Skip to main content
Download PDF
Download ePub

Sex-specific association of per- and polyfluoroalkyl substances (PFAS) exposure with vitamin D concentrations in older adults in the USA: an observational study

Environmental Health volume 23, Article number: 100 (2024) Cite this article

Abstract

Background

Per- and polyfluoroalkyl substances (PFAS) are commonly utilized in consumer products. While earlier studies have suggested potential impacts of certain PFAS on serum concentrations of vitamin D, these investigations were constrained to a limited set of conventional PFAS. Moreover, they did not specifically focus on populations with longer duration of PFAS exposure and potentially higher blood PFAS levels, such as older adults, and lacked adequate evidence to examine sex-related disparities.

Methods

This observational investigation utilized cross-sectional data obtained from the U.S. NHANES spanning the years 2003 to 2018. Survey-weighted multiple regression models were employed to evaluate the relationship between PFAS exposure and vitamin D concentrations. Multi-pollutant models were employed to evaluate the association between PFAS mixtures and vitamin D concentrations. Subsequently, environmental risk scores (ERS) were constructed to gauge associations with vitamin D concentrations. ERS was computed through a weighted linear combination of PFAS, utilizing calculations from ridge regression and adaptive elasticity network (adENET) methodologies. All analyses were stratified by sex.

Results

The study encompassed 3,853 older adults. Our analysis revealed a negative association between PFOA, PFOS, PFNA, and MeFOSAA and serum vitamin D concentrations. In analyses examining mixed exposures, various models consistently indicated an inverse association between PFAS mixed exposure and vitamin D concentrations. Moreover, an increase in ERS of PFAS across the interquartile range was associated with a decrease in vitamin D concentrations (Q4 vs. Q1, adENET: β: -0.083, 95% CI: -0.117, -0.048; ridge regression: β: -0.077, 95% CI: -0.111, -0.042). Notably, these associations were exclusively observed within the female population.

Conclusions

Our study indicates that heightened exposure to PFAS correlates with diminished serum vitamin D concentrations in females aged 60 years and older, evident in both single and mixed exposures. These findings find support in in vitro mechanistic studies, suggesting that PFAS may impact the metabolism of 25(OH)D, consequently affecting vitamin D concentrations.

Peer Review reports

Introduction

Per- and polyfluoroalkyl substances (PFAS) are aliphatic compounds predominantly composed of carbon and fluorine atoms [1]. The exceptional stability of the carbon-fluorine bond contributes to the prolonged half-lives of long-chain PFAS, which can range from three years to several decades [1]. Due to their impressive chemical and thermal stability, PFAS find extensive use in various applications, including kitchen utensils, food containers, medical appliances, and flame retardant foams [2]. Perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS), principal monomers of PFAS, were listed as persistent organic pollutants in the Stockholm Convention. Governmental agreements in the U.S. since the turn of the century aim to curtail PFAS production and use by companies [3]. Nonetheless, PFOS degradation in the environment remains slow to negligible. This persistence allows them to be transported over extensive distances through air and water, ultimately entering diverse ecosystems’ food webs [4]. Consequently, exposure to PFAS can endure over an extended period.

Vitamin D, an essential nutrient facilitating calcium and phosphate absorption, is implicated in various diseases, including immune disorders, cancer, neurocognitive disorders, and cardiovascular ailments [5]. A longitudinal study on aging has revealed that older adults with vitamin D deficiency often exhibit heightened levels of inflammation [6]. Furthermore, ethnically diverse older adults, particularly African Americans and Hispanics with a high prevalence of vitamin D deficiency, are prone to accelerated cognitive decline in the presence of low vitamin D status [7]. Hence, sustaining adequate levels of vitamin D is crucial for overall well-being. Global estimates indicate that one billion individuals across all age groups suffer from vitamin D deficiency (VDD), defined as serum 25-hydroxyvitamin D [25(OH)D] levels below 50 nmol/L (20 ng/mL) [8]. In the U.S, nearly 20% of middle-aged and older adults exhibit vitamin D deficiency [9]. Age-related alterations are evident in both the production and metabolism of vitamin D; a recent investigation indicated a 13% decline in age-related vitamin D production per decade, resulting in a 50% reduction by age 70 compared to age 20 [10]. Consequently, older individuals face elevated odds of insufficient vitamin D concentrations.

In addition to its synthesis in the epidermis via sunlight exposure, vitamin D can be sourced from daily dietary intake or supplements. Various factors impact vitamin D concentrations, encompassing genetic factors, obesity, ethnicity, chronic diseases, and variations in sunlight exposure due to outdoor activity differences [11]. Environmental pollutants, including heavy metals and endocrine disruptors (EDCs) such as polychlorinated biphenyls (PCBs), are believed to disrupt the human vitamin D endocrine system. These pollutants may lower vitamin D concentrations by inhibiting the transcription and activity of cytochrome P450 mixed-function oxidases (CYPs), or indirectly through disorders in parathyroid hormone (PTH) and calcium homeostasis [12]. An investigation utilizing data from the 2003–2010 NHANES revealed a link between serum PFOS concentrations and decreased vitamin D concentrations, while perfluorohexane sulfonic acid (PFHxS) showed an association with increased vitamin D concentrations [13]. Given the widespread presence of PFAS contaminants in the environment and their potential to disrupt endocrine function [14], there is an imperative to examine the influence of PFAS exposure on the vitamin D status among humans, particularly within vulnerable populations. Notably, the heterogeneity in the association between PFAS exposure and vitamin D concentrations, as identified in the Etzel TM study, persisted across age, sex, and race. The authors also found that PFOA and perfluorononanoic acid (PFNA) showed no significant associations with vitamin D concentrations [13]. However, the study did not account for dietary sources of vitamin D intake, including milk consumption, which are recognized as primary dietary contributors to vitamin D concentrations [15]. Another analysis, encompassing 421 individuals, explored the relationship between various pollutants and vitamin D concentrations during pregnancy and did not identify any PFAS linked to diminished vitamin D concentrations [16]. The limited sample size might have constrained the precision of this finding. Importantly, individuals are frequently exposed not merely to a singular PFAS but to multiple PFAS simultaneously. Given the notable correlation among certain PFAS arising from common sources [17], neglecting the consideration of PFAS mixtures could introduce bias into the estimates of individual PFAS.

The concentration of PFAS tends to rise with age. Prior research conducted in New Zealand has established a positive correlation between advancing age and serum PFAS concentrations [18]. Consistently, our study observed significantly higher concentrations of all six PFAS compounds in the elderly population (P < 0.05, Table S3). Furthermore, a cross-sectional study conducted in China revealed a significant association between PFAS exposure and hypertension risk solely among individuals aged 60 years or older, with no such association observed in the younger group [19]. Consequently, the elderly population exhibits heightened susceptibility to PFAS compared to younger and middle-aged cohorts, making them particularly suitable for investigating the long-term health implications of PFAS exposure. In our investigation, we compiled cross-sectional data spanning eight NHANES cycles covering the period from 2003 to 2018. The aim was to explore the association between exposure to the six PFAS, both individually and in combination, and serum vitamin D concentrations in an aging U.S. population. Furthermore, we delved into sex-specific correlations between PFAS exposure and vitamin D concentrations, considering previous research indicating heightened sensitivity to PFAS among females [19], along with the well-established association between estrogen levels and vitamin D metabolism [20].

Methods

Study population

This study integrated cross-sectional data from eight NHANES cycles, covering the period from 2003 to 2018. NHANES, is a longstanding national survey designed to assess the health and nutritional status of the U.S. population. The research protocol obtained ethical clearance from the National Center for Health Statistics. In this investigation, participants aged 60 years and older were included, providing comprehensive data on serum PFAS concentrations, vitamin D serum biomarkers (25(OH)D), and pertinent covariates, as elaborated in the covariates section.

PFAS measurement in serum

The NHANES official website provides comprehensive documentation on the methods and procedures employed in this study. In summary, serum samples were preserved at a temperature of -20 °C. Following formic acid dilution, a 100 µL aliquot of serum underwent injection into a commercial column switching system for analyte concentration and chromatography, culminating in quantitative detection of serum PFAS via turbo ion spray ionization-tandem mass spectrometry (TIS-MS/MS). Serum PFAS concentration measurements were conducted on approximately one-third of the random subsamples in NHANES. The NHANES laboratory analyzed 14 different serum levels of PFAS. NHANES measured the PFOA and PFOS isomers in the 2013-18 cycle, and consequently, serum concentrations of PFOS and PFOA were determined as the combined sum of their respective isomer concentrations. For this study’s final analysis, PFAS compounds with detection rates exceeding 70% were considered, namely: PFOA, PFOS, PFHxS, perfluorodecanoic acid (PFDeA), PFNA, and 2-(N-methylperfluorooctane sulfonamido) acetate (MeFOSAA). The limit of detection (LOD) for the six PFAS analyzed in this study exhibited slight variation across cycles. Concentrations falling below the LOD in each cycle were expressed as the LOD divided by the square root of two. The detection rates and concentration distributions of PFAS in serum are presented in Table S1, with detection rates exceeding 70% for all six PFAS, with PFOS exhibiting the highest concentrations.

Vitamin D measurement

The standard indicator used to assess vitamin D status is the total serum concentration of 25(OH)D, which comprises the combined levels of 25-hydroxyvitamin D2 and 25-hydroxyvitamin D3. Notably, the methodology for measuring serum 25(OH)D has evolved across various NHANES survey cycles. Specifically, assessments were conducted using the DiaSorin RIA kit between 2003 and 2006, transitioning to standardized liquid chromatography-tandem mass spectrometry from 2007 to 2018 [21]. To mitigate potential methodological bias, 25(OH)D concentrations from 2003 to 2006 were transformed into LC-MS/MS equivalents using a regression equation in accordance with official NHANES recommendations [22].

Covariates

The selection of covariates was guided by prior investigations into the association between environmental pollutant exposure and vitamin D concentrations [23]. Covariates were obtained through questionnaires and encompassed age (continuous), sex (categorical), race (categorized as Mexican American, non-Hispanic Black, non-Hispanic White, other Hispanic, other race), marital status (categorized as unmarried or separated, married), education levels (highest grade; high school and below, above high school), body mass index (BMI, continuous), income-to-poverty ratio (PIR, categorized as ≤ 1.30, 1.31–3.50, > 3.50), vitamin D supplement use (yes/no), serum cotinine (a marker of exposure to environmental tobacco smoke, continuous), milk product consumption (categorized as never, rarely (less than weekly), sometimes (weekly or more often), often (once a day or more)), physical activity level (categorized as vigorous or moderate, low or none) and NHANES inspection time period (six-month examination period), which was employed as a proxy variable for season, distinguishing between November to April and May to October.

Statistical analysis

In this study, the comparison of vitamin D concentrations among diverse participant subgroups was conducted using analysis of variance (ANOVA). Given the skewed distribution of both serum vitamin D and PFAS concentrations, a natural logarithmic transformation was implemented to mitigate the influence of outliers. Vitamin D concentrations were incorporated as continuous variables in all our models. All models were adjusted for variables including age, sex, race, marital status, education, BMI, PIR, serum cotinine, vitamin D supplement use, milk product consumption, physical activity level, and six-month examination period. Furthermore, subgroup analyses based on sex were conducted, and potential sex interactions were evaluated using the likelihood ratio test. This test compares the model fit between those with and without interaction terms, thereby elucidating sex-specific associations between PFAS exposure and vitamin D concentrations.

Given the complexity of the multi-stage sampling design of NHANES, our analyses incorporated sub-sample weights in accordance with official NHANES recommendations. Specifically, the PFAS sample weights were calculated using the formula: WTSB16YR = 1/8 * WTSB2YR (where WTSB2YR represents the PFAS sample weights for each cycle, and WTSB16YR denotes the final weight utilized in the analysis). Due to the non-normal distribution of PFAS concentrations (Table S1), a natural logarithmic transformation was applied. To explore the association between individual PFAS concentrations and vitamin D concentrations, we employed survey-weighted multivariate linear regression models to estimate the incremental change in vitamin D concentrations for each 2.7-fold increase in PFAS concentration (i.e., for each unit increase in natural logarithmically transformed PFAS concentration). Furthermore, to investigate potential dose-response patterns, we categorized PFAS concentrations into quartiles, assessing the variance between the 2nd through 4th quartiles and the 1st quartile. A linear trend test was conducted across quartiles. Briefly, the regression model coefficients for the groups were evaluated for significant deviation from zero by calculating the Wald statistic and the corresponding p-value to assess the trend. Additionally, we employed a restricted cubic spline curve (RCS) with four nodes to examine potential nonlinear relationships between serum PFAS levels and vitamin D concentrations.

Given the evident correlation among most PFAS (Figure S1) and their prevalent co-occurrence in the environment, we employed a weighted quantile sum (WQS) regression model to assess the potential influence of PFAS mixtures on vitamin D concentrations. WQS regression quantifies the chemical composition and constructs weighted indices of the chemical mixtures, effectively addressing overfitting and collinearity. The WQS index was developed in this study using quartiles of mixtures of serum PFAS. We randomly assigned 40% of the samples to the test dataset and the remaining 60% to the validation dataset. Due to the non-parametric distribution of the constructed WQS index, corresponding regression coefficients and statistical inferences were derived through 1000 bootstrap samplings. Additionally, recognizing that the WQS regression inherently places directional constraints on individual pollutant effects, we adopted a quantile g-computation (qgcomp) to investigate joint effects using parametric inference. By integrating the inferential structure of WQS with the adaptive characteristics of g-computation, addressed the constraint of directional homogeneity [24]. This approach facilitates the association of various components with the target outcome in diverse directions, thereby discerning both negative and positive weights of each PFAS linked to vitamin D concentrations. It is worth noting that when the health effects of multiple pollutants are nonlinear and nonadditive, the health effect estimates and associated variances derived by qgcomp are unbiased, whereas those derived by WQS may be biased [24]. Each statistical method used to assess the health effects of exposure to mixtures of multiple pollutants has unique strengths and limitations [25]. Therefore, it is necessary to use a combination of applicable statistical methods and implement multiple comparison strategies.

To evaluate the cumulative effect of PFAS exposure on vitamin D, we introduced the Environmental Risk Score (ERS), a personalized metric aimed at assessing the collective risk associated with exposure to chemical mixtures [26]. The ERS was formulated using adaptive elasticity nets (adENET) and ridge regression. In the adENET process, we employed two adjustment parameters (lambda 1 and lambda 2) for predictor variable selection. Lambda 1 narrowed down coefficients of less influential predictor variables to zero, while lambda 2 facilitated stable variable selection amid highly associated PFAS. However, adENET’s variable selection, dependent on other predictor variables removed from the model, could be unstable. To address this, we incorporated ridge regression, a method similar to adENET but utilizing only one adjustment parameter. Ridge regression narrowed unimportant predictor variables to values close to zero (but not precisely zero), ensuring a more stable variable selection compared to adENET. The adjustment parameters for both techniques were refined using quintuple cross-validation and evaluated based on cross-validated prediction errors. The beta coefficients (β) derived from both ridge regression and adENET were applied to each participant’s PFAS concentration matrix to yield weighted concentrations of individual PFAS, subsequently summed to calculate individual ERS, with the ERS being based on the sum of ln-transformed weighted concentrations of PFAS. These ERS were then categorized into quartiles, followed by the utilization of multiple linear regressions to explore the relationships between ascending ERS quartiles and vitamin D concentrations.

To bolster the reliability of our findings, we executed multiple sensitivity analyses. Firstly, we operationally defined VDD as a serum 25(OH)D level below 50 nmol/L [8, 27]. We then explored the association between PFAS exposure and VDD using weighted multiple logistic regression models along with mixed exposure models (WQS, qgcomp). Secondly, while our study focused on individuals aged over 60, we conducted additional analyses involving the younger population aged 20–59 to provide context. Therefore, we added analyses of the association between PFAS exposure and vitamin D concentrations in the 20–59 years old population. Additionally, we utilized Bayesian kernel machine regression (BKMR) to evaluate the cumulative impact of PFAS on vitamin D concentrations. This method estimates exposure-response relationships for multiple pollutants by considering the covariance among their exposure profiles, allowing for a comprehensive fit to potentially complex nonlinear relationships and yielding more robust results [28]. However, ERS constructed using adENET remain the primary method in this study due to their adept variable selection capabilities and ability to accommodate a large number of covariates. In contrast, BKMR is a non-parametric method based on kernel smoothing and recursive partitioning, which can pose challenges when modeling a substantial number of predictors and is susceptible to overfitting [29]. Finally, considering the accumulation of PFAS and vitamin D in adipose tissue, we performed subgroup analyses by obesity level (BMI < 30, BMI ≥ 30) in the female population.

A value of P < 0.05 signifies a statistically significant difference. All statistical computations were executed using R (4.2.3).

Results

Population characteristics

This study initially involved 80,312 NHANES participants from 2003 to 2018. After excluding 11,642 participants due to missing data on six PFAS serum concentrations and serum 25(OH)D concentrations, and further excluding 208 participants with missing covariates and 7,581 individuals younger than 60 years of age, the final analysis included 3,853 participants. Table 1 delineates the primary characteristics of the elderly participants alongside the distribution of serum vitamin D concentrations (nmol/L) based on demographic attributes. The average age of these older individuals stood at 70.4 ± 7.3 years. Furthermore, serum vitamin D concentrations exhibited a lower value among males relative to females (68.5 ± 25.6 vs. 72.3 ± 31.5, P < 0.05). Additionally, certain subgroups, such as non-Hispanic whites, the married population, those with high school education or higher, participants with a PIR > 3.5, users of vitamin D supplements, those with more frequent consumption of milk products, those engaging in vigorous or moderate levels of physical activity, and subjects sampled between May and October, demonstrated significantly higher serum vitamin D concentrations (P < 0.05).

Table 1 Basic characteristics of participants: individuals aged 60 years and older, NHANES 2003–2018
Full size table

Table S1 presents the distribution of concentrations for the six PFAS in the serum of the elderly population, all with detection rates exceeding 70%. PFOS exhibited the highest average concentration (17.31 ng/mL), succeeded by PFOA (3.58 ng/mL), while the mean values for other PFAS ranged between 0.41 and 2.59 ng/mL. Pearson’s correlations among the six PFAS were weak to moderate (r = 0.19 to 0.71) (see Figure S1).

Individual PFAS and vitamin D status

In single chemical models, elevated concentrations of all four PFAS, with the exception of PFHxS and PFDeA, were associated with reduced serum vitamin D concentrations. All four PFAS showed statistical significance in continuous metric: PFOS (β: -0.03, 95% CI: -0.05, -0.02), PFOA (β: -0.03, 95% CI: -0.05, -0.01), PFNA (β: -0.03, 95% CI: -0.05, -0.01), and MeFOSAA (β: -0.03, 95% CI: -0.04, -0.01). Additionally, we observed that the highest quartile (Q4) of PFOA, PFOS, PFNA, MeFOSAA was associated with lower concentrations of vitamin D compared to the lowest quartile (Q1) (refer to Table 2), and a linear trend was evident for all four PFAS in the quartile-based linear trend test (P < 0.05). In sex-based subgroup analyses, the relationship between PFAS exposure and vitamin D concentrations differed between males and females. Findings suggested that increased concentrations of the five PFAS (PFOA, PFOS, PFNA, PFDeA and MeFOSAA) were associated with decreased serum vitamin D concentrations in the female elderly population. The same trend was observed in different quartiles (see Table 2). In contrast, no significant association between individual PFAS exposures and vitamin D concentrations was observed in the male population (P > 0.05). Moreover, sex interaction analysis using the likelihood ratio test revealed a significant interaction between the five PFAS compounds (PFOA, PFOS, PFNA, PFDeA, and MeFOSAA) in the sex subgroups (P value for interaction < 0.05, Table 2).

Table 2 Association of PFAS with vitamin D concentrations (adjusted difference in total 25(OH)D per unit increase in PFAS): individuals aged 60 years and older, NHANES 2003–2018
Full size table

The results of the RCS spline regression, depicted in Fig. 1, reveal that serum concentrations of PFOA, PFOS, and PFDeA exhibit a nonlinear association with serum vitamin D concentrations (P-value < 0.01 for all; P-values for nonlinearity: <0.001, < 0.001, and 0.012, respectively, Fig. 1A, B, E). In contrast, PFNA and MeFOSAA demonstrate linear relationships with serum vitamin D concentrations (P-values of 0.002 and < 0.001, respectively; nonlinearity P-values stood at 0.139 and 0.371, respectively, Fig. 1D, F).

Fig. 1
figure 1

Dose-response relationship between ln-transformed levels of PFAS and serum vitamin D concentrations were estimated by RCS models in female population. (A) PFOA, (B) PFOS, (C) PFHxS, (D) PFNA, (E) PFDeA, (F) MeFOSAA. Models were adjusted for age, race, marital status, education, BMI, PIR, serum cotinine, vitamin D supplement use, milk product consumption, activity and six-month examination period. β: beta coefficient; Solid line: beta coefficient; grey-shade: 95%CI

Full size image

PFAS mixtures and vitamin D status

The WQS model results suggest a negative association between PFAS mixtures and vitamin D concentrations, indicating that concurrent exposure to PFAS correlates with reduced vitamin D concentrations (β: -0.02, 95% CI: -0.04, -0.00, see Fig. 2A), individual PFAS were weighted more heavily (see Figure S2 A-C). In sex-stratified subgroup analyses, a pronounced association emerged between PFAS co-exposure and reduced vitamin D concentrations among females (β: -0.04, 95% CI: -0.07, -0.01), while in the male population, PFAS co-exposure showed no significant association with vitamin D concentrations (β: 0.01, 95% CI: -0.01, 0.03). Within the qgcomp modeling framework, which avoids presupposing a particular direction of PFAS exposure, the estimated exposure weights spanned both negative and positive values. These trends were echoed in the qgcomp analysis, highlighting association between PFAS co-exposure and diminished vitamin D concentrations (β: -0.02, 95% CI: -0.04, -0.00; see Fig. 2B). Subgroup analyses within the female population showed a more pronounced association (β: -0.04, 95% CI: -0.07, -0.02), and the individual PFAS weights, both positive and negative, are illustrated in Figure S2 D-F.

Fig. 2
figure 2

Association of co-exposure to PFAS mixtures with vitamin D concentrations in both the total population and sex subgroups, as evaluated by WQS (A) and qgcomp (B) modeling. Models were adjusted for age, sex, race, marital status, education, BMI, PIR, serum cotinine, vitamin D supplement use, milk product consumption, activity and six-month examination period. β: beta coefficient

Full size image

Table S2 displays the PFAS screened in adENET and ridge regression that influenced vitamin D concentrations and their corresponding weights. In both the total and female, PFOS, PFHxS, and MeFOSAA emerged as stronger predictors in both adENET and ridge regression. However, in the male subgroup, both adENET and ridge regression produced weaker weights. In the linear regression model, PFAS-ERS served as the exposure variable to investigate the association between PFAS mixture exposure and vitamin D concentrations. Table 3 outlines the association between PFAS-ERS quartiles and vitamin D concentrations. Notably, PFAS-ERS in the highest quartile (Q4) exhibited a significant negative association with vitamin D concentrations compared to the lowest quartile (Q1) (adENET: β: -0.083, 95% CI: -0.117, -0.048; ridge regression: β: -0.077, 95% CI: -0.111, -0.042). In the female subgroup, this association was more pronounced (adENET: β: -0.168, 95% CI: -0.218, -0.117; ridge regression: β: -0.177, 95% CI: -0.227, -0.126), and all revealed a significant linear trend (P < 0.05) based on quartile results. Interestingly, although not significant in the male population, the ERS for PFAS was significantly and negatively associated with vitamin D concentrations in the third quartile (Q3) compared to the Q1 (adENET: β = -0.046, 95% CI: -0.090, -0.002).

Table 3 The association of quartiles of environmental risk score (ERS) with vitamin D concentrations: adjusted difference in total 25(OH)D per unit increase in ERS
Full size table

Sensitivity analysis

The results were as follows. Firstly, the odds of VDD prevalence were assessed using odds ratios (OR) and 95%CI. A significant increase in VDD odds was observed solely in the female subgroup in the PFOS single-chemical exposure analyses, particularly with exposures to PFOS, PFNA, PFDeA, and MeFOSAA. Specifically, we observed a 30% increase in VDD odds for each 2.7-fold increase in PFOS (OR = 1.30, 95% CI: 1.09–1.55), a 27% increase for each 2.7-fold rise in PFNA (OR = 1.27, 95% CI: 1.06–1.54), a 25% increase for each 2.7-fold increase in PFDeA (OR = 1.25, 95% CI: 1.02–1.53), and a 24% increase for every 2.7-fold rise in MeFOSAA (OR = 1.24, 95% CI: 1.06–1.45). This association persisted in the analysis of mixed PFAS exposure, but no significant associations were noted in the overall or male subgroups (see Table 4). Secondly, stratified analyses by age groups revealed lower vitamin D concentrations and levels of six serum PFAS in individuals aged 20–59 years compared to those aged 60 years and older (P < 0.05, see Table S3). However, in the analysis of the association between PFAS exposure and vitamin D concentrations, some PFAS exposures were positively associated with vitamin D concentrations in the 20–59 years old population, including PFOA, PFOS, PFHxS, and PFDeA, contrary to results in the elderly population (Table S4). In addition, the BKMR results aligned with previous analyses, indicating a significant negative association between mixed PFAS exposure and serum vitamin D concentrations. Notably, this association was markedly stronger in the female population and not significantly different in the male population (see Figure S3). Additionally, the BKMR results indicated that PFOS and MeFOSAA were the primary contributors to the association between mixed PFAS exposure and vitamin D concentrations in both the overall and female populations, aligning with the findings of WQS and qgcomp analyses (Figure S2, S3). Finally, Subgroup analyses based on obesity levels revealed that the association between PFAS exposure and vitamin D concentrations differed between individuals with BMI < 30 and those with BMI ≥ 30, with a more pronounced reduction in vitamin D concentrations among those with BMI ≥ 30. However, interaction analyses showed a significant interaction only for PFDeA in the obesity level subgroups (P value for interaction < 0.05, Table S5).

Table 4 Associations of PFAS with vitamin D deficiency: survey-weighted logistic regression analysis for continuous variables and mixture of WQS and qgcomp analysis
Full size table

Discussion

In this study, we identified a robust negative association between the exposure to PFOA, PFOS, PFNA, and MeFOSAA alone, and serum vitamin D concentrations within an elderly population. RCS modeling provided support for both linear and nonlinear relationships observed in the data. Furthermore, PFAS mixture exposure models showed a significant negative correlation between higher PFAS levels and vitamin D concentrations. Notably, these associations were exclusively observed in older females. Furthermore, our supplementary examination of vitamin D deficiency indicated that exposure to PFAS, either individually or in combination, significantly heightens the likelihood of vitamin D deficiency among older females.

Human exposure to PFAS is extensive, and despite the existence of up to 4000 commercially reported PFAS, PFOA and PFOS remain the primary compounds detected in humans, with PFHxS and PFNA closely following [30]. Limited epidemiological studies have suggested associations between PFAS and total 25(OH)D concentrations. A neonatal birth cohort study in Wuhan, China, revealed that exposure to PFNA, PFUnDA, PFTrDA, PFHxS, and PFOS in umbilical cord blood was linked to elevated total 25(OH)D concentrations in male infants [31]. Similarly, among African American women, a positive association was observed between 25(OH)D concentrations and PFHxS, PFOS, PFOA, and PFDeA, while PFPeA showed a negative association specifically in pregnancies with male fetuses [32]. Conversely, in a cross-sectional study of obese children, PFNA exposure was significantly associated with adverse skeletal parameters, with no observed relationship between the four PFAS exposures and 25(OH)D concentrations [33]. Furthermore, an analysis based on NHANES data indicated that a two-fold increase in serum PFOS was associated to a 0.9 nmol/L decrease in total 25(OH)D concentration, with a more pronounced effect observed in the elderly [13]. Conversely, a two-fold increase in PFHxS was associated to a 0.8 nmol/L increase in total 25(OH)D. Notably, PFOA and PFNA exhibited no significant association with total 25(OH)D concentrations, and this lack of association remained consistent across age, sex, and race [13]. Conversely, our investigation produced distinct findings, revealing significant negative correlations between PFOS, PFOA, PFNA, and MeFOSAA exposures and serum total 25(OH)D concentrations in the elderly population. Moreover, employing various statistical methods to analyze mixed PFAS exposures consistently supported the conclusion that exposure to a PFAS mixture significantly diminishes serum total 25(OH)D concentrations in the elderly. Notably, our sensitivity analyses indicated positive associations between PFOA, PFOS, PFHxS, and PFDeA exposures and serum total 25(OH)D concentrations among individuals aged 20–59 years, aligning with findings from previous studies [31, 32].

We provide the following explanations for some of the disparities observed compared to previously reported findings. Firstly, our results’ robustness is significantly bolstered by the inclusion of the extended NHANES cycle (2003–2018), the broadening of the study population pool, and the incorporation of additional covariates, including dietary conditions, and physical activity. Secondly, Etzel’s earlier NHANES-based analyses did not integrate subsample weights for PFAS in the multiple linear regression analyses, although cluster and strata weights in the population dataset were considered [13]. In contrast, our study results were derived from a survey-weighted multiple linear regression model. The study suggests that within the NHANES dataset, the unweighted model’s performance was exaggerated compared to the weighted model, with the latter offering a more accurate reflection of model predictions in the target population [34]. This discrepancy might also underlie some of the inconsistencies observed in our study’s outcomes.

The mechanism by which PFAS affects vitamin D concentrations in the body is not fully understood; however, some studies propose that endocrine-disrupting chemicals could influence vitamin D homeostasis [16]. We observed a significant negative correlation between higher levels of PFAS exposure and vitamin D concentrations in an elderly population. Given that vitamin D is pivotal for mineral and bone homeostasis, its depletion in the body intensifies with age, particularly as the elderly encounter heightened bone loss. Exposure to PFAS may disrupt the metabolic profile of vitamin D, exacerbating depletion levels. Inadequate vitamin D concentrations impede optimal calcium absorption, resulting in compensatory mechanisms such as elevated PTH levels (secondary hyperparathyroidism), this, in turn, stimulates bone resorption, hastening bone loss and potentially contributing to osteoporosis [35]. A prospective study from the POUNDS-LOST trial demonstrated that higher plasma PFAS concentrations were associated not only with lower baseline bone mineral density (BMD) but also with a faster decline in BMD in a bariatric trial setting. The authors further noted that, with the exception of PFHxS, all four PFAS compounds were negatively correlated with BMD at multiple skeletal sites [36]. In our study, we similarly observed a significant negative correlation between five PFAS compounds (excluding PFHxS) and vitamin D concentrations. Additionally, prospective analyses from two child cohorts suggest that bone may be a target tissue for PFAS, as higher exposure levels were linked to lower BMD Z-scores [37, 38]. Another multi-cohort study found that PFOS is associated with long-term changes in BMD, with exposure during adolescence and young adulthood leading to reduced BMD [39]. Intriguingly, in a case-control study conducted in Saudi Arabia, exposures to PFAS were linked to an elevated risk of osteoporosis [40]. Analogously, findings from a Swedish cohort study highlighted the detrimental impact of PFAS on osteoporosis, particularly pronounced in individuals aged 50 and above [41]. Furthermore, a study utilizing molecular docking and free energy binding calculations explored the potential binding sites of PFAS to the VDR, revealing that 83 out of 256 PFAS exhibited stronger binding to the VDR than PFOA [42]. Intriguingly, our study identified four PFAS (PFOA, PFOS, PFNA, and MeFOSAA) significantly and negatively correlated with serum vitamin D concentrations, all of which were among the 83 PFAS highlighted in the aforementioned investigation. Furthermore, research has shown that the activation of VDR can lead to the transcriptional upregulation of 25-dihydroxyvitamin D-24-hydroxylase (CYP24A1), a key cytochrome P450 enzyme involved in the metabolism of 25(OH)D in the liver [43]. PFAS, through binding to VDR, may induce abnormal activation of CYP24A1, resulting in decreased serum vitamin D concentrations.

Remarkably, our study revealed a sex-specific effect in the relationship between PFAS exposure and vitamin D concentrations, with females demonstrating a greater sensitivity to these associations. Furthermore, it’s worth noting that menstruation serves as a significant pathway for PFAS elimination in premenopausal women [44]. Animal studies also suggest potential sex-specific clearance rates for PFAS in males and females [45]. While epidemiological studies have not extensively explored this sex disparity in the impact of PFAS on vitamin D concentrations, a U.S. population-based study indicated associations between elevated serum PFAS concentrations and reduced osteoporosis and bone mineral density, predominantly observed among women [46]. Reproductive hormones, such as estrogen, may contribute to this phenomenon, as studies have indicated associations between PFAS concentrations and lower estrogen levels in women [47]. In vitro studies have further demonstrated that PFAS attenuates estradiol-stimulated luciferase activity at concentrations comparable to those found in human serum [48]. Importantly, research indicates that estrogen enhances the functionality of vitamin D, facilitating its accumulation and elevating the expression of VDR. In contrast, vitamin D reduces the expression of aromatase in immune cells, resulting in diminished estrogen levels [20]. This intricate interplay may render females more sensitive to PFAS toxicity. Additionally, in the ERS constructed based on PFAS mixtures, we observed interesting results: in males within the Q4 of ERS, we observed a negative association with vitamin D concentrations. Although this correlation did not attain statistical significance, and the linear trend test based on ERS quartiles was also not statistically significant (P = 0.061), suggesting that the male population might possess a higher threshold for PFAS exposure. Additional studies are needed to clarify the sex-specific effects of PFAS exposure on vitamin D concentrations.

In sensitivity analyses, age-specific associations between PFAS exposure and 25(OH)D concentrations were observed, particularly positive associations between PFOA, PFOS, PFHxS, and PFDeA exposures and serum total 25(OH)D concentrations in individuals aged 20–59. While residual confounders could contribute, reverse causality appears to be a more plausible explanation. Epidemiological studies have demonstrated a positive association between PFAS (PFOA, PFOS, PFDeA, PFHxS, and PFNA) exposure and serum calcium levels [49]. In vitro experiments have further indicated that PFOS-induced autophagy leads to increased mitochondrial calcium accumulation and elevated intracellular calcium ion levels [50]. Notably, heightened serum calcium levels inhibit PTH release via the calcium-sensitive receptor (CaSR), subsequently inhibiting the activity of CYP27B1—the primary enzyme responsible for converting 25(OH)D to its active metabolite, 1,25(OH)2D, in the kidney [51]. Negative feedback regulation due to heightened serum calcium levels results in increased circulating 25(OH)D concentrations. However, studies indicate that in the elderly population, serum calcium levels are typically lower and tend to decline with age [52]. Consequently, the elevation of serum calcium induced by PFAS exposure may not be as pronounced in the elderly, thereby masking its effects, partially elucidating the age-specific association between PFAS exposure and 25(OH)D concentrations.

A notable strength of our study is its extensive sample size combined with national representativeness. Rigorous quality control measures within the NHANES framework ensured the precision of both exposure and outcome assessments. Furthermore, our findings exhibited robustness, as evidenced by consistent results across multiple sensitivity analyses, including the relationship between PFAS exposure and VDD. Furthermore, this study stands as the most comprehensive investigation to date on the relationship between exposure to six PFAS mixtures and vitamin D concentrations within the elderly U.S. population. We employed multiple mixture modeling approaches, including WQS, qgcomp, and ERS construction, thereby offering a thorough assessment of mixing effects. Lastly, the ample sample size empowered us to discern sex-specific associations between PFAS exposure and vitamin D concentrations, a facet less explored in preceding studies. Several limitations warrant consideration in interpreting our findings. Firstly, the cross-sectional design of our study precludes the establishment of temporal relationships between exposure and outcomes, thereby limiting our ability to infer causality. Secondly, despite our efforts to control for potential confounders by adjusting for numerous covariates, the absence of data on sunscreen usage and duration of outdoor sun exposure among older adults in the NHANES questionnaire may introduce bias into our analyses. Thirdly, due to the design limitations of the multipollutant model, we were unable to incorporate NHANES subsample weights in our modeling, which may have introduced bias. However, we enhanced the robustness of our results by comparing multiple statistical methods. Lastly, it’s worth noting that real-world human exposure extends beyond the six common PFAS compounds we examined; notably, the NHANES dataset provided scant information on short-chain PFAS. Consequently, future research should explore a broader array of PFAS compounds and their combinations in diverse human exposure contexts.

Conclusion

In summary, our study revealed an inverse association between serum levels of four PFAS compounds (PFOA, PFOS, PFNA, and MeFOSAA) and vitamin D concentrations among individuals aged 60 years and older in the U.S. Moreover, exposure to PFAS mixtures associated with diminished serum vitamin D concentrations. Notably, this association was exclusively observed in females. These findings are pertinent for assessing the potential health risks posed by PFAS, particularly in vulnerable populations such as women aged 60 years and older. Moreover, they may furnish epidemiological evidence crucial for crafting more targeted regulatory policies. Further prospective and experimental investigations are essential to confirm or dispute the causal link between PFAS exposure and serum vitamin D concentrations and to elucidate the underlying mechanisms.

Data availability

No datasets were generated or analysed during the current study.

References

  1. Buck RC, Franklin J, Berger U, Conder JM, Cousins IT, de Voogt P, Jensen AA, Kannan K, Mabury SA, van Leeuwen SP. Perfluoroalkyl and polyfluoroalkyl substances in the environment: terminology, classification, and origins. Integr Environ Assess Manag. 2011;7(4):513–41.

    Article  CAS  Google Scholar 

  2. Gluge J, Scheringer M, Cousins IT, DeWitt JC, Goldenman G, Herzke D, Lohmann R, Ng CA, Trier X, Wang Z. An overview of the uses of per- and polyfluoroalkyl substances (PFAS). Environ Sci Process Impacts. 2020;22(12):2345–73.

    Article  CAS  Google Scholar 

  3. Cotruvo JA, Goldhaber SB, Cohen AJB. EPA’s Unprecedented Interim Drinking Water Health Advisories for PFOA and PFOS. Ground Water. 2023;61(3):301–3.

    Article  CAS  Google Scholar 

  4. Koch A, Jonsson M, Yeung LWY, Karrman A, Ahrens L, Ekblad A, Wang T. Per- and polyfluoroalkyl-contaminated freshwater impacts adjacent riparian food webs. Environ Sci Technol. 2020;54(19):11951–60.

    Article  CAS  Google Scholar 

  5. Amrein K, Scherkl M, Hoffmann M, Neuwersch-Sommeregger S, Kostenberger M, Tmava Berisha A, Martucci G, Pilz S, Malle O. Vitamin D deficiency 2.0: an update on the current status worldwide. Eur J Clin Nutr. 2020;74(11):1498–513.

    Article  CAS  Google Scholar 

  6. Laird E, O’Halloran AM, Molloy AM, Healy M, Bourke N, Kenny RA. Vitamin D status & associations with inflammation in older adults. PLoS ONE. 2023;18(6):e0287169.

    Article  CAS  Google Scholar 

  7. Miller JW, Harvey DJ, Beckett LA, Green R, Farias ST, Reed BR, Olichney JM, Mungas DM, DeCarli C. Vitamin D Status and Rates of Cognitive decline in a multiethnic cohort of older adults. JAMA Neurol. 2015;72(11):1295–303.

    Article  Google Scholar 

  8. Giustina A, Bouillon R, Dawson-Hughes B, Ebeling PR, Lazaretti-Castro M, Lips P, Marcocci C, Bilezikian JP. Vitamin D in the older population: a consensus statement. Endocrine. 2023;79(1):31–44.

    Article  CAS  Google Scholar 

  9. Wang TY, Wang HW, Jiang MY. Prevalence of vitamin D deficiency and associated risk of all-cause and cause-specific mortality among middle-aged and older adults in the United States. Front Nutr. 2023;10:1163737.

    Article  Google Scholar 

  10. Chalcraft JR, Cardinal LM, Wechsler PJ, Hollis BW, Gerow KG, Alexander BM, Keith JF, Larson-Meyer DE. Vitamin D synthesis following a single Bout of Sun exposure in older and younger men and women. Nutrients 2020, 12(8).

  11. Thacher TD, Clarke BL. Vitamin D insufficiency. Mayo Clin Proc. 2011;86(1):50–60.

    Article  CAS  Google Scholar 

  12. Mousavi SE, Amini H, Heydarpour P, Amini Chermahini F, Godderis L. Air pollution, environmental chemicals, and smoking may trigger vitamin D deficiency: evidence and potential mechanisms. Environ Int. 2019;122:67–90.

    Article  CAS  Google Scholar 

  13. Etzel TM, Braun JM, Buckley JP. Associations of serum perfluoroalkyl substance and vitamin D biomarker concentrations in NHANES, 2003–2010. Int J Hyg Environ Health. 2019;222(2):262–9.

    Article  CAS  Google Scholar 

  14. Radke EG, Wright JM, Christensen K, Lin CJ, Goldstone AE, Lemeris C, Thayer KA. Epidemiology evidence for Health effects of 150 per- and polyfluoroalkyl substances: a systematic evidence map. Environ Health Perspect. 2022;130(9):96003.

    Article  CAS  Google Scholar 

  15. Schmid A, Walther B. Natural vitamin D content in animal products. Adv Nutr. 2013;4(4):453–62.

    Article  CAS  Google Scholar 

  16. Berger K, Bradshaw PT, Poon V, Kharrazi M, Eyles D, Ashwood P, Lyall K, Volk HE, Ames J, Croen LA, et al. Mixture of air pollution, brominated flame retardants, polychlorinated biphenyls, per- and polyfluoroalkyl substances, and organochlorine pesticides in relation to vitamin D concentrations in pregnancy. Environ Pollut. 2024;340(Pt 2):122808.

    Article  CAS  Google Scholar 

  17. Zhan W, Qiu W, Ao Y, Zhou W, Sun Y, Zhao H, Zhang J. Environmental exposure to Emerging Alternatives of Per- and polyfluoroalkyl substances and polycystic ovarian syndrome in women diagnosed with infertility: a mixture analysis. Environ Health Perspect. 2023;131(5):57001.

    Article  CAS  Google Scholar 

  18. Coakley J, Bridgen P, Mueller J, Douwes J, t Mannetje A. Polybrominated diphenyl ethers and perfluorinated alkyl substances in blood serum of New Zealand adults, 2011–2013. Chemosphere. 2018;208:382–9.

    Article  CAS  Google Scholar 

  19. Mi X, Yang YQ, Zeeshan M, Wang ZB, Zeng XY, Zhou Y, Yang BY, Hu LW, Yu HY, Zeng XW, et al. Serum levels of per- and polyfluoroalkyl substances alternatives and blood pressure by sex status: isomers of C8 health project in China. Chemosphere. 2020;261:127691.

    Article  CAS  Google Scholar 

  20. Dupuis ML, Pagano MT, Pierdominici M, Ortona E. The role of vitamin D in autoimmune diseases: could sex make the difference? Biol Sex Differ. 2021;12(1):12.

    Article  CAS  Google Scholar 

  21. Tai SS, Bedner M, Phinney KW. Development of a candidate reference measurement procedure for the determination of 25-hydroxyvitamin D3 and 25-hydroxyvitamin D2 in human serum using isotope-dilution liquid chromatography-tandem mass spectrometry. Anal Chem. 2010;82(5):1942–8.

    Article  CAS  Google Scholar 

  22. Yetley EA, Pfeiffer CM, Schleicher RL, Phinney KW, Lacher DA, Christakos S, Eckfeldt JH, Fleet JC, Howard G, Hoofnagle AN, et al. NHANES monitoring of serum 25-hydroxyvitamin D: a roundtable summary. J Nutr. 2010;140(11):S2030–45.

    Article  Google Scholar 

  23. Johns LE, Ferguson KK, Cantonwine DE, McElrath TF, Mukherjee B, Meeker JD. Urinary BPA and phthalate metabolite concentrations and plasma vitamin D levels in pregnant women: a repeated measures Analysis. Environ Health Perspect. 2017;125(8):087026.

    Article  Google Scholar 

  24. Keil AP, Buckley JP, O’Brien KM, Ferguson KK, Zhao S, White AJ. A quantile-based g-Computation Approach to addressing the effects of exposure mixtures. Environ Health Perspect. 2020;128(4):47004.

    Article  Google Scholar 

  25. Yu L, Liu W, Wang X, Ye Z, Tan Q, Qiu W, Nie X, Li M, Wang B, Chen W. A review of practical statistical methods used in epidemiological studies to estimate the health effects of multi-pollutant mixture. Environ Pollut. 2022;306:119356.

    Article  CAS  Google Scholar 

  26. Jiang M, Zhao H. Joint association of heavy metals and polycyclic aromatic hydrocarbons exposure with depression in adults. Environ Res. 2023;242:117807.

    Article  Google Scholar 

  27. Holick MF. Vitamin D deficiency. N Engl J Med. 2007;357(3):266–81.

    Article  CAS  Google Scholar 

  28. Bobb JF, Claus Henn B, Valeri L, Coull BA. Statistical software for analyzing the health effects of multiple concurrent exposures via bayesian kernel machine regression. Environ Health. 2018;17(1):67.

    Article  Google Scholar 

  29. Park SK, Zhao Z, Mukherjee B. Construction of environmental risk score beyond standard linear models using machine learning methods: application to metal mixtures, oxidative stress and cardiovascular disease in NHANES. Environ Health. 2017;16(1):102.

    Article  Google Scholar 

  30. Kannan K, Corsolini S, Falandysz J, Fillmann G, Kumar KS, Loganathan BG, Mohd MA, Olivero J, Van Wouwe N, Yang JH, et al. Perfluorooctanesulfonate and related fluorochemicals in human blood from several countries. Environ Sci Technol. 2004;38(17):4489–95.

    Article  CAS  Google Scholar 

  31. Liu H, Huang Y, Pan Y, Cheng R, Li X, Li Y, Lu S, Zhou A, Dai J, Xu S. Associations between per and polyfluoroalkyl ether sulfonic acids and vitamin D biomarker levels in Chinese newborns. Sci Total Environ. 2023;866:161410.

    Article  CAS  Google Scholar 

  32. Chang CJ, Barr DB, Zhang Q, Dunlop AL, Smarr MM, Kannan K, Panuwet P, Tangpricha V, Shi L, Liang D, et al. Associations of single and multiple per- and polyfluoroalkyl substance (PFAS) exposure with vitamin D biomarkers in African American women during pregnancy. Environ Res. 2021;202:111713.

    Article  CAS  Google Scholar 

  33. Khalil N, Ebert JR, Honda M, Lee M, Nahhas RW, Koskela A, Hangartner T, Kannan K. Perfluoroalkyl substances, bone density, and cardio-metabolic risk factors in obese 8–12 year old children: a pilot study. Environ Res. 2018;160:314–21.

    Article  CAS  Google Scholar 

  34. MacNell N, Feinstein L, Wilkerson J, Salo PM, Molsberry SA, Fessler MB, Thorne PS, Motsinger-Reif AA, Zeldin DC. Implementing machine learning methods with complex survey data: lessons learned on the impacts of accounting sampling weights in gradient boosting. PLoS ONE. 2023;18(1):e0280387.

    Article  CAS  Google Scholar 

  35. Di Nisio A, Rocca MS, De Toni L, Sabovic I, Guidolin D, Dall’Acqua S, Acquasaliente L, De Filippis V, Plebani M, Foresta C. Endocrine disruption of vitamin D activity by perfluoro-octanoic acid (PFOA). Sci Rep. 2020;10(1):16789.

    Article  Google Scholar 

  36. Hu Y, Liu G, Rood J, Liang L, Bray GA, de Jonge L, Coull B, Furtado JD, Qi L, Grandjean P et al. Perfluoroalkyl substances and changes in bone mineral density: a prospective analysis in the POUNDS-LOST study. Environ Res 2019, 179(Pt A):108775.

  37. Blomberg A, Mortensen J, Weihe P, Grandjean P. Bone mass density following developmental exposures to perfluoroalkyl substances (PFAS): a longitudinal cohort study. Environ Health. 2022;21(1):113.

    Article  CAS  Google Scholar 

  38. Hojsager FD, Sigvaldsen A, Andersen MS, Juul A, Nielsen F, Moller S, Christesen HBT, Grontved A, Grandjean P, Jensen TK. Prenatal and early postnatal exposure to perfluoroalkyl substances and bone mineral content and density in the Odense Child Cohort. Environ Int. 2023;181:108264.

    Article  Google Scholar 

  39. Beglarian E, Costello E, Walker DI, Wang H, Alderete TL, Chen Z, Valvi D, Baumert BO, Rock S, Rubbo B, et al. Exposure to perfluoroalkyl substances and longitudinal changes in bone mineral density in adolescents and young adults: a multi-cohort study. Environ Res. 2024;244:117611.

    Article  CAS  Google Scholar 

  40. Banjabi AA, Li AJ, Kumosani TA, Yousef JM, Kannan K. Serum concentrations of perfluoroalkyl substances and their association with osteoporosis in a population in Jeddah, Saudi Arabia. Environ Res. 2020;187:109676.

    Article  CAS  Google Scholar 

  41. Xu Y, Hansson E, Andersson EM, Jakobsson K, Li H. High exposure to perfluoroalkyl substances in drinking water is associated with increased risk of osteoporotic fractures - A cohort study from Ronneby, Sweden. Environ Res. 2023;217:114796.

    Article  CAS  Google Scholar 

  42. Azhagiya Singam ER, Durkin KA, La Merrill MA, Furlow JD, Wang JC, Smith MT. The vitamin D receptor as a potential target for the toxic effects of per- and polyfluoroalkyl substances (PFASs): an in-silico study. Environ Res. 2023;217:114832.

    Article  CAS  Google Scholar 

  43. Jones G, Prosser DE, Kaufmann M. 25-Hydroxyvitamin D-24-hydroxylase (CYP24A1): its important role in the degradation of vitamin D. Arch Biochem Biophys. 2012;523(1):9–18.

    Article  CAS  Google Scholar 

  44. Wong F, MacLeod M, Mueller JF, Cousins IT. Enhanced elimination of perfluorooctane sulfonic acid by menstruating women: evidence from population-based pharmacokinetic modeling. Environ Sci Technol. 2014;48(15):8807–14.

    Article  CAS  Google Scholar 

  45. Betts KS. Perfluoroalkyl acids: what is the evidence telling us? Environ Health Perspect. 2007;115(5):A250–256.

    Article  CAS  Google Scholar 

  46. Khalil N, Chen A, Lee M, Czerwinski SA, Ebert JR, DeWitt JC, Kannan K. Association of Perfluoroalkyl Substances, Bone Mineral density, and osteoporosis in the U.S. Population in NHANES 2009–2010. Environ Health Perspect. 2016;124(1):81–7.

    Article  CAS  Google Scholar 

  47. Liu Y, Calafat AM, Chen A, Lanphear BP, Jones NY, Cecil KM, Rose SR, Yolton K, Buckley JP, Braun JM. Associations of prenatal and postnatal exposure to perfluoroalkyl substances with pubertal development and reproductive hormones in females and males: the HOME study. Sci Total Environ. 2023;890:164353.

    Article  CAS  Google Scholar 

  48. Li J, Cao H, Feng H, Xue Q, Zhang A, Fu J. Evaluation of the Estrogenic/Antiestrogenic Activities of Perfluoroalkyl Substances and their interactions with the human estrogen receptor by combining in Vitro assays and in Silico Modeling. Environ Sci Technol. 2020;54(22):14514–24.

    Article  CAS  Google Scholar 

  49. Cakmak S, Lukina A, Karthikeyan S, Atlas E, Dales R. The association between blood PFAS concentrations and clinical biochemical measures of organ function and metabolism in participants of the Canadian Health measures Survey (CHMS). Sci Total Environ. 2022;827:153900.

    Article  CAS  Google Scholar 

  50. Mokra K, Kaczmarska I, Bukowska B. Perfluorooctane sulfonate (PFOS) and its selected analogs induce various cell death types in peripheral blood mononuclear cells. Chemosphere. 2024;354:141664.

    Article  CAS  Google Scholar 

  51. Goltzman D, Mannstadt M, Marcocci C. Physiology of the calcium-parathyroid hormone-vitamin D Axis. Front Horm Res. 2018;50:1–13.

    Article  CAS  Google Scholar 

  52. Roof BS, Piel CF, Hansen J, Fudenberg HH. Serum parathyroid hormone levels and serum calcium levels from birth to senescence. Mech Ageing Dev. 1976;5(4):289–304.

    Article  CAS  Google Scholar 

Download references

Acknowledgements

Not applicable.

Funding

This work was supported by grants from the National Natural Science Foundation of China (82073655, 82373672).

Author information

Author notes

  1. Hui Zhao, Yuxin Ren and Jianping Ni contributed equally to this work and should be considered co-first authors.

Authors and Affiliations

  1. Department of Epidemiology and Biostatistics, School of Public Health, Anhui Medical University, 81 Meishan Rd, Hefei, Anhui, 230032, China

    Hui Zhao, Yuxin Ren, Jianping Ni, Lanlan Fang, Tao Zhang, Mengmeng Wang, Guoqi Cai, Yubo Ma & Faming Pan

  2. The Inflammation and Immune Mediated Diseases Laboratory of Anhui Province, Anhui Medical University, 81 Meishan Road, Hefei, Anhui, 230032, China

    Hui Zhao, Yuxin Ren, Jianping Ni, Lanlan Fang, Tao Zhang, Mengmeng Wang, Guoqi Cai, Yubo Ma & Faming Pan

Authors
  1. Hui Zhao
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Yuxin Ren
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Jianping Ni
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Lanlan Fang
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Tao Zhang
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Mengmeng Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Guoqi Cai
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Yubo Ma
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Faming Pan
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

Study concept and design: Z-H, M-YB and P-FM. Acquisition of data: F-LL. Analysis of data: Z-T, W-MM, and C-GQ. Interpretation of data: All authors. Drafting of the manuscript: Z-H, R-YX and N-JP. Critical revision of the manuscript for important intellectual content: All authors.

Corresponding authors

Correspondence to Yubo Ma or Faming Pan.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zhao, H., Ren, Y., Ni, J. et al. Sex-specific association of per- and polyfluoroalkyl substances (PFAS) exposure with vitamin D concentrations in older adults in the USA: an observational study. Environ Health 23, 100 (2024). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12940-024-01140-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12940-024-01140-9

Keywords

Environmental Health

ISSN: 1476-069X