The Relationship Between Perfluoroalkyl Substances Concentrations and Thyroid Function in Early Childhood: A Prospective Cohort Study

Abstract

Background:

Exposure to perfluoroalkyl substances (PFAS) has been suggested to affect thyroid function; however, data on early-life exposure and thyroid function in early childhood are scarce. We investigated the cross-sectional and longitudinal relationships of early-life exposure to PFAS with thyroid function at 2, 4, and 6 years of age.

Methods:

This study used data on PFAS exposure and thyroid function from the Environment and Development of Children (EDC) cohort study. A total of 660 children who visited at least once at 2, 4, or 6 years of age (381 children aged 2 years, 569 children aged 4 years, and 511 children aged 6 years) were included in this study. Serum thyrotropin (TSH) levels were measured at 2, 4, and 6 years of age. The relationship of serum PFAS (sPFAS) concentrations with TSH levels at the three time points was assessed by repeated-measure analysis using linear mixed models. The serum levels of free thyroxine (fT4) and triiodothyronine (T3) were measured once (at 6 years of age). The relationship of sPFAS with fT4 and T3 levels at 6 years of age was investigated by linear regression analyses.

Results:

None complained of hyper- or hypothyroid symptoms with normal fT4 and T3 levels. Repeated-measure analysis showed that TSH levels at 2, 4, and 6 years of age were inversely associated with serum perfluorononanoic acid (sPFNA), after adjusting for age, sex, and/or dietary iodine intake (p < 0.05). When stratified by sex, TSH levels were inversely associated with serum perfluorooctanoic acid (sPFOA) in boys and sPFNA in girls (p < 0.05 for both). fT4 levels at 6 years of age were positively related to sPFNA and serum perfluorohexane sulfonic acid at 2 years of age and sPFOA at 6 years of age, and T3 levels at 6 years of age showed positive relationships with serum perfluorodecanoic acid and serum perfluorooctane sulfonic acid at 6 years of age (p < 0.05 for all). When stratified by sex, similar positive relationships for sPFAS with fT4 and T3 levels were significant among boys only.

Conclusions:

A significant relationship was found between early-life exposure to PFAS and thyroid function. Early-life exposure to PFAS was associated with decreased TSH and increased fT4 or T3 levels among preschool-age children.

Introduction

Perfluoroalkyl substances (PFAS) are a group of highly stable synthetic chemicals consisting of a fully fluorinated alkyl chain with a terminal functional group. PFAS have been widely used to make products resistant to stains, grease, and water, including in food packaging, cookware, carpets, clothing fabrics, and fire extinguishers (1). Human exposure to PFAS occurs through the ingestion of contaminated water or food, the absorption of the compounds by the skin or mucous membranes, and/or the inhalation of indoor dust or outdoor air (2). PFAS are ubiquitous and not easily degraded, so they stably persist in the environment. In humans, the half-lives of serum perfluorooctane sulfonic acid (sPFOS), perfluorooctanoic acid (sPFOA), and perfluorohexane sulfonic acid (sPFHxS) were reported to be 5.4, 3.8, and 8.5 years, respectively (3). Therefore, PFAS can accumulate in the human body for several years (3,4) and can cause long-term health problems.

PFAS can disrupt thyroid hormone homeostasis in humans (5,6). A recent meta-analysis reported a negative relationship between blood PFAS (PFOS, PFOA, and PFHxS) concentrations and total thyroxine (T4) levels in the adult population without performing a sex-specific analysis (5). Meanwhile, age-specific (7) and sex-specific (8,9) relationships between PFAS concentrations and thyroid hormone levels have been reported. According to the 2007–2010 data from the United States National Health and Nutrition Examination Survey (US NHANES), thyroid hormone levels were positively associated with serum PFAS (sPFAS) concentrations; total triiodothyronine (T3) with sPFOA and sPFHxS and total T4 with sPFHxS in females, whereas free thyroxine (fT4) was inversely related to sPFHxS in males (8). In residents exposed to PFOA for >50 years from a mid-Ohio River Valley chemical plant, high exposure to PFOA was related to hyperthyroidism and hypothyroidism in females but hypothyroidism in males (9).

Several studies have been conducted on the effect of maternal exposure to PFAS during pregnancy on thyroid function in cord blood or neonates (6,10). However, only three cross-sectional studies have reported the relationship between postnatal exposure to PFAS and thyroid function in the pediatric age group (7,11,12), and only one study included prepubertal children (12). Since early childhood is a period of rapid development and vulnerable to endocrine disruptors (13), a prospective study to evaluate the association between early-life exposure to PFAS and subsequent thyroid function is warranted. In the present study, we aimed to evaluate longitudinal changes in sPFAS at 2, 4, and 6 years of age in a prospective Korean pediatric cohort. Then, we investigated the cross-sectional and longitudinal relationships of sPFAS at 2, 4, and 6 years of age with the current and subsequent thyroid function in early childhood.

Methods

Subjects

This study used data on PFAS exposure and thyroid function from the Environment and Development of Children (EDC) cohort study. The EDC cohort was followed up at two-year intervals to prospectively investigate the influence of early-life environmental exposures on physical and neurobehavioral development as described previously (14). A total of 726 children, consisting of 2-year-old (n = 425) and 4-year-old (n = 301) children, were initially enrolled during 2012–2015. Among the 425 children initially enrolled at 2 years of age, 343 were followed up at 4 years of age (follow-up rate: 80.7%) and 305 were followed up at 6 years of age (follow-up rate from the first visit: 71.8%). Among the 301 children initially enrolled at 4 years, 252 were followed up at 6 years of age (follow-up rate: 83.7%). After excluding multiple births (n = 63) and participants without data on sPFAS concentrations or thyroid function tests (n = 25), a total of 660 children who visited at least once at 2, 4, or 6 years of age (381 children aged 2 years, 569 children aged 4 years, and 511 children aged 6 years) were included in the present analysis (Supplementary Fig. S1). Although the gestational age (p = 0.018) and proportion of nulliparity (p = 0.015) were lower among the 660 subjects included than among the 25 singleton subjects who were not included in this study, no significant differences in age at delivery, mode of delivery, maternal education level, smoking during pregnancy, and offspring's sex and birth weight were found between the two groups (Supplementary Table S1). Written informed consent was obtained from the parents of the participants before enrollment. The study protocol was approved by the institutional review board (IRB) of the SNUH (IRB No. 1201-010-392).

Anthropometric measurements and questionnaires

Height (cm) was measured using a Harpenden Stadiometer (Holtain Ltd., Crymych, Wales, United Kingdom), and weight (kg) was measured using a digital scale (150 A; Cas Co. Ltd., Seoul, Korea). Body mass index (BMI) was calculated as weight (kg) divided by height squared (m²). The Z-scores for height, weight, and BMI were assigned on the basis of the 2007 Korean National Growth Charts (15). An extensive questionnaire including demographics, medical history, socioeconomic status, lifestyle, and environmental exposure was completed by the parents at every visit. Information on the dietary patterns was also gathered among the participants when they were 4 and 6 years old. Dietary iodine intake was estimated by the Computer-Aided Nutritional Analysis Program 4.0 for Professionals using dietary information collected from the participating children's mothers using a semiquantified food frequency questionnaire (Korean Society of Nutrition, Seoul, Republic of Korea).

Measurement of thyroid function

Serum thyrotropin (TSH) levels were measured at 2, 4, and 6 years of age. The serum levels of fT4 and T3 were measured in all children aged 6 years and in selected children with abnormal TSH levels at 2 and 4 years of age. A chemiluminescent microparticle immune assay was used for each measurement and was measured on an ARCHITECT i2000 System (Abbott Korea, Seoul, Korea). The normal reference ranges of fT4, T3, and TSH levels were 0.70–1.48 ng/dL (9.01–19.05 pmol/L), 58–159 ng/dL (0.89–2.44 nmol/L), and 0.38–4.94 μIU/mL, respectively. Subclinical hyperthyroidism was defined as TSH <0.38 μIU/mL with normal fT4 levels, and subclinical hypothyroidism was defined as TSH >4.94 μIU/mL with normal fT4 levels. Clinical thyroid disease was defined by hyper- or hypothyroid symptoms with abnormal levels of both TSH and thyroid hormones (fT4 and T3).

Measurement of sPFAS concentrations

Blood serum samples were collected in the morning after an overnight fast and stored at −70°C until analysis. We measured 14 sPFAS (in ng/mL), which included perfluoropentanoic acid (PFPeA), perfluorohexanoic acid (PFHxA), perfluoroheptanoic acid (PFHpA), PFOA, perfluorononanoic acid (PFNA), perfluorodecanoic acid (PFDA), perfluoroundecanoic acid (PFUnDA), perfluorododecanoic acid (PFDoDA), perfluorotridecanoic acid (PFTrDA), perfluorotetradecanoic acid (PFTeDA), perfluorobutane sulfonate (PFBS), PFHxS, PFOS, and perfluorodecane sulfonate (PFDS). Fourteen native PFAS compounds and isotope-labeled internal standards (¹³C₄-PFHxA, ¹³C₄-PFOA, ¹³C₄-PFNA, ¹³C₄-PFDA, ¹³C₄-PFUnDA, and ¹³C₄-PFDoDA, ¹³C₄-PFHxS, and ¹³C₄-PFOS) were purchased from Wellington Laboratories (Guelph, ON, Canada), and ammonium acetate, ammonium hydroxide, and formic acid were purchased from Sigma–Aldrich (St. Louis, MO, USA). Acetonitrile, methanol, and water (Burdick & Jackson, Muskegon, MI, USA) were used for sample preparation and liquid chromatography tandem-mass spectrometry analysis.

Blood serum samples were prepared according to previously reported methods (16), with minor modifications. Briefly, 20 μL of labeled internal standards was added to 200 μL serum and 500 μL of 0.1% formic acid diluted in water. Next, solid-phase extraction (SPE) was performed using oasis weak anion-exchange cartridges (WAX; 1 cm³, 30 mg; Waters, Milford, MA, USA). The resulting extract of PFAS was eluted using 4 mL of 0.1% ammonium hydroxide in methanol, evaporated to dry, and reconstituted in 0.2 mL acetonitrile. Quality control samples and calibration standards were subjected to SPE using bovine serum samples that had been previously spiked with the standard mixture and internal standards.

PFAS were analyzed using a high-performance liquid chromatography Series 1100 (Agilent Technologies, Palo Alto, CA, USA) with a 2.0 × 150 mm, 3 m YMC C18 column (Waters). The injection volume was 3 μL, and the flow rate was 200 μL/min using a gradient mode, with 70% mobile phase A (5 mM ammonium acetate with 0.02% formic acid in water) and 30% B (methanol) to 100% B within 10 minutes, which was maintained for 7 minutes. The identification and quantification of analytes were accomplished using an API 4000 triple quadrupole mass spectrometer (Applied Biosystems, Foster City, CA, USA) as detailed previously (16).

The limit of detection (LOD) for each PFAS was as follows: 0.076 ng/mL for PFPeA, 0.180 ng/mL for PFHxA, 0.157 ng/mL for PFHpA, 0.078 ng/mL for PFOA, 0.050 ng/mL for PFNA, 0.059 ng/mL for PFDA, 0.078 ng/mL for PFUnDA, 0.052 ng/mL for PFDoDA, 0.146 ng/mL for PFTrDA, 0.095 ng/mL for PFTeDA, 0.227 ng/mL for PFBS, 0.160 ng/mL for PFHxS, 0.113 ng/mL for PFOS, and 0.104 ng/mL for PFDS. For concentrations below the LOD, we used a value equal to the LOD divided by the square root of 2 (8).

Assumed causal pathway and covariate selection

We used a directed acyclic graph (DAG) (Supplementary Fig. S2) to identify potential confounding variables for the causal relationship of PFAS exposure and thyroid function, based on previous studies (6,12,17). Age and sex, which can affect exposure and/or outcome, were included in model 1. Dietary iodine intake was additionally adjusted in model 2 since excessive iodine status was related to subclinical hypothyroidism in children (18). However, BMI was not included as covariate in the model. As shown in the DAG (Supplementary Fig. S2), both PFAS concentrations (exposure) (19,20) and thyroid function (outcome) can affect BMI (21), and BMI can be considered as a collider rather than a confounder; adjustment for BMI could therefore bias the models (22). We constructed model 1 (age- and sex-adjusted) and model 2 (age-, sex-, and iodine intake-adjusted) as the final models.

Statistical analysis

Characteristics of the study participants were analyzed using Student's t-test for continuous variables and the chi-square test for categorical variables. The levels of PFAS, fT4, T3, and TSH and the dietary iodine intake were natural log-transformed because of their skewed distribution. PFAS that were detected in more than 90% of the samples were used for statistical analysis. The association of sPFAS and subclinical hypothyroidism at 6 years of age was investigated by binary logistic regression analysis. For TSH levels measured at multiple time points (at 2, 4, and 6 years of age), the relationship between sPFAS and TSH levels was analyzed by repeated-measure analysis using linear mixed models with a random effect for each participant and adjusted for covariates. For fT4 and T3 levels measured once (at 6 years of age), the relationships of sPFAS with fT4 and T3 levels were investigated by linear regression analyses after adjusting for covariates. We constructed multivariate-adjusted model 1 (age- and sex-adjusted) and model 2 (age-, sex-, and dietary iodine intake-adjusted) for the relationships of sPFAS with subclinical hypothyroidism, TSH, fT4, and T3 levels.

For a sensitivity analysis, we performed subgroup analyses to evaluate the relationship between sPFAS and thyroid function in 269 children who were studied at all 3 time points (at 2, 4, and 6 years of age), with the same covariates used in the main analyses. The interaction of sex with the effect of sPFAS on thyroid function was evaluated using product terms of sex and each PFAS concentration. Since we found heterogeneity among the associations between some sPFAS and thyroid function by sex with p-values for interaction <0.10, we also performed sex-stratified analyses for the associations between each PFAS concentration and subclinical hypothyroidism or thyroid function. SAS version 9.4 (SAS Institute, Inc., Cary, NC, USA) and SPSS for Windows (version 19.0; SPSS, Inc., Chicago, IL, USA) were used, and p-values <0.05 were considered statistically significant.

Results

Characteristics of the subjects

Table 1 shows the baseline characteristics of the 660 participants. The mean age was 23.3 ± 0.8 months at the 2-year-old visit (200 boys and 181 girls), 47.3 ± 1.8 months at the 4-year-old visit (299 boys and 270 girls), and 71.1 ± 1.6 months at the 6-year-old visit (268 boys and 243 girls). The mean gestational age at delivery was 38.8 ± 1.4 weeks, and the mean birth weight was 3.3 ± 0.4 kg. More than 80% of the parents were educated to the college level or above. The median dietary iodine intake at 4 and 6 years of age was 204.2 (137.4–320.8) and 221.8 (156.7–312.4) μg/day, respectively. The median (interquartile range) TSH levels at 2, 4, and 6 years of age were 2.2 (1.6–3.1), 2.2 (1.6–3.0), and 2.2 (1.6–3.0) μIU/mL, respectively. The median fT4 and T3 levels at 6 years of age were 1.1 (1.1–1.2) and 147.3 (135.4–160.0) ng/dL, respectively.

Table 1.

Characteristics of the Study Participants

Participants	Variables	Total	Boys	Girls
Mothers	No. of participants	660	348	312
	Age (years) at delivery	31.3 ± 3.6	31.5 ± 3.8	31.0 ± 3.5
	Gestational age (weeks)	38.8 ± 1.4	38.8 ± 1.4	38.9 ± 1.5
	Mode of delivery (vaginal)	440 (66.7)	221 (63.7)	219 (70.2)
	Birth weight (offspring) (kg)	3.3 ± 0.4	3.3 ± 0.4	3.2 ± 0.4^*
	Nulliparous	373 (56.5)	202 (58.0)	171 (54.8)
	Education > high school	542 (82.1)	287 (82.7)	255 (81.7)
	Smoking during pregnancy (never-/ever-/smoking)	352/254/39 (53.3/38.5/5.9)	175/142/20 (51.9/42.1/5.9)	177/112/19 (57.5/36.4/6.2)
Children aged 2 years	No. of children	381	200	181
	Age (months)	23.3 ± 0.8	23.3 ± 0.8	23.3 ± 0.7
	Height (cm)	86.3 ± 3.0	87.0 ± 3.0	85.6 ± 2.8^*
	Weight (kg)	12.3 ± 1.4	12.7 ± 1.4	11.9 ± 1.3^*
	BMI (kg/m²)	16.5 ± 1.4	16.8 ± 1.5	16.2 ± 1.3^*
	Height Z-score	0.0 ± 0.8	0.1 ± 0.8	−0.0 ± 0.8
	Weight Z-score	0.0 ± 1.0	0.1 ± 1.0	−0.0 ± 0.9
	BMI Z-score	−0.2 ± 0.9	−0.1 ± 1.0	−0.2 ± 0.9
	TSH (μIU/mL)	2.2 (1.6–3.1)	2.2 (1.6–3.3)	2.2 (1.6–3.1)
Children aged 4 years	No. of children	569	299	270
	Age (months)	47.3 ± 1.8	47.2 ± 1.7	47.5 ± 1.9^*
	Height (cm)	102.0 ± 3.8	102.5 ± 3.8	101.6 ± 3.7^*
	Weight (kg)	16.3 ± 1.9	16.5 ± 1.8	16.1 ± 1.9^*
	BMI (kg/m²)	15.6 ± 1.3	15.7 ± 1.2	15.6 ± 1.3
	Height Z-score	0.2 ± 0.8	0.2 ± 0.9	0.2 ± 0.8
	Weight Z-score	0.1 ± 1.0	0.02 ± 0.94	0.1 ± 1.0
	BMI Z-score	−0.1 ± 1.0	−0.2 ± 1.0	−0.1 ± 1.0
	Iodine intake (μg/day)	204.2 (137.4–320.8)	207.0 (134.0–341.2)	196.1 (139.9–299.7)
	TSH (μIU/mL)	2.2 (1.6–3.0)	2.2 (1.6–3.1)	2.2 (1.6–3.0)
Children aged 6 years	No. of children	511	268	243
	Age (months)	71.1 ± 1.6	71.1 ± 1.6	71.2 ± 1.5
	Height (cm)	115.7 ± 4.4	116.3 ± 4.6	115.1 ± 4.1^*
	Weight (kg)	21.2 ± 3.3	21.4 ± 3.2	21.1 ± 3.3
	BMI (kg/m²)	15.8 ± 1.8	15.8 ± 1.7	15.8 ± 1.9
	Height Z-score	0.3 ± 1.0	0.3 ± 1.0	0.3 ± 0.9
	Weight Z-score	0.1 ± 1.0	0.02 ± 1.0	0.2 ± 1.0^*
	BMI Z-score	−0.1 ± 1.0	−0.2 ± 1.0	0.01 ± 1.1^*
	Iodine intake (μg/day)	221.8 (156.7–312.4)	218.9 (161.2–305.6)	221.8 (147.8–322.2)
	TSH (μIU/mL)	2.2 (1.6–3.0)	2.2 (1.5–3.0)	2.2 (1.7–3.0)^*
	fT4 (ng/dL)	1.1 (1.1–1.2)	1.2 (1.1–1.2)	1.1 (1.1–1.2)
	T3 (ng/dL)	147.3 (135.4–160.0)	147.7 (136.1–159.6)	146.4 (134.9–160.9)
	Subclinical hypothyroidism^a	24 (4.7)	13 (4.9)	11 (4.5)
	Subclinical hyperthyroidism^b	1 (0.2)	1 (0.4)	0 (0)

Data are expressed as mean ± SD, median (25th percentile–75th percentile), or number (%).

Defined as TSH >4.94 μIU/mL and fT4 was within the reference range.

Defined as TSH <0.38 μIU/mL and fT4 was within the reference range.

p < 0.05 between boys and girls.

BMI, body mass index; fT4, free thyroxine; SD, standard deviation; T3, triiodothyronine; TSH, thyrotropin.

PFAS exposure in children aged 2, 4, and 6 years

Table 2 shows the distribution of sPFAS (ng/mL) at 2, 4, and 6 years of age. Among the 14 PFAS analyzed, sPFOS, sPFOA, sPFHxS, serum PFDA (sPFDA), and serum PFNA (sPFNA) had a detection frequency of more than 90% at all studied ages. In particular, sPFOS, sPFOA, sPFHxS, and sPFNA were detected in >95% of the children at all studied ages (Table 2). No significant differences were found in the concentration of each PFAS at 2, 4, and 6 years of age by sex (data not shown), except for the sPFOS levels at 6 years of age (4.399 ± 1.761 among boys vs. 4.015 ± 1.603 among girls; p = 0.038).

Table 2.

Summary of Serum Perfluoroalkyl Substances Concentrations (ng/mL) in Children Aged 2, 4, and 6 Years

	Age (years)	n	LOD (ng/mL)	Detection frequency (%)	Mean	SD	25th	50th	75th	Range (min, max)
PFPeA	2	381	0.076	126 (33.1)	0.206	0.476	0.050	0.050	0.160	0.050, 7.780
	4	569	0.076	202 (35.5)	0.165	0.230	0.054	0.054	0.171	0.050, 1.590
	6	511	0.076	74 (13.2)	0.069	0.064	0.054	0.054	0.054	0.054, 0.960
PFHxA	2	381	0.179	88 (23.1)	0.206	0.291	0.127	0.127	0.127	0.127, 5.250
	4	569	0.179	7 (1.2)	0.128	0.015	0.127	0.127	0.127	0.010, 0.271
	6	511	0.179	6 (1.1)	0.130	0.039	0.127	0.127	0.127	0.127, 0.920
PFHpA	2	381	0.157	282 (74.0)	0.456	0.476	0.111	0.331	0.594	0.110, 5.800
	4	569	0.157	227 (39.9)	0.174	0.110	0.111	0.111	0.220	0.032, 0.948
	6	511	0.157	96 (17.1)	0.135	0.083	0.111	0.111	0.114	0.017, 1.540
PFOA	2	381	0.078	381 (100)	4.984	3.016	2.730	4.390	6.350	0.820, 21.200
	4	569	0.078	567 (99.6)	4.024	1.879	2.867	3.650	4.930	0.055, 21.700
	6	511	0.078	562 (100)	4.132	1.683	2.960	3.830	4.740	0.846, 12.600
PFNA	2	381	0.050	381 (100)	1.996	2.433	0.850	1.410	2.180	0.150, 22.000
	4	569	0.050	568 (99.8)	1.214	0.890	0.728	0.992	1.380	0.040, 7.560
	6	511	0.050	542 (96.4)	0.811	0.673	0.417	0.695	1.028	0.030, 9.220
PFDA	2	381	0.058	354 (92.9)	0.397	0.238	0.200	0.370	0.560	0.040, 1.250
	4	569	0.058	559 (98.2)	0.362	0.168	0.257	0.337	0.447	0.029, 1.270
	6	511	0.058	559 (99.5)	0.372	0.480	0.245	0.324	0.431	0.040, 10.400
PFUnDA	2	381	0.077	341 (89.5)	0.461	0.324	0.225	0.400	0.647	0.050, 1.800
	4	569	0.077	564 (99.1)	0.498	0.242	0.325	0.458	0.629	0.055, 1.660
	6	511	0.077	529 (94.1)	0.461	0.489	0.240	0.394	0.582	0.030, 8.380
PFDoDA	2	381	0.052	186 (48.8)	0.120	0.134	0.037	0.037	0.156	0.037, 0.811
	4	569	0.052	306 (53.8)	0.068	0.038	0.037	0.056	0.087	0.037, 0.245
	6	511	0.052	214 (38.1)	0.060	0.062	0.037	0.037	0.068	0.008, 1.120
PFTrDA	2	381	0.146	70 (18.4)	0.128	0.073	0.100	0.100	0.100	0.100, 0.600
	4	569	0.146	136 (23.9)	0.128	0.060	0.103	0.103	0.121	0.046, 0.622
	6	511	0.146	103 (18.3)	0.125	0.059	0.103	0.103	0.103	0.017, 0.681
PFTeDA	2	381	0.095	0 (0.0)	0.070	0.001	0.070	0.070	0.070	0.067, 0.070
	4	569	0.095	5 (0.9)	0.068	0.005	0.067	0.067	0.067	0.067, 0.135
	6	511	0.095	14 (2.5)	0.071	0.037	0.067	0.067	0.067	0.067, 0.804
PFBS	2	381	0.227	193 (50.7)	0.374	0.349	0.160	0.240	0.515	0.160, 4.610
	4	569	0.227	25 (4.4)	0.167	0.033	0.160	0.160	0.160	0.160, 0.528
	6	511	0.227	15 (2.7)	0.176	0.152	0.160	0.160	0.160	0.160, 2.790
PFHxS	2	381	0.160	379 (99.5)	1.391	0.885	0.750	1.190	1.810	0.110, 5.980
	4	569	0.160	569 (100)	1.130	0.545	0.780	0.103	1.350	0.382, 8.070
	6	511	0.160	561 (99.8)	1.371	0.588	0.988	1.270	1.616	0.110, 6.600
PFOS	2	381	0.113	381 (100)	5.256	3.388	2.720	4.530	6.795	0.490, 22.800
	4	569	0.113	569 (100)	4.432	2.050	2.970	4.050	5.510	0.397, 16.900
	6	511	0.113	562 (100)	4.216	1.761	2.910	3.980	5.130	1.070, 13.000
PFDS	2	380	0.104	30 (7.9)	0.078	0.030	0.070	0.070	0.070	0.070, 0.340
	4	569	0.104	8 (1.4)	0.074	0.010	0.073	0.073	0.073	0.070, 0.260
	6	511	0.104	1 (0.2)	0.076	0.061	0.073	0.073	0.073	0.073, 1.460

LOD, limit of detection; PFBS, perfluorobutane sulfonate; PFDA, perfluorodecanoic acid; PFDoDA, perfluorododecanoic acid; PFDS, perfluorodecane sulfonate; PFHpA, perfluoroheptanoic acid; PFHxA, perfluorohexanoic acid; PFHxS, perfluorohexane sulfonate; PFNA, perfluorononanoic acid; PFOA, perfluorooctanoic acid; PFOS, perfluorooctane sulfonate; PFPeA, perfluoropentanoic acid; PFTeDA, perfluorotetradecanoic acid; PFTrDA, perfluorotridecanoic acid; PFUnDA, perfluoroundecanoic acid.

Association of sPFAS with subclinical thyroid disease

None of the participants complained of hyper- or hypothyroid symptoms associated with clinical thyroid disease. At 2 and 4 years of age, serum fT4 and T3 levels were selectively evaluated in cases with abnormal TSH levels. fT4 and T3 levels were within the normal reference range, although TSH levels were elevated in 19 (5.0%) and 27 (4.7%) cases at 2 and 4 years of age, respectively, and suppressed in 1 (0.3%) case at 2 years of age. At 6 years of age, all children had normal fT4 and T3 levels. Subclinical hyperthyroidism was defined in 1 boy (0.2%), and subclinical hypothyroidism was present in 24 (4.7%) children (Table 1). The associations of sPFAS with subclinical hypothyroidism at 6 years of age are presented in Supplementary Table S2. The age- and sex-adjusted odds ratio [95% confidence interval] of subclinical hypothyroidism was 0.34 [0.13–0.85] for sPFHxS and 0.37 [0.14–0.97] for sPFOS, which remained significant after further adjustment for dietary iodine intake (both p < 0.05).

Repeated-measure analyses for the relationship of sPFAS with TSH levels at 2, 4, and 6 years of age

Table 3 shows the relationship of sPFAS with TSH levels at 2, 4, and 6 years of age using linear mixed models. Overall, TSH levels were inversely associated with sPFAS, which was statistically significant for sPFNA after adjusting for age and sex (β = −0.053, p = 0.014 in model 1). The inverse association between sPFNA and TSH levels remained significant after additionally adjusting for iodine intake (β = −0.057, p = 0.028 in model 2). When stratified by sex, TSH levels were inversely related to sPFOA in boys and sPFNA in girls (p < 0.05 for boys and girls in model 1). This inverse relationship of sPFNA with TSH levels remained significant after additional adjustment for iodine intake in girls (p = 0.026 in model 2, Table 3). In the sensitivity analysis of the group of 269 children who were studied at all 3 time points (at 2, 4, and 6 years of age), the inverse association of TSH levels with sPFAS was similar to the results for the 660 children (Supplementary Table S3).

Table 3.

Repeated-Measure Analyses of Serum Perfluoroalkyl Substances Concentrations with Thyrotropin Levels at 2, 4, and 6 Years of Age (n = 660)

	PFAS	Model 1				Model 2
		All	Boys	Girls	p (sex difference)	All	Boys	Girls	p (sex difference)
		β (SE)	β (SE)	β (SE)	p (sex difference)	β (SE)	β (SE)	β (SE)	p (sex difference)
TSH	PFOA	−0.031 (0.041)	−0.069 (0.034)^*	−0.032 (0.041)	0.530	−0.029 (0.057)	−0.052 (0.042)	−0.034 (0.055)	0.783
	PFNA	−0.053 (0.021)^*	−0.010 (0.020)	−0.054 (0.022)^*	0.150	−0.057 (0.026)^*	−0.002 (0.024)	−0.057 (0.026)^*	0.132
	PFDA	−0.044 (0.029)	−0.041 (0.026)	−0.048 (0.029)	0.969	−0.010 (0.037)	−0.030 (0.039)	−0.013 (0.036)	0.739
	PFHxS	−0.018 (0.039)	−0.055 (0.035)	−0.019 (0.039)	0.509	0.020 (0.049)	−0.037 (0.044)	0.015 (0.048)	0.414
	PFOS	−0.015 (0.037)	−0.039 (0.033)	−0.018 (0.037)	0.568	−0.060 (0.050)	−0.002 (0.043)	−0.062 (0.049)	0.414

Model 1 adjusted for age and sex; model 2 adjusted for age, sex, and dietary iodine intake.

p < 0.05.

PFAS, perfluoroalkyl substances; SE, standard error; TSH, thyrotropin.

Relationship of sPFAS with fT4 and T3 levels at 6 years of age

Table 4 shows the relationships of sPFAS (at 2, 4, and 6 years of age) with fT4 and T3 levels (at 6 years of age) using linear regression analysis. fT4 levels were positively associated with sPFNA and sPFHxS at 2 years of age and sPFOA at 6 years of age after adjusting for age and sex (p < 0.05 for all in model 1). These positive relationships of sPFNA at 2 years of age and sPFOA at 6 years of age with fT4 levels remained significant after additional adjustment for iodine intake (p < 0.05 for both in model 2). T3 levels were positively related to sPFDA and sPFOS at 6 years of age after adjusting for age and sex (p < 0.05 for both in model 1), which remained significant after additional adjustment for iodine intake (p < 0.05 for both in model 2). When stratified by sex, similar positive associations for sPFAS with fT4 and T3 levels were observed, which were statistically significant among boys only. In boys, sPFNA at 2 years of age and sPFOA at 6 years of age were positively associated with fT4 levels (p < 0.05 for all in models 1 and 2), and sPFNA at 4 years of age and sPFOS at 6 years of age were positively associated with T3 levels (p < 0.05 for all in models 1 and 2, Table 4). In the sensitivity analysis of the group of 269 children who were studied at all 3 time points (at 2, 4, and 6 years of age), the positive associations of fT4 levels with sPFAS were consistent with the results of the 660 children (Supplementary Table S4).

Table 4.

Relationships of Serum Perfluoroalkyl Substances Concentrations with Free Thyroxine or Triiodothyronine Levels at 6 Years of Age (n = 660)

	PFAS	Model 1				Model 2
		All	Boys	Girls	p (sex difference)	All	Boys	Girls	p (sex difference)
		β (SE)	β (SE)	β (SE)	p (sex difference)	β (SE)	β (SE)	β (SE)	p (sex difference)
fT4 (6 years)	PFOA (2 years)	0.010 (0.009)	0.006 (0.013)	0.012 (0.014)	0.598	0.010 (0.009)	0.006 (0.013)	0.012 (0.014)	0.573
	PFNA (2 years)	0.019 (0.007)^**	0.026 (0.009)^**	0.008 (0.010)	0.290	0.019 (0.007)^**	0.026 (0.010)^**	0.007 (0.011)	0.267
	PFDA (2 years)	0.013 (0.007)	0.016 (0.009)	0.008 (0.011)	0.648	0.013 (0.007)	0.016 (0.009)	0.008 (0.011)	0.697
	PFHxS (2 years)	0.018 (0.009)^*	0.022 (0.013)	0.014 (0.012)	0.790	0.018 (0.009)	0.022 (0.013)	0.013 (0.012)	0.780
	PFOS (2 years)	0.015 (0.009)	0.019 (0.012)	0.010 (0.012)	0.838	0.015 (0.009)	0.018 (0.012)	0.011 (0.012)	0.904
	PFOA (4 years)	0.000 (0.009)	0.001 (0.012)	−0.001 (0.016)	0.885	0.000 (0.009)	0.000 (0.012)	−0.002 (0.016)	0.917
	PFNA (4 years)	0.003 (0.007)	0.009 (0.010)	−0.005 (0.011)	0.397	0.003 (0.007)	0.007 (0.010)	−0.005 (0.011)	0.437
	PFDA (4 years)	0.000 (0.008)	−0.001 (0.011)	0.002 (0.011)	0.827	0.001 (0.008)	−0.002 (0.011)	0.002 (0.011)	0.747
	PFHxS (4 years)	0.002 (0.011)	0.021 (0.015)	−0.018 (0.016)	0.086	0.003 (0.011)	0.020 (0.015)	−0.020 (0.016)	0.100
	PFOS (4 years)	0.003 (0.009)	0.011 (0.011)	−0.008 (0.014)	0.336	0.003 (0.009)	0.009 (0.011)	−0.008 (0.014)	0.381
	PFOA (6 years)	0.036 (0.015)^*	0.034 (0.015)^*	0.004 (0.016)	0.130	0.037 (0.015)^*	0.035 (0.015)^*	0.003 (0.016)	0.112
	PFNA (6 years)	0.006 (0.006)	0.006 (0.006)	0.006 (0.007)	0.895	0.006 (0.007)	0.006 (0.007)	0.005 (0.008)	0.973
	PFDA (6 years)	0.004 (0.012)	0.003 (0.012)	0.011 (0.012)	0.685	0.003 (0.012)	0.002 (0.012)	0.011 (0.012)	0.649
	PFHxS (6 years)	0.018 (0.014)	0.016 (0.014)	0.009 (0.015)	0.650	0.018 (0.014)	0.016 (0.014)	0.007 (0.015)	0.621
	PFOS (6 years)	0.016 (0.013)	0.014 (0.013)	−0.002 (0.015)	0.335	0.016 (0.013)	0.014 (0.013)	−0.003 (0.015)	0.306
T3 (6 years)	PFOA (2 years)	0.010 (0.013)	0.006 (0.019)	0.016 (0.018)	0.782	0.009 (0.013)	0.007 (0.018)	0.017 (0.018)	0.836
	PFNA (2 years)	0.014 (0.010)	0.024 (0.014)	0.003 (0.014)	0.233	0.014 (0.010)	0.027 (0.014)	0.003 (0.014)	0.169
	PFDA (2 years)	0.005 (0.010)	0.010 (0.013)	−0.002 (0.014)	0.492	0.004 (0.010)	0.010 (0.013)	−0.002 (0.014)	0.473
	PFHxS (2 years)	0.004 (0.012)	0.014 (0.019)	−0.005 (0.016)	0.376	0.003 (0.012)	0.014 (0.019)	−0.006 (0.016)	0.331
	PFOS (2 years)	0.009 (0.012)	0.013 (0.018)	0.004 (0.016)	0.592	0.009 (0.012)	0.015 (0.018)	0.004 (0.016)	0.551
	PFOA (4 years)	0.019 (0.013)	0.019 (0.017)	0.019 (0.020)	0.981	0.019 (0.013)	0.019 (0.017)	0.018 (0.020)	0.993
	PFNA (4 years)	0.011 (0.010)	0.028 (0.014)^*	−0.010 (0.014)	0.053	0.011 (0.010)	0.029 (0.014)^*	−0.010 (0.014)	0.048
	PFDA (4 years)	0.003 (0.011)	0.022 (0.015)	−0.016 (0.014)	0.076	0.003 (0.011)	0.022 (0.016)	−0.017 (0.015)	0.069
	PFHxS (4 years)	0.025 (0.015)	0.032 (0.022)	0.019 (0.020)	0.620	0.025 (0.015)	0.032 (0.022)	0.018 (0.021)	0.605
	PFOS (4 years)	0.013 (0.012)	0.018 (0.016)	0.007 (0.018)	0.629	0.013 (0.012)	0.002 (0.013)	0.004 (0.014)	0.609
	PFOA (6 years)	0.025 (0.020)	0.026 (0.021)	−0.004 (0.020)	0.324	0.027 (0.020)	0.028 (0.021)	−0.005 (0.021)	0.292
	PFNA (6 years)	0.012 (0.009)	0.012 (0.009)	−0.008 (0.009)	0.124	0.014 (0.009)	0.014 (0.009)	−0.009 (0.010)	0.081
	PFDA (6 years)	0.032 (0.016)^*	0.033 (0.017)	−0.018 (0.015)	0.025	0.033 (0.016)^*	0.033 (0.017)	−0.019 (0.015)	0.023
	PFHxS (6 years)	0.005 (0.019)	0.006 (0.020)	0.012 (0.019)	0.804	0.004 (0.019)	0.005 (0.020)	0.011 (0.019)	0.795
	PFOS (6 years)	0.037 (0.017)^*	0.039 (0.018)^*	0.004 (0.019)	0.218	0.038 (0.017)^*	0.040 (0.018)^*	0.003 (0.019)	0.206

Model 1 adjusted for age and sex; model 2 adjusted for age, sex, and dietary iodine intake.

p < 0.05, ^** p < 0.01.

Discussion

When children were serially followed up at 2, 4, and 6 years of age, sPFOS, sPFOA, sPFHxS, sPFDA, and sPFNA were detected in >90% of the children at all studied ages. None of the children complained of hyper- or hypothyroid symptoms with normal fT4 and T3 levels. Repeated-measure analysis showed that TSH levels at 2, 4, and 6 years of age were inversely associated with sPFNA, after adjusting for age, sex, and dietary iodine intake. After adjusting for covariates, fT4 levels at 6 years of age were positively related to sPFNA and sPFHxS at 2 years of age and sPFOA at 6 years of age, and T3 levels at 6 years of age showed a positive relationship with sPFDA and sPFOS at 6 years of age.

During the follow-up period from 2 years of age (in 2012/2013), 4 years of age (in 2013/2015), and 6 years of age (in 2015/2017), the serum concentrations of each PFAS had not changed. When the mean values of three PFAS measurements at 2, 4, and 6 years of age were arranged from the highest to lowest, the order was sPFOS (4.2 ng/mL), sPFOA (4.0 ng/mL), sPFNA (1.0 ng/mL), sPFHxS (0.9 ng/mL), and sPFDA (0.3 ng/mL). According to the 2011–2012 US NHANES, sPFOS (4.6 ng/mL), sPFNA (0.8 ng/mL), and sPFHxS (1.3 ng/mL) concentrations in US 12- to 19-year-old adolescents were similar to those of our preschool children, with the exception of lower sPFOA levels (1.9 ng/mL) among US adolescents (7). Additionally, the serum concentration of each PFAS in preschool children residing in developed countries was comparable to that of our children (23 –25). When we compared our results with Korean pediatric reports (17,26), sPFAS were different according to age and residing area (27). sPFAS in our 2-, 4-, and 6-year-old children were similar to those of adolescents aged 12–19 years residing in the city of Siheung in Gyeonggi-do, except there were higher sPFOA levels among our children (17), but the levels were lower than those of children aged 5–13 years living in the city of Daegu, where many wastewater treatment plants are located (26).

Previous research on the association between PFAS and clinical or subclinical thyroid dysfunction has been limited. According to a US general adult population study (28), higher sPFOA and sPFOS were associated with self-reported thyroid disease, although detailed information on hyper- or hypothyroidism was limited. In the pediatric population aged 1–17 years, Lopez-Espinosa et al. (12) have reported a positive association between sPFOA and self-reported thyroid disease, mostly hypothyroidism, although sPFOA was not significantly related to subclinical thyroid disease. In this study, we found an inverse relationship of sPFHxS and sPFOS with subclinical hypothyroidism at 6 years of age. However, because the present study was limited by a small proportion of subclinical hypothyroidism (4.7%), further studies with larger sample sizes are warranted to confirm these results in preschool-age children.

This study showed a significant relationship between childhood exposure to PFAS and cross-sectional and longitudinal thyroid function among preschool-age children, implying that there was a positive relationship with fT4 or T3 levels and an inverse association with TSH levels. Three pediatric studies have reported the effect of postnatal PFAS exposure on thyroid function (6), although limited by a cross-sectional design (Supplementary Table S5). For the adolescent age group, Lewis et al. (7) and Lopez-Espinosa et al. (12) reported a positive relationship between sPFOS or sPFNA and TSH levels in adolescent boys, although Lin et al. (11) did not observe any significant association. For prepubertal children, Lopez-Espinosa et al. (12) reported positive relationships between sPFOS, sPFOA, and sPFNA and total T4 levels among girls at school age (6–10 years old). The positive relationship of PFAS concentrations with thyroid hormone levels among preschool boys (12) was comparable to our study results.

In the present study conducted among preschool-age children, we found significant sex interactions in some but not all of the associations between individual sPFAS and fT4 and T3 levels. This sex-specific relationship has been reported in adults (7 –9) and children (7,12). One possible explanation for sex-specific effects of PFAS on health outcomes is the shorter half-lives of PFAS in females than in males, partly due to menstruation (29). However, this explanation is not applicable to our preschool children, and the sex-specific relationship during and after puberty needs to be further investigated.

PFAS can interfere with the binding of free thyroid hormone to T4-binding globulin or transthyretin (30), and this competition could result in increased levels of circulating free thyroid hormone levels (31). In male zebrafish, long-term PFNA exposure significantly elevated plasma T3 levels by inducing histopathological changes in the thyroid follicles, leading to the increased synthesis/secretion of thyroid hormone, and inhibiting the expression of the sulfate and glucuronide conjugation pathway, leading to the reduced biliary elimination of thyroid hormone (32). Meanwhile, decreased TSH levels in our study may result from negative feedback by elevated thyroid hormone levels rather than by the direct suppression of PFAS on TSH secretion. Since the results and mechanisms observed in animal studies cannot be generalized in humans due to different exposure levels and interspecies differences in the modes of action, further human studies are required to elucidate the mechanisms of thyroid disruption of PFAS during critical windows of development.

This study had several limitations. First, only TSH levels were measured at 2 and 4 years of age due to the limited blood volume. Since fT4 and T3 levels were only measured at 6 years of age, we could evaluate subclinical thyroid dysfunction in 6-year-old children. Also, as we could not perform repeated-measure analysis for fT4 and T3 levels measured once (at 6 years of age), we conducted linear regression analysis, in contrast to TSH levels measured at multiple time points (at 2, 4, and 6 years of age). Hence, for the relationship of sPFAS with fT4 and T3 levels, we could not exclude the possibility of inflated type 1 errors for the multiple comparisons. Further longitudinal studies are warranted to reproduce our results on fT4 and T3 levels. Second, this study is limited by a lack of information on thyroid autoantibodies. Although the modifying effect of thyroid peroxidase antibodies (TPOAb) on associations between sPFAS and thyroid function has been reported in the adult population (33), the positive rate of TPOAb was very low in a Korean pediatric population (18). Since our study population was a preschool-age group, the possible effect of positive TPOAb on the relationship between sPFAS and thyroid function may be negligible. Third, urinary iodine concentration was not evaluated in this study; however, dietary iodine intake was used to adjust individual iodine status affecting thyroid function. Finally, the thyroid effects of other chemicals were not considered in the present study. However, since previous studies have reported a low correlation between sPFAS and other environmental chemical levels, such as polychlorinated biphenyls, organochlorines, phthalates, and bisphenols (34,35), there is a low possibility that the observed associations between sPFAS and thyroid function were confounded by other chemicals. Nevertheless, this study is strengthened by the use of a prospective study design to evaluate the effect of early-life PFAS exposure on childhood thyroid function.

In conclusion, sPFOS, sPFOA, sPFHxS, sPFDA, and sPFNA were consistently detected in >90% of Korean children aged 2, 4, and 6 years. PFAS were significantly associated with decreased TSH and increased fT4 or T3 levels. Whether this effect of PFAS on thyroid function in preschool-age children is sustained in school-age children or whether it gives rise to other health problems in childhood and adolescence needs to be further evaluated.

Footnotes

Acknowledgments

The authors thank Kyung-shin Lee, Jin-A Park, Ji-Young Lee, and Yumi Choi for their assistance with data collection.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This study was initially supported by grants from the Environmental Health Center funded by the Korean Ministry of Environment and a grant from the Ministry of Food and Drug Safety in 2018 (18162MFDS121) and Center for Environmental Health through the Ministry of Environment. This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Ministry of Education, Science and Technology (MEST) of the Korean government (No. 2018R1D1A1B07049806 and 2018R1D1A1B07043446).

Supplementary Material

Supplementary Figure S1

Supplementary Figure S2

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

Supplementary Table S4

Supplementary Table S5

References

Lau

, Anitole

, Hodes

, Lai

, Pfahles-Hutchens

, Seed

. 2007. Perfluoroalkyl acids: a review of monitoring and toxicological findings. Toxicol Sci, 99:366–394.

Giesy

, Kannan

. 2001. Global distribution of perfluorooctane sulfonate in wildlife. Environ Sci Technol, 35:1339–1342.

Olsen

, Burris

, Ehresman

, Froehlich

, Seacat

, Butenhoff

, Zobel

. 2007. Half-life of serum elimination of perfluorooctanesulfonate, perfluorohexanesulfonate, and perfluorooctanoate in retired fluorochemical production workers. Environ Health Perspect, 115:1298–1305.

Conder

, Hoke

, De Wolf

, Russell

, Buck

. 2008. Are PFCAs bioaccumulative? A critical review and comparison with regulatory criteria and persistent lipophilic compounds. Environ Sci Technol, 42:995–1003.

Kim

, Moon

, Oh

, Jung

, Ji

, Choi

, Park

. 2018. Association between perfluoroalkyl substances exposure and thyroid function in adults: a meta-analysis. PLoS One, 13:e0197244.

Ballesteros

, Costa

, Iñiguez

, Fletcher

, Ballester

, Lopez-Espinosa

. 2017. Exposure to perfluoroalkyl substances and thyroid function in pregnant women and children: a systematic review of epidemiologic studies. Environ Int, 99:15–28.

Lewis

, Johns

, Meeker

. 2015. Serum biomarkers of exposure to perfluoroalkyl substances in relation to serum testosterone and measures of thyroid function among adults and adolescents from NHANES 2011–2012. Int J Environ Res Public Health, 12:6098–6114.

Wen

, Lin

, Su

, Chen

, Lin

. 2013. Association between serum perfluorinated chemicals and thyroid function in US adults: the National Health and Nutrition Examination Survey 2007–2010. J Clin Endocrinol Metabol, 98:E1456–E1464.

Winquist

, Steenland

. 2014. Perfluorooctanoic acid exposure and thyroid disease in community and worker cohorts. Epidemiology, 25:255–264.

10.

Rappazzo

, Coffman

, Hines

. 2017. Exposure to perfluorinated alkyl substances and health outcomes in children: a systematic review of the epidemiologic literature. Int J Environ Res Public Health, 14:E691.

11.

Lin

, Wen

, Lin

, Wen

, Lien

, Hsu

, Chien

, Liao

, Sung

, Chen

, Su

. 2013. The associations between serum perfluorinated chemicals and thyroid function in adolescents and young adults. J Hazard Mater, 244–245:637–644.

12.

Lopez-Espinosa

, Mondal

, Armstrong

, Bloom

, Fletcher

. 2012. Thyroid function and perfluoroalkyl acids in children living near a chemical plant. Environ Health Perspect, 120:1036–1041.

13.

Bruckner

. 2000. Differences in sensitivity of children and adults to chemical toxicity: the NAS panel report. Regul Toxicol Pharmacol, 31:280–285.

14.

Kim

, Lim

, Shin

, Lee

, Kim

, Hwang

, Suh

, Hong

. 2018. Cohort profile: The Environment and Development of Children (EDC) study: a prospective children's cohort. Int J Epidemiol, 47:1049–1050f.

15.

Moon

, Lee

, Nam

, Choi

J-M

, Choe

B-K

, Seo

J-W

, Oh

, Jang

M-J

, Hwang

S-S

, Yoo

, Kim

, Lee

. 2008. 2007 Korean National Growth Charts: review of developmental process and an outlook. Korean J Pediatr, 51:1–25.

16.

Kwon

, Shin

, Kim

, Shah-Kulkarni

, Park

, Kho

, Park

, Kim

, Ha

. 2016. Prenatal exposure to perfluorinated compounds affects birth weight through GSTM1 polymorphism. J Occup Environ Med, 58:e198–e205.

17.

, Kim

, Kho

, Paek

, Sakong

, Ha

, Kim

, Choi

. 2012. Serum concentrations of major perfluorinated compounds among the general population in Korea: dietary sources and potential impact on thyroid hormones. Environ Int, 45:78–85.

18.

Kang

, Hwang

, Chung

. 2018. Excessive iodine intake and subclinical hypothyroidism in children and adolescents aged 6–19 years: results of the sixth Korean National Health and Nutrition Examination Survey, 2013–2015. Thyroid, 28:773–779.

19.

Halldorsson

, Rytter

, Haug

, Bech

, Danielsen

, Becher

, Henriksen

, Olsen

. 2012. Prenatal exposure to perfluorooctanoate and risk of overweight at 20 years of age: a prospective cohort study. Environ Health Perspect, 120:668–673.

20.

Johnson

, Sutton

, Atchley

, Koustas

, Lam

, Sen

, Robinson

, Axelrad

, Woodruff

. 2014. The Navigation Guide—evidence-based medicine meets environmental health: systematic review of human evidence for PFOA effects on fetal growth.. Environ Health Perspect, 122:1028–1039.

21.

Pearce

. 2012. Thyroid hormone and obesity. Curr Opin Endocrinol Diabetes Obes, 19:408–413.

22.

Cole

, Platt

, Schisterman

, Chu

, Westreich

, Richardson

, Poole

. 2010. Illustrating bias due to conditioning on a collider. Int J Epidemiol, 39:417–420.

23.

Fromme

, Mosch

, Morovitz

, Alba-Alejandre

, Boehmer

, Kiranoglu

, Faber

, Hannibal

, Genzel-Boroviczény

, Koletzko

, Völkel

. 2010. Pre-and postnatal exposure to perfluorinated compounds (PFCs). Environ Sci Technol, 44:7123–7129.

24.

Koponen

, Winkens

, Airaksinen

, Berger

, Vestergren

, Cousins

, Karvonen

, Pekkanen

, Kiviranta

. 2018. Longitudinal trends of per-and polyfluoroalkyl substances in children's serum. Environ Int, 121:591–599.

25.

Papadopoulou

, Sabaredzovic

, Namork

, Nygaard

, Granum

, Haug

. 2016. Exposure of Norwegian toddlers to perfluoroalkyl substances (PFAS): the association with breastfeeding and maternal PFAS concentrations. Environ Int, 94:687–694.

26.

Kim

, Lee

, Oh

. 2014. Perfluorinated compounds in serum and urine samples from children aged 5–13 years in South Korea. Environ Pollut, 192:171–178.

27.

Cho

, Lam

, Cho

, Kannan

, Cho

. 2015. Concentration and correlations of perfluoroalkyl substances in whole blood among subjects from three different geographical areas in Korea. Sci Total Environ, 512–513:397–405.

28.

Melzer

, Rice

, Depledge

, Henley

, Galloway

. 2010. Association between serum perfluorooctanoic acid (PFOA) and thyroid disease in the U.S. National Health and Nutrition Examination Survey. Environ Health Perspect, 118:686–692.

29.

Wong

, MacLeod

, Mueller

, Cousins

. 2014. Enhanced elimination of perfluorooctane sulfonic acid by menstruating women: evidence from population-based pharmacokinetic modeling. Environ Sci Technol, 48:8807–8814.

30.

Ren

, Qin

, Cao

, Zhang

, Yang

, Wan

, Guo

. 2016. Binding interactions of perfluoroalkyl substances with thyroid hormone transport proteins and potential toxicological implications. Toxicology, 366–367:32–42.

31.

Chang

, Thibodeaux

, Eastvold

, Ehresman

, Bjork

, Froehlich

, Lau

, Singh

, Wallace

, Butenhoff

. 2008. Thyroid hormone status and pituitary function in adult rats given oral doses of perfluorooctanesulfonate (PFOS). Toxicology, 243:330–339.

32.

Liu

, Wang

, Fang

, Zhang

, Dai

. 2011. The thyroid-disrupting effects of long-term perfluorononanoate exposure on zebrafish (Danio rerio). Ecotoxicology, 20:47–55.

33.

Webster

, Rauch

, Marie

, Mattman

, Lanphear

, Venners

. 2016. Cross-sectional associations of serum perfluoroalkyl acids and thyroid hormones in U.S. adults: variation according to TPOAb and iodine status (NHANES 2007–2008). Environ Health Perspect, 124:935–942.

34.

Fisher

, Arbuckle

, Liang

, LeBlanc

, Gaudreau

, Foster

, Haines

, Davis

, Fraser

. 2016. Concentrations of persistent organic pollutants in maternal and cord blood from the maternal-infant research on environmental chemicals (MIREC) cohort study. Environ Health, 15:59.

35.

Robinson

, Basagaña

, Agier

, de Castro

, Hernandez-Ferrer

, Gonzalez

, Grimalt

, Nieuwenhuijsen

, Sunyer

, Slama

, Vrijheid

. 2015. The pregnancy exposome: multiple environmental exposures in the INMA-Sabadell birth cohort. Environ Sci Technol, 49:10632–10641.