The present study demonstrated a positive association between perchlorate and the risk of PTC and perchlorate can disturb the homeostasis of thyroid function, but did not find an association between perchlorate and NG and PTMC. In addition, no association was found between iodine and the risk of thyroid tumors and nodular goiter, and thyroid function. The association of endocrine-disrupting compounds and iodide inhibitors with thyroid diseases has received worldwide attention in recent years. However, few population-based studies have examined the relationship between iodination inhibitors and thyroid diseases. To our knowledge, the present study is the first population-based study to simultaneously examine the association of perchlorate exposure with three thyroid neoplasma types (NG, PTMC, and PTC).
Our results showed a positive association between perchlorate exposure and the risk of PTC (OR = 1.058, 95% CI: 1.009, 1.110) with non-linear dose–response relationship (P-non-linear < 0.05). This is consistent with another small population study conducted in China, in which the association of perchlorate exposure and PTC was found (OR = 2.27, 95% CI: 1.03–5.03) . The risk of perchlorate on PTC that we found was marginal and the effect of small sample size needs to be considered. In addition, humans are generally exposed to multiple endocrine disruptors in the external environment simultaneously , thus it is worth considering whether perchlorate act synergistically or antagonistically when combined with them on the thyroid. For example, Zhang et al., 2018  included both perchlorate and thiocyanate and obtained a higher OR for the risk association than our study (2.27 > 1.058), so the combined effect of other thyroid disruptors on thyroid tumors cannot be excluded. It is worth noting that the initial concentration of perchlorate associated with the PTC risk was between the median value and the P75 level in the controls from the present study, which provided a statistical reference for the human exposure threshold. Animals studies have shown that exposure to perchlorate increases the size and number of thyroid follicles, which leads to structural changes in the thyroid gland . However, previous studies have not determined whether perchlorate is involved in the initiating or the promoting mechanism in thyroid tumors. In the present study, we found a positive association between perchlorate and NG in univariate model, but this association disappeared in multifactorial model. It is possible that perchlorate is indeed not associated with NG, or the association may have been masked by some confounding factors not taken into account in our study. A report on perchlorate drinking water exposure test in rats from 2 weeks prior to cohabitation to lactation day 10 found that thyroid weight increased in female pups at 1.0 mg/kg/d ammonium perchlorate (AP) exposure compared to controls on day 22 of lactation . Moreover, increased thyroid weight and decreased colloid in female rats and male/female pups were observed at 30.0 mg/kg/d AP exposure, further follicular hypertrophy was induced in female rats, suggesting that perchlorate may impair thyroid cells and alter thyroid morphology with gender sensitivity . NG is also an irreversible pathological change that can be transformed from a diffuse goiter, thus the association of perchlorate with NG can be further investigated. In addition, no statistical association of perchlorate with the risk of PTMC was found either. Compared with PTC, PTMC is clinically asymptomatic, smaller in diameter (< 10 mm), and has a much lower mortality rate . Given that we found that perchlorate exposure was associated with the risk of PTC but not PTMC, perchlorate exposure likely influences tumor size and probably acts as a promoter of TC rather than an initiator.
Furthermore, BMI was found to be a protective factor for thyroid tumors and NG, while drinking was a risk factor in this study. Similar findings were reported in Mijović T et al., 2009  and 2010 to 2011 Korean National Health and Nutrition Examination Survey (KNHANES) . However, recent studies show that BMI is positively correlated with papillary and follicular carcinomas of the thyroid, and this may be related to elevate TSH induced by obesity, chronic inflammatory response and oxidative stress, of which elevated TSH is considered a risk factor for TC . Moreover, the study by Yeo et al. showed that smoking and drinking were negatively correlate with TC, which may be related to the lowering of TSH induced by smoking and drinking, as well as the anti-estrogenic effects . These above findings suggest that confounding factors such as BMI, smoking, and drinking may affect the development of thyroid tumors and have been controlled in the following multivariate analysis of this study.
The effects of perchlorate on the indicators of thyroid function were assessed in this study, and the results showed that perchlorate was negatively correlated with FT3 and positively correlated with TgAb, and only positively correlated with TSH when iodine at adequate levels. According to the National Health and Nutrition Examination Survey (NHANES) in the United States, perchlorate was positively associated with TSH only in women aged above 12 years, regardless of whether iodine was adequate or not; in contrast, perchlorate was negatively associated with TT4 only in iodine deficiency status, suggesting an association between perchlorate and hypothyroidism . However, perchlorate has been found to be positively associated with TSH in pregnant women only when co-exposed with thiocyanate and nitrate rather than independently . A subchronic toxicity test study in rats showed that AP exposure was negatively associated with TT3 and TT4 and positively associated with TSH, which occurred only at doses of up to 10 mg/kg/day of AP exposure . However, no significant changes in thyroid function were found in both trials from 9 healthy male volunteers for 2-week exposure at perchlorate dose 10 mg/d and 13 healthy volunteers for 6-month at exposure levels of 3 mg/d and 0.5 mg/d, and the small sample size for the above two studies may influence the statistical power. Nevertheless, effects of high-dose perchlorate exposure or lower doses of more prolonged perchlorate exposure on thyroid function haven’t been demonstrated in the general population till now [33, 34]. This indicates that perchlorate interference with TSH is also influenced by gender, exposure dose, and combined effects with other NIS inhibitors. We found the negative correlation of perchlorate and FT3 but not the positive correlation between perchlorate and TSH at the same iodine level, which may be related to sample size limitation or other confounding factors. In fact, it has been reported that TSH can stimulate the transcription and biosynthesis of NIS . The above results might suggest the following possible mechanism: perchlorate competes with iodine for NIS to reduce iodine uptake in the thyroid gland, thereby inducing a decrease in FT3, which can reflect TT3 or TT4 decrease when not accompanied by thyroid hormone binding protein decrease, and increasing the secretion of TSH under the regulation of a negative feedback mechanism, while TSH restores the iodide level in thyroid cells by upregulation of NIS. In addition, the association of thyroid hormones and thyroid autoimmune antibodies with the risk of thyroid tumors has been reported [35, 36]. It has been shown that TSH can activate the phospholipase C/protein kinase C (PLC-PKC) signaling pathway to stimulate the secretion of vascular endothelial growth factor (VEGF) and induce neoangiogenesis, thereby promoting the proliferation of TC cells , which suggests that elevated TSH due to perchlorate exposure may have a role in promoting TC.
Both TPOAb and TgAb are markers of thyroid autoimmunity and can assist in the prediction of thyroid disease, such as hypothyroidism and prognosis of hyperthyroidism treatment . Our study found a positive correlation between perchlorate and TgAb, suggesting that perchlorate may be associated with thyroid cell damage. However, since autoimmune thyroid disease usually requires a combination of positive TgAb and TPOAb for diagnosis, the effect of perchlorate on thyroid autoimmunity needs to be further verified. In particular, TgAb can also be used as a surveillance indicator for differentiated thyroid cancer (TDC) and has important implications for predicting the prognosis of TDC.
For iodine, we found no association with the risk of thyroid tumors and NG. Wang et al., 2014 pointed out that high iodine levels significantly correlate with the occurrence of benign and malignant thyroid tumors and the aggressiveness of PTC . However, the median urinary iodine concentration in their study was 331.33 and 466.23 μg/L for thyroid nodules and TC, among which the iodine excess accounted for 62.75 and 66.99%, respectively, and much higher than that in our study . In the present study, the median urinary iodine concentration was 130.19, 148.34 and 126.97 μg/L for NG, PTMC and PTC, respectively, while only 12.50 and 7.14% of patients with NG and TC (including PTMC and PTC) were at iodine excess levels (Supplementary Material, Table 4). Thus, the result of correlation between iodine and thyroid tumor from high iodine levels at iodine excess was no comparability with the present study. Several previous studies have demonstrated that the relationship between iodine and thyroid tumors is more evident at ultra-low and ultra-high intakes. For instance, experimental research revealed that ultra-low (1.0 × 10− 6 mol/L) and ultra-high dose (1.0 × 10− 3 mol/L) of iodine induced high expression of the SPANXA1 gene and promoted TC cell proliferation and apoptosis . Furthermore, meta-analyses of animal and human studies have shown that both chronic iodine insufficiency and excess were risk factors for TC, but iodine intake was not a significant independent factor in the development of thyroid disease . This is consistent with two epidemiological studies suggesting that high iodine intake may play a role in the tumor size and capsular invasion compared with the occurrence of PTC [42, 43].
Iodine plays an important role in the synthesis of thyroid hormones. However, no correlation between iodine levels and the thyroid function indicators was found in this study. This is similar to a previous cross-sectional study in China, which found no significant association between iodine levels and TSH, FT3, or FT4 . Similarly, a large population study of over 7000 participants (NHANES) found no correlation between iodine levels and TSH and TT4 . In contrast, studies based on complete food frequency surveys to assess iodine levels among young people residing in the mountains of the western United States have shown that iodine levels positively correlate with TSH . The great variation among the studies may reflect the physiological regulation of normal human thyroid, where thyroid hormones are not significantly affected unless the thyroid gland is overloaded . In addition, Wang et al., 2019 showed that both iodine deficiency and iodine excess are risk factors for autoimmune thyroid diseases . Although both TgAb and TPOAb levels in the present study increased with the increase in iodine nutrition level, their variance at different iodine nutrition levels was not statistically significant and the antibody levels were still within the normal range (Supplementary Material, Table 5). It has been suggested that iodine is more likely to induce thyroid autoimmunity through unmasking a cryptic epitope on thyroglobulin . The results of above study were only obtained from a single measurement. In fact, as individual iodine levels are closely related to the diurnal variation of dietary iodine content , it is even more important to observe fluctuations in thyroid hormones in the population by monitoring iodine nutrition levels over time.
The source of perchlorate exposure for the recruited population was analyzed in this study. The median perchlorate concentration in this study (11.27 μg/L) was notably higher than that of 6044 participants of the general population included in the NHANES (3.9 μg/L) ; however, the US study included not only multiregional and multiethnic populations but also adolescent population, suggesting that demographic and geographical variations may contribute to the differences in perchlorate exposure. Human perchlorate exposure has been related to the living environment and dietary intake. Studies of people living in fireworks production sites have shown that the mean concentration of perchlorate in urine is 16 times higher than that in the present study . Research has supported that food intake contributes 86% to perchlorate exposure . Although there have been no reports of perchlorate contamination in Shenzhen foods, limited studies have shown that perchlorate residues were found in vegetables, fruits, grains, and dairy products, especially in leafy vegetables (spinach) . Moreover, the study of perchlorate concentration in bottled water and groundwater from 15 cities in China showed that the highest concentration (54.4 μg/L) was detected in Hengyang, Hunan Province—a fireworks production site, while Shenzhen was at the lowest concentration (0.14 μg/L) , suggesting that drinking water may not be the primary source of perchlorate exposure of the recruited population. Thus, the main source of perchlorate exposure in Shenzhen residents was more likely to be food intake. Further investigation on perchlorate exposure for Shenzhen residents is needed in the future.
There are some strengths and limitations in this study. First, cases with three pathological types of thyroid tumor and corresponding controls were matched by gender and age and recruited based on strict inclusion and exclusion criteria. Second, the authenticity and accuracy of data were ensured under strict quality control measures. The confounding variables were adjusted for using a multifactorial model. This study was the first to compare the effects of perchlorate exposure on the three types of thyroid neoplasms. Thus, the findings provide important etiological evidence for thyroid tumor and nodular goiter in order to improve thyroid disease prevention in future. However, there are some limitations to this study. First, the sample size in different pathological types was small, especially in the PTMC group. Second, the levels of perchlorate and iodine in urine were based on a single measurement as it is difficult to reflect long-term human burden levels in a cross-sectional study. However, the present study found the association of perchlorate with PTC and first indicated that perchlorate has no statistical association with the risk of NG and PTMC. In future, population-based cohort studies with precisely monitored iodine intake through long-term follow-up in the exposed populations from perchlorate-contaminated areas should be conducted.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.