The Association between Ambient Fine Particulate Matter and Oral Neoplasm among Smokers and Betel Quids Chewers

Received: March 21, 2021
Revised: June 29, 2021
Accepted: July 23, 2021

 Copyright The Author(s). This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are cited.

Cite this article:

Ku, M.S., Siewhcaisakul, P., Yen, A.M.F., Liu, C.Y. (2021). The Association between Ambient Fine Particulate Matter and Oral Neoplasm among Smokers and Betel Quids Chewers. Aerosol Air Qual. Res. 21, 210060. https://doi.org/10.4209/aaqr.210060

HIGHLIGHTS

• We examined the effects of PM_2.5 on OPMD/OC using the nationwide screening data.
• An increased risk of OPMD/OC and OC was noted in the areas of higher PM_2.5 levels.
• We observed a concentration-dependent relationship between PM_2.5 and OC risk.

Keywords: Particulate matter, Oral cancer, Nationwide screening program, Taiwan

1 INTRODUCTION

It is inevitable that air pollution has become environmental burden worldwide. Among all air pollutants, fine particulate matters (particulate matter with an aerodynamic diameter ≤ 2.5 µm; PM_2.5) are thought to be the most important toxicants in urban air (Kaiser 2005). In 2016, World Health Organization (WHO) reported that approximately 92% of the world’s population were residing in places where level of the PM_2.5 exceeds WHO air quality guideline (AQG), an annual average concentration of 10 µg m^–3 (WHO, 2006). Although studies have indicated a decreasing trend in PM_2.5 concentrations in many countries, PM_2.5 is still a significant public concern in Taiwan, due to many of the people in Taiwan still live in areas with the PM_2.5 exceeds 10 µg m^–3 and the related adverse health effects such as cardiovascular disease, respiratory disease and cancers have been reported (Cheng and Hsu, 2019; Yang et al., 2018).

Oral cancer (OC) is the fifth leading cause of cancer death and has been the fourth most common cancer in men for more than 10 consecutive years since 2003 in Taiwan (Ministry of Health and Welfare, 2018). In order to prevent the disease onset and to cure or slow the disease progression, the Taiwanese government has implemented the biennial nationwide OC screening focusing on adults at a higher risk (betel quid chewers or smokers) since 2004 to detect oral potentially malignant disorder (OPMD) and OC (Chuang et al., 2017). In addition, since betel quid chewing has been identified as a risk factor of OPMD and OC independently (Amarasinghe et al., 2010; Ko et al., 1995; Merchant et al., 2000), and in conjunction with smoking and alcohol drinking (Yen et al., 2007a, 2008; Yen et al., 2008b), the Taiwanese government has reinforced the betel quid chewing prevention and control in recent decades. The betel quid chewing rate among male adults over 18 has been declined from 2007 to 2018 by 59.2% in Taiwan (Health Promotion Administration, 2019). Although age-standardized OC incidence rate increased continuously in both sexes before 2009, it has become flatten since then (Siewchaisakul et al., 2020a).

The other risk factors of OPMD and OC include exposure to heavy metals (arsenic, nickel, and chromium) at high concentration (Chu et al., 2019; Kaldor et al., 1984; Su et al., 2010; Yuan et al., 2011), human papillomavirus (HPV) infection (Chaturvedi, 2012), presence of metabolic syndrome (MetS) (Yen et al., 2011), and age younger than 70 years (Siewchaisakul et al., 2020a).

Recent studies have shown that the level of PM_2.5 was associated with an increased risk in cardiovascular inflammatory marker such as C-reactive protein (CRP) in adults with MetS (Dabass et al., 2018). Other inflammatory markers including T lymphocytes, macrophages, transforming growth factor (TGF)-β1, tumor necrosis factor (TNF)-α, and interleukin (IL)-6 were also found as underpinning the link between MetS and OPMD (Chiang et al., 2002; Hsu et al., 2014; Ujpál et al., 2004). This led to the hypothesis that PM_2.5 may also associate with OPMD.

In 2013, PM_2.5 was noted by International Agency for Research on Cancer (IARC) as carcinogenic to human (Loomis et al., 2013). Carcinomas that have been found to be associated with PM_2.5 are lung, leiomyoma, bladder cancer, and OC (Chu et al., 2019; Hamra et al., 2014; Mahalingaiah et al., 2014). Previous study has revealed that the long-term PM_2.5 concentrations and daily mean concentrations over 35 µg m^–3 varied by meteorological conditions and regions in Taiwan (Cheng and Hsu, 2019). Cheng and Hsu’s (2019) study prompted us to determine the cut-off point of PM_2.5 and also take into account the heterogenous PM_2.5 exposure effect on OC between areas in current study. However, few population-based studies investigated the association between PM_2.5 and OC, especially OPMD.

We therefore aimed to elucidate the association between PM_2.5 and OPMD/OC, taking heterogeneity between areas into account based on the longitudinal data of air quality monitoring and nationwide OC screening in Taiwan.

 
2 METHODS

2.1 Study Population and Data Collection

2.1.1 Data source for oral neolplasm

A retrospective cohort design was conducted in this study. Population in this study was based on the nationwide OC screening program between 2006 and 2016. The program has been launched by Health Promotion Administration (HPA), Taiwan, since 2004. The details have been described by Chuang et al. (2017). The biennial OC screening program with visual inspection was done by the trained dentists and physicians and was targeting high risk population aged 30 years and above who had the habit of either cigarette smoking or betel quid chewing. Those with suspicious lesions would be invited by public health nurses to complete the confirmatory follow-up examination. We excluded those who had diagnosed with OC before screening and those cancer-free subjects who were diagnosed as epithelial dysplasia in oral cavity.

The definition of oral cancer was according to International Classification of Diseases, 9^th Revision‐Clinical Modification (codes 140–141, 143–146, and 148–149) and from the International Classification of Diseases, 10^th Revision (codes C00–C06, C09–C10, C12–C14) including lip, tongue, gingival, floor of mouth, palate, other parts of mouth, oropharyngeal, hypopharyngeal and unspecified pharyngeal cancer. OPMD was defined when individuals were clinical diagnosed as leukoplakia, erythroleukoplakia, erythroplakia, oral submucous fibrosis, and verrucous hyperplasia.

Data on age at screening, sex, residential area (county/city), unhealthy oral habits (betel quid and cigarettes exposure) were retrieved. Subjects may attend the screening program multiple times. We kept only the latest record of subjects upon the detection of OPMD/OC at screening or before the clinical diagnosis of OC, or the last records of subjects who were free from OPMD/OC. The study was approved by the Research Ethics Committee of National Taiwan University Hospital with informed consent pursuant to the regulations of the Institutional Review Board.

 
2.1.2 Exposure measurements of PM_2.5

Air quality data measured by Taiwan’s Environmental Protection Administration (TEPA) were used to estimate PM_2.5 exposures. Since 1993, Taiwan Air Quality Monitoring Network (AQMN) has been established by TEPA to monitor nationwide air quality by 73 monitoring sites distributed in 22 cities/counties (TWEPA, 2019). The complete record of daily continuous monitoring of PM_2.5 were available since August, 2005. Daily PM_2.5 concentrations of all monitoring stations in AQMA from 2006 to 2016 were used for the whole year-round measurements and were grouped by city/county.

To investigate the association between PM_2.5 and OPMD/OC, these spatially averaged PM_2.5 data were chronically matched to study subjects’ screening date by month and the city/county the subjects lived in to represent their PM_2.5 exposures.

 
2.2 Statistical Analysis

Descriptive information is reported as frequency and percentage. We first categorized the PM_2.5 concentration into dichotomous variable (low and high) using cut-off point of 35 µg m^–3, the third quartile of the monthly average PM_2.5 concentration of all the 22 cities/counties over eleven years. In addition to the dichotomous, we also categorized PM_2.5 by 10 µg m^–3 into 5 groups (< 10, 10–19, 20–29, 30–39, and 40+ µg m^–3). The lowest PM_2.5 concentration group (< 10 µg m^–3) was used as the reference group.

Since PM_2.5 measurement was based on the level of county/city on monthly basis, we examined the associations between PM_2.5 and both OPMD/OC, OPMD or OC using the Bayesian univariable random-effects logistic regression models, treating county-specific PM_2.5 effects as random-effect variables. Age (30–49, 50–69, 70+), sex (male, female), and unhealthy oral habits (cigarette smoking only, betel quid chewing only and betel quid chewing and cigarette smoking) are the potential individual risk factors for oral cancer. We further conducted the multivariable analyses controlling the significant confounding factors based on the univariable model. Notably, we did model the PM_2.5 as dichotomous and polychotomous in different multivariable models. All tests assumed a two-sided type I error of 0.05 and were performed with SAS version 9.4 (SAS Institute Inc., Cary, NC).

 
3 RESULTS AND DISCUSSION

There were 154,030 OPMD, 23,286 OC cases, and 3,686,729 subjects free from OPMD/OC included in the analyses. Frequency of demographic information and environmental risk exposure of betel quid, cigarette smoking is shown in Table 1. Most of the study population was diagnosed with OPMD in the age group of 30 to 49 years and with OC in the age group of 50 to 69 years. Men who had the habit of cigarette smoking and betel quid chewing were dominant in both OPMD and OC groups.

Fig. 1 shows the average concentrations of PM_2.5 by county/city in Taiwan. The highest average concentration was in Chiayi City with the mean of 34.05 µg m^–3 (standard deviation [SD]: 14.52 µg m^–3). The lowest average concentration was in Taitung County with the mean of 11.82 µg m^–3 (SD: 3.88 µg m^–3) (Table S1).

Fig. 1. Geographically gradients of monthly average PM_2.5 concentration by county/city. Quintile was used for classification of pattern.

 
3.1 PM_2.5 and Risk of OPMD/OC

Fig. 2(a) shows the proportions of diagnosed OPMD and OC combined cases by county/city. The county/city with the highest proportion of diagnosed cases was Chiayi City (8.88%), the same city with the highest average PM_2.5 concentration, and the lowest proportion was in Taichung City (1.47%) (Table S2).

 Fig. 2. Geographically gradients of oral neoplasms by county/city. (a) rate (%) of OPMD and OC cases (b) rate (%) of OPMD cases (c) rate (%) of OC cases. Quintile was used for classification of pattern.

The effects of PM_2.5 and other related factors of OPMD/OC are elucidated in Table 2. In the crude univariable analysis, PM_2.5 with concentration 35 µg m^–3 and above showed a significantly increased risk of OPMD/OC by 4% (RR = 1.04, 95% CI: 1.03–1.06) compared to concentration lower than 35 µg m^–3. In the multivariable analysis after adjusting for sex, age, betel quid chewing and cigarette smoking, the significant effect of PM_2.5 on OPMD/OC still remains with an increased adjusted risk of 11% (adjusted RR (aRR) = 1.11, 95% CI: 1.09^–1.13). When categorized PM_2.5 by 10 µg m^–3, the risks of OPMD/OC were significantly higher in the groups in the univariable and the multivariable analyses. The aRRs of OPMD/OC were 1.20 (95% CI: 1.16–1.24), 1.36 (95% CI: 1.31–1.40), 1.39 (95% CI: 1.34–1.44), and 1.52 (95% CI: 1.46–1.57) in the PM_2.5 concentration groups of 10–19, 20–29, 30–39, ≥ 40 µg m^–3, respectively, after adjusting for age, sex, betel quid chewing and cigarettes smoking (Table 2).

Dose-response relationships of oral habits and OPMD or OC has been demonstrated in previous studies and we chose some typical papers with comparable attributes of oral habits in Taiwan or India (Madathil et al., 2016; Yang et al., 2010; Yen et al., 2007; Znaor et al., 2003). As shown in previous studies, the higher the quantity of cigarette smoking or betel quid chewing, the higher the risk of OPMD/OC. As shown in Fig. 3(a), the concentration-dependent effect of PM_2.5 can be observed in comparison with magnitude of effects of traditional risk factors (betel quid and cigarette smoking).

Fig. 3. Adjusted odds/relative risk ratio of PM_2.5 and other risk factors for (a) OPMD or oral cancer (b) OPMD and (c) OC.

Many studies have reported that exposure to PM_2.5 increased the risk of lung cancer or respiratory diseases (Ghazipura et al., 2019; Hamra et al., 2014; Hopke et al., 2019; Raaschou-Nielsen et al., 2016). Undoubtedly, oral is one of the exposure routes to these ultrafine particles. Biological evidence has been established for plausible mechanisms between PM_2.5 and carcinogenesis, such as increased oxidative stress, induced DNA damages and DNA mutations, inflammatory responses, and may lead to the occurrence of malignant tumor (Feng et al., 2016; Ghio et al., 2012). Furthermore, OC/OPMD is mostly developed from the squamous cell, in that accounted to 80–90% of oral malignancy (Johnson et al., 2011). Thus, inflammatory response maybe be one of the mechanisms underpinning the association between PM_2.5 and OPMD/OC.

 
3.2 PM_2.5 and Risk of OPMD

The highest average PM_2.5 concentration was noted in Chiayi City with the mean of 34.00 µg m^–3 (SD: 14.50 µg m^–3) and the lowest average concentration was found in Taitung County with the mean of 11.81 µg m^–3 (SD: 3.88 µg m^–3) (Table S3). The proportion of diagnosed OPMD cases by county/city was shown in Fig. 2(b). The highest proportion of diagnosed cases was found in Chiayi City (7.83%), followed by Penghu County (6.48%) and Changhua County (6.26%) and the lowest proportion was found in Hsinchu County (1.23%), see Table S4.

The effect of PM_2.5 and others related factors on OPMD are elucidated in Table 3. In the univariable analysis, PM_2.5 exposure was insignificant associated with OPMD. However, higher PM_2.5 significantly increased the risk of OPMD by 4% (aRR = 1.04, 95% CI: 1.03–1.06) in the multivariable analysis. When categorized PM_2.5 by 10 µg m^–3, the risks of OPMD/OC were significantly higher in the PM_2.5 > 10 µg m^–3 concentration groups in the univariable and the multivariable analyses. These concentration-dependent relationships can be observed, and compared to the magnitude of effects of traditional risk factors (betel quid and cigarette smoking) as shown in Fig. 3(b).

Recent study has shown one of the cardiovascular related-inflammatory markers, the CRP was increased with high level of PM_2.5 in adults with MetS (Dabass et al., 2018). MetS has been found to be associated with OPMD and several cancers (Braun et al., 2011; Ku et al., 2019; Siewchaisakul et al., 2020; Yen et al., 2011). Pro-inflammatory markers, including TNF-α, IL-6 and CRP were shown to be the underlying mechanism between MetS and OPMD as well (Hsu et al., 2014).

3.3 PM_2.5 and Risk of OC

The highest average PM_2.5 concentration was noted in Chiayi City with the mean of 33.98 µg m^–3 (SD: 14.48 µg m^–3) and the lowest average concentration was found in Taitung County with the mean of 11.83 µg m^–3 (SD: 3.88 µg m^–3) (Table S5). The proportion of diagnosed OC cases by county/city was shown in Fig. 2(c). The highest proportion of diagnosed cases was found in Chiayi City (1.23%), followed by Changhua County (1.23%) and Yunlin County (1.00%) and the lowest proportion was found in Lianjiang County with no case found in our study, see Table S6.

The effect of PM_2.5 and others related factors on OC are elucidated in Table 4. In the univariable analysis, PM_2.5 with concentration greater than or equal to 35 µg m^–3 shown significantly increased risk of OC by 44% (RR = 1.44, 95% CI: 1.39–1.50) compared to those lower than 35 µg m^–3. In the multivariable analysis, higher PM_2.5 exposure statistically increased the risk of OC by 55% (aRR = 1.55, 95% CI: 1.49–1.60) compared to low PM_2.5 exposure. When categorized PM_2.5 by 10 µg m^–3 in the multivariable analysis, PM_2.5 still increased risk of OC with higher concentrations. PM_2.5 with concentration of 10-19 µg m^–3 has greater risk of OC by 1.68-fold compared to those lower than 10 µg m^–3 (aRR = 1.68, 95% CI: 1.52–1.85). This result is consistent with the concentrations of 20–29, 30–39, ≥ 40 µg m^–3 compared to < 10 µg m^–3 in which the aRR were 2.58 (95% CI: 2.33–2.85), 3.12 (95% CI: 2.81–3.45), and 4.21 (95% CI: 3.78–4.68), respectively (Table 4).

These concentration-dependent relationships were commensurate with the magnitude of effects of traditional risk factors (betel quid and cigarette smoking) as shown in Fig. 3(c). The higher the quantity of cigarette smoking or betel quid chewing, the higher the risk of OC. The same trend could also be found for the association between PM_2.5 exposure and risk of OC, consistent with previous study reported by Chu et al. (2019).

The components of PM_2.5, such as arsenic, nickel, chromium, asbestos, and polycyclic aromatic hydrocarbons (PAHs), have been found to be associated with an increased OC risk (Gutiérrez-Castillo et al., 2006; Paget-Bailly et al., 2012; Su et al., 2010; Yuan et al., 2011). The carcinogenetic potential of chronic PM_2.5 exposure has been widely demonstrated in animal studies, with genome wide epigenetic and RNA transcription changes and malignant cancer cell behaviors such as cross-talk between epithelial-mesenchymal transition (EMT) and cancer stem cells (CSCs) properties which may lead to enhanced tumorigenicity and promote cancer cell progression (Yang and Xiao, 2018; Lee et al., 2020). These together may explain the association of PM_2.5 and OPMD/OC, as well the reason why the association between PM_2.5 and OC is more profound than OPMD in this study.

 
3.4 Comparison between Studies

To our knowledge, only limited epidemiological studies have reported PM_2.5 and OC risk. A previous study in Taiwan has reported that an increased risk of OC among men who were exposed to PM_2.5 concentration over 40.37 µg m^–3 (aOR = 1.43, 95% CI: 1.17–1.74), compared with 26.74 µg m^–3 < PM_2.5 (Chu et al., 2019). We targeted smokers and/or betel quid chewers, the high-risk population of OPMD/OC and found that PM_2.5 with concentration of 35 µg m^–3 and higher has a significant increased risk of OC, and the effect magnitude in associated with OC was even higher compared to others. We also found that PM_2.5 with concentration of 10 µg m^–3 and higher has a significant increased risk of OPMD. The novelties of this study could be emphasized by comparing to the previous studies. First, our study is a longitudinal cohort study with 11 years followed-up of OPMD/OC nationwide screening data. This makes allowance for the identification of OPMD cases, the precursor of invasive oral cancer, as well as the information on both sexes and unhealthy oral habits. Second, we took into account the geographic variations when evaluating the OC risk in association with PM_2.5 in our study. We consider monthly average PM_2.5 concentration as a random effect in our model since this exposure indicator is a hierarchical factor rather than an individual one. Last but not least, an ecological study by Su et al. (2019) presented the geographical variations on the association between PM_2.5 and cancer incidence in Taiwan. Su et al. (2019) also reported a significant dose-response relationship between PM_2.5 and OC incidence with spearman correlation ranged between 0.56–0.59 in men. However, several potential risk factors such as habitual smoking or betel quid chewing were not considered in their study.

3.5 Limitation

Some limitations in this study should be noted. First, we do not have exact data on individuals’ daily PM_2.5 exposures. Therefore, we estimated monthly PM_2.5 concentration in average by areas. Second, we used data based on nationwide OC screening program in which targeted population were those who only experienced betel quid chewing or smoking behavior. Thus, generalization of our study results would limit to smokers and/or betel quid chewers. Third, the PM_2.5 exposure duration and the exposure history has not been considered in this study. Those who were diagnosed as oral cancer might have been continuously exposed to high level of PM_2.5. Finally, there are still numerous variables related to OPMD/OC risk such as genetic, family history of OC, history of chronic diseases and other potential carcinogenic compounds, which we haven’t considered in this study.

 
4 CONCLUSIONS

To our knowledge, this is the first study to investigate the association between PM_2.5 and OPMD/OC at nationwide level. Based on the longitudinal nationwide OC data, we found that PM_2.5 exposure was linked to an increased risk of OPMD/OC among smokers and/or betel quid chewers. Our results also demonstrated the spatial variations for PM_2.5 concentration and OPMD/OC. This information would beneficial to OPMD/OC prevention policy in some specific counties or cities. Future studies are warranted to investigate the effect of personal PM_2.5 exposure on OPMD/OC risk.

 
ACKNOWLEDGMENTS

We thank the Research Ethics Committee of National Taiwan University Hospital approved this project, and granted a waiver for informed consent (202002091W) pursuant to the regulations of the Institutional Review Board. This work was financially supported by the “National Taiwan University Higher Education Sprout Project (NTU-110L8810)” within the framework of the Higher Education Sprout Project by the Ministry of Education (MOE) in Taiwan.