The Association of Combined Per- and Polyfluoroalkyl Substances and Metals with Allostatic Load Using Bayesian Kernel Machine Regression

**Background:** The exposome concept recognizes that cumulative environmental exposures affect health, yet little is understood about how mixtures of per- and polyfluoroalkyl substances (PFAS) and metals interact to influence the stress response system. Allostatic load (AL) is a composite biological measure of chronic physiological stress, reflecting wear and tear on the body from repeated activation of the hypothalamic-pituitary-adrenal (HPA) axis. This study hypothesized that combined exposure to PFAS and metals is associated with elevated AL.

**Methods:** Data were drawn from the National Health and Nutrition Examination Survey (NHANES) 2007–2014, including adults aged 20 years and older. AL was calculated from 10 biomarkers: systolic blood pressure, diastolic blood pressure, total cholesterol, HDL cholesterol, glycosylated hemoglobin (HbA1c), albumin, triglycerides, body mass index, creatinine clearance, and C-reactive protein. High-risk quartiles (top 25% for most markers; bottom 25% for albumin, creatinine clearance, and HDL) were summed to create an AL index out of 10, with values ≥3 classified as elevated. PFAS were quantified in serum using high-performance liquid chromatography–tandem mass spectrometry; metals were measured in whole blood using inductively coupled plasma mass spectrometry. Bayesian Kernel Machine Regression (BKMR) with hierarchical variable selection was used to model the joint exposure–response relationship, adjusting for sex, age, smoking, physical activity, ethnicity, occupation, income, alcohol consumption, education, birthplace, and time in the U.S. Posterior inclusion probabilities (PIPs) were calculated to quantify variable importance.

**Key Results:** The six highest PIPs were observed for cesium (PIP = 1.000), molybdenum (PIP = 1.000), PFHS (PIP = 0.854), PFNA (PIP = 0.824), mercury (PIP = 0.807), and PFOA (PIP = 0.754). The overall effect estimate for the BKMR model was 0.680. Univariate exposure-response functions showed positive associations with AL for PFNA, PFUA, PFOA, PFHS, mercury, cesium, thallium, tungsten, and uranium. Mean AL levels varied by ethnicity and age: non-Hispanic Blacks had the highest AL means across all age groups (3.32 for ages 20–39, 3.92 for ages 40–59, 3.83 for ages 60+), while the Other and Asian group had the lowest (2.63, 3.08, and 3.13, respectively). The strongest correlation among exposures was between cesium and mercury (r = 0.438). Mean contaminant levels also differed by ethnicity: Asians had the highest molybdenum (65.9), cesium (5.37), and PFNA (0.26); Blacks had the highest mercury (0.66); and Whites had the highest PFOA (0.97) and PFHS (0.82).

**Clinical Implications:** This study provides evidence that combined exposure to PFAS and metals is associated with elevated allostatic load, suggesting that environmental contaminant mixtures may contribute to chronic stress-related health outcomes. The findings highlight cesium, molybdenum, mercury, PFHS, and PFNA as priority contaminants in mixture analyses. Ethnic and age differences in both exposure levels and AL underscore the potential role of environmental exposures in health disparities. The authors note that AL may serve as a mediator between environmental exposures and adverse health outcomes such as cardiovascular disease, metabolic syndrome, and obesity. Limitations include the cross-sectional design (precluding causal inference), the use of secondary data, and the lack of laboratory studies on combined PFAS–metal toxicity. Longitudinal studies and experimental research are needed to confirm these findings and elucidate underlying mechanisms.