Association between ambient temperature and chronic rhinosinusitis

1 INTRODUCTION

Chronic rhinosinusitis (CRS) is an inflammatory condition of the nose and paranasal sinuses, diagnosed by the presence of at least two of the following cardinal symptoms for at least 12 consecutive weeks and confirmed by endoscopic or radiographic findings¹: nasal obstruction/congestion/blockage; anterior or posterior (mucopurulent) nasal drainage; loss or decreased sense of smell; facial pressure/pain/fullness. Studies have shown that CRS is associated with depression, impaired cognitive function, sleep disturbances, and poor quality of life.¹ CRS varies by country, with a prevalence of 6.9%–27.1% in Europe,² 2.1%–13.8% in the United States,^3-6 and 4.8%–9.7%⁷ in China. Recently, our group and others have demonstrated that environmental particulate matter air pollution, specifically PM_2.5, is associated with the development of CRS.⁸ These studies highlight the increasing role of environmental factors in the development of CRS.

With the progress of global warming, the health effects of high ambient temperature are receiving increasing attention.^9-13 A meta-analysis of 266 studies showed that a 1°C increase was positively associated with increased cardiovascular disease-related mortality across all diagnoses considered.⁹ A meta-analysis of over 1.7 million mental health-related deaths demonstrated that a 1°C temperature rise was associated with a significant increase in morbidity of, such as, mood disorders, organic mental disorders, schizophrenia, neurotic, and anxiety disorders.¹⁰ In addition, studies have indicated that high ambient temperatures increase the risk of asthma and chronic obstructive pulmonary disease (COPD).^{11, 12} However, little is known about the impact of high ambient temperature on sinonasal diseases.

In this large outpatient-based case–control study, we focused on investigating the association between high ambient temperature exposure and the risk of developing CRS in patients who visited the Johns Hopkins hospitals in the Baltimore–Washington metropolitan area. We hypothesized that a positive association exists between high ambient temperature exposure and the development of CRS.

2 MATERIALS AND METHODS

3 RESULTS

Our analysis included a total of 4752 participants, 56.9% of whom were women, with an average age of 51.8 (16.8) years (Table 1). The year distribution of these 2367 CRS cases was similar to the trend of average daily maximum temperature from 2013 to 2022 (Figure 1). After matching, the age, sex, and race/ethnicity composition of the control group was identical with that of the case group. In terms of other characteristics, patients with CRS were more likely to be overweight and obese and live in areas with higher household incomes. However, they were less likely to smoke or drink alcohol or have been diagnosed with hypertension or diabetes.

The average daily maximum temperature of the case group and the control group were 27.8°C (82.0°F) and 27.6°C (81.7°F), respectively (Table 1). We found that the cumulative effects of temperature over 21 days on the risk of CRS were significant at both high and low temperatures, compared with the MMT at 25.3 (Figure 2). Specifically, there was a significant effect of extreme heat (35.0°C, 95th percentiles of temperature distribution), with an OR of 2.37 (95% CI, 1.60–3.50). According to the overall lag effects of extreme heat (35.0°C) on the risks of CRS, we found it evident that the temperature–CRS association was nonlinear (Table 2 and Figure S1(a)). Generally, the effect of high temperature reached a small peak after 3 days of exposure, and there was a significant increase in the risk of CRS exacerbation after approximately 2 weeks. For the secondary outcome, we found that cumulative temperature exposure was associated with a higher risk of maxillary sinusitis at both high and low temperatures (Figures S1 and S2), with a significant effect after 1 week of exposure compared with the MMT.

We conducted subgroup analyses and found that young and middle-aged CRS patients, patients with abnormal weight, and patients without nasal polyps (CRSsNP) were more likely to be affected by ambient temperature (Figure 3). Other factors did not significantly affect the association between ambient temperature and CRS (Figure S3). In the sensitivity analysis, we changed the lag period from 21 days to 14 and 28 days, and the results were consistent (Figure S4). To identify the robustness of the association, we changed the number of internal knots to model the exposure–response association and the lag–response association. We found that the overall lag structures using different DLNM model specifications showed a similar tendency to our main results (Figure S5). When we measured the exposure using daily minimum and mean temperatures instead of maximum temperatures, the results showed a similar association, although the association for daily minimum temperatures was not significant (Figure S6).

4 DISCUSSION

We found a U-shaped association between ambient temperature exposure levels and the risk of developing CRS, with a significant effect of extreme heat during a 21-day lag. The current findings have broad public health implications in the setting of global warming. This study benefits from strengths including a robust patient population (objective testing, otolaryngologist diagnosis) and a novel matching strategy for controls, who had clear sinus imaging.

Few studies have explored the relationship between ambient temperature and CRS. Only a recent study found that the associations between mean/highest/lowest temperature, temperature range, and CRS were not statistically significant but did not account for lag effects.³⁰ Existing studies focus more on the effect of ambient temperature on one of the cardinal symptoms of CRS, loss or decreased sense of smell. Animal experiments from blowflies,³¹ drosophilas,³² and turtles³³ have shown a nonlinear association, and high ambient temperature can lead to loss or decreased sense of smell. Evidence from the population is relatively lacking, and no positive association has been found so far.^34-36 However, the sample sizes of the existing studies were small, involving only approximately 10 or 50 participants, and did not take the lag effect of ambient temperature on the sense of smell into account.

We found the effect of high temperature on CRS reached a small peak after 3 days of exposure and became significant after 14 days of exposure, suggesting that the effect may be indirect. Elevated ambient temperature may increase PM_2.5 levels and thereby exacerbate sinonasal symptoms. The mechanism of the lag effect is unclear, but there may be different pathways for different phenotypes.³⁷ For bacterial CRS and allergic fungal RS, the ambient temperature may impact CRS by affecting the growth of bacteria, fungi, and viruses. For example, the literature indicates that allergic fungal RS occurs mainly in areas with high temperature and high humidity that are conducive to fungal growth.¹ For CRS associated with other respiratory diseases, such as asthma, aspirin sensitivity, idiopathic bronchiectasis, and cystic fibrosis, the ambient temperature may influence CRS by affecting the onset of other respiratory diseases. For instance, high temperatures have been shown to increase the risk of adult asthma hospitalizations.^{11, 38}

Subgroup analysis showed that the young and middle-aged group (18–64 years) may have a higher risk of CRS attributable to ambient temperatures than the elderly. This age group difference in susceptibility was consistent with some studies^{11, 39} and contrary to others.^{40, 41} Young and middle-aged people were more likely to be exposed to extreme temperatures while working or exercising outdoors, but they also changed their activity habits because of the uncomfortable temperatures. More research is needed to determine the exact differences in temperature sensitivity between age groups. We found that CRS patients with abnormal weight were more susceptible to the effect of ambient temperatures, suggesting that body weight and metabolism may modify the association between ambient temperature and CRS. Further studies are called for to explore the mechanism of this potential modifying effect. We also found that patients without nasal polyps (CRSsNP) were more affected, which needs to be confirmed by studies with a larger sample size of CRSwNP. Among the anatomy-based CRS subtypes, maxillary sinusitis had the highest prevalence due to the ostium being situated high up on the medial wall.⁴² Our secondary outcome results demonstrated that maxillary sinusitis was most susceptible to extreme heat, suggesting that the effect of ambient temperature on CRS is of great concern.

There are several limitations to our study. First, our study has inherent weaknesses in its retrospective design regarding coding accuracy, potential sampling bias, or inability to determine causal direction. Further experimental studies are needed to confirm the effects. Second, our model did not include participants’ activity patterns (time spent indoors and outdoors) or air conditioner use behavior due to data inaccessibility. Failure to adjust for these variables may have biased the results. To address this shortcoming, we adjusted for household income, which may reflect the overall SES of patients. In addition, changes in participants’ residential address during the study period were not applicable. Therefore, our exposure measurement relied only on ambient temperature mapping to the participants’ initial residential address. Lastly, data on occupational or environmental exposure to allergens or the use of medications were not available.

5 CONCLUSION

In this cross-sectional study, short-term exposure to high ambient temperature was associated with a lag exacerbation of CRS symptoms, suggesting that the effects may be indirect. This finding has broad implications for the proactive response to global warming.

AUTHOR CONTRIBUTIONS

R.D. and Z.Z. designed the study and cleaned the data. R.D. and W.J. performed the analyses. J.M. and Q.Z. checked the codes. S.S. and Z.L. provided technical support for statistical model specifications selection. R.D. and Z.Z. drafted the manuscript. R.D., Z.Z., and M.R. revised the manuscript. All authors reviewed and contributed to the article and approved the submitted version.

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no competing interests.

AVAILABILITY OF DATA AND MATERIALS

The hyperlinks of publicly available exposure data are included throughout the literature cited. The datasets of patients used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

ETHICS APPROVAL AND CONSENT TO PARTICIPATE

Our study protocol was approved by the Johns Hopkins University School of Medicine Institutional Review Board. Informed consent was not required because, except for zip codes, no patient-identifying information was collected.