Background
Ambient air pollution contains known human carcinogens and is a problem worldwide, and has recently become a major public concern in Korea [
1]. A report from the Organization for Economic Cooperation and Development (OECD) documented that outdoor air pollution in South Korea was estimated at 359 premature deaths per million individuals in 2010. If this situation continues, it will rise to more than 1100 per million individuals by 2060 [
2]. Most premature deaths related to air pollution were linked to cardiovascular diseases, including stroke and myocardial infarction [
3,
4].
In recent years, emerging evidence has shown the harmful effects of ambient air pollution on thyroid hormone levels [
5‐
9]. Experimental findings have shown how their toxic effects on environmental chemicals affect thyroid hormone levels [
10,
11]. A more recent study has suggested that obese status mediates the association between ambient air pollution and thyroid function [
1]. Although the mechanisms underlying this relationship are not understood completely, several possible hypotheses have been proposed. Obese status, especially abdominal obesity, is linked to the thyroid functions related to oxidation of fatty acids, lipid metabolism, and so on [
12,
13]. Abdominal fat distribution, as measured by the subcutaneous and visceral fat area, plays a role in the high risk of diseases, including metabolic diseases, cardiovascular diseases, and hypertension [
14,
15].
It is essential to understand the role of regional abdominal fat traits to understand these associations and mechanisms more fully. Excessive visceral adipose tissue (VAT) may contribute to elevated oxidative stress or oxidative damage [
16,
17]. VAT secretes large amounts of cytokines that induce inflammation, including tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6), compared to subcutaneous fat [
18,
19]. In fact, a previous study suggested an increase in VAT as the best predictor of thyrotropin (TSH) in obesity [
20]. By contrast, subcutaneous fat may affect the change of thyroid hormone in specific tissue [
21]. The thyroid hormone receptor is also more activated in subcutaneous adipose tissue (SAT) than in VAT [
22].
In this context, we reasoned that abdominal fat may mediate the relationship between air pollution and thyroid hormone and that this may be reflected in differences according to adiposity traits. However, previous studies of the effects of obesity on the relationship between environmental pollutants and thyroid hormones have mostly used body mass index (BMI) as a proxy variable for obesity [
1,
23]. This indirect indicator of obesity cannot accurately distinguish certain depots of abdominal fat. Therefore, the precise estimation of fat mass using computed tomography (CT) is essential to examine these potential associations further. To our knowledge, no studies have evaluated this hypothesis. In this study, we assessed the relationship of exposure to air pollutants and thyroid hormone in Korean adults and whether these relationships are mediated by abdominal fat type.
Methods
Participants
The subjects included in our study were recruited at two health screening hospitals run by a university Hospital in Seoul from 2009 to 2015, (1) the Seoul National University Hospital Health Promotion Center (site A), and (2) the Seoul National University Hospital Healthcare System Gangnam Center (site B). Korean adults receive regular health examinations for the prevention or early diagnosis of illness. There is a large difference in the abdominal fat mass between both sexes, and the accumulation of visceral fat tissue in Korean adults is more pronounced in men than in women [
24,
25]. For this reason, mainly adult men were selected and the proportion of women was very low (
n = 2560 and 344 for men and women, respectively). In the analyses, when we classified the subjects as two groups by adiposity level (normal vs. high), the difference in the abdominal adiposity distribution between male and female made it difficult to apply identical cutoff points. Therefore, we included only adult men in this study.
A total of 2560 adults underwent the physical examinations, including abdominal CT, during the period noted above. We excluded 120 who met the following exclusion criteria from the final analysis: (1) those whose air pollution exposure concentration could not be estimated because of missing postal code information; (2) those with missing information for thyroid-related outcomes, such as serum TSH and free thyroxine (FT4) concentrations; (3) those whose these hormone levels could not be measured accurately because of thyroid surgery or medications that may affect thyroid hormones; and (4) those whose lifestyle data such as frequency of alcohol use, smoking status, and moderate activity were incomplete. A total of 2440 adults were used for the final statistical analysis.
The need for informed consent was waived by our institutional review board, because this was a retrospective study using de-identified data. The protocol for this study was approved by the institutional review boards of the Seoul National University Hospital and National Cancer Center.
Assessment of the exposure to air pollution
Atmospheric monitoring data to obtain 24-h concentrations were obtained from a total of 328 nationwide monitoring stations operated by the Ministry of the Environment of South Korea. The data included real-time measurements of various air pollutants, including particulate matter with an aerodynamic diameter of ≤ 10 μm (PM10), sulfur dioxide (SO2), nitrogen dioxide (NO2), and carbon monoxide (CO). Each participant’s residential postal code was extracted from the internal hospital database to estimate exposure levels corresponding to each individual. We checked the monitoring point nearest to each participant’s residence using the postal code information and calculated the annual average values of all air pollutants. The individual exposure concentration was estimated by matching the annual mean concentration based on the medical examination year at the monitoring site closest to the residence of each individual. Of the total monitoring sites, 302 monitoring subsets were finally used. The concentrations of all four air pollutants were interrelated, and the correlation levels ranged from 0.16 to 0.69 (data not shown).
Assessment of the abdominal adiposity
Obesity indicators, including several abdominal fat-related traits, were measured when individuals visited for health checkups. Waist circumference (WC) was measured with the subject wearing thin clothing. We classified the subjects into two groups according to the criteria for abdominal obesity in Asian men as follows: normal (WC < 90 cm) and obese (WC ≥ 90 cm). We measured abdominal adiposity using a CT scanner (Somatom Sensation 16 CT scanner, Siemens AG, Erlangen, Germany). The cross-sectional area of each adipose tissue site, such as VAT and SAT, was calculated using Rapidia software (version 2.8; Infinitt, Seoul, South Korea). The visceral fat area was estimated by drawing a boundary with the parietal peritoneum or transversalis fascia, except for the vertebrae and spinal muscles. The subcutaneous fat area was defined by deducting the visceral fat area from the total fat area, excluding the vertebrae and spinal muscles. We included the visceral-to-subcutaneous fat ratio (VSR). Participants were divided into two groups (normal adiposity and high adiposity) using a cutoff criterion of 100 cm2 for VAT and SAT. For the VSR, a cutoff criterion of 1.0 was used.
Assessment of thyroid function and other variables
We obtained data to evaluate the thyroid hormone level from standard blood tests. Thyroid hormone included serum TSH and FT4 concentrations. TSH and FT4 were measured using immunoradiometric assay kit and radioimmunoassay kits (RIAKEY, Shinjin Medics Inc. Seoul, Korea), respectively. The manufacturer reference ranges for TSH and FT4 were 0.3–5.0 μIU/mL and 0.7–1.8 ng/dL, respectively. To test the associations between exposure to air pollution and thyroid hormones, we controlled the analysis for potential confounding variables, including the site of recruitment, age, status of alcohol use (never, former, or current drinker), cigarette smoking (never, former, or current smoker), and physical activity (yes or no). Physical activity was assessed with the short form of the International Physical Activity Questionnaires (IPAQ), and was defined as engaging moderate (i.e., carrying light loads, bicycling at a regular pace, or doubles tennis) or vigorous intensity activities (i.e., heavy lifting, digging, aerobics, or fast bicycling) for at least 10 min a week. Lifestyle behaviors were assessed using questionnaires on the day of the medical examination.
Statistical analysis
The distribution of TSH and FT4 concentrations were checked prior to analyses of association. Because both TSH and FT4 followed a nonnormal distribution, we used the best transformation approaches, and the square root and natural logarithm transformations were applied to the TSH and FT4 levels, respectively. To test the difference in participant characteristics between groups according to each abdominal fat level, a t test and a Chi-square test were performed according to the type of each variable (i.e., continuous or categorical variables). We considered multiple linear regression analysis to assess the relationship between air pollution exposure and thyroid hormone. In our data, the correlations between all four air pollutants were significant, even though the value of variance inflation factor for multicollinearity was less than 10. Therefore, we finally did not include other air pollutants as confounders in our model. In adjusted model, the results of the association analyses were controlled for the recruitment center, age, smoking status, status of alcohol use, and physical activity [Y = β0 + β1 recruitment center + β2 age + β3 smoking status + β4 status of alcohol use + β5 physical activity + β6 each air pollutant + e]. We also conducted a stratified analysis for the two subgroups according to abdominal fat traits, including WC, VAT, SAT, and VSR. Beta coefficients (β) and standard errors (SEs) were converted to the interquartile range (IQR) (9.1 μg/m3 for PM10, 13.8 ppb for NO2, 1.5 ppb for SO2, and 0.2 ppm for CO). All statistical analyses were performed using SAS (version 9.4; SAS Institute, Cary, NC, USA).
Discussion
We assessed the associations of exposure to air pollution and thyroid hormones, including TSH and FT4 concentration, in a Korean population and whether these relations are modified by VAT and SAT areas. We found that persistent exposure to air pollution was significantly related to increased TSH and decreased FT4 concentrations. In the stratified analyses according to abdominal adiposity traits, PM10, NO2, and CO exposure showed a stronger association with TSH and FT4 concentrations in the subjects with high adiposity. We identified the significant effects of PM10 and CO exposure on thyroid function in all adiposity groups, as measured by WC, VAT and SAT areas, and VSR. Among the abdominal fat-related traits, the impact of air pollution exposure on the thyroid hormone level was the strongest for VSR in the subjects with high adiposity. Our findings suggest that people with high adiposity may be more likely to develop disturbances in thyroid hormones caused by air pollution, especially PM10 and CO.
Epidemiological researches have shown associations between exposure to air pollutants and thyroid hormone [
5,
6,
8,
9,
26], but most of these studies have shown limited results in pregnant women [
5,
6,
9]. A study of children across southern California found that prenatal PM exposure was significantly associated with increased newborn thyroxine concentration [
6]. Two epidemiological studies have reported an inverse relationship between PM
2.5 and NO
2 and FT
4 level during pregnancy, although the relationship with TSH level was not significant [
5,
9]. A recent cross-sectional, population-based study in China reported that a 10 μg/m
3 increase in PM
2.5 exposure decreased 0.12 pmol/L in FT
4 concentration [
26]. Another study reported that exposure to each of the air pollutants including PM
2.5, PM
10, NO
2, and SO
2 was significantly and linearly associated with the risk of thyroid nodules in Chinese adults [
8]. We found that FT
4 concentration decreased and TSH concentration increased with increasing exposure to air pollutants. Our results were robust even after stratifying according to abdominal adiposity traits. However, these associations require confirmation by prospective cohort studies.
Obesity, especially abdominal adiposity, is closely related to thyroid hormone levels [
13]. A recent study of the association between thyroid function and distribution of abdominal adiposity suggests that SAT is independently and positively associated with elevated triiodothyronine (FT
3) concentration in a China population (male: OR = 0.183, 95% CI = 0.094 to 0.272, and female: OR = 0.089, 95% CI = 0.007–0.171) [
22]. Another study also identified the positive association between abdominal subcutaneous fat and TSH or FT
3 level [
27]. In contrast, a cross-sectional study reported that the excessive adipose tissue, particularly VAT, plays the crucial role in thyroid dysfunction [
20]. In our study, both VAT and SAT are associated with a decrease of FT
4 concentration. These different results may be partially explained by several factors, such as study design and population.
With regard to modifying the effect of obese status on the association between exposure to air pollutants and thyroid function, our research team in 2020 provided the first clue that air pollution, including NO
2 and CO, is strongly linked to thyroid function in people with overweight or obesity [
1]. However, it was only a result of showing the modifying effects of overall obesity defined by BMI. To date, to our knowledge, no studies have evaluated the relationship between exposure to ambient air pollution and thyroid functions, considering the abdominal fat distribution in the general population. We found for the first time that the relations between air pollution exposure and thyroid functions were more strongly associated in the subjects with high adiposity, as defined by high VAT, SAT, and VSR areas.
The pathophysiological mechanism linking air pollution, abdominal fat level, and thyroid function is unclear. However, possible hypotheses involving the adipocyte have been proposed. The oxidative stress reaction is the most plausible mechanism. Thyroid hormones regulate the antioxidant system via the promotion of oxidative stress and reactive oxygen species [
28,
29]. Both type of hypothyroidism and hyperthyroidism have been shown to be related to oxidative damage to cell structures [
28]. The toxic effect of air pollutant exposure also leads to the production of other reactive oxygen species and oxidative injury [
30]. Similarly, abdominal fat, particularly the visceral fat accumulation, is responsible for activation of oxidative stress or oxidative injury [
18]. In addition, inflammatory response can be a plausible mechanism. Systemic inflammation through cytokine changes induces thyroid dysfunction [
28]. Obesity is related to an increased production of cytokines for the pro-inflammatory response (e.g., leptin, TNF-α, and IL-6) and a decreased secretion of adipokines that protect inflammatory response (e.g., adiponectin and IL-10). Such inflammatory cytokines are mainly secreted from visceral adipocytes [
18,
19]. Besides, exposure to air pollution is implicated in systemic inflammation [
31]. Therefore, our results stratified by VAT or VSR may be interpreted by the synergistic effects of exposure to air pollution and abdominal fat on the oxidative stress or inflammation. However, additional studies are needed to better understand the clear mechanisms.
The strength of our study is the observation, for the first time, that the relationship between air pollution exposure and thyroid hormone is stronger in people with high adiposity, as indicated by WC, VAT and SAT areas, and VSR. Our study has some limitations. First, this study is a cross-sectional type, which cannot be used to identify a causal relation between exposure to air pollutant, abdominal fat distribution, and thyroid hormone. Second, our study did not conduct analysis of possible sex differences in abdominal adiposity distribution because of insufficient data on women. Health outcomes caused by air pollution or obesity may differ between two sexes due to differences in patterns of physical activity, sex hormone levels, occupational features, and the way of life, as well as differences in adiposity distribution. Third, to assess individual air pollution exposure levels, we considered the exposure assessment at the community level using residential postal code districts instead of a more accurate exposure estimate for each participant, because there is not enough relevant data in this population. This method of data collection did not allow us to examine the potential roles of multiple factors, such as occupational characteristics or exposure to air pollution in indoor spaces, residence history, nearness to road traffic conditions, local topography, and change in climate. In this context, each subject's level of exposure to air pollution may have been overestimated or underestimated, which may have contributed to potential misclassification of air pollution exposure. Fourth, it is not possible to determine the temporal consistency between exposure and effects for association analysis. In addition, iodine nutrition, which is known to affect thyroid hormones, could not be included as a confounder due to lack of relevant information. Finally, further research on the short-term effect via lag analysis or association with air pollution in FT3 is needed.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.