Exposome-wide association study of thyroid function using U.S. National Health and Nutrition Examination Survey data

Abstract

Previous epidemiologic studies examining thyroid function and chemical exposures have typically focused on a single or a limited number of chemical classes, often neglecting the effects of chemical mixtures. This study addressed this gap by exploring the associations between exposure to hundreds of chemicals and thyroid function using an exposome-wide association study (ExWAS) approach and National Health and Nutrition Examination Survey (NHANES) data. We analyzed data from three NHANES cycles (2007-2008, 2009-2010, and 2011-2012), which include measures of thyroid function (free and total triiodothyronine [T3], free and total thyroxine [T4], thyroid-stimulating hormone [TSH]) and chemical biomarker concentrations from 9,082 participants. For adolescents (aged 12-19 years) and adults (aged ≥20 years), we employed multiple regression by accounting for survey weights to identify biomarkers associated with thyroid function test levels and used Bayesian group weighted quantile sum (BGWQS) regression to assess the effects of chemical mixtures on these measurements. After adjusting for multiple comparisons, we found in single exposure scenarios that 44 and 67 biomarkers were associated with at least one thyroid function measure in adolescents and adults, respectively (adjusted p-value <0.05). In scenarios involving mixed chemical exposures, groups such as pesticides, sodium/iodide symporter (NIS) inhibitors, and metals were associated with alterations in thyroid hormones or TSH across both age groups. Volatile organic compounds were specifically linked to lower T4 levels in adolescents, whereas phenols and parabens were associated with lower TSH levels exclusively in adults. Although limited by the cross-sectional data, this study identified chemical biomarkers linked to thyroid function.

Keywords: Adolescents; Adults; Environmental chemicals; ExWAS; NHANES; Thyroid function.

Fig. 1.

Volcano plots showing the associations between chemical biomarkers and thyroid function measurements, including free and total triiodothyronine (T3), free and total thyroxine (T4) and thyroid-stimulating hormone (TSH), based on the ExWAS approach applied to NHANES 2007–2012 data: (A) adolescents aged 12–19 years and (B) adults aged ≥20 years. The colored circles above the dashed line represent statistically significant chemical biomarkers for thyroid function measurements in single-exposure scenarios (adjusted p-value <0.05). Detailed values are provided in Table S3 of the Supplementary Material. Abbreviations: triiodothyronine (T3), thyroxine (T4), thyroid-stimulating hormone (TSH), smoking-related compounds (SRCs), metals and metalloids (Metals), perchlorate and other anions (NIS inhibitors), per- and polyfluoroalkyl substances (PFAS), pesticides, insecticides, and herbicides (Pesticides), phenols and parabens (PPs), phthalates and plasticizers’ metabolites (phthalates), polycyclic aromatic hydrocarbons (PAHs), and volatile organic compounds (VOCs)

Fig. 2.

Associations between thyroid function measurements and groups of chemical biomarkers by age group (adolescents aged 12–19 years and adults aged ≥20 years). The forest plots display linear regression coefficients (β) from the Bayesian group weighted quantile sum (BGWQS) model, showing the strength and direction of the relationships between thyroid function measurements and each chemical group. Each coefficient represents the impact of the specific chemical group on hormone levels, with corresponding confidence intervals. The shade area indicates statistical significance among chemical groups (p-value <0.05). Detailed values are provided in Table S4 of the Supplementary Material. Abbreviations: triiodothyronine (T3), thyroxine (T4), thyroid-stimulating hormone (TSH), smoking-related compounds (SRCs), metals and metalloids (Metals), perchlorate and other anions (NIS inhibitors), per- and polyfluoroalkyl substances (PFAS), pesticides, insecticides, and herbicides (Pesticides), phenols and parabens (PPs), phthalates and plasticizers metabolites (Phthalates), polycyclic aromatic hydrocarbons (PAHs), and volatile organic compounds (VOCs)

Fig. 3.

Circular bar plots of the relative importance or weight of chemical biomarkers within each chemical group from the Bayesian group weighted quantile sum (BGWQS) model. Each chemical group is represented by a distinct color, and the sum of the weights within each group equal to 1. If only one chemical biomarker from a specific group is included in the co-exposure scenario, its weight is set to 1. The density of the color reflects the type of thyroid function measurement, as indicated in the legend. For example, in the adolescents aged 12–19 years exposed to a combination of 29 chemical biomarkers, the weights of urinary lead (Pb, urine) concentrations among the eight metals are 0.048, 0.105, 0.079, 0.074, and 0.044 for free T3, total T3, free T4, total T4, and TSH, respectively. Detailed values are provided in Table S5 of the Supplementary Material. Abbreviations: triiodothyronine (T3), thyroxine (T4), thyroid-stimulating hormone (TSH), smoking-related compounds (SRCs), metals and metalloids (Metals), perchlorate and other anions (NIS inhibitors), per- and polyfluoroalkyl substances (PFAS), pesticides, insecticides, and herbicides (Pesticides), phenols and parabens (PPs), phthalates and plasticizers metabolites (Phthalates), polycyclic aromatic hydrocarbons (PAHs), and volatile organic compounds (VOCs), cotinine (COT), selenium (Se), mercury, total (Hg, blood total), cadmium (Cd, urine), sum of arsenic metabolites (sum arsenics), antimony (Sb), dimethylarsonic acid (DMA), uranium (U), tungsten (W), lead (Pb, urine), lead (Pb, blood), arsenic, total (As, total), barium (Ba), cobalt (Co), cesium (Cs), molybdenum (M), thiocyanate (SCN), perfluorohexane sulfonic acid (PFHxS), perfluorooctane sulfonic acid (PFOS), 3,5,6-trichloropyridinol (3,5,6-TCPy), dimethyl thiophosphate (DMTP), paranitrophenol (pNP), 2,5-dichlorophenol (2,5-DCP), 2,4-dichlorophenol (2,4-DCP), bisphenol A (BPA), propyl paraben (PrP), sum of metabolites of DEHP (DEHP), mono(carboxyoctyl) phthalate (MCOP), mono-(3-carboxypropyl) phthalate (MCPP), mono-ethyl phthalate (MEP), equol (Equol), enterodiol (ETD), enterolactone (ENL), 2-hydroxynapthalene (2-NAP), 3-hydroxyphenanthrene (3-PHE), 1-hydroxyphenanthrene (1-PHE), 2-hydroxyphenanthrene (2-PHE), 9-hydroxyfluorene (9-FLU), bromodichloromethane (BDCM), m-/p-xylene (m-/p-Xylene), 3-methylhippuric acid & 4-methylhippuric acid (3MHA and 4MHA), n-acetyl-s-(n-methylcarbamoyl)-l-cysteine (AMCC), n-acetyl-s-(n-propyl)-l-cysteine (BPMA), n-acetyl-s-2-carboxyethyl-l-cysteine (CEMA), n-acetyl-s-(2-cyanoethyl)-l-cysteine (CYMA), n-acetyl-s-(3,4-dihydroxybutyl)-l-cysteine (DHBMA), n-acetyl-s-(2-hydroxypropyl)-l-cysteine (2HPMA), n-acetyl-s-(3-hydroxypropyl)-l-cysteine (3HPMA), mandelic acid (MA), phenylglyoxylic acid (PGA), and n-acetyl-s-(3-hydroxypropyl-1-methyl)-l-cysteine (HPMMA)

Fig. 4.

Sensitivity analysis: volcano plots showing the associations between chemical biomarkers and thyroid function measurements, including free and total triiodothyronine (T3), free and total thyroxine (T4) and thyroid-stimulating hormone (TSH), based on the ExWAS approach applied to NHANES 2007–2012 data: (A) adolescents aged 12–19 years and (B) adults aged ≥20 years. This sensitivity ExWAS analysis was restricted to individuals with an abnormal range of thyroid function measurements. The colored circles above the dashed line represent statistically significant chemical biomarkers for thyroid function measurements in single-exposure scenarios (adjusted p-value <0.05). Chemical biomarkers with a detection rate of ≥70% are marked with an asterisk (*). Detailed values are provided in Table S6 and S7 of the Supplementary Material. Abbreviations: triiodothyronine (T3), thyroxine (T4), thyroid-stimulating hormone (TSH), smoking-related compounds (SRCs), metals and metalloids (Metals), perchlorate and other anions (NIS inhibitors), per- and polyfluoroalkyl substances (PFAS), pesticides, insecticides, and herbicides (Pesticides), phenols and parabens (PPs), phthalates and plasticizers’ metabolites (phthalates), polycyclic aromatic hydrocarbons (PAHs), volatile organic compounds (VOCs), thiocyanate (SCN), tungsten (W), n-acetyl-s-(n-propyl)-l-cysteine (BPMA), n-acetyl-s-(3-hydroxypropyl)-l-cysteine (3HPMA), perfluoroheptanoic acid (PFHpA), perfluoroundecanoic acid (PFUnDA), mercury (Hg), enterodiol (ETD), and enterolactone (ENL)