Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jul 15.
Published in final edited form as: Environ Res. 2025 Apr 3;277:121543. doi: 10.1016/j.envres.2025.121543

Neighborhood Deprivation and Iodine Levels Influence Air Pollution Effects on Maternal Thyroid Function in Early Pregnancy

Xiaoran Yang a, Sandrah P Eckel a, Zhongzheng Niu a,b, Roxana Khalili a, Tingyu Yang a, Xinci Chen a, Mario Vigil a, Trevor E Angell c, Fred Lurmann d, Patrick J Parsons e,f, Christopher D Palmer e,f, Jill Johnston a, Shohreh F Farzan a, Rima Habre a,g, Theresa M Bastain a, Carrie Breton a
PMCID: PMC12124949  NIHMSID: NIHMS2073207  PMID: 40187391

Abstract

Exposure to air pollution during pregnancy that disrupts thyroid function can lead to adverse health outcomes in mother and child. We evaluated the overall effect and critical exposure window of residential ambient air pollution exposures on thyroid function in the MADRES pregnancy cohort. We also investigated whether these associations varied by iodine deficiency status and neighborhood deprivation. Early pregnancy (6–20 weeks) serum thyroid stimulating hormone (TSH) and free thyroxine (FT4) were measured for 217 mothers. Daily residential ambient air pollution exposures (PM2.5, PM10, NO2 and O3 8hr max) were estimated using inverse-distance squared spatial interpolation from regulatory monitors. We used linear regression to assess effects of single ambient air pollutants on thyroid function, including exploration of effect modification by iodine deficiency and neighborhood deprivation (Area Deprivation Index and Gini Index of income inequality, dichotomized at the median). Distributed lag models (DLM) were used to assess critical windows of exposure for ambient air pollutants from 12 weeks preconception to first trimester. We found that one SD increase in PM2.5 (2.4 μg/m3) and PM10 (5.8 μg/m3) were associated with 18.9% (95% CI: 2.7, 37.8%) and 16.8% (95% CI: 0.7, 35.6%) higher TSH levels, respectively, with significant windows of susceptibility in the first trimester (GW 5–8 or 6–8). These associations were also modified by neighborhood deprivation, and iodine status. Our findings indicate that relatively low levels of PM exposures in early pregnancy are associated with increased TSH levels particularly among women with replete iodine levels and women living in neighborhoods with greater deprivation.

Graphical Abstract

graphic file with name nihms-2073207-f0001.jpg

1. Introduction

Thyroid function is important throughout pregnancy, and it is particularly critical during early pregnancy for both the mother and the fetus (Glinoer, 1999; Haddow et al., 2016; Krassas et al., 2010). Thyroid dysfunction (e.g., elevated/decreased thyroid stimulating hormone [TSH] or thyroxine [T4]) during early pregnancy has been associated with miscarriage (Benhadi et al., 2009) and maternal pregnancy complications (Krassas et al., 2010). It may also lead to impaired cognition and neurodevelopmental disorders in fetuses and children, and in some cases, has resulted in fetal death, since the fetus is heavily reliant on circulating maternal thyroid hormones during early pregnancy due to an immature fetal thyroid gland (Ghassabian et al., 2014; Haddow et al., 2016; Levie et al., 2018). Maternal thyroid disorders affect up to 5–7% of pregnancies (Dong and Stagnaro-Green, 2019). Among all the gestational thyroid disorders, subclinical hypothyroidism is more prevalent: it occurs in about 3.5% (95% CI: 2.9–4.1%) of women during pregnancy, compared to subclinical hyperthyroidism, which only occurs in 1.8% (95% CI: 0.8–2.9%) of pregnant women (pooled prevalence rates from a meta-analysis) (Dong and Stagnaro-Green, 2019). TSH, which is secreted by the pituitary gland, is the most sensitive indicator of thyroid function during pregnancy while measurements of free thyroxine (FT4) and triiodothyronine (T3) can establish the degree of abnormal thyroid function (Kim and Ladenson, 2012; Park, 2018). Serum TSH level controls the thyroidal production and secretion release of thyroid hormones (T4 and T3) (12). Circulating thyroid hormones are mostly bound to plasma proteins (>99%), and TSH is modulated by circulating free T4 and free T3 concentrations in a negative feedback loop (Brent, 2012). Although T3 is the metabolically active form of thyroid hormone, T4 is the predominant circulating form of thyroid hormone, and assessment of thyroid function can be derived largely from T4 alone (Brent, 2012; Ekins, 1993).

Inadequate iodine intake is the most common cause of thyroid disorders globally (Zoeller et al., 2012), but a growing body of evidence has suggested that exposure to environmental toxicants (e.g., endocrine disruptors, certain pesticides, heavy metals, and air pollution) might contribute to thyroid disorders (Mulder et al., 2019; Qiu et al., 2022; Wang et al., 2020). Some studies have examined the association of exposure to ambient air pollution with thyroid function in fetuses, newborns, and pregnant women in moderate to high air pollution environments (Janssen et al., 2017; Howe et al., 2018; Zhao et al., 2019; Qiu et al., 2022; Zhang et al., 2022). The majority of these studies have suggested that exposure to air pollution during early pregnancy, mainly particulate matter (PM), decreases the level of T4 and increases the level of TSH. However, critical exposure windows from the preconception period to early pregnancy for maternal thyroid function have not been extensively studied. Meanwhile, many existing studies have been conducted in Non-Hispanic White and Asian populations (Mulder et al., 2019; Qiu et al., 2022; Wang et al., 2020) with limited studies among Hispanic populations in the US. Hispanic populations may have higher risk of thyroid disorders (Aoki et al., 2007) as well as higher exposure to PM pollution (Letellier et al., 2022) compared to non-Hispanic White populations, thus putting them at increased risk.

Other important and potentially influential factors such as individual iodine deficiency status and neighborhood deprivation, have rarely been considered and may be important modifying factors. For example, women living in communities with higher income inequality, as measured by the Gini index, have been shown to have higher risks of thyroid disorders and thyroid cancer (Harari et al., 2014; Mouseli et al., 2023) and may be more susceptible to air pollution exposure (Bell and Ebisu, 2012; Hajat et al., 2015). However, the Gini index cannot distinguish between different income distributions within the same level of Gini index score. Therefore, we also utilized the California state level area deprivation index (ADI) to assess socioeconomic disadvantage at the neighborhood level (Kind and Buckingham, 2018) to further evaluate potential population vulnerability.

In this study, we aimed to evaluate (1) the association between ambient air pollution exposures with maternal thyroid function (TSH and FT4) during preconception and early pregnancy; (2) whether these associations are modified by iodine deficiency status and neighborhood deprivation; (3) identify critical windows of exposure from preconception period to early pregnancy in a low-income Hispanic population in Los Angeles.

2. Methods

2.1. Study population and study design

The current study focused on a subset of mothers who participated in the Maternal and Developmental Risks from Environmental and Social Stressors (MADRES) pregnancy cohort study. From November 2015 through April 2023, 1,065 women (and 914 live births) were recruited from urban Los Angeles, and the study’s methodology and protocols have been previously outlined (Bastain et al., 2019). For this sub-study, 234 mothers had their thyroid function measured during an in-person visit at early pregnancy (<20 weeks). After data processing and quality control, we excluded 17 mothers with missing ambient air pollution exposure at both the preconception period (12 weeks prior to conception) and early pregnancy (≤ 12 weeks gestation). The final analytic sample included 217 mothers. The flow chart illustrating the sample selection is shown in Figure S1.

2.2. Residential Ambient air pollution and temperature exposure

Daily ambient air pollutant concentrations of particulate matter less than 2.5 (PM2.5) and 10 (PM10) microns in diameter, nitrogen dioxide (NO2) and 8-hour maximum ozone (O3) exposures were estimated using inverse-distance squared spatial interpolation from United States Environmental Protection Agency (EPA) Air Quality System monitors and were assigned to the residential locations. Twenty-four-hour daily averages were used for PM2.5, PM10, and NO2 while 8-hour maximum value was used for O3. At every contact point, occupancy dates, and historical and prospective residential address were collected for each participant by residential history forms which were used to compute daily residential histories and to capture all residential mobility or relocations (Bastain et al., 2019). For the overall effect of ambient air pollution during early pregnancy, we used two periods of average daily exposure to ambient air pollution from 12 weeks preconception to the start of pregnancy, and the start of pregnancy through the date of thyroid function measurement, or until gestational week 13, whichever occurred first for each participant. The exposure period during gestation consists of a mean (standard deviation [SD]) of 11.6 (1.9) weeks exposure time. Weekly and early pregnancy averages of air pollutants were standardized to 1-SD of their means prior to inclusion in statistical models. We examined preconception windows up to 12 weeks given evidence to suggest windows of exposure effect on menstrual cycles and ovulation prior to conception (Gershon and Dekel, 2020) as well as similar time periods used in previous literature for hypothyroidism (Sun et al., 2023).

Additionally, to assess the confounding effect of temperature, daily temperature in degrees Celsius was calculated as the average of daily minimum and maximum temperatures extracted from a high-resolution (4 km × 4 km) gridded surface meteorological dataset (Abatzoglou, 2013). Weekly averages were further calculated from daily estimates of temperature 12 weeks prior to conception to gestational week 12 to be included in the analysis with consistent time frame with the ambient air pollution exposure.

2.3. Thyroid function

Maternal thyroid function during early pregnancy was measured by TSH (mIU/L) and free thyroxine (FT4) (ng/dL), which were quantified from maternal blood samples collected at an in-person clinical visit during early pregnancy (average gestational weeks at collection: 11.6 ± 1.9 weeks). The most widely employed measure for T4 is an immunoassay to measure FT4 based on American Thyroid Association guidelines since T4 bound to plasma proteins might be affected by other disorders or conditions and not as accurately reflect thyroid function (Kim and Ladenson, 2012; Van Uytfanghe et al., 2023). Maternal blood samples were collected by a trained phlebotomist using standard venipuncture protocols (Bastain et al., 2019). All samples were transported on ice, processed within 2 hours, and later stored at −80°C. Details regarding how serum TSH and FT4 levels were measured are described in the Supplements. Briefly, serum TSH and FT4 levels were analyzed by automated immunoassay analyzer platform (Roche Cobas e411) with the limits of detection (LOD) for TSH and FT4 as 0.01 mIU/L and 0.10 ng/dL, respectively. No mothers had FT4 under the LOD and TSH values were imputed by dividing by the square root of 2 for three mothers with TSH under the LOD. Overt hyperthyroidism, overt hypothyroidism, subclinical hyperthyroidism and subclinical hypothyroidism were defined based on literature, current healthcare practice, and American Thyroid Association’s guideline (Alexander et al., 2017; McNeil and Stanford, 2015; Stagnaro-Green et al., 2011). Detailed definitions are described in the Supplements. Thyroperoxidase antibody (TPO-Ab) level was examined as a potential confounder in sensitivity analyses. It was measured using the Access TPO-Ab assay (Beckman DxI 600), an LOD of 5.0 IU/mL in our study. TPO-Ab status was further dichotomized with values under 5.0 IU/mL as negative and over 5.0 IU/mL as positive status in the sensitivity analysis.

2.4. Urinary Iodine level

Urinary iodine (UI) level was measured during early pregnancy (average gestational weeks at collection: 13.3 ± 4.5 weeks). Urine samples were shipped on dry ice to the Trace Elements Laboratory at the New York State Department of Health’s Wadsworth Center, Albany NY. Wadsworth is one of several specialized laboratories funded by NIH under the Human Health Exposure Analysis Resource (HHEAR). Urine samples were analyzed for iodine content using a PerkinElmer NexION® 300D Inductively Coupled Plasma Mass Spectrometer (ICP-MS). The Wadsworth ICP-MS method is a modified version of one used by the US Centers for Disease Control and Prevention (CDC) to analyze urine for iodine and mercury for the National Health and Nutrition Examination Survey (NHANES) (Caldwell et al., 2003; Inorganic and Radiation Analytical Toxicology Branch, Division of Laboratory Services, n.d.). Method details and performance parameters are described in the Supplements.

Lastly, the UI level was dichotomized by 150 ng/mL as iodine replete group and iodine deficient group for effect modification analysis based on WHO suggested criteria for pregnant women (Secretariat et al., 2007).

2.5. Covariates and effect modifiers

Covariates were visualized with a directed acyclic graph (DAG) using Dagitty (Textor et al., 2016). The minimal sufficient adjustments set for estimating the total effect of ambient air pollution exposure on maternal thyroid function during early pregnancy consisted of race/ethnicity (White and non-Hispanic, Black and non-Hispanic, Hispanic, Other = multiracial and non-Hispanic mothers), level of attained education (less than high school, high school diploma or equivalent, more than high school), maternal age (continuous in years), pre-pregnancy BMI (kg/m2), parity (nulliparous and multiparous), and pre-pregnancy smoking status (yes and no). (Figure. S2). Household income, recruitment site, gestational age at thyroid measurements collections, TPO-Ab status and averaged temperature at early pregnancy were evaluated for potential confounding as listed in the DAG, but they were not included in the final model to have a parsimonious model since they did not significantly change the effect estimate.

The variables listed above, except for pre-pregnancy BMI, gestational age at thyroid measurements collections, and TPO-Ab status, were self-reported via interviewer-administered questionnaires in either English or Spanish. Pre-pregnancy BMI in kg/m2 was calculated using self-reported pre-pregnancy weight and we measured height with a stadiometer. Gestational week at thyroid measurement collection was calculated by subtracting the date of thyroid measurements and gestational age at birth which was calculated and standardized by a hierarchy of methods established (Committee on Obstetric Practice, 2017). A first trimester (<14 weeks gestation) ultrasound measurement of crown-rump length was ideal if available (54%), but, if missing, a second trimester (<28 weeks gestation) ultrasound measurement of fetal biparietal diameter was used (22%). If no measurements from an early ultrasound were available, gestational age at birth was established from a physician’s best clinical estimate from the EMR (24%).

Lastly, neighborhood deprivation, measured by Gini index of income inequality and area deprivation index (ADI), (Blesch et al., 2022; “Income inequality measures,” 2007; Maroko, 2016), were used for effect modification analysis. The Gini index of income inequality was obtained from American Community Survey (ACS) 2015–2019 5-year estimates (Bureau, n.d., n.d.) at the block group level within a subset of current population (n=215). It is one of the most commonly used measures of income inequality at the neighborhood level, although it cannot distinguish between levels of income just their inequality within the same administrative unit boundary (“Income inequality measures,” 2007). To address the disadvantages of Gini index, we also additionally adjusted for individual income level to account for individual income status. ADI (on a scale of 1–10, where 1 = least deprived) at California state level from the Neighborhood Atlas (Kind and Buckingham, 2018) was used to assess socioeconomic disadvantage at the neighborhood level by integrating Social Determinants of Health (SDOH) in California. This index helps identify areas with higher levels of deprivation, which may require additional resources for prevention and treatment interventions (Maroko, 2016). These indices were assigned to participants by 2010 Census block group boundary in which their neighborhoods were located. We dichotomized these indices by their median (median: Gini index=0.43, ADI=7) to use in our primary analysis.

2.6. Statistical analysis

Frequencies were used to summarize categorical variables (level of attained education, race/ethnicity, income, pre-pregnancy smoking status, parity, TPO-Ab status, overt hyperthyroidism, overt hypothyroidism, subclinical hyperthyroidism subclinical hypothyroidism, and iodine deficiency status). Means and standard deviations were used for continuous variables (maternal age, pre-pregnancy BMI, ambient air pollution, TSH, and FT4).

Single pollutant linear regression models were used to evaluate the relationship between overall effects of ambient air pollution and each maternal thyroid function measurement (TSH and FT4) during early pregnancy. Regression models were not conducted for other clinical and subclinical thyroid outcomes due to small frequencies in our study population (Table 1). All ambient air pollution concentrations (PM2.5, PM10, NO2, and 8-hour maximum O3) and thyroid function measurements were modeled continuously. TSH and FT4 were log-transformed due to their skewed distribution. Multi-pollutant models with 2-pollutant and 3-pollutant models were evaluated to ascertain the independent effects of ambient pollutants. Since PM2.5 and PM10, were highly correlated, (Figure S3), we did not include these ambient air pollutants in the same model. Variance inflation factors were calculated for each pollutant to identify potential collinearity in multipollutant models. Potential interaction between ambient air pollution and Gini index, ADI, or iodine deficiency status were separately evaluated in all single pollutant models. Household income level (i.e., <$30,000, ≥$30,000, and don’t know) was adjusted in the effect modification analysis of Gini index, or ADI to account for individual income information in the analysis.

Table 1.

Descriptive statistics of maternal characteristics (n = 217)

Variables N (%)/Mean (SD)
Maternal age (years) 28.8 (6.1)
Pre-pregnancy BMI (Kg/m 2 ) 28.9 (6.3)
Pre-pregnancy smoker, yes 6 (2.8%)
Race/ethnicity
 White, non-Hispanic 14 (6.5%)
 Black, non-Hispanic 25 (11.5%)
 Hispanic 170 (78.3%)
 Other1 8 (3.7%)
Parity, multiparous 141 (65.0%)
Highest Education Level
 Less than high school 56 (25.8%)
 High school diploma or equivalent 58 (26.7%)
 More than high school 103 (47.5%)
Income
 <$30,000 94 (43.3%)
 ≥$30,000 61 (28.1%)
 Don’t know 62 (28.6%)
Gini index (median [q1, q3]) 2,3 0.43 (0.4, 0.47)
Area deprivation index (median [q1, q3]) 3 7 (5, 8)
Thyroid function during early pregnancy
 Gestational age at collection (weeks) 11.9 (1.9)
 Overt hyperthyroidism 2 (0.9%)
 Overt hypothyroidism 1 (0.5%)
 Subclinical hyperthyroidism 3 (1.4%)
 Subclinical hypothyroidism 0 (0%)
 TSH (mIU/L) 0.9 (3.1)
 FT4 (ng/dL) 1.1 (1.2)
Iodine deficiency (<150 ng/mL), yes 4 47 (60.3%)
Ambient air pollution during early pregnancy 5
 Exposure time (weeks) 11.9 (1.9)
 PM2.5 (<2.5 μm) (μg/m3)
  Mean (SD) 12.0 (2.4)
  Median (IQR) 11.5 (10.4–13.2)
 PM10 (<10 μm) (μg/m3)
  Mean (SD) 29.7 (5.9)
  Median (IQR) 30.0 (25.0–33.1)
 NO2 (ppb)
  Mean (SD) 16.6 (5.9)
  Median (IQR) 14.7 (12.0–20.9)
 8 hour max O3 (ppb)
  Mean (SD) 43.1 (7.4)
  Median (IQR) 44.7 (37.8–47.5)
Open in a new tab

Abbreviations: SD=standard deviation; BMI=body mass index; ADI = area deprivation index; TSH=thyroid stimulating hormone; FT4=free thyroxine; IQR=interquartile range; PM2.5=Particulate Matter (<2.5 μm); PM10=Particulate Matter (<10 μm); NO2=Nitrogen Dioxide; O3=Ozone.

1

Other race/ethnicity include multiracial, non-Hispanic mothers and other, non-Hispanic mothers.

2

Subset of mothers (n=215).

3

No unit.

4

Subset of mothers (n=78), average gestational weeks at collection: 13.3 ± 4.5 weeks.

5

Average daily exposure to ambient air pollution from the start of pregnancy through the date of thyroid function measurement, or until gestational week 13, whichever occurred first for each participant

We used distributed lag models (DLMs), which could account for both current and past values of the exposure, to evaluate critical windows of exposure only for air pollutants that remained significantly associated with TSH or FT4 in the single pollutant linear regression models (Gasparrini, 2011). We first examined the linearity of the relationship between weekly ambient air pollution exposure and TSH or FT4 level in generalized additive models from 12 weeks prior to conception to gestational week 12. Most models showed non-significance for the spline term, suggesting a linear relationship between weekly ambient air pollution exposure and TSH or FT4 level. Thus, we used linear regression models for weekly exposure-outcome relationship. Since biological mechanism studies on the relationship between air pollution and thyroid function during early pregnancy were limited, we evaluated several one-basis natural cubic spline functions with 2, 3, 4, and 5 degrees of freedom (df) for lag-outcome relationship and compared with Akaike information criterion (AIC). Four dfs (2 knots) were selected with lowest AIC among all the tested dfs. Afterwards, AICs were compared for all the combination of 2 knots and knots placed at 4 weeks prior to conception and at conception were selected as final model with lowest AIC.

In sensitivity analyses, we assess the confounding effects of variables, identified by the DAGs but were not included in the primary analysis. In single pollutant models examining early pregnancy averages of each air pollutant, we additional adjusted for (1) household income, recruitment site, and gestational age at thyroid measurement collection; (2) TPO-Ab status; (3) averaged temperature at early pregnancy. We also restricted to only Hispanic mothers (78% of our sample) to specifically examine the effects of air pollution in this population. For the critical windows analysis, we examined gestational weeks up to 6 to create distinct temporal trends between ambient air pollution and thyroid function in sensitivity analysis as the DLM analyses in R can only be conducted among observations with non-missing weekly air pollution exposure data and the earliest thyroid function measurement was done at gestational week 6. Additionally, the weekly temperature of a cross-basis function was also included in final DLMs as another sensitivity analysis. The linear relationship for both dose-response function and the lag-response function were selected for cross-basis function of temperature with lowest AIC.

Data management and linear regression models were conducted in SAS Version 9.4. DLMs were conducted using dlnm package in R. All models met linear regression modeling assumptions, and all tests used two-sided hypotheses with α= 0.05.

3. Results

3.1. Participant characteristics

The characteristics of study participants are shown in Table 1. The majority of mothers were Hispanic (170, 78.3%), multiparous (141, 65.0%), and had a high school diploma or less education (114, 52.5%). Participants had mean (SD) maternal age of 28.8 (6.1) years with 72.4% of the population overweight or obese prior to pregnancy (mean [SD] pre-pregnancy BMI of 28.9 (6.3) kg/m2). The current study sample was similar to the full MADRES cohort on key demographic characteristics including maternal age, race and ethnicity, education, income, and recruitment site (Table S1).

The mean (SD) FT4 and TSH measurements were 1.1 (1.2) ng/dL and 0.9 (3.1) mIU/L, respectively, which were similar to what has been observed in other populations with slightly lower level of FT4 (Ghassabian et al., 2019; Zhao et al., 2019). We did not observe any subclinical hypothyroidism in the current population. We observed subclinical hyperthyroidism occurring in 1.4% of mothers which is similar to existing literature among the Hispanic population (1.6%) although the participants from the cited study were at a later stage of gestation (Casey et al., 2006). Average ambient air pollutant exposures levels were similar from 12-weeks preconception to early pregnancy (Figure S4). The mean (SD) PM2.5 and PM10 were 12.0 (2.4) μg/m3 and 29.7 (5.9) μg/m3, respectively,

3.2. Single pollutant models for maternal thyroid function during early pregnancy

In single pollutant models, after adjusting for hypothesized confounders, we observed 18.9% (95% CI: 2.7, 37.8%) higher and 16.8% (95% CI: 0.7, 35.6%) higher TSH levels were associated with a 1-SD increase in PM2.5 (equivalent to 2.4 μg/m3) and PM10 (equivalent to 5.9 μg/m3) during early pregnancy, respectively. O3 showed a positive association with TSH levels after adjusting for hypothesized confounders, although the magnitude of the association was smaller compared to PM. In contrast, the association between NO2 and TSH concentrations showed a slightly negative trend but was closer to the null. For FT4, we observed an inverse association with all pollutants during early pregnancy but the associations were generally closer to the null (Table 2). Similarly, no overall effect of ambient air pollution exposures on TSH or FT4 were found at the period of 12 weeks preconception to the start of gestation (Table S2) except for a similar impact of O3 on TSH levels observed during early pregnancy. Specifically, 1-SD increase in O3 was associated with 16.3% (95% CI: 0.3, 34.9%) higher TSH level. Associations between PM2.5 or PM10 exposure and maternal TSH concentrations during early pregnancy remained similar in magnitude in sensitivity analyses additionally adjusted for other potential confounders and restricted only to Hispanic mothers (Table S2 and S3).

Table 2.

Association between residential ambient air pollution exposure and maternal thyroid function in early pregnancy per SD increase in ambient air pollution level (n=217)

TSH (mIU/L) FT4 (ng/dL)
Crude % diff (95% CI)a p Adjusted % diff (95% CI)a,b p Crude % diff (95% CI)a p Adjusted % diff (95% CI)a,b p
PM2.5 20.1 (3.7, 39.0) 0.01 * 18.9 (2.7, 37.8) 0.02 * −1.8 (−4.1, 0.5) 0.13 −2.0 (−4.3, 0.3) 0.09
PM10 18.0 (1.5, 37.2) 0.03 * 16.8 (0.7, 35.6) 0.04 * −1.3 (−3.7, 1.1) 0.27 −1.5 (−3.8, 0.9) 0.21
NO2 0.8 (−13.4, 17.3) 0.92 −1.8 (−16.1, 14.8) 0.82 −0.6 (−3.0, 1.8) 0.62 −0.6 (−3.1, 1.9) 0.62
O3 10.6 (−4.8, 28.6) 0.19 14.0 (−2.1, 32.7) 0.09 −1.1 (−3.4, 1.3) 0.38 −1.1 (−3.5, 1.3) 0.35
Open in a new tab

Abbreviations: SD=standard deviation; TSH=thyroid stimulating hormone; FT4=free thyroxine; BMI=body mass index; CI=confidence interval; diff=difference; PM2.5=Particulate Matter (<2.5 μm); PM10=Particulate Matter (<10 μm); NO2=Nitrogen Dioxide; O3=Ozone.

*

p<0.05

a

SD of ambient air pollution - PM2.5: 2.4 μg/m3; PM10: 5.9 μg/m3; NO2: 5.9 ppb; O3: 7.4 ppb.

b

All estimates adjusted for race/ethnicity, pre-pregnancy BMI, education, maternal age, parity, and pre-pregnancy smoking status.

When including iodine in this relationship as an effect modifier (n=78), the associations between ambient air pollution and TSH (p-for-interaction, PM10: 0.43, O3: 0.76) or FT4 (p-for-interaction, PM2.5: 0.38, PM10: 0.43) varied by iodine status, although the interaction p-values were mostly not statistically significant. (Figure 1 and Table S4) The magnitude of the effect of all the ambient pollutants were stronger among mothers with replete iodine. Specifically, among mothers with replete iodine, a 1-SD increase in PM10 or O3 was associated with 45.1% (95% CI: −4.5, 120.3%) higher and 33.4% (95% CI: 0.04, 78.0%) higher levels in TSH, respectively. Among mothers with iodine deficiency, a 1-SD increase in PM10 or O3 was associated with 17.3% (95% CI: −12.2, 56.6%) higher and 23.9% (95% CI: −8.3, 67.2%) higher levels in TSH, respectively.

Figure 1. Association between residential ambient air pollution exposure and maternal thyroid function in early pregnancy per SD increase in ambient air pollution level, stratified by iodine deficiency status.

Figure 1.

Open in a new tab

Abbreviations: BMI=body mass index; CI=confidence interval; FT4=free thyroxine; NO2=Nitrogen Dioxide; O3=Ozone; PM2.5=Particulate Matter (<2.5 μm); PM10=Particulate Matter (<10 μm); SD=standard deviation; TSH=thyroid stimulating hormone. Notes: (1) % of difference in TSH level and 95% confidence intervals are shown for the association between ambient pollutants and TSH level or FT4 in early pregnancy for a-SD increased in ambient pollutants level. (2) All estimates adjusted for race/ethnicity, pre-pregnancy BMI, education, maternal age, parity, and pre-pregnancy smoking status. (3) SD of ambient air pollution - PM2.5: 2.4 μg/m3; PM10: 5.9 μg/m3; NO2: 5.9 ppb; O3: 7.4 ppb.

We observed suggestive interaction effects between the Gini index and ADI, and PM2.5 or PM10 with TSH level, although interaction p-values were not statistically significant for either metrics (p-for-interaction, Gini index - PM2.5: 0.41, PM10: 0.36; ADI- PM2.5: 0.59, PM10: 0.54). Among individuals with higher income inequality (Gini index > 0.43), a 1-SD increase in PM2.5 or PM10 was associated with 27.5% (95% CI: 2.7, 58.3%) higher and 27.4% (95% CI: 1.0, 60.6%) higher levels in TSH, respectively. Among individuals with lower income inequality (Gini index ≤0.43), a 1-SD increase in PM2.5 or PM10 was associated with 12.2% (95% CI: −9.3, 38.8%) higher and 10.3% (95% CI: −10.2, 35.5%) higher levels in TSH, respectively (Figure 2; Table S5). Similar results were observed for pregnant mothers with higher ADI compared to pregnant mothers with lower ADI (Figure 2; Table S6). ADI and income inequality did not modify the associations between NO2 or O3 and TSH level or between ambient air pollution and FT4 (Table S5 and Table S6).

Figure 2. Association between residential ambient air pollution exposure and maternal thyroid function in early pregnancy per SD increase in ambient air pollution level, stratified by 50th percentile of Gini index or ADI.

Figure 2.

Open in a new tab

Abbreviations: ADI = area deprivation index; BMI=body mass index; CI=confidence interval; FT4=free thyroxine; NO2=Nitrogen Dioxide; O3=Ozone; PM2.5=Particulate Matter (<2.5 μm); PM10=Particulate Matter (<10 μm); SD=standard deviation; TSH=thyroid stimulating hormone. Notes: (1) % of difference in TSH level and 95% confidence intervals are shown for the association between ambient pollutants and TSH level or FT4 in early pregnancy for a-SD increased in ambient pollutants level. (2) All estimates adjusted for race/ethnicity, pre-pregnancy BMI, education, maternal age, parity, household income, and pre-pregnancy smoking status. (3) SD of ambient air pollution - PM2.5: 2.4 μg/m3; PM10: 5.9 μg/m3; NO2: 5.9 ppb; O3: 7.4 ppb.

3.3. Multi-pollutant models for maternal thyroid function during early pregnancy

Variance inflation factors were smaller for 2-pollutant models compared to 3-pollutant models, particularly for models including both NO2 and O3 (Table S7). Results were similar comparing model estimates from multipollutant models and from single pollutant models (Table S8). For example, in the two pollutant models of PM2.5 and O3, we also observed that higher PM2.5 or O3 concentrations were statistically significantly associated with increased TSH.

3.4. Distributed lag models for weekly PM exposures

Results of DLMs for PM2.5 or PM10 on TSH for 12 weeks prior to conception to gestational week 12 are shown in Figure 3. Exposure to PM2.5 or PM10 during gestational weeks 5 to 8 or 6 to 8 of pregnancy was positively associated with TSH in early pregnancy. The strongest association between PM2.5 or PM10 and TSH was observed for exposure at week 6 or 7 of pregnancy (a 1-SD increased in PM2.5: % difference in TSH = 1.6%; 95% CI: 0.1, 3.0%; PM10: 1.8%, 95% CI: 0.3, 3.3%). Results were similar when the 2 knots were placed at 4 weeks prior to conception and at conception for PM2.5 and PM10, respectively, based on AIC (PM2.5 AIC = 668.05 and PM10 AIC = 666.62) rather than 2 knots were equally placed through 12-week of preconception to gestational week 12 (PM2.5 AIC = 668.36 and PM10 AIC = 666.20) (Figures S5). The critical window of exposure to PM2.5 or PM10 for maternal TSH concentrations during early pregnancy remained similar in sensitivity analyses with from 12 weeks prior to conception to gestational week 6 (Figures S6–7) and with weekly temperature of a cross-basis function included in the final model (Figures S8).

Figure 3. Distributed lag model results for PM2.5 and PM10.

Figure 3.

Open in a new tab

Abbreviations: BMI=body mass index; CI=confidence interval; df=degree of freedom; DLM=distributed lag model; FT4=free thyroxine; PM2.5=Particulate Matter (<2.5 μm); PM10=Particulate Matter (<10 μm); SD=standard deviation; TSH=thyroid stimulating hormone. Note: (1) DLM result for PM2.5 and PM10 were depicted with 4-dfs and 2 knots placed at 4 weeks prior to conception and at conception; (2) % of difference in TSH level and 95% confidence intervals are shown for the association between PM2.5 or PM10 and TSH level in early pregnancy for a-SD increased in PM2.5 or PM10. (3) All estimates adjusted for race/ethnicity, pre-pregnancy BMI, education, maternal age, parity, and pre-pregnancy smoking status. (4) Week 0 indicates conception. (5) SD of ambient air pollution - PM2.5: 2.4 μg/m3; PM10: 5.9 μg/m3; NO2: 5.9 ppb; O3: 7.4 ppb.

4. Discussion

In the current study, exposures to PM2.5 and PM10 but not NO2 or O3 during early pregnancy were associated with higher maternal TSH concentrations. Two-pollutant models also suggested PM2.5 was associated with lower FT4 when including adjustment for O3, although the association was not observed in other single pollutant and 2-pollutant models. DLM analysis revealed that the positive association between PM2.5 or PM10 with maternal TSH emerged in the first 5–8 gestational weeks. Moreover, stronger effects of most pollutants on TSH were observed among iodine replete mothers and among pregnant mothers living in neighborhoods with higher income inequality and greater neighborhood deprivation.

In our study, we observed air pollutant concentrations lower than or closer to the lower limit of both the EPA National Ambient Air Quality Standards (NAAQS) (US EPA, 2014) and levels reported in previous research on this topic (Ghassabian et al., 2019; Wang et al., 2019; Zhang et al., 2022; Zhao et al., 2019). Specifically, the observed average PM2.5 levels during early pregnancy was below or close to the NAAQS primary and secondary standard (24-hour average: 35 μg/m3 and annal average: 9 μg/m3 and 15 μg/m3). Similarly, the observed average PM10 during early pregnancy was lower than the 24-hour average of NAAQS primary and secondary standard (150 μg/m3) (US EPA, 2014). These levels were also lower than those reported in previous studies, including research conducted in China (NO2: 40.1 ppb; PM2.5: 56.6 μg/m3) (Zhao et al., 2019) and a European study analyzing data from four cohorts (PM2.5: 11.5–20.6 μg/m3; PM10: 27.3–34.6 μg/m3) (Ghassabian et al., 2019). Despite being located in Los Angeles, our study setting allowed us to address a research gap by examining the relationship between relatively low levels of air pollutants and thyroid function, providing insights into whether health effects on thyroid may still be evident and whether current regulatory standards are sufficiently protective.

The majority of epidemiological studies have observed a positive relationship between ambient air pollution and maternal TSH, and inverse relationships between ambient air pollution and maternal FT4, which is consistent with the observed results of the current study (Ghassabian et al., 2019; Wang et al., 2019; Zhang et al., 2022; Zhao et al., 2019). We were not able to directly compare the magnitude of the ambient air pollution effect on TSH level due to the use of different outcomes in different studies (e.g., dichotomized TSH level as binary variable, hypothyroxinemia). However, our results are consistent with a skewing of thyroid hormone programming toward hypothyroidism (higher TSH with lower FT4 levels) under ambient pollution exposure (Ghassabian et al., 2019; Zhao et al., 2019). This is important because maternal hypothyroidism has been associated with detrimental neurocognitive outcomes in children and increased risk of pregnancy complications and cardiovascular risks in women (Krassas et al., 2010). Although the overall effect of ambient air pollution during pregnancy on thyroid function were similar, the iodine concentration in the urine was not considered in those studies and most of the studies did not additionally examine or adjust for ambient exposure at previous time points including preconception time and pregnancy time.

In addition to evaluating the effect of average air pollutant levels on thyroid function, we observed that the association between PM2.5 or PM10 and higher maternal TSH was strongest around gestational weeks 5–8 in DLMs. In the primary analysis, we modeled weekly ambient air pollution exposures from 12 weeks prior to conception to gestational week 12 in DLMs. We was able to isolate the direct effects of air pollution on maternal thyroid function with critical exposure window identified because the fetal thyroid gland is fully developed after gestational week 12 and thereafter interacts with maternal thyroid function (Fisher, 1989). Early gestation is an extremely important developmental period for fetal organogenesis, and also occurs at a time when most women do not yet realize they are pregnant. Intervention studies that have tried to intervene in pregnancy with T4 supplementation have not always had the desired effect of improving offspring IQ (Scholz et al., 2024), in part because they may not be able to intervene early enough in pregnancy. Overall, air pollution exposure during pregnancy has been consistently linked to adverse neurodevelopmental outcomes in offspring (Morgan et al., 2023; Perera et al., 2024). Given that thyroid hormones play an important role in fetal brain development as well, the impact of air pollution on maternal thyroid function during early pregnancy may represent one biological pathway through which air pollutants affect fetal neurodevelopment. Therefore, recognizing and mitigating the effect of air pollution during critical periods of fetal brain development should be an important goal for ensuring both optimal thyroid health and future fetal brain function.

One potential biological mechanism that could play a part is the negative feedback loop of the thyroid hormone regulation which leads to an increase in TSH level during early pregnancy (Brent, 2012) (Figure 4). A disturbance in thyroid hormone biosynthesis of T4 and T3 can lead to increased TSH level. As pregnancy begins, thyroid function is altered to meet the fetus’s need until the fetus begins producing its own supply of T4 during late first trimester (Obregon et al., 2007), T4 levels are also supposed to be stimulated by human chorionic gonadotropin (hCG) which causes a corresponding decrease in TSH level (Glinoer et al., 1990; Haddow et al., 2008). These changes makes the early first trimester a particularly vulnerable period for the thyroid function among mothers (Obregon et al., 2007). However, TSH level can also be affected by other factors in addition to pregnancy itself. An animal study has suggested that PM2.5 exposure decreases the level of Thyroperoxidase which in turn decreases the adequate utilization of iodine (Dong et al., 2021). Iodine plays an important role in thyroid hormone biosynthesis and deficiency in iodine can leading to a reduction in circulating T4 and T3 levels which might further lead to an increase in TSH (Dong et al., 2021). In terms of the effect of O3 on thyroid function, limited studies have provided biological mechanisms. One animal study showed that TSH level and thyroid weight increased after short-term O3 exposure (Clemons and Garcia, 1980) while another study suggested that higher TSH levels are reacted to promote the destruction-repair cycle of thyroid follicular epithelium cells to maintain thyroid homeostasis under long-term O3 exposure (He et al., 2022). Future mechanistic studies should be conducted to understand how ambient air pollutants affect thyroid function in humans.

Figure 4. Potential pathways of residential ambient air pollution exposure impact on maternal thyroid function during pregnancy.

Figure 4.

Open in a new tab

Abbreviations: FT4=free thyroxine; PM2.5=Particulate Matter (<2.5 μm); PM10=Particulate Matter (<10 μm); TSH=thyroid stimulating hormone. Note: (1) arrow with dotted line represent the typical processes without any exposure while the arrow with solid line represent the processes after exposing to PM exposure; (2) ① depicts how pregnancy can affect thyroid function by increasing thyroid volume with hCG level increased, and this process remain the same. Thus, the arrows are solid. (3) ② depicts a potential biological process: the residential ambient air pollution exposures disturb thyroid hormone biosynthesis activity of T4 and T3 through iodine utilization which can lead to an increased TSH level; (4) This figure was created in BioRender. Yang, X. (2025) https://BioRender.com/i66j769.

Although inadequate iodine intake is the most common cause of thyroid disorders globally (Zoeller et al., 2012), no specific studies have directly examined the role of iodine in the relationship between air pollution and thyroid hormone synthesis to date. Only one study suggested that iodine sufficiency may positively alter the thyroid gland’s sensitivity to environmental stressors (Grossklaus et al., 2023). Additionally, iodine requirements increase substantially during pregnancy due to increased physiological demands, and when combined with environmental stressors like air pollution and socioeconomic factors, pregnant women may become particularly vulnerable to thyroid dysfunction, underscoring the importance of examining these interactions in our study (Grossklaus et al., 2023; Pearce, 2017). In our study, we found suggestive stronger associations between residential exposure to all ambient air pollutants and maternal TSH among iodine replete women, which suggests that iodine-replete individuals might be more vulnerable to the thyroid-disrupting effects of air pollutants, or conversely, that lack of iodine in iodine-deficient individuals may already contribute to increased thyroid dysfunction that would mask any additional effects of pollutants. However, we did not observe statistically significant associations except for O3 due to limited sample size. These results indicate the complex interactions between iodine status and air pollution. Although we dichotomized urinary iodine levels at 150 ng/mL based on WHO’s clinically meaningful criteria for insufficient iodine intake in pregnant women (Secretariat et al., 2007), this binary categorization may simplify a potentially more complex interaction between iodine intake and the ambient air pollution. For example, elevated urinary iodine levels are positively associated with both hypothyroidism and hypothyroxinemia among pregnant women in a large Chinese study (Shi et al., 2015) suggesting that both insufficient and excessive iodine intake may affect thyroid health. Urinary iodine measurements were taken before or around the time of thyroid hormone measurements and were available only for a subset of participants (n=78). We also assumed that urinary iodine levels could reflect recent individual iodine intake during early pregnancy (Hlucny et al., 2021; König et al., 2011; Pearce and Caldwell, 2016). Further research is needed to understand the underlying mechanisms and confirm our findings with both urinary and serum iodine level, particularly given the small sample size we had for the iodine analyses.

We examined two measures, the Gini index of income inequality and the ADI (Blesch et al., 2022; “Income inequality measures,” 2007; Maroko, 2016), as potential measures of neighborhood deprivation that might modify air pollution effects on thyroid. We observed that pregnant mothers living in neighborhoods with higher income inequality (Gini index>0.43) or higher area deprivation (ADI>7) had larger increase in maternal TSH level associated with increased PM2.5 or PM10. To our knowledge, this is the first study to investigate how neighborhood-level characteristics may influence air pollution effects on maternal thyroid function during early pregnancy. Our results are plausible given the existing evidence relating socioeconomic disparities, measure by Gini index, or Yost’s index, or annual income, to healthcare inequities such as poor healthcare access (Harari et al., 2014; Hsu et al., 2023) as well as higher burden of multiple environmental exposures, especially ambient air pollution (Bell and Ebisu, 2012). These initial results illustrate the potential complexities in relationships between neighborhood level characteristics, exposures and individual factors, all of which ultimately may affect thyroid health in pregnancy. Greater investigations into their component and joint effects are thus needed in larger populations.

The results of the current study are notable for several reasons. First, there has been limited research investigating the relationship between ambient air pollution and maternal thyroid function during pregnancy in low-income Hispanic participants. A notable strength of our cohort is the opportunity to address this research gap by studying air pollution’s impact on maternal thyroid function in an understudied population that often faces disproportionate environmental exposures. However, as we recruited predominantly Hispanic participants (78%) from the urban Los Angeles area, we acknowledge that our findings may have limited generalizability to other race and ethnic groups, the broader United States population, or rural areas. Second, we investigated both the overall effect of ambient air pollution during early pregnancy and applied DLMs to understand critical windows for air pollution exposures from preconception to early pregnancy, on maternal thyroid function. The identified critical window, very early pregnancy, is particularly important as this is a crucial period for fetal development in which thyroid hormones play an essential role. The identification of this early critical window provides an opportunity for timely intervention, allowing healthcare providers and pregnant women to take preventive actions, such as reducing outdoor activities during high pollution days or using air purification systems during early pregnancy. According to the pregnancy-specific thyroid hormone range from American Thyroid Association guidelines (Alexander et al., 2017), the observed air pollution exposure in our study could potentially shift approximately 15% of pregnant women in our cohort toward the upper limit of TSH level (>2.5 mIU/L) during the first trimester, which is the threshold value of subclinical hypothyroidism under normal range of FT4, based on our main effect estimates. Moreover, in the identified potentially vulnerable subgroups from our effect modification analyses, this percentage could increase to approximately 20% of mothers potentially shifting towards upper limit of TSH level. This is particularly concerning given the nature of air pollution exposure, suggesting that even modest individual-level effects could have substantial population-level impacts, especially in vulnerable communities with higher exposure burdens. Lastly, this is one of the few studies to explore how individual and neighborhood characteristics modify the associations of ambient exposures and maternal thyroid function during early pregnancy.

The current study has several limitations. First, only FT4 and TSH measures were available for the current study participants. Therefore, other indicators of thyroid function and related parameters (e.g., TT4, T3, TBG) could not be examined in relation to ambient air pollution, which limits the ability to interpret the observed associations between ambient air pollution and maternal thyroid function during early pregnancy. Another limitation is that the sample size of our study was relatively small, thereby restricting the power to detect true associations. Meanwhile, possible selection bias may be introduced as the current study used a subsample of the whole cohort. However, our analyses demonstrated that participant characteristics among this subset were very similar to the full MADRES cohort across various key demographic characteristics including race/ethnicity, income, and education. This similarity suggests that our subsample is likely representative of the full cohort, though we acknowledge that some unmeasured differences may exist. Using urinary iodine level instead of serum iodine level can be another limitation. Lastly, although several sensitivity analyses were performed to assess the robustness of our results, we acknowledge that residual confounding remains possible to explain the associations seen with increased TSH and decreased FT4.

5. Conclusion

Our findings indicate that relatively low levels of PM2.5 and PM10 exposures in early pregnancy are associated with increased TSH levels during early pregnancy, particularly among women with replete iodine levels, and women living in neighborhoods with greater deprivation. Because significantly higher TSH levels have been associated with adverse maternal and neonatal outcomes, it will be important to further define the potential excess risk attributed to air pollution exposures.

Supplementary Material

1

Highlights.

  • Relatively low levels of air pollution may affect thyroid health in early pregnancy.

  • Higher first trimester PM2.5 and PM10 were associated with higher maternal TSH.

  • Early first trimester is crucial for ambient air pollution’s impact on maternal TSH.

  • Iodine level and neighborhood deprivation modified the effect of PM and O3 on TSH.

  • No associations found between ambient air pollution and FT4 in early pregnancy.

Acknowledgment

We sincerely thank the families, nurses, midwives, physicians, and staff at our study sites for their participation and work in the MADRES study. We also would like to acknowledge Ms. Charelle Trim for validating the modified method for this project and analyzing the urine samples for iodine.

Funding

This work was supported by the National Center on Minority Health and Health Disparities [grant numbers P50MD015705] and National Institute of Environmental Health Sciences [grant numbers R01ES027409, U2CES026542].

Declaration of interests

Carrie Breton reports financial support was provided by National Institute on Minority Health and Health Disparities. Carrie Breton reports financial support was provided by National Institute of Environmental Health Sciences. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Glossary

ADI

Area deprivation index

FT4

Free thyroxine

TSH

Thyroid stimulating hormone

T3

Triiodothyronine

T4

Thyroxine

UI

Urinary iodine

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Data Availability Statement

Restrictions apply to the availability of all data generated or analyzed during this study because they were used under license. The corresponding author will on request detail the restrictions and any conditions under which access to some data may be provided.

Reference

  1. Abatzoglou JT, 2013. Development of gridded surface meteorological data for ecological applications and modelling. Int. J. Climatol 33, 121–131. 10.1002/joc.3413 [DOI] [Google Scholar]
  2. Alexander EK, Pearce EN, Brent GA, Brown RS, Chen H, Dosiou C, Grobman WA, Laurberg P, Lazarus JH, Mandel SJ, Peeters RP, Sullivan S, 2017. 2017 Guidelines of the American Thyroid Association for the diagnosis and management of thyroid disease during pregnancy and the postpartum. Thyroid® 27, 315–389. 10.1089/thy.2016.0457 [DOI] [PubMed] [Google Scholar]
  3. Aoki Y, Belin RM, Clickner R, Jeffries R, Phillips L, Mahaffey KR, 2007. Serum TSH and total T4 in the United States population and their association with participant characteristics: National Health and Nutrition Examination Survey (NHANES 1999–2002). Thyroid Off. J. Am. Thyroid Assoc 17, 1211–1223. 10.1089/thy.2006.0235 [DOI] [PubMed] [Google Scholar]
  4. Bastain TM, Chavez T, Habre R, Girguis MS, Grubbs B, Toledo-Corral C, Amadeus M, Farzan SF, Al-Marayati L, Lerner D, Noya D, Quimby A, Twogood S, Wilson M, Chatzi L, Cousineau M, Berhane K, Eckel SP, Lurmann F, Johnston J, Dunton GF, Gilliland F, Breton C, 2019. Social determinants of depression among hispanic women study design, protocol and profile of the maternal and developmental risks from environmental and social stressors (MADRES) pregnancy cohort: a prospective cohort study in predominantly low-income hispanic women in urban. BMC Pregnancy Childbirth 19, 189. 10.1186/s12884-019-2330-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bell ML, Ebisu K, 2012. Environmental inequality in exposures to airborne particulate matter components in the United States. Environ. Health Perspect 120, 1699–1704. 10.1289/ehp.1205201 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Benhadi N, Wiersinga WM, Reitsma JB, Vrijkotte TGM, Bonsel GJ, 2009. Higher maternal TSH levels in pregnancy are associated with increased risk for miscarriage, fetal or neonatal death. Eur. J. Endocrinol 160, 985–991. 10.1530/EJE-08-0953 [DOI] [PubMed] [Google Scholar]
  7. Blesch K, Hauser OP, Jachimowicz JM, 2022. Measuring inequality beyond the Gini coefficient may clarify conflicting findings. Nat. Hum. Behav 6, 1525–1536. 10.1038/s41562-022-01430-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Brent GA, 2012. Mechanisms of thyroid hormone action. J. Clin. Invest 122, 3035–3043. 10.1172/JCI60047 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bureau UC, n.d. Gini Index. Census.gov. URL https://www.census.gov/topics/income-poverty/income-inequality/about/metrics/gini-index.html (accessed 8.23.24a).
  10. Bureau UC, n.d. ACS Income Inequality Data Tables. Census.gov. URL https://www.census.gov/topics/income-poverty/income-inequality/data/data-tables/acs-data-tables.html (accessed 9.5.24b).
  11. Caldwell KL, Maxwell CB, Makhmudov A, Pino S, Braverman LE, Jones RL, Hollowell JG, 2003. Use of inductively coupled plasma mass spectrometry to measure urinary iodine in NHANES 2000: comparison with previous method. Clin. Chem 49, 1019–1021. 10.1373/49.6.1019 [DOI] [PubMed] [Google Scholar]
  12. Casey BM, Dashe JS, Wells CE, McIntire DD, Leveno KJ, Cunningham FG, 2006. Subclinical Hyperthyroidism and pregnancy outcomes. Obstet. Gynecol 107, 337. 10.1097/01.AOG.0000197991.64246.9a [DOI] [PubMed] [Google Scholar]
  13. Clemons GK, Garcia JF, 1980. Changes in thyroid function after short-term ozone exposure in rats. J. Environ. Pathol. Toxicol 4, 359–369. [PubMed] [Google Scholar]
  14. Committee on Obstetric Practice, 2017. Committee Opinion No 700: methods for estimating the due date. Obstet. Gynecol 129, e150–e154. 10.1097/AOG.0000000000002046 [DOI] [PubMed] [Google Scholar]
  15. Dong AC, Stagnaro-Green A, 2019. Differences in diagnostic criteria mask the true prevalence of thyroid disease in pregnancy: a systematic review and meta-analysis. Thyroid Off. J. Am. Thyroid Assoc 29, 278–289. 10.1089/thy.2018.0475 [DOI] [PubMed] [Google Scholar]
  16. Dong X, Wu W, Yao S, Li H, Li Z, Zhang L, Jiang J, Xu J, Zhang F, 2021. PM2.5 disrupts thyroid hormone homeostasis through activation of the hypothalamic-pituitary-thyroid (HPT) axis and induction of hepatic transthyretin in female rats 2.5. Ecotoxicol. Environ. Saf 208, 111720. 10.1016/j.ecoenv.2020.111720 [DOI] [PubMed] [Google Scholar]
  17. Ekins R, 1993. Analytic measurements of free thyroxine. Clin. Lab. Med., Pathophysiology of thyroid disease 13, 599–630. 10.1016/S0272-2712(18)30428-1 [DOI] [PubMed] [Google Scholar]
  18. Fisher DA, 1989. Development of fetal thyroid system control, in: DeLong GR, Robbins J, Condliffe PG (Eds.), Iodine and the Brain. Springer; US, Boston, MA, pp. 167–176. 10.1007/978-1-4613-0765-5_15 [DOI] [Google Scholar]
  19. Gasparrini A, 2011. Distributed Lag Linear and Non-Linear Models in R: the package dlnm. J. Stat. Softw 43, 1–20. [PMC free article] [PubMed] [Google Scholar]
  20. Gershon E, Dekel N, 2020. Newly identified regulators of ovarian folliculogenesis and ovulation. Int. J. Mol. Sci 21, 4565. 10.3390/ijms21124565 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Ghassabian A, El Marroun H, Peeters RP, Jaddoe VW, Hofman A, Verhulst FC, Tiemeier H, White T, 2014. Downstream effects of maternal hypothyroxinemia in early pregnancy: nonverbal IQ and brain morphology in school-age children. J. Clin. Endocrinol. Metab 99, 2383–2390. 10.1210/jc.2013-4281 [DOI] [PubMed] [Google Scholar]
  22. Ghassabian A, Pierotti L, Basterrechea M, Chatzi L, Estarlich M, Fernández-Somoano A, Fleisch AF, Gold DR, Julvez J, Karakosta P, Lertxundi A, Lopez-Espinosa M-J, Mulder TA, Korevaar TIM, Oken E, Peeters RP, Rifas-Shiman S, Stephanou E, Tardón A, Tiemeier H, Vrijheid M, Vrijkotte TGM, Sunyer J, Guxens M, 2019. Association of exposure to ambient air pollution with thyroid function during pregnancy. JAMA Netw. Open 2, e1912902. 10.1001/jamanetworkopen.2019.12902 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Glinoer D, 1999. What happens to the normal thyroid during pregnancy? Thyroid Off. J. Am. Thyroid Assoc 9, 631–635. 10.1089/thy.1999.9.631 [DOI] [PubMed] [Google Scholar]
  24. Glinoer D, Nayer PP, Bourdoux P, Lemone M, ROBYN C, STEIRTEGHEM AV, KINTHAERT J, LEJEUNE B, 1990. Regulation of maternal thyroid during pregnancy. J. Clin. Endocrinol. Metab 71, 276–287. 10.1210/jcem-71-2-276 [DOI] [PubMed] [Google Scholar]
  25. Grossklaus R, Liesenkötter K-P, Doubek K, Völzke H, Gaertner R, 2023. Iodine deficiency, maternal hypothyroxinemia and endocrine disrupters affecting fetal brain development: A scoping review. Nutrients 15, 2249. 10.3390/nu15102249 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Haddow JE, Craig WY, Neveux LM, Palomaki GE, Lambert-Messerlian G, Malone FD, D’Alton ME, 2016. Free thyroxine during early pregnancy and risk for gestational diabetes. PLoS ONE 11, e0149065. 10.1371/journal.pone.0149065 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Haddow JE, McClain MR, Lambert-Messerlian G, Palomaki GE, Canick JA, Cleary-Goldman J, Malone FD, Porter TF, Nyberg DA, Bernstein P, D’Alton ME, First and Second Trimester Evaluation of Risk for Fetal Aneuploidy Research Consortium, 2008. Variability in thyroid-stimulating hormone suppression by human chorionic gonadotropin during early pregnancy. J. Clin. Endocrinol. Metab 93, 3341–3347. 10.1210/jc.2008-0568 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hajat A, Hsia C, O’Neill MS, 2015. Socioeconomic disparities and air pollution exposure: a global review. Curr. Environ. Health Rep 2, 440–450. 10.1007/s40572-015-0069-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Harari A, Li N, Yeh MW, 2014. Racial and socioeconomic disparities in presentation and outcomes of well-differentiated thyroid cancer. J. Clin. Endocrinol. Metab 99, 133–141. 10.1210/jc.2013-2781 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. He Q, Wu M, Shi Q, Tan H, Wei B, Tang N, Chen J, Liu M, Duan S, Chang S, Huang P, 2022. Association of Ozone exposures with the risk of thyroid nodules in Hunan Province: a population-based cohort study. Environ. Health 21, 65. 10.1186/s12940-022-00874-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Hlucny K, Alexander BM, Gerow K, Larson-Meyer DE, 2021. Reflection of dietary iodine in the 24 h urinary iodine concentration, serum iodine and thyroglobulin as biomarkers of iodine status: a pilot study. Nutrients 13, 2520. 10.3390/nu13082520 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Howe CG, Eckel SP, Habre R, Girguis MS, Gao L, Lurmann FW, Gilliland FD, Breton CV, 2018. Association of prenatal exposure to ambient and traffic-related air pollution with newborn thyroid function: findings from the Children’s Health Study. JAMA Netw. Open 1, e182172. 10.1001/jamanetworkopen.2018.2172 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Hsu Y-C, Cheng SY-H, Chien M-N, Cheng S-P, 2023. Impact of social and economic factors on global thyroid cancer incidence and mortality. Eur. Arch. Otorhinolaryngol 280, 4185–4193. 10.1007/s00405-023-07992-0 [DOI] [PubMed] [Google Scholar]
  34. Income inequality measures, 2007. J. Epidemiol. Community Health 61, 849–852. 10.1136/jech.2006.052969 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Inorganic and Radiation Analytical Toxicology Branch, Division of Laboratory Services, N.C. for E.H., n.d. Iodine and Mercury in urine by ICP-DRC-MS, Method No. 3002.7–05. URL https://wwwn.cdc.gov/nchs/data/nhanes/2017-2018/labmethods/UIO-J-UHG-J-MET-508.pdf (accessed 7.30.24).
  36. Janssen BG, Saenen ND, Roels HA, Madhloum N, Gyselaers W, Lefebvre W, Penders J, Vanpoucke C, Vrijens K, Nawrot TS, 2017. Fetal thyroid function, birth weight, and in utero exposure to fine particle air pollution: a birth cohort study. Environ. Health Perspect 125, 699–705. 10.1289/EHP508 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Kim M, Ladenson P, 2012. 233 - Thyroid, in: Goldman L, Schafer AI (Eds.), Goldman’s Cecil Medicine (Twenty Fourth Edition). W.B. Saunders, Philadelphia, pp. e64–e77. 10.1016/B978-1-4377-1604-7.00560-1 [DOI] [Google Scholar]
  38. Kind AJH, Buckingham WR, 2018. Making Neighborhood-Disadvantage Metrics Accessible - The Neighborhood Atlas. N. Engl. J. Med 378, 2456–2458. 10.1056/NEJMp1802313 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. König F, Andersson M, Hotz K, Aeberli I, Zimmermann MB, 2011. Ten repeat collections for urinary iodine from spot samples or 24-hour samples are needed to reliably estimate individual iodine status in women. J. Nutr 141, 2049–2054. 10.3945/jn.111.144071 [DOI] [PubMed] [Google Scholar]
  40. Krassas GE, Poppe K, Glinoer D, 2010. Thyroid function and human reproductive health. Endocr. Rev 31, 702–755. 10.1210/er.2009-0041 [DOI] [PubMed] [Google Scholar]
  41. Letellier N, Zamora S, Spoon C, Yang J-A, Mortamais M, Carrasco Escobar G, Sears DD, Jankowska MM, Benmarhnia T, 2022. Air pollution and metabolic disorders: Dynamic versus static measures of exposure among Hispanics/Latinos and non-Hispanics. Environ. Res 209, 112846. 10.1016/j.envres.2022.112846 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Levie D, Korevaar TIM, Bath SC, Dalmau-Bueno A, Murcia M, Espada M, Dineva M, Ibarluzea JM, Sunyer J, Tiemeier H, Rebagliato M, Rayman MP, Peeters RP, Guxens M, 2018. Thyroid function in early pregnancy, child IQ, and autistic traits: a meta-analysis of individual participant data. J. Clin. Endocrinol. Metab 103, 2967–2979. 10.1210/jc.2018-00224 [DOI] [PubMed] [Google Scholar]
  43. Maroko AR, 2016. Integrating social determinants of health with treatment and prevention: a new tool to assess local area deprivation. Prev. Chronic. Dis 13. 10.5888/pcd13.160221 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. McNeil AR, Stanford PE, 2015. Reporting thyroid function tests in pregnancy. Clin. Biochem. Rev 36, 109–126. [PMC free article] [PubMed] [Google Scholar]
  45. Morgan ZEM, Bailey MJ, Trifonova DI, Naik NC, Patterson WB, Lurmann FW, Chang HH, Peterson BS, Goran MI, Alderete TL, 2023. Prenatal exposure to ambient air pollution is associated with neurodevelopmental outcomes at 2 years of age. Environ. Health Glob. Access Sci. Source 22, 11. 10.1186/s12940-022-00951-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Mouseli A, Sharafi M, Mastaneh Z, Shiri MS, 2023. Contrasting socioeconomic inequality with noncommunicable diseases: insights from a population-based survey using the concentration index in Kong cohort study. Health Sci. Rep 6, e1682. 10.1002/hsr2.1682 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Mulder TA, van den Dries MA, Korevaar TIM, Ferguson KK, Peeters RP, Tiemeier H, 2019. Organophosphate pesticides exposure in pregnant women and maternal and cord blood thyroid hormone concentrations. Environ. Int 132, 105124. 10.1016/j.envint.2019.105124 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Obregon MJ, Calvo RM, Escobar del Rey F, Morreale de Escobar G, 2007. Ontogenesis of thyroid function and interactions with maternal function. 10.1159/000106821 [DOI] [PubMed] [Google Scholar]
  49. Park C-E, 2018. Evaluation of pregnancy and thyroid function. Korean J. Clin. Lab. Sci 50, 1–10. 10.15324/kjcls.2018.50.1.1 [DOI] [Google Scholar]
  50. Pearce EN, 2017. Iodine supplementation during pregnancy. URL https://www.who.int/tools/elena/commentary/iodine-pregnancy (accessed 2.22.25).
  51. Pearce EN, Caldwell KL, 2016. Urinary iodine, thyroid function, and thyroglobulin as biomarkers of iodine status123. Am. J. Clin. Nutr 104, 898S–901S. 10.3945/ajcn.115.110395 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Perera F, Miao Y, Ross Z, Rauh V, Margolis A, Hoepner L, Riley KW, Herbstman J, Wang S, 2024. Prenatal exposure to air pollution during the early and middle stages of pregnancy is associated with adverse neurodevelopmental outcomes at ages 1 to 3 years. Environ. Health 23, 95. 10.1186/s12940-024-01132-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Qiu L, Shen W, Ye C, Wu J, Zheng S, Lou B, Chen Z, Xu P, Xu D, Wang X, Feng B, 2022. Association of exposure to PM2.5-bound metals with maternal thyroid function in early pregnancy. Sci. Total Environ 810, 151167. 10.1016/j.scitotenv.2021.151167 [DOI] [PubMed] [Google Scholar]
  54. Scholz A, McNabb CB, Bloomfield L, Bhargava R, Hales C, Dayan CM, Taylor PN, Lazarus JH, Okosieme O, Ludgate M, Jones DK, Rees DA, 2024. Controlled Antenatal Thyroid Screening Study III: effects of gestational thyroid status on adolescent brain morphology. J. Clin. Endocrinol. Metab dgae338. 10.1210/clinem/dgae338 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Secretariat WHO, Andersson M, de Benoist B, Delange F, Zupan J, 2007. Prevention and control of iodine deficiency in pregnant and lactating women and in children less than 2-years-old: conclusions and recommendations of the Technical Consultation. Public Health Nutr. 10, 1606–1611. 10.1017/S1368980007361004 [DOI] [PubMed] [Google Scholar]
  56. Shi X, Han C, Li Chenyan, Mao J, Wang W, Xie X, Li Chenyang, Xu B, Meng T, Du J, Zhang S, Gao Z, Zhang X, Fan C, Shan Z, Teng W, 2015. Optimal and safe upper limits of iodine intake for early pregnancy in iodine-sufficient regions: a cross-sectional study of 7190 pregnant women in China. J. Clin. Endocrinol. Metab 100, 1630–1638. 10.1210/jc.2014-3704 [DOI] [PubMed] [Google Scholar]
  57. Stagnaro-Green A, Abalovich M, Alexander E, Azizi F, Mestman J, Negro R, Nixon A, Pearce EN, Soldin OP, Sullivan S, Wiersinga W, American Thyroid Association Taskforce on Thyroid Disease During Pregnancy and Postpartum, 2011. Guidelines of the American Thyroid Association for the diagnosis and management of thyroid disease during pregnancy and postpartum. Thyroid Off. J. Am. Thyroid Assoc 21, 1081–1125. 10.1089/thy.2011.0087 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Sun Q, Chen Y, Ye F, Liu J, Liu D, Ao B, Hui Q, Zhang Q, 2023. Association of hypothyroidism during pregnancy with preconception and early pregnancy exposure to ambient particulate matter. Environ. Sci. Pollut. Res. Int 30, 88084–88094. 10.1007/s11356-023-28683-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Textor J, van der Zander B, Gilthorpe MS, Liskiewicz M, Ellison GT, 2016. Robust causal inference using directed acyclic graphs: the R package “dagitty”. Int. J. Epidemiol 45, 1887–1894. 10.1093/ije/dyw341 [DOI] [PubMed] [Google Scholar]
  60. US EPA, O., 2014. National Ambient Air Quality Standards. URL https://www.epa.gov/criteria-air-pollutants/naaqs-table (accessed 2.21.25).
  61. Van Uytfanghe K, Ehrenkranz J, Halsall D, Hoff K, Loh TP, Spencer CA, Köhrle J, 2023. Thyroid stimulating hormone and thyroid hormones (triiodothyronine and thyroxine): an American Thyroid Association-Commissioned review of current clinical and laboratory status. Thyroid® 33, 1013–1028. 10.1089/thy.2023.0169 [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Wang X, Liu C, Zhang M, Han Y, Aase H, Villanger GD, Myhre O, van Donkelaar A, Martin RV, Baines EA, Chen R, Kan H, Xia Y, 2019. Evaluation of maternal exposure to PM2.5 and its components on maternal and neonatal thyroid function and birth weight: a cohort study. Thyroid® 29, 1147–1157. 10.1089/thy.2018.0780 [DOI] [PubMed] [Google Scholar]
  63. Wang X, Sun X, Zhang Y, Chen M, Dehli Villanger G, Aase H, Xia Y, 2020. Identifying a critical window of maternal metal exposure for maternal and neonatal thyroid function in China: A cohort study. Environ. Int 139, 105696. 10.1016/j.envint.2020.105696 [DOI] [PubMed] [Google Scholar]
  64. Zhang X, Huels A, Makuch R, Zhou A, Zheng T, Xia W, Gaskins A, Makuch J, Zhu Z, Zhu C, Qian Z, Xu S, Li Y, 2022. Association of exposure to ambient particulate matter with maternal thyroid function in early pregnancy. Environ. Res 214, 113942. 10.1016/j.envres.2022.113942 [DOI] [PubMed] [Google Scholar]
  65. Zhao Y, Cao Z, Li H, Su X, Yang Y, Liu C, Hua J, 2019. Air pollution exposure in association with maternal thyroid function during early pregnancy. J. Hazard. Mater 367, 188–193. 10.1016/j.jhazmat.2018.12.078 [DOI] [PubMed] [Google Scholar]
  66. Zoeller RT, Brown TR, Doan LL, Gore AC, Skakkebaek NE, Soto AM, Woodruff TJ, Vom Saal FS, 2012. Endocrine-disrupting chemicals and public health protection: a statement of principles from The Endocrine Society. Endocrinology 153, 4097–4110. 10.1210/en.2012-1422 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

1
NIHMS2073207-supplement-1.docx (1MB, docx)

Data Availability Statement

Restrictions apply to the availability of all data generated or analyzed during this study because they were used under license. The corresponding author will on request detail the restrictions and any conditions under which access to some data may be provided.

RESOURCES