Air Pollution Exposure and the Development of Chronic Rhinosinusitis in the Active Duty Population Free

ABSTRACT

Introduction

It has been shown that combat environment exposure, including burn pits that produce particulate matter 2.5 (PM_2.5), is associated with lower respiratory tract disease in the military population with increased hypothetical risk of upper respiratory disease, but no study has been done that examines the effects of non-combat environmental exposures on the development of chronic rhinosinusitis (CRS) in the active duty population. The primary goal of this study is to evaluate how air pollution exposure correlates to the development of CRS in active duty service members in the United States.

Methods

The military electronic medical record was queried for active duty service members diagnosed with CRS by an otolaryngologist between January 2016 and January 2018, who have never deployed, stationed in the United States from 2015 to 2018 (n = 399). For each subject, the 1-year mean exposure of PM_2.5, particulate matter 10 (PM₁₀), nitrogen dioxide (NO₂), and ozone was calculated. The control group was comprised of the same criteria except these patients were diagnosed with cerumen impaction and matched to the case group by age and gender (n = 399). Pollution exposure was calculated based on the Environmental Protection Agency’s data tables for each subject. Values were calculated using chi-square test for categorical variables and the Mann–Whitney U-test for continuous variables.

Results

Matched cases and controls (n = 399) with 33.1% male showed a statistically significant odds ratio (OR) of 5.99 (95% CI, 2.55-14.03) for exposure of every 5 µg/m³ of PM_2.5 increase and the development of CRS when controlling for age, gender, and diagnosis year. When further adjusting for smoking status, the OR was still statistically significant at 3.15 (95% CI, 1.03-9.68). Particulate matter 10, ozone, and NO₂ did not show any statistical significance. Odds ratios remained statistically significant when further adjusting for PM₁₀ and ozone, but not NO₂. Dose-dependent curves largely did not show a statistical significance; however, they did trend towards increased exposure of PM_2.5 leading to an elevated OR.

Conclusion

This study showed that PM_2.5 exposure is a major independent contributor to the development of CRS. Exposure to elevated levels produced statistically significant odds even among smokers and remained significant when controlling for other measured pollutants. There is still much to be understood about the genesis of CRS. From a pollution exposure perspective, a prospective cohort study would better elucidate the risk of the development of CRS among those exposed to other pollutants.

INTRODUCTION

Chronic rhinosinusitis (CRS) is characterized by mucosal inflammation of the nose and paranasal sinuses that lasts at least 12 consecutive weeks in duration.¹ CRS affects millions of Americans, and in 2014, healthcare costs related to the management of CRS were estimated at 22 billion USD.¹ Chronic rhinosinusitis is also associated with a significant decrease in health-related quality of life with increased rates of anxiety and depression.² Chronic rhinosinusitis is multifactorial in its etiology likely with a significant environmental contribution.³ Recent studies have shown that air pollutants and tobacco smoke may play a significant role in disease exacerbations and potentially initiation.^3–7 Specifically, animal studies have demonstrated an increase in type 2 eosinophilic sinonasal inflammation with long-term exposure to PM_2.5.⁸ There is also data to suggest that other pollutants such as nitrogen dioxide (NO₂) and particulate matter 10 (PM₁₀) also increase the odds of developing CRS after prolonged exposure.⁹ The United States Environmental Protection Agency (EPA) has identified several criteria pollutants, including PM_2.5, PM₁₀, NO₂, and ozone, which have been shown to have adverse health effects, specifically in respiratory illnesses.¹⁰

Particulate matter is comprised of solid and liquid droplets in the air composed of organic chemicals, metals, and soil or dust particles that are emitted as a result of complex chemical reactions associated with fires, automobiles, power plants, and factories in addition to matter aerosolized in construction sites and unpaved roads. Similarly, NO₂ and ozone are also emitted via complex reactions associated with industrialized environments. Recent studies have suggested that increased PM_2.5 exposure may worsen CRS disease progression and symptoms, but few, if any, studies have examined PM_2.5 as an initiating exposure for well-characterized CRS patients.⁵

Studies have also shown that combat environment exposure, including burn pits that produce PM_2.5, is associated with lower respiratory tract disease in the military population, with increased hypothetical risk of upper respiratory disease.¹¹ The active duty military population is a unique population that is younger and has fewer comorbidities than the general population as many diseases that may be implicated in respiratory tract pathology precludes a potential recruit from military service.^12,13 Although the association of ambient air pollution on the progression or exacerbation of CRS has been studied, to the authors’ knowledge, it has not been examined as an initiator of disease across multiple geographic areas, specifically in the active duty population. With active duty military service members (ADSMs) stationed in the United States and having the same unified electronic health record system, this population was selected to further elucidate the effects of pollution on the development of CRS. The primary goal of this study is to evaluate how air pollution exposure correlates to the development of CRS in ADSMs in the United States.

METHODS AND MATERIALS

This protocol was approved by the internal review board of the host institution. The Military Health Systems Mart (M2) is the database that holds de-identified data regarding patients in the Military Health System that can be queried for research purposes. From an M2 query, a list of patients was generated for those who were diagnosed with ICD 10 codes of all types of CRS with and without nasal polyposis from January 2015 to January 2018 in any otolaryngology clinic at all military treatment facilities and who have never before had a CRS diagnosis by ICD 10 or 9 in their military health record. This list was then reduced by filtering out those who were not ADSMs, those who were stationed outside the United States, and those who had been deployed in that time. A list of controls was generated using the same method but using the ICD 10 code for cerumen impaction. M2 was then used to extract the following data about each subject: date of birth of the subjects, Periodic Health Assessment (PHA) dates (yearly physical), PHA locations for each year, date of first diagnosis, location of first diagnosis, smoking status from PHA, and sex. Controls were matched to cases by age and gender (case–control ratio 1:1). Duty station location and living quarters were assumed to be in the same county as where the ADSMs did PHAs.

For each participant, we calculated the 12-month mean PM₁₀, PM₁₀, NO₂, and ozone concentration before the diagnosis date of CRS using the EPA pre-generated air pollution values. Participants were excluded when the service time at each PHA location was shorter than 12 months. We used 5-unit increases in concentrations as the exposure for PM_2.5, PM₁₀, and NO₂ air pollution, and used one interquartile range (IQR) incease as the units for ozone air pollution because of the narrow-range distribution.

Crude associations were calculated using chi-square test for categorical variables and the Mann–Whitney U-test for continuous variables. Adjusted ORs and associated 95% CIs were obtained by using a conditional logistical regression model. Model 1 adjusted for age, gender, and diagnosis year, while model 2 further adjusted for smoking status. To evaluate nonlinear dose–response relationships between all pollutant exposure (PM_2.5, PM₁₀, NO₂, and ozone) and CRS, we modeled pollutant exposure variables using restricted cubic splines with knots at the fifth, 35th, 65th, and 95th percentiles of the distribution of air pollution concentrations. Statistical analyses were conducted using STATA (version 16.0; Stata Corporation) and R (version 4.1; R Development Core Team).

RESULTS

In the years studied, there were 1,066 new cases of chronic rhinosinusitis among ADSMs who had never deployed and were stationed in the United States. After cross referencing their location with available EPA data, there were 399 cases and 399 controls. These were matched by age and gender with most of the cases diagnosed in females with the average age of diagnosis at 29.76 (SD 7.94) in controls and 30.98 (SD 8.85) in cases. In the CRS group, 22.8% had CRS with nasal polyps, while 0% of the controls had nasal polyposis. The average pollution exposure was similar between both cases and controls for NO₂ and PM₁₀; however, PM_2.5 and ozone concentrations were higher in the CRS group. 31.6% of controls and 30.8% of cases identified as smokers, which is greater than the number of smokers in the U.S. population (approximately 13.7%). The Northeast had the fewest number of cases and controls while the southwest had the largest. Most of the cases and controls were diagnosed in 2017.

In the adjusted model, the ORs (95% CI) for risk of CRS associated with each 5 unit or IQR increase in 1-year mean PM_2.5, PM₁₀, NO₂, and ozone were 3.15 (95% CI, 1.09-9.68), 1.19 (95% CI, 0.65-2.16), 1.22 (95% CI, 0.53-2.82), respectively (Table II).

The OR for risk of CRS associated with an IQR increase in ozone was 1.66 (95% CI, 0.73-3.74). Effect estimates further adjusted for gaseous air pollutants (NO₂ and ozone) in the two-pollutant model analyses by adding PM_2.5 in the same model were consistent with the estimates in the single-pollutant models (Table III). Patients with nasal polyps did not show a higher risk of disease when exposed to a higher level of air pollutants in the subgroup analyses (Table IV).

Spline regression analyses confirmed that increasing 1-year mean PM_2.5, PM10, NO₂, and ozone concentration was associated with an increased risk of CRS (Supplemental Materials). The dose–response curve was calculated using restricted cubic splines with knots at the 10th, 50th, and 90th percentiles of the distribution of 1-year air pollution concentration. The reference exposure level was set at the 10th percentile of the distribution of 12-month PM_2.5 (5.74 µg/m³), PM₁₀ (11.33 µg/m³), NO₂ (5.88 ppb), and ozone (0.027 ppb). ORs were adjusted for age, gender, diagnosis year, and smoking status.

DISCUSSION

Although CRS research has evolved tremendously over the past decade, the initiating triggers that may influence pathogenesis remain largely unclear. Environmental exposures have always been considered a likely factor in CRS initiation and progression, yet there have been few powered studies in appropriately diagnosed CRS patients with either CT scans or nasal endoscopy to demonstrate this relationship. Bhattacharyya et al. reported that an improvement in air quality was associated with decreased prevalence in sinusitis based on self-reported national surveys.¹⁴ In addition, Putman et al. reported a higher rate of CRS development in post-9/11 rescue workers at the World Trade Center that were involved with digging and rescue work, implicating the possible role for dust and heavy metals compared to emergency medical service works who were not.¹⁵ Mady et al. have reported that exposure to increased levels of PM_2.5 is associated with more symptomology and disease progression in already diagnosed CRS patients without nasal polyps.⁵ In both human in vitro and animal in vivo models, PM_2.5 has been shown to cause sinonasal ciliary disruption, increased mucous secretion, sinonasal epithelial barrier dysfunction, and type 2 eosinophilic inflammation.^8,16,17

This study adds to the growing body of literature demonstrating the relationship between air pollution and CRS. This study found a statistically significant OR of 5.99 for the development of chronic sinusitis for every 5 µg/m³ of average PM_2.5. This OR decreased to 3.15, although still statistically significant, when adjusting for smoking status, which has also been implicated in the pathogenesis of CRS.⁴ This finding suggests that PM_2.5 plays a major role in the development of CRS but also corroborates other studies implicating smoking is a major contributing factor to disease development by virtue of a decrease in the OR by nearly half.^4,18 There was no statistical significance between PM₁₀, NO₂, or ozone and CRS incidence.

Cerumen impaction is one of the most common reasons people seek medical care from an otolaryngologist.¹⁹ The control group was made up of patients who were diagnosed with cerumen impaction by an otolaryngologist. This patient population was used for several reasons. Being a common complaint in an otolaryngology clinic, the researchers believed that it would yield a sample size sufficiently large and diverse to make an effective matching process for the statistical calculations. The authors are not aware of any literature that has shown an association between CRS and cerumen impaction and so believe that these two pathologies are not correlated and so cerumen impaction would be a good control group.

The development of CRS is complex, and there are several theories that have been proposed to characterize its pathogenesis. Aeroallergens likely have a component in the development of CRS, and they have been shown to play a part in nasal polyposis.^20,21 This study did not find any statistically significant difference in region, year, or demographics in those who had CRS with or without nasal polyposis.

Since air pollution composition may differ with regional variation, we utilized multi-pollutant models to examine the effects of different combined pollutants on CRS incidence. PM_2.5 was found to remain a statistically significant OR when adjusting for PM₁₀ and ozone, but not for NO₂. The CIs also increased when adjusting for these pollutants, with the largest CI found for NO₂. This is likely due to the substantially fewer data collection sites for NO₂ when compared to the other pollutants, which limited the final numbers used in our analysis.

The primary strength of this study is the patient population. We used active duty service members, which eliminates many of the confounding variables that a similar study may encounter outside of the military healthcare system. Prospective military recruits are disqualified for military service with previously diagnosed medical conditions including mucociliary dysfunction, immunodeficiency, hypoimmunoglobulinemia, and chronic lower respiratory tract. These diseases often include asthma, thus eliminating some of the confounding variables that may increase susceptibility to CRS, especially with environmental exposures.⁸ The active duty military population also tends to skew toward a younger group and are less likely to be diagnosed with other chronic medical conditions that are commonly seen in older Americans. In our study, the average age of diagnosis for both cases and controls was approximately 30 years. Another strength of this study is that all patients with CRS or CRS with nasal polyps were diagnosed by a board-certified otolaryngologist presumably using either nasal endoscopy or sinus CT scans, eliminating some of the uncertainties of self-reported diagnoses seen in national health surveys or diagnoses made purely by history.

This study also examined the effects of air pollution exposure on the development of CRS in multiple regions and across multiple years. As demonstrated in Table I, which reports results for regions and years of diagnosis, there were no differences between regions that were examined. Some regions had very few subjects in this study due to exclusion criteria or lack of EPA data for the county or bordering county in which they conducted their PHA.

There are several limitations to this study. Air pollution exposure was estimated using EPA data instead of measuring individual exposure. This study generalized the exposure over a county-wide area which may or may not fully represent the actual pollution exposure of each patient. Also, CRS is a disease that is thought to develop over a longer course of time. In addition, not every county collects EPA air pollution data, and many do not report all of the pollutants that were of interest in this study. These two factors made the sample size of this study smaller than what the authors would prefer and may account for the lack of statistical significance in the OR of PM_2.5 when controlling for NO₂. Lastly, there are some limitations to the case–control design of our study, which should be further explored in larger retrospective or prospective cohort studies.

CONCLUSION

In conclusion, this study demonstrates that PM_2.5 exposure is a major independent risk factor to the development of CRS in the active duty military population. Exposure to elevated levels of PM_2.5 produced statistically significant odds even among smokers and remained significant when controlling for other measured pollutants and is evidence of the importance of PM_2.5 in the causal pathway for CRS. These findings, although preliminary, suggest a role for tighter federal regulation of air quality to reduce the burden of sinonasal disease, especially in the active duty military population. Further research regarding the mechanisms of PM-induced rhinosinusitis in addition to larger retrospective studies or prospective cohort studies are needed.

SUPPLEMENTARY MATERIAL

SUPPLEMENTARY MATERIAL is available at Military Medicine online.

FUNDING

None declared.

CONFLICT OF INTEREST STATEMENT

The opinions and assertions contained herein are those of the authors and do not reflect those of the Uniformed Services University or the Department of Defense.

REFERENCES

Author notes