOM.0b013e3181965d80. [DOI] [PubMed] [Google Scholar]
  11. Davis A.S., Viera A.J., Mead M.D. Leukemia: An overview for primary care. Am. Fam. Physician. 2014;89:731–738. [PubMed] [Google Scholar]
  12. DeWitt, J.C., Shnyra, A., Badr, M.Z., Loveless, S.E., Hoban, D., Frame, S.R., Cunard, R., Anderson, S.E., Meade, B.J., Peden-Adams, M.M., Luebke, R.W., Luster, M.I., 2009. Immunotoxicity of perfluorooctanoic acid and perfluorooctane sulfonate and the role of peroxisome proliferator-activated receptor alpha. Crit. Rev. Toxicol. https://doi.org/10.1080/10408440802209804. [DOI] [PubMed]
  13. Dhingra R., Winquist A., Darrow L.A., Klein M., Steenland K. A study of reverse causation: Examining the associations of perfluorooctanoic acid serum levels with two outcomes. Environ. Health Perspect. 2017;125:416–421. doi: 10.1289/EHP273. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dong G.H., Tung K.Y., Tsai C.H., Liu M.M., Wang D., Liu W., Jin Y.H., Hsieh W.S., Lee Y.L., Chen P.C. Serum polyfluoroalkyl concentrations, asthma outcomes, and immunological markers in a case-control study of Taiwanese children. Environ. Health Perspect. 2013;121:507–513. doi: 10.1289/ehp.1205351. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. EFSA [European Food Safety Authority] EFSA CONTAM Panel (EFSA Panel on Contaminants in the Food Chain), Knutsen HK, Alexander J, Barregard L, Bignami M, Bruschweiler B, Ceccatelli S, Cottrill B, Dinovi M, Edler L, Grasl-Kraupp B, Hogstrand C, Hoogenboom LR, Nebbia CS, Oswald IP, Petersen A. R. FSA J. 2018;16:5194. doi: 10.2903/j.efsa.2018.5194. [DOI] [Google Scholar]
  16. Emmett E.A., Zhang H., Shofer F.S., Freeman D., Rodway N.V., Desai C., Shaw L.M. Community exposure to perfluorooctanoate: Relationships between serum levels and certain health parameters. J. Occup. Environ. Med. 2006;48:771–779. doi: 10.1097/01.jom.0000233380.13087.37. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fitz-Simon N., Fletcher T., Luster M.I., Steenland K., Calafat A.M., Kato K., Armstrong B. Reductions in serum lipids with a 4-year decline in serum perfluorooctanoic acid and perfluorooctanesulfonic acid. Epidemiology. 2013;24:569–576. doi: 10.1097/EDE.0b013e31829443ee. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Flaherty J.M., Connolly P.D., Decker E.R., Kennedy S.M., Ellefson M.E., Reagen W.K., Szostek B. Quantitative determination of perfluorooctanoic acid in serum and plasma by liquid chromatography tandem mass spectrometry. J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 2005;819:329–338. doi: 10.1016/j.jchromb.2005.03.002. [DOI] [PubMed] [Google Scholar]
  19. Frisbee S.J., Brooks A.P., Maher A., Flensborg P., Arnold S., Fletcher T., Steenland K., Shankar A., Knox S.S., Pollard C., Halverson J.A., Vieira V.M., Jin C., Leyden K.M., Ducatman A.M. The C8 Health Project: Design, methods, and participants. Environ. Health Perspect. 2009;117:1873–1882. doi: 10.1289/ehp.0800379. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gaylord A., Berger K.I., Naidu M., Attina T.M., Gilbert J., Koshy T.T., Han X., Marmor M., Shao Y., Giusti R., Goldring R.M., Kannan K., Trasande L. Serum perfluoroalkyl substances and lung function in adolescents exposed to the World Trade Center disaster. Environ. Res. 2019;172:266–272. doi: 10.1016/j.envres.2019.02.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. George-Gay B., Parker K. Understanding the complete blood count with differential. J. Perianesthesia Nurs. 2003;18:96–117. doi: 10.1053/jpan.2003.50013. [DOI] [PubMed] [Google Scholar]
  22. Giorgi J.V., Cheng H.L., Margolick J.B., Bauer K.D., Ferbas J., Waxdal M., Schmid I., Hultin L.E., Jackson A.L., Park L., Taylor J.M.G. Quality control in the flow cytometric measurement of T-lymphocyte subsets: The Multicenter AIDS Cohort Study experience. Clin. Immunol. Immunopathol. 1990;55:173–186. doi: 10.1016/0090-1229(90)90096-9. [DOI] [PubMed] [Google Scholar]
  23. Grandjean P., Andersen E.W., Budtz-Jørgensen E., Nielsen F., Mølbak K.R., Weihe P., Heilmann C. Serum vaccine antibody concentrations in children exposed to perfluorinated compounds. JAMA. 2012;307:391–397. doi: 10.1001/jama.2011.2034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Grandjean P., Heilmann C., Weihe P., Nielsen F., Mogensen U.B., Budtz-Jørgensen E. Serum vaccine antibody concentrations in adolescents exposed to perfluorinated compounds. Environ. Health Perspect. 2017;125 doi: 10.1289/EHP275. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Haase H., Fahlenkamp A., Schettgen T., Esser A., Gube M., Ziegler P., Kraus T., Rink L. Immunotoxicity monitoring in a population exposedto polychlorinated biphenyls. Int. J. Environ. Res. Public Health. 2016;13:1–13. doi: 10.3390/ijerph13030295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. International Pollutants Elimination Network PFAS pollution across the Middle East and Asia [WWW Document] 2019. https://www.documentcloud.org/documents/5980916-Pfas-Pollution-Across-the-Middle-East-and-Asia.html accessed 6.8.20.
  27. Kato K., Basden B.J., Needham L.L., Calafat A.M. Improved selectivity for the analysis of maternal serum and cord serum for polyfluoroalkyl chemicals. J. Chromatogr. A. 2011;1218:2133–2137. doi: 10.1016/j.chroma.2010.10.051. [DOI] [PubMed] [Google Scholar]
  28. Kielsen K., Shamim Z., Ryder L.P., Nielsen F., Grandjean P., Budtz-Jørgensen E., Heilmann C. Antibody response to booster vaccination with tetanus and diphtheria in adults exposed to perfluorinated alkylates. J. Immunotoxicol. 2016;13:270–273. doi: 10.3109/1547691X.2015.1067259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Knudsen A.K.S., Long M., Pedersen H.S., Bonefeld-Jørgensen E.C. Persistent organic pollutants and haematological markers in Greenlandic pregnant women: the ACCEPT sub-study. Int. J. Circumpolar Health. 2018;77:1456303. doi: 10.1080/22423982.2018.1456303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kvalem H.E., Nygaard U.C., Lødrup Carlsen K.C., Carlsen K.H., Haug L.S., Granum B. Perfluoroalkyl substances, airways infections, allergy and asthma related health outcomes – implications of gender, exposure period and study design. Environ. Int. 2020;134 doi: 10.1016/j.envint.2019.105259. [DOI] [PubMed] [Google Scholar]
  31. Lau C., Anitole K., Hodes C., Lai D., Pfahles-Hutchens A., Seed J. Perfluoroalkyl acids: A review of monitoring and toxicological findings. Toxicol. Sci. 2007;99:366–394. doi: 10.1093/toxsci/kfm128. [DOI] [PubMed] [Google Scholar]
  32. Leonardi G.S., Houthuijs D., Steerenberg P.A., Fletcher T., Armstrong B., Antova T., Lochman I., Lochmanova A., Rudnai P., Erdei E., Musial J., Jazwiec-Kanyion B., Niciu E.M., Durbaca S., Fabianova E., Koppova K., Lebret E., Brunekreef B., Van Loveren H. Immune biomarkers in relation to exposure to particulate matter: A cross-sectional survey in 17 cities of Central Europe. Inhal. Toxicol. 2000;12:1–14. doi: 10.1080/089583700750019486. [DOI] [PubMed] [Google Scholar]
  33. Levey A.S., Coresh J., Greene T., Stevens L.A., Zhang Y., Hendriksen S., Kusek J.W., Van Lente F. Using standardized serum creatinine values in the Modification of Diet in Renal Disease study equation for estimating glomerular filtration rate. Ann. Intern. Med. 2006;145:247–254. doi: 10.7326/0003-4819-145-4-200608150-00004. [DOI] [PubMed] [Google Scholar]
  34. Long J.S., Ervin L.H. Using heteroscedasticity consistent standard errors in the linear regression model. Am. Stat. 2000;54:217–224. doi: 10.1080/00031305.2000.10474549. [DOI] [Google Scholar]
  35. Looker C., Luster M.I., Calafat A.M., Johnson V.J., Burleson G.R., Burleson F.G., Fletcher T. Influenza vaccine response in adults exposed to perfluorooctanoate and perfluorooctanesulfonate. Toxicol. Sci. 2014;138:76–88. doi: 10.1093/toxsci/kft269. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Martí A., Marcos A., Martínez J.A. Obesity and immune function relationships. Obes. Rev. 2001;2:131–140. doi: 10.1046/j.1467-789x.2001.00025.x. [DOI] [PubMed] [Google Scholar]
  37. Matilla-Santander N., Valvi D., Lopez-Espinosa M.J., Manzano-Salgado C.B., Ballester F., Ibarluzea J., Santa-Marina L., Schettgen T., Guxens M., Sunyer J., Vrijheid M. Exposure to perfluoroalkyl substances and metabolic outcomes in pregnant women: Evidence from the Spanish INMA birth cohorts. Environ. Health Perspect. 2017;125 doi: 10.1289/EHP1062. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Olsen G.W., Burris J.M., Burlew M.M., Mandel J.H. Epidemiologic assessment of worker serum perfluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA) concentrations and medical surveillance examinations. J. Occup. Environ. Med. 2003;45:260–270. doi: 10.1097/01.jom.0000052958.59271.10. [DOI] [PubMed] [Google Scholar]
  39. Oulhote Y., Shamim Z., Kielsen K., Weihe P., Grandjean P., Ryder L.P., Heilmann C. Children’s white blood cell counts in relation to developmental exposures to methylmercury and persistent organic pollutants. Reprod. Toxicol. 2017;68:207–214. doi: 10.1016/j.reprotox.2016.08.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Park L.P., Margolick J.B., Giorgi J.V., Ferbas J., Bauer K., Kaslow R., Muñoz A. Influence of HIV-1 infection and cigarette smoking on leukocyte profiles in homosexual men. The Multicenter AIDS Cohort Study. J. Acquir. Immune Defic. Syndr. 1992;5:1124–1130. [PubMed] [Google Scholar]
  41. Peden-Adams M.M., Keller J.M., Eudaly J.G., Berger J., Gilkeson G.S., Keil D.E. Suppression of humoral immunity in mice following exposure to perfluorooctane sulfonate. Toxicol. Sci. 2008;104:144–154. doi: 10.1093/toxsci/kfn059. [DOI] [PubMed] [Google Scholar]
  42. Pedersen K.M., Çolak Y., Ellervik C., Hasselbalch H.C., Bojesen S.E., Nordestgaard B.G. Smoking and increased white and red blood cells. Arterioscler. Thromb. Vasc. Biol. 2019;39:965–977. doi: 10.1161/ATVBAHA.118.312338. [DOI] [PubMed] [Google Scholar]
  43. Qin, X. Di, Qian, Z. (Min), Dharmage, S.C., Perret, J., Geiger, S.D., Rigdon, S.E., Howard, S., Zeng, X.W., Hu, L.W., Yang, B.Y., Zhou, Y., Li, M., Xu, S.L., Bao, W.W., Zhang, Y.Z., Yuan, P., Wang, J., Zhang, C., Tian, Y.P., Nian, M., Xiao, X., Chen, W., Lee, Y.L., Dong, G.H., 2017. Association of perfluoroalkyl substances exposure with impaired lung function in children. Environ. Res. 155, 15–21. https://doi.org/10.1016/j.envres.2017.01.025. [DOI] [PubMed]
  44. R Core Team . Austria; Vienna: 2020. R: A language and Environment for Statistical Computing. R Foundation for Statistical Computing. [Google Scholar]
  45. Schenker E.L., Hultin L.E., Bauer K.D., Ferbas J., Margolick J.B., Giorgi J.V. Evaluation of a dual-color flow cytometry immunophenotyping panel in a multicenter quality assurance program. Cytometry. 1993;14:307–317. doi: 10.1002/cyto.990140311. [DOI] [PubMed] [Google Scholar]
  46. Son H.Y., Lee S., Tak E.N., Cho H.S., Shin H.I., Kim S.H., Yang J.H. Perfluorooctanoic acid alters T lymphocyte phenotypes and cytokine expression in mice. Environ. Toxicol. 2009;24:580–588. doi: 10.1002/tox.20459. [DOI] [PubMed] [Google Scholar]
  47. Sopori M. Effects of cigarette smoke on the immune system. Nat. Rev. Immunol. 2002;2:372–377. doi: 10.1038/nri803. [DOI] [PubMed] [Google Scholar]
  48. Steenland K., Jin C., MacNeil J., Lally C., Ducatman A., Vieira V., Fletcher T. Predictors of PFOA levels in a community surrounding a chemical plant. Environ. Health Perspect. 2009;117:1083–1088. doi: 10.1289/ehp.0800294. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Sunderland E.M., Hu X.C., Dassuncao C., Tokranov A.K., Wagner C.C., Allen J.G. A review of the pathways of human exposure to poly- and perfluoroalkyl substances (PFASs) and present understanding of health effects. J. Expo. Sci. Environ. Epidemiol. 2019;29:131–147. doi: 10.1038/s41370-018-0094-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Tryphonas H. Approaches to detecting immunotoxic effects of environmental contaminants in humans. Environ. Health Perspect. 2001;109:877–884. doi: 10.1289/ehp.01109s6877. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. United States Department of Health and Human Services-National Toxicology Program [USDHHS-NTP] Systematic review of immunotoxicity associated with exposure to perfluorooctanoic acid (PFOA) or perfluorooctane sulfonate (PFOS) [WWW Document] 2016. https://ntp.niehs.nih.gov/ntp/ohat/pfoa_pfos/pfoa_pfosmonograph_508.pdf accessed 6.8.20.
  52. Van Leeuwen S.P.J., Karrman A., Van Bavel B., De Boer J., Lindstrom G. Struggle for quality in determination of perfluorinated contaminants in environmental and human samples. Environ. Sci. Technol. 2006;40:7854–7860. doi: 10.1021/es061052c. [DOI] [PubMed] [Google Scholar]
  53. Watkins D.J., Josson J., Elston B., Bartell S.M., Shin H.M., Vieira V.M., Savitz D.A., Fletcher T., Wellenius G.A. Exposure to perfluoroalkyl acids and markers of kidney function among children and adolescents living near a chemical plant. Environ. Health Perspect. 2013;121:625–630. doi: 10.1289/ehp.1205838. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Yang Q., Xie Y., Eriksson A.M., Nelson B.D., DePierre J.W. Further evidence for the involvement of inhibition of cell proliferation and development in thymic and splenic atrophy induced by the peroxisome proliferator perfluoroctanoic acid in mice. Biochem. Pharmacol. 2001;62:1133–1140. doi: 10.1016/S0006-2952(01)00752-3. [DOI] [PubMed] [Google Scholar]
  55. Zhu Y., Qin X. Di, Zeng X.W., Paul G., Morawska L., Su M.W., Tsai C.H., Wang S.Q., Lee Y.L., Dong G.H. Associations of serum perfluoroalkyl acid levels with T-helper cell-specific cytokines in children: By gender and asthma status. Sci. Total Environ. 2016;559:166–173. doi: 10.1016/j.scitotenv.2016.03.187. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary data 1
mmc1.pdf (2.3MB, pdf)

RESOURCES