Abstract
Background:
The association between exposure to air pollution and papillary thyroid carcinoma (PTC) is unknown. We sought to estimate the relationship between long-term exposure to fine (diameter <2.5 μm) particulate matter (PM2.5) component of air pollution and risk of papillary thyroid cancer (PTC).
Methods:
Adult (age≥18) patients with newly diagnosed PTC between January 1, 2013 and December 31, 2016 across a single health system were identified using electronic medical records. Data from 1,990 patients with PTC were compared to 3,980 age and gender matched controls without any evidence of thyroid disease. Cumulative PM2.5 exposure was estimated by incorporating patients’ residential zip codes into a deep learning neural networks model, which utilizes both meteorological and satellite-based measurements. Conditional logistic regression was performed to assess for association between PTC and increasing PM2.5 concentrations over one, two, and three years of cumulative exposure preceding PTC diagnosis.
Results:
An increased odds of developing PTC was associated with a 5 μg/m3 increase of PM2.5 concentrations over two years (aOR=1.18, 95%CI: 1.00–1.40) and three years (aOR=1.23, 95%CI: 1.05–1.44) of exposure. This risk differed by smoking status (pinteraction=0.04). Among current smokers (n=623), the risk of developing PTC was highest (aOR=1.35, 95%CI: 1.12–1.63).
Conclusions:
Increasing concentration of PM2.5 in air pollution is significantly associated with the incidence of PTC with two and three years of exposure. Our novel findings provide additional insight into the potential associations between risk factors and PTC, and warrant further investigation, specifically in areas with high levels of air pollution both nationally and internationally.
Article Summary
Rising concentrations of fine particulate matter in air pollution are significantly associated with the incidence of PTC with two and three years of exposure. Our novel findings provide additional insight into the potential pathophysiology of PTC and warrant further investigation both nationally and internationally.
Introduction
Since the 1980s, there has been a gradual rise in the incidence of thyroid cancer that is not solely attributable to increased detection.1 While known risk factors for thyroid cancer include female sex, history of ionizing radiation exposure, and family history, more recent studies have focused on increased risk due to patient-related factors, such as obesity and smoking status, and environmental factors.2–8 Although chemical pollutants including phthalates, bisphenol A, polychlorinated biphenyls, perfluorinated compounds, and perchlorates, impact thyroid function, their association with thyroid cancer is less clear.9 Air pollution, however, as an environmental risk factor and potential contributor to the rise of thyroid cancer has been suggested by a population-based study from Shanghai, which demonstrated that air pollution from waste gas emissions was significantly correlated with incident thyroid cancer of all types.10 To our knowledge, the relationship between air pollution and papillary thyroid carcinoma has not been explored in the United States.
In 2013, the International Agency for Research on Cancer Working Group classified air pollution as carcinogenic.11 Air pollution is comprised of a myriad of substances, including a category of ambient particulate matter (PM) with an aerodynamic diameter of ≤2.5 μm (PM2.5). Although the Environmental Protection Agency (EPA) breaks down components of PM10 into coarse or ‘primary’ particles such as dust and carbon combustion, it defines the composition of PM2.5 as composed of ‘secondary’ particles which are formed in the atmosphere from the chemical reactions of gaseous emissions, and cannot be identified further.12 These particles can be inhaled and enter the bloodstream. PM2.5 is often used as an indicator of pollution, and an annual average upper limit concentration of 10 μg/m3 PM2.5 has been provided as a guideline value by the World Health Organization (WHO).13 High levels of PM2.5 are associated with well-documented impacts on cardiovascular diseases, cognitive decline, and overall mortality.14 Although exposure to PM2.5 has been linked to lung cancer, the association between exposure to fine particulate air pollution and PTC is unknown.15 Therefore, we sought to estimate the relationship between long-term exposure to PM2.5 in air pollution and risk of PTC for patients living in the United States.
Methods
Setting and participants
Under IRB approval, adult (age ≥18) patients with newly diagnosed PTC between January 1, 2013 and December 31, 2016 across the Johns Hopkins Health System were identified using a large volume electronic medical record data extract. Data from 1,990 patients with PTC were compared to 3,980 age and gender matched healthy controls without any evidence of thyroid disease or pathology (Figure 1). Patient demographics, body mass index (BMI), smoking status, comorbidities and diagnosis date of PTC were extracted from medical records. Patients were only included if cumulative PM2.5 exposure preceded the diagnosis of PTC. In order to minimize misclassification of control cases, patients were excluded if they had a past medical history of thyroid cancer, thyroid disorders (ICD10 E07.9) or thyroid nodules (ICD10 E04.1) as an undiagnosed thyroid cancer may have been present. They were also excluded if their medical records were missing information regarding potential risk factors or confounders (BMI, smoking status, alcohol consumption, medium household income), or if they were residing outside of the United States (U.S.). The US Census Bureau’s American Community Survey was used to determine the medium household income by patient’s residence zip code tabulation area (ZCTA).16 The medium ZCTA household income was inflation adjusted to 2016 US dollars.
Figure 1.
Flow Chart for Cohort Selection Including Cases of Papillary Thyroid Cancer (PTC) and Age and Gender Matched Controls.
Exposure assessment
Cumulative PM2.5 exposure was estimated by a novel technique incorporating patients’ residential zip codes into a deep learning neural networks model, which utilizes both meteorological and satellite-based measurements. Air pollution monitoring data were acquired from the U.S. Environmental Protection Agency (EPA) Air Quality System (AQS) (1,928 monitors for PM2.5). The neural network was fit with monitoring data from the EPA AQS. Daily PM2.5 concentrations were then predicted from 2000 to 2016 nationwide. Models were created because most agency monitors that take direct measurements are located in cities or polluted areas (such as power plants). However, all subjects in this cohort do not live in locations with agency monitors. For example, some participants may live in highly polluted areas (i.e. close proximity to a highway), and others may live in a rural area. Using models, we can adjust the geographic information (population density, sea level, meteorological data) and estimate the exposure level in locations without agency air pollution monitors.
Cross-validation indicated that the models had a high predictive accuracy across the entire study area. The coefficient of determination (R2) for PM2.5 was 0.86 with a variation between 0.71 and 0.93; the mean square error between the measurements and predicted values for PM2.5 was 1.50 μg/m3. A variety of exposure metrics were created as appropriate to examine different potential important periods of exposure, including 12-, 24-, and 36- month mean PM2.5 concentration before diagnosis date. This model has been validated in China and used in several South Korean studies, with cross-validation results indicating that the model may only underestimate monthly-average PM2.5 levels when observed PM2.5 concentrations are very high.17–19 It has additionally been utilized in a study in the United States evaluating chronic rhinosinusitis.20 Thus, it is more accurate than using EPA monitors alone.
Statistical analysis
Descriptive analyses were performed using mean (standard deviation) or frequency (percentage). Values were calculated using the chi-square test for categorical variables and the Mann-Whitney U-test for continuous variables. Conditional logistic regression was performed to assess for association between PTC and increasing average PM2.5 concentrations over 12-months, 24-months, and 36-months of cumulative exposure preceding PTC diagnosis and reported as adjusted odds ratios (ORs) with associated 95% CIs. Several models were fit to adjust for varying confounders. Model 1 was adjusted for age, gender, and race. Model 2 was further adjusted for BMI, current alcohol use, medium household income and current smoking status. Finally, model 3 was further adjusted for existing hypertension, diabetes, chronic obstructive pulmonary disease (COPD), and asthma. Additional analyses were performed to test for effect modification by patient characteristics, including race (white, African American, Hispanic/Latino, and other), BMI (underweight, normal, overweight, and obese), current smoking status (never, current, and former smoking), hypertension, diabetes, COPD, and asthma. For each potential effect modifier, we evaluated effect modification by likelihood ratio tests comparing models that included an interaction term between air pollution exposure and the effect modifier versus models without the interaction term. Stratum-specific odds ratios were obtained from the same interaction model by using the appropriate coefficients and variance-covariance matrix.
To evaluate nonlinear dose-response relationships between PM2.5 exposure and risk of PTC, we modeled PM2.5 air pollution exposure variables using restricted cubic splines with knots at the fifth, 10th, 50th, and 90th percentiles of the distribution of PM2.5 air pollution concentrations. Statistical analyses were conducted using STATA (version 16.0; Stata Corporation) and R (version 4.1; R Development Core Team).
Results
Characteristics of Study Population
A total of 5,970 patients were analyzed, including 1,990 with incident PTC and 3,980 age and gender matched controls. The majority of cases resided in the northeastern United States, followed by the South and Midwest (Figure 2). The mean age of both PTC patients and controls was 51 ±17.5 (Table 1). The majority of patients in the PTC and control groups were female (75.5% versus 73.5%, respectively), white (75% versus 62.5%), and had a BMI ≥25 (65.4% versus 60.1%). In the PTC group, the most prevalent comorbidities were BMI ≥25 (65.4%) followed by hypertension (15.6%) and diabetes (5.5%), whereas in the controls, the most prevalent comorbidities were BMI ≥25 (60.1%) followed by hypertension (30.5%), and asthma (12.8%). The matched controls were more likely to have hypertension, diabetes, COPD, asthma, nasal polyps, and environmental allergies compared to those with PTC. Average 12- and 24-month PM2.5 concentrations prior to diagnosis were lower in the PTC group compared to matched controls, while 36-month concentrations were similar between the two groups (Table 1). PTC patients were noted to have a higher medium household income compared to controls.
Figure 2.
Geographic Distribution of the Exposure Group (n= 5,970).
Table 1.
Characteristics of Patients Stratified by Cases of Papillary Thyroid Cancer (PTC) and Matched Controls.
| Characteristics | Controls (n = 3,980) | Papillary Thyroid Cancer (n = 1,990) | p - Value |
|---|---|---|---|
| Age (year) | 51.04 (17.50) | 51.16 (14.95) | 0.81 |
| Male sex | 1066 (26.5) | 492 (24.5) | 0.09 |
| Race | <0.001 | ||
| White | 2513 (62.5) | 1507 (75.0) | |
| African American | 1125 (28.0) | 189 (9.4) | |
| Hispanic /Latino ethnicity | 134 (3.3) | 141 (7.0) | |
| Other | 248 (6.2) | 173 (8.6) | |
| 12-month PM2.5 average (μg/m3) | 10.16 (1.84) | 10.01 (1.77) | 0.002 |
| 24-month PM2.5 average (μg/m3) | 10.41 (1.93) | 10.32 (1.89) | 0.07 |
| 36-month PM2.5 average (μg/m3) | 10.68 (2.02) | 10.62 (1.97) | 0.26 |
| Body mass index, kg/m2 | <0.001 | ||
| Underweight (<18.5) | 133 (3.3) | 22 (1.1) | |
| Normal weight (18.5 ≤ BMI < 25) | 1471 (36.6) | 672 (33.4) | |
| Overweight (25 ≤ BMI < 30) | 1173 (29.2) | 674 (33.5) | |
| Obese (≥ 30) | 1243 (30.9) | 642 (31.9) | |
| Current smoking status (%) | <0.001 | ||
| Never smoker | 2480 (61.7) | 1428 (71.0) | |
| Current smoker | 394 (9.8) | 106 (5.3) | |
| Former smoker | 1146 (28.5) | 476 (23.7) | |
| Current alcohol consumption | 1618 (40.2) | 704 (35.0) | <0.001 |
| Medium household income (US dollars) | 74,544.74 (31,201.17) | 90,864.95 (35,290.20) | <0.001 |
| Hypertension | 1226 (30.5) | 313 (15.6) | <0.001 |
| Diabetes mellitus | 425 (10.6) | 111 (5.5) | <0.001 |
| COPD | 119 (3.0) | 12 (0.6) | <0.001 |
| Asthma | 515 (12.8) | 78 (3.9) | <0.001 |
| Nasal Polyps | 78 (1.9) | 0 (0.0) | <0.001 |
| Environmental allergy | 109 (2.7) | 16 (0.8) | <0.001 |
Values are mean (SD) or n (%). Values were calculated using chi-square test for categorical variables and the Mann-Whitney U-test for continuous variables. COPD = Chronic obstructive pulmonary disease.
Association between PM2.5 and PTC
After adjusting for confounders, conditional multivariable modeling demonstrated that a 5 μg/m3 increase of PM2.5 concentrations over 12-months of exposure was not associated with incident PTC (Table 2). However, the odds of developing PTC after a 5-μg/m3 increase in 24-month PM2.5 exposure was 18% (aOR=1.18, 95%CI: 1.00, 1.40) and increased to 23% after 36-months of exposure (aOR=1.23, 95%CI: 1.05, 1.44). Thus the odds of developing PTC increased with longer duration of PM2.5 exposure. Furthermore, fully adjusted spline regression analyses illustrated a rising risk of PTC with increasing levels of PM2.5 concentrations (Figure 3).
Table 2.
Association (Odds Ratios) between Long-Term Fine Particulate Matter Exposure and Incident Papillary Thyroid Cancer (PTC) Utilizing Adjusted Conditional Logistic Regression Analyses.
| Air Pollution | Model 1a OR (95% CI) |
Model 2b OR (95% CI) |
Model 3c OR (95% CI) |
|---|---|---|---|
| PM2.5 1 year | 1.04 (0.89, 1.22) | 1.06 (0.90, 1.26) | 1.11 (0.94, 1.32) |
| PM2.5 2 years | 1.15 (0.98, 1.33) | 1.14 (0.97, 1.34) | 1.18 (1.00, 1.40)* |
| PM2.5 3 years | 1.20 (1.04, 1.40)* | 1.20 (1.02, 1.40)* | 1.23 (1.05, 1.44)* |
Model 1, adjusted for age, gender, race.
Model 2, additionally adjusted for body-mass index, current alcohol consumption status, current smoking status, medium household income.
Model 3, additionally adjusted for the medical history of hypertension, diabetes, chronic obstructive pulmonary disease, and asthma.
Figure 3.
Risk of Incident Papillary Thyroid Cancer (PTC) by the Level of Exposure to PM2.5 Concentration in Each Exposure Period and Percentage of Population Exposed.
A: 1-year exposure period; B: 2-year exposure period; C: 3-year exposure period. The dose-response curve was calculated using restricted cubic splines with knots at the 10th, 50th, and 90th percentiles of the distribution of 1-, 2-, and 3-year PM2.5 concentrations. The reference exposure level was set at the 10th percentile of the distribution of PM2.5 concentrations (7.70 μg/m3 for 1-year PM2.5 concentration, 7.88 μg/m3 for 2-year PM2.5 concentration, 8.06 μg/m3 for 3-year PM2.5 concentration,). Odds ratios were adjusted for age, gender, race, body-mass index, current alcohol consumption status, current smoking status, medical history of hypertension, diabetes, chronic obstructive pulmonary disease, and asthma.
An exploratory subgroup analysis demonstrated that PTC risk differed by smoking status (pinteraction=0.04) and diabetes (pinteraction=0.05). At 36 months of exposure to rising PM2.5 concentrations, the odds of developing PTC was 35% higher in current smokers compared to never-smokers [aOR=1.35 (95% CI: 1.12–1.63)] (Table 3). The odds of developing PTC was 27% higher in non-diabetics compared to diabetics [aOR=1.27 (95% CI: 1.08, 1.49)]. There was no significant difference in risk by race, BMI, hypertension, COPD, or asthma.
Table 3.
Risk of Incident Papillary Thyroid Carcinoma Associated with a 5 μg/m3 Increase in 36-month PM2.5 Concentrations by Race, BMI, Smoking Status, and Co-morbidities.
| Variables | No. of Participants | Odds Ratio | P value for interaction |
|---|---|---|---|
| Race | 0.50 | ||
| White | 3,995 | 1.21 (1.01, 1.45) | |
| African American | 1,309 | 1.39 (0.82, 2.33) | |
| Hispanic/Latino ethnicity | 274 | 1.73 (0.87, 3.43) | |
| Others | 392 | 1.03 (0.59, 1.80) | |
| BMI | |||
| Underweight | 172 | 3.39 (0.92, 12.56) | 0.54 |
| Normal weight | 2,449 | 1.32 (1.03, 1.71) | |
| Overweight | 2,178 | 1.21 (0.93, 1.58) | |
| Obese | 2,272 | 1.07 (0.80, 1.43) | |
| Smoking | |||
| Never | 4,540 | 0.57 (0.30, 1.11) | 0.04 |
| Current | 623 | 1.35 (1.12, 1.63) | |
| Former | 1,908 | 1.12 (0.83, 1.51) | |
| Hypertension | |||
| No | 5,137 | 1.24 (1.04, 1.47) | 0.93 |
| Yes | 1,934 | 1.21 (0.82, 1.78) | |
| Diabetes | |||
| No | 6,438 | 1.27 (1.08, 1.49) | 0.05 |
| Yes | 633 | 0.67 (0.36, 1.27) | |
| COPD | |||
| No | 6,913 | 1.24 (1.06, 1.46) | 0.12 |
| Yes | 158 | 0.28 (0.04, 1.85) | |
| Asthma | |||
| No | 6,355 | 1.23 (1.05, 1.45) | 0.94 |
| Yes | 716 | 1.20 (0.59, 2.42) |
Model adjusted for age, gender, race, body-mass index, current alcohol consumption status, current smoking status, medium household income, medical history of hypertension, diabetes, chronic obstructive pulmonary disease, and asthma. BMI= body mass index; COPD= chronic obstructive pulmonary disease
Discussion
This large study, to our knowledge, is the first to explore the relationship between ambient air pollution and PTC, demonstrating that both rising concentrations of PM2.5 and duration of exposure to elevated concentrations for as little as two years increases the risk for PTC. The association between rising air pollution concentrations and PTC becomes more pronounced after extensive adjustment of confounders. Current smoking status conferred an even higher risk in patients exposed to rising PM2.5 concentrations.
In contrast to prior literature, this study shows that smoking in an environment of rising PM2.5 exposure confers the highest risk for PTC. Smoking is a well-known risk factor for several cancers, including those of the lung, bladder, and pancreas. However, it has not been shown to be a consistent risk factor in thyroid carcinoma. A meta-analysis of 25 case-control and 6 cohort studies examining the relationship between smoking and thyroid carcinoma demonstrated that smoking may in fact be protective, with the risk of PTC being 26% lower in current smokers compared to never smokers.21 Cigarette smoke, however, is comprised of a plethora of chemicals. It may be that certain chemicals have a protective effect on the development of thyroid cancer, while others induce carcinogenesis. The effect of these latter chemicals may only be appreciated or “outweigh” the protective chemicals in the presence of another influence, such as air pollution. However, additional studies need to be performed.
The proportion of the world exposed to air pollution levels above the recommended WHO limit of 10 μg/m3 fell from 94.2% in 2010 to 90% in 2016.22 This decrease is in large part due to significant drops in air pollution levels in North America and Europe achieved through environmental policy implementation. In the U.S., two measures successfully decreased ambient pollution levels over time: the Clean Air Act in 1963 and the Convention on Long-range Transboundary Air Pollution protocols of the 1980s.23 Presently, the majority of the world’s air pollution burden falls on South, East, and Southeast Asia. PM2.5 levels in China and India alone increased by 52.7% and 69.8%, respectively, since the 1960s and now account for 55% of global deaths due to PM2.5.24 These trends raise concerns in regions where the health burden due to air pollution is high but access to health care is limited, such as South Asia.
While the current study shows an association between fine particulate matter air pollution and PTC, it does not demonstrate causation or the underlying mechanism for this association. In the multistep pathway of thyroid carcinogenesis, alteration of growth factors causes neoplasia. Subsequent mutations in proto-oncogenes (BRAF, RET/PTC, ras) and other factors result in invasive properties.25 Carcinogens and other risk factors induce genomic instability which can propel tumorigenesis forward. Therefore, it may be plausible that air pollution induces genomic instability. Various air pollutants have also been linked to other diseases of systemic inflammation.26, 27 For example, inhaled polycyclic aromatic hydrocarbons (PAHs)- a group of carcinogens found in air pollution- are absorbed by tissues where they are transformed into reactive intermediates that can bind to DNA and cause mutations.28 It is by this mechanism that PAH exposure can initiate the development of lung cancer. Air pollution, thus, may not directly cause cancer but might create an environment more favorable to its progression.
There are several limitations in this study in addition to those inherent in large electronic medical record extractions including coding errors. Geographic models were used to deduce ambient pollution levels, and, although these models are usually validated by environmental measurements, it’s possible they do not correlate at either the environmental or personal level. At the personal level, for example, air pollution models do not account for occupational exposure nor do they differentiate between indoor, outdoor, or ingested pollutant exposure. It is not uncommon for indoor pollution to exceed outdoor pollution. That being said, the substantial size and case-control design diminish the effects of individual exposures. Misclassification is another potential limitation, as it is possible that people were diagnosed with PTC while residing in one location and then moved to a separate location. The only way to precisely track exposure and travel would be to apply personal tracking monitors. Alternatively, patients could be surveyed and asked if they have moved within the time frame of the study. Unfortunately, the former option would be a nearly impossible endeavor and the latter is not available in our electronic medical system (Epic). Moreover, a survey would have drastically extended data collection time and limited our sample size. Other risk factors of thyroid cancer could not be incorporated such as family history of thyroid cancer and history of childhood ionizing radiation. In a 2016 pooled analysis of 12 studies by Veiga et al, the relative risk of PTC after childhood exposure to external beam radiation ranged from 2.4 to 22.4 depending on the dose of radiation.29 This relative risk range translates to an absolute risk of 1.8% to 27.8% compared to 1.3% in the general population. Positive family history has been associated with a 3-fold to 6-fold increased risk of PTC depending on the affected family member.30 These risks range widely, and are dependent on variables (e.g. total dose of radiation, specific family member affected) that could not be reliable extracted through a large electronic medical record search. Differentiation and stage of cancer were also not included. Similarly, PTC variant information and tumor size could not be captured from available patient records. It is never clear how long cancers have been present until they are diagnosed and, thus, the PTCs in this study may have been preexisting. However, while fine ambient pollution may not be causing PTC, it may be making it more clinically evident. The inclusion of tumor size in this study would have helped distinguish the difference. For example, if the majority of cancers in the exposure group were microscopic-PTC, the increased incidence could be due to a confounder: lung disease or symptoms leading to incidental thyroid findings on computed tomography of the chest. Larger tumors are more likely to be true positives as they are typically palpable at presentation. Most patients in this study resided in the East Coast, with some distributed throughout the U.S. limiting generalizability to the entire U.S. and also globally. Finally, while this study demonstrates an association, it does not demonstrate causality.
The strengths of this study include its large sample size and cross-validated modeling of air pollution data with high predictive accuracy across the entire study area. The air pollution modeling data utilized in this study is more accurate than using EPA monitors alone. The dataset included comprehensive clinical and demographic information, minimizing recall bias and confounding. Geographically, patients were widely distributed throughout the U.S. with some predilection towards the East Coast.
The findings in this study have not been previously described. Air pollution is a modifiable risk factor for cancer, and a growing body of parallel literature will hopefully have an impact on environmental policy. There are regions of the U.S. where air pollution monitoring systems are not in place, and, therefore, cannot be as accurately studied for health outcomes. These regions, both locally and globally, represent a void of environmental health knowledge that needs to be addressed in future studies as well.
Conclusion
Rising concentrations of fine particulate matter in air pollution are significantly associated with the incidence of PTC with two and three years of exposure. Our novel findings provide additional insight into the potential associations between risk factors and PTC and warrant further investigation both nationally and internationally.
Acknowledgments
Funding/Financial Support: K23AG053429 (PI: Mathur)
R01AI143731 (PI: Ramanathan)
K24AI144954, R01DK111233 (PI: Segev)
R01DK120518, R01AG055781, and R01DK114074 (PI: McAdams-DeMarco)
U01ES026721, R01DE0263031, R01CA206155 (PI: Biswal).
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.
Oral Podium Presentation at the 41st Annual Meeting of the American Association of Endocrine Surgeons, Virtual, April 25–27, 2021
Ethical approval: The study was approved by the institutional review board of the Johns Hopkins University School of Medicine.
COI/Disclosures: All authors have reported that they have no relationships relevant to the contents of this paper to disclose. The authors declare they have no actual or potential competing financial interests.
References
- 1.Surveillance E, and End Results Program Cancer Stat Facts: Thyroid Cancer [Available from: https://seer.cancer.gov/statfacts/html/thyro.html.
- 2.An SY KS, Oh DJ, Min C, Sim S, Choi HG. Obesity is positively related and tobacco smoking and alcohol consumption are negatively related to an increased risk of thyroid cancer. Sci Rep. 2020;10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kabat GC, Kim MY, Wactawski-Wende J, Rohan TE. Smoking and alcohol consumption in relation to risk of thyroid cancer in postmenopausal women. Cancer Epidemiol. 2012;36(4):335–40. [DOI] [PubMed] [Google Scholar]
- 4.Kitahara CM, Linet MS, Beane Freeman LE, Check DP, Church TR, Park Y, et al. Cigarette smoking, alcohol intake, and thyroid cancer risk: a pooled analysis of five prospective studies in the United States. Cancer Causes Control. 2012;23(10):1615–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kitahara CM, Platz EA, Freeman LE, Hsing AW, Linet MS, Park Y, et al. Obesity and thyroid cancer risk among U.S. men and women: a pooled analysis of five prospective studies. Cancer Epidemiol Biomarkers Prev. 2011;20(3):464–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Lee JH, Youn S, Jung S, Kim K, Chai YJ, Chung YS, et al. A national database analysis for factors associated with thyroid cancer occurrence. Sci Rep. 2020;10(1):17791. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Schmid D, Ricci C, Behrens G, Leitzmann MF. Adiposity and risk of thyroid cancer: a systematic review and meta-analysis. Obes Rev. 2015;16(12):1042–54. [DOI] [PubMed] [Google Scholar]
- 8.Zhao ZG, Guo XG, Ba CX, Wang W, Yang YY, Wang J, et al. Overweight, obesity and thyroid cancer risk: a meta-analysis of cohort studies. J Int Med Res. 2012;40(6):2041–50. [DOI] [PubMed] [Google Scholar]
- 9.Fiore M, Oliveri Conti G, Caltabiano R, Buffone A, Zuccarello P, Cormaci L, et al. Role of Emerging Environmental Risk Factors in Thyroid Cancer: A Brief Review. Int J Environ Res Public Health. 2019;16(7). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Cong X Air pollution from industrial waste gas emissions is associated with cancer incidences in Shanghai, China. Environ Sci Pollut Res. 2018;25:13067–78. [DOI] [PubMed] [Google Scholar]
- 11.Loomis D, Grosse Y, Lauby-Secretan B, El Ghissassi F, Bouvard V, Benbrahim-Tallaa L, et al. The carcinogenicity of outdoor air pollution. Lancet Oncol. 2013;14(13):1262–3. [DOI] [PubMed] [Google Scholar]
- 12.Particulate Matter Emissions. United States Environment Protection Agency; 2018.
- 13.Brauer M, Amann M, Burnett RT, Cohen A, Dentener F, Ezzati M, et al. Exposure assessment for estimation of the global burden of disease attributable to outdoor air pollution. Environ Sci Technol. 2012;46(2):652–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Al-Kindi SG, Brook RD, Biswal S, Rajagopalan S. Environmental determinants of cardiovascular disease: lessons learned from air pollution. Nat Rev Cardiol. 2020;17(10):656–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Raaschou-Nielsen O, Andersen ZJ, Beelen R, Samoli E, Stafoggia M, Weinmayr G, et al. Air pollution and lung cancer incidence in 17 European cohorts: prospective analyses from the European Study of Cohorts for Air Pollution Effects (ESCAPE). Lancet Oncol. 2013;14(9):813–22. [DOI] [PubMed] [Google Scholar]
- 16.Bureau USC. American Community Survery (ACS). 2016. [Google Scholar]
- 17.Zhang ZWJ, Hart J, Laden F, Zhao C, Li T, et al. National scale spatiotemporal land-use regression model for PM2.5, PM10 and NO2 concentration in China. Atmos Environ 2018;192:48–54. [Google Scholar]
- 18.Zhang Z, Kang J, Hong YS, Chang Y, Ryu S, Park J, et al. Long-Term Particulate Matter Exposure and Incidence of Arrhythmias: A Cohort Study. J Am Heart Assoc. 2020;9(22):e016885. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Zhang Z, Zhao D, Hong YS, Chang Y, Ryu S, Kang D, et al. Long-Term Particulate Matter Exposure and Onset of Depression in Middle-Aged Men and Women. Environ Health Perspect. 2019;127(7):77001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.RNRZZR M. Long-term Exposure to Particulate Matter Air Pollution and Anosmia: A Case-Control Study conditionally accepted to JAMA 2021.
- 21.Cho YA, Kim J. Thyroid cancer risk and smoking status: a meta-analysis. Cancer Causes Control. 2014;25(9):1187–95. [DOI] [PubMed] [Google Scholar]
- 22.Shaddick GTM, Mudu P, Ruggeri G, Gumy S. Half the world’s population are exposed to increasing air pollution. npj Climate and Atmospheric Science 2020;3. [Google Scholar]
- 23.The Byrne A. 1979 convention on long-range transboundary air pollution: assessing its effectiveness as a multilateral environmental regime after 35 years. Transnatl Environ Law 2015;4:37–67. [Google Scholar]
- 24.Butt EW TS, Rigby R, Reddington CL, Yoshioka M, Johnson JS, et al. Global and regional trends in particulate air pollution and attributable health burden over the past 50 years. Environ Res Lett. 2017;12(10). [Google Scholar]
- 25.Parameswaran R, Brooks S, Sadler GP. Molecular pathogenesis of follicular cell derived thyroid cancers. Int J Surg. 2010;8(3):186–93. [DOI] [PubMed] [Google Scholar]
- 26.Brook RD, Rajagopalan S, Pope CA 3rd, Brook JR, Bhatnagar A, Diez-Roux AV, et al. Particulate matter air pollution and cardiovascular disease: An update to the scientific statement from the American Heart Association. Circulation. 2010;121(21):2331–78. [DOI] [PubMed] [Google Scholar]
- 27.Hart JE, Laden F, Puett RC, Costenbader KH, Karlson EW. Exposure to traffic pollution and increased risk of rheumatoid arthritis. Environ Health Perspect. 2009;117(7):1065–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Castano-Vinyals G, D’Errico A, Malats N, Kogevinas M. Biomarkers of exposure to polycyclic aromatic hydrocarbons from environmental air pollution. Occup Environ Med. 2004;61(4):e12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Veiga LH, Holmberg E, Anderson H, Pottern L, Sadetzki S, Adams MJ, et al. Thyroid Cancer after Childhood Exposure to External Radiation: An Updated Pooled Analysis of 12 Studies. Radiat Res. 2016;185(5):473–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Hemminki K, Eng C, Chen B. Familial risks for nonmedullary thyroid cancer. J Clin Endocrinol Metab. 2005;90(10):5747–53. [DOI] [PubMed] [Google Scholar]