Assessing the Complex Impact of Smoking Habits on Allergic Rhinitis: A National Cross-Sectional Study
Article information
Abstract
Objectives.
Allergic rhinitis (AR) significantly impacts quality of life and incurs socioeconomic costs. The influence of smoking habits, including the use of conventional cigarettes (CCs) and electronic cigarettes (ECs), on the prevalence and management of AR remains a subject of debate. This study aims to explore the association between smoking status (CC and EC use) and the prevalence and management of AR among Koreans by analyzing data from the Korea National Health and Nutrition Examination Survey (KNHANES) VII (2018) and VIII (2019–2021).
Methods.
This cross-sectional study involved 22,290 participants aged 19 years and older from the KNHANES. Participants self-reported their smoking status, and urinary cotinine levels were measured to assess nicotine exposure. We employed statistical analyses, including logistic regression, to examine the relationships between smoking status, cotinine levels, and the prevalence and management of AR.
Results.
In univariable logistic regression analysis, EC users exhibited a 35.8% increased risk of AR compared to non-smokers, whereas CC users experienced a 27.7% reduced risk. Multivariable logistic regression analysis showed a 20.3% lower risk of AR among CC users; however, no significant association was observed for EC users. Higher cotinine levels (>500 ng/mL) were associated with a lower prevalence of AR. Specifically, heavy CC users with high cotinine levels demonstrated a 35% reduced risk of AR. Nonetheless, after adjusting for confounders, this association was no longer significant, indicating that other variables might influence this relationship.
Conclusion.
Smoking status is associated with the prevalence of AR in Koreans. Notably, heavy use of CCs is negatively correlated with the prevalence of AR.
INTRODUCTION
Allergic rhinitis (AR) is an inflammatory condition triggered by allergic reactions within the nasal mucosa, presenting symptoms like nasal congestion, sneezing, and rhinorrhea. It not only imposes a socioeconomic burden but also significantly deteriorates the quality of life of affected individuals [1,2]. Although the exact pathophysiology of AR is still not fully understood, it is recognized as a multifactorial disease influenced by genetic factors, environmental exposures, and lifestyle choices, such as smoking habits [3].
Inconsistent findings have been reported regarding the association between smoking and AR. Some research suggests that smoking may increase both the prevalence and severity of AR. For instance, studies focusing on adolescents have shown a significantly higher prevalence of AR among smokers than among non-smokers [4]. However, a cross-sectional study from Sweden reported a surprisingly low prevalence of AR among male smokers. This study also found that the incidence of sensitization to common airborne allergens was lower among smokers or those with a history of smoking compared to non-smokers. Furthermore, in groups characterized by heavy smoking, there was no significant reduction in markers of atopic inflammation or quality of life measures related to upper airway inflammatory diseases, although a decrease in interleukin (IL)-33 levels was noted [5]. Nicotine, a primary constituent of tobacco smoke, has been shown to suppress inflammatory immune responses. Earlier studies have indicated a reduction in T helper 2 (Th2) cytokines (IL-4, IL-5, and IL-13), which may diminish allergic antigen sensitization by inhibiting pulmonary Th2 responses in smokers [6-8].
This study analyzed data from the Korea National Health and Nutrition Examination Survey (KNHANES) VII (2018) and VIII (2019–2021) to explore the relationship between smoking status, including the use of conventional cigarettes (CCs) and electronic cigarettes (ECs), and the prevalence and management of AR among Koreans. By examining a range of sociodemographic factors, health behaviors, and clinical indicators, such as urinary cotinine levels, the study aims to elucidate the nuanced effects of different smoking methods on AR. Additionally, we assessed cotinine levels, a biomarker for nicotine exposure, to investigate the dose-response relationship between smoking intensity and AR risk.
We also examined the correlation between smoking status, which includes both CC and EC use, and the prevalence of AR among Koreans. While the link between CC use and AR is well-documented, the impact of EC on AR prevalence remains unclear. With the growing popularity of ECs among younger demographics and their perceived safety compared to traditional smoking, some research suggests that ECs are relatively less harmful than CCs. However, other studies have raised concerns about the potential risks associated with the aerosol compounds and additives found in ECs [9,10]. In this study, we compared the use of CCs and ECs to better understand their potential effects on the prevalence and symptoms of AR.
MATERIALS AND METHODS
Survey participants
The data used in this study were derived from the KNHANES VII (2018) and VIII (2019–2021). The KNHANES is a nationwide cross-sectional survey conducted by the Korean Centers for Disease Control and Prevention. The inclusion criterion was age of ≥19 years. Meanwhile, participants without records of any parameter (age, body mass index [BMI], sex, alcohol consumption, area of residence, income, education level, diabetes mellitus [DM], hypertension [HTN], AR, asthma, atopic dermatitis, and urine cotinine level) were excluded from the study. Of the 30,551 study participants, 25,180 were aged ≥19 years. Among them, 2,890 participants without information for parameters were excluded. Therefore, data from 22,290 participants were used in the final statistical analysis (Fig. 1). As this study utilized only publicly available statistical data, it did not require institutional review board approval and informed consent from patients.
A flowchart of survey participants. Among a total of 30,551 participants from Korea National Health and Nutrition Examination Survey (KNHANES) 2018–2021, those aged under 19 years and with missing values for the variables (age, body mass index, sex, alcohol consumption, area of residence, income, education level, diabetes mellitus, hypertension, allergic rhinitis, asthma, atopic dermatitis, and urine cotinine level) were excluded. After applying these criteria, 22,290 participants were deemed eligible for inclusion in the study.
Defining sociodemographic variables
Data on age, sex, BMI, smoking, alcohol consumption, area of residence, income, education level, and comorbidities were collected. The income level of the respondents was categorized by separating them into two parts based on the average monthly income of their households. The educational level of the respondents was dichotomized based on the level of education they had completed. The alcohol consumption was classified as “none,” “less than once a month,” and “once a month or more,” based on whether excessive alcohol consumption occurred more than once a month in the last year.
Defining AR, current smoking status, and urine cotinine levels
Participants were considered to have physician-diagnosed AR if they answered “yes” to the question, “Have you ever been diagnosed with AR by a physician?” The current smoking status for CC, EC, and heated tobacco products was recorded through self-reported surveys. Participants’ smoking status was classified into three groups: non-smokers (control), CC (conventional cigarette) users, and EC (electronic cigarette) users. CC users referred to individuals who smoked only CC, while EC users included both EC users and hybrid users (those who used both EC and CC). In the KNHANES, smoking status was categorized as either “daily smoking” or “occasional smoking.” CC users were those who checked “daily smoking” or “occasional smoking” for CC. EC users were defined as individuals who checked “daily smoking” or “occasional smoking” for EC (including heated tobacco products and liquid-type ECs). Additionally, EC users included dual users, who checked “occasional smoking” for CC but “daily smoking” for EC.
Urine cotinine was collected through urine test, and urine cotinine levels were categorized into ranges: ≤10 ng/mL, 11–30 ng/mL, 31–500 ng/mL, and >500 ng/mL. AR management refers to whether patients with AR have ever used allergy medications, such as nasal sprays or antihistamines, and is one of the items included in the KNHANES survey used in this study. AR management was defined based on responses from a health questionnaire pertaining to the present therapeutic status, with 0 indicating no ongoing treatment and 1 indicating ongoing treatment.
Statistical analysis
Statistical analysis was performed using the Statistical Analysis System version 9.4 (SAS Institute Inc.). According to the smoking type of the participants, baseline characteristics were assessed using analysis of variance and Pearson chi-square test. Data are presented as means±standard errors and frequencies (percentages) for continuous and categorical variables, respectively. Statistical significance was set at P<0.05. Using a univariable logistic regression model, unadjusted odd ratios (ORs) with a 95% CI were calculated for the variables that indicated significant differences between the AR and non-AR groups. For the multivariable logistic regression analysis, confounding factors including demographic factors (age, sex, area of residence, income, and education), comorbidities (HTN, DM, and atopic dermatitis), health-related factors (alcohol consumption and current smoking status), laboratory factors (BMI and urine cotinine levels), and ear, nose, and throat-related factor (asthma) were used to confirm the adjusted association between current smoking status and prevalence of AR.
RESULTS
General characteristics of the study population according to smoking status
Altogether, 22,290 participants were evaluated, selected from the 30,551 respondents to the KNHANES VII and VIII surveys conducted between 2018 and 2021. A detailed flowchart illustrating the participant enrollment process is shown in Fig. 1. Participants were divided into three groups based on their smoking status: non-smokers (control), CC smokers (CC users), and EC smokers (EC users). Of the total participants, a significant majority, 81.2%, were classified as non-smokers, while the remaining 18.8% were identified as smokers. Within the smoking group, a substantial majority (75.13%) were CC users, and 24.87% were EC users.
EC users had a mean age of 37.6±11.14 years, which was significantly younger than the mean ages of 49.2±15.26 years and 52.9±17.06 years for CC users and non-smokers, respectively. An analysis of urine cotinine levels, a quantitative marker of smoking intensity, indicated no significant difference between EC users (1,325.3±847.13 ng/mL) and CC users (1,277.4±826.79 ng/mL), suggesting comparable nicotine intake between these groups. The distribution of sex among the smoking population did not show significant differences between CC and EC users. However, compared to the non-smoking group, a higher proportion of men were smokers (CC+EC; 34.8%) compared to women, who made up only 6%. Among smokers, men constituted 82.2%, while women accounted for 17.8% (P<0.001). There was also a noticeable trend of increased alcohol consumption among smokers, with EC users reporting more frequent alcohol consumption than CC users (81.46% vs. 74.28%, P<0.001).
Socioeconomic and lifestyle characteristics varied significantly among the groups. A higher percentage of EC users (88.47%) lived in neighborhoods or districts, compared to CC users (78.44%) and non-smokers (80.18%) (P<0.001). Additionally, EC users generally had higher incomes, with 71.09% falling into the middle-high or high income brackets, and higher education levels, with 53.7% holding college degrees or higher. These figures substantially exceed those of CC users (55.61% and 31.38%, respectively) and non-smokers (56.97% and 38.25%, respectively) in these socioeconomic indicators (P<0.001).
The prevalence of DM and HTN was significantly lower in the EC user group than in other groups, with rates of 4.61% and 10.09%, respectively (P<0.001). No significant relationships were found between smoking status and asthma prevalence (P=0.796). However, the prevalence of AR and atopic dermatitis showed notable trends. AR rates were lower among CC users (11.61%) and higher among EC users (19.79%), compared to non-smokers (15.38%). Additionally, the occurrence of atopic dermatitis was higher among EC users than in the other groups (P<0.001) (Table 1).
Characteristics of the study population according to smoking status and type
Univariable logistic regression analysis of AR: demographic, clinical, and lifestyle associations
In the univariable logistic regression analysis that explored the relationship between AR and various demographic and clinical factors, multiple significant associations emerged. The analysis indicated that the likelihood of AR decreased with increases in age and BMI. Specifically, the odds of developing AR dropped by 3% for each additional year of age (odds ratio [OR], 0.97; 95% CI, 0.968–0.972; P<0.001). Similarly, for every 1-kg/m2 increase in BMI, the odds of AR occurrence decreased by approximately 2.3% (OR, 0.977; 95% CI, 0.967–0.988; P<0.001). There were also notable sex differences; women were 43.3% more likely to develop AR compared to men (OR, 1.433; 95% CI, 1.328–1.546; P<0.001). Furthermore, an analysis of urine cotinine levels showed that for every 100-ng/mL increase in cotinine levels, the probability of AR occurrence decreased by 1.3% (OR, 0.987; 95% CI, 0.980–0.995; P=0.001).
Environmental and lifestyle factors were also significantly associated with the presence of AR. Consuming alcohol more than once per month was linked to a slight increase in the risk of AR (OR, 1.082; 95% CI, 1.005–1.165; P<0.001). Living in a neighborhood or district, as opposed to a town or township, was found to be protective against AR, reducing the odds by approximately 29.4% (OR, 0.706; 95% CI, 0.638–0.780; P<0.001). Higher income levels were associated with an increased risk of AR, with individuals in the middle-high or high-income brackets having 41.3% higher odds of AR compared to those with low or middle-low income (OR, 1.413; 95% CI, 1.309–1.525; P<0.001). Educational attainment further influenced the incidence of AR, with college graduates and those with higher education levels exhibiting a 64.9% increased likelihood of AR compared to those with a high school education or lower (OR, 1.649; 95% CI, 1.531–1.775; P<0.001). Comorbid conditions such as DM and HTN were inversely related to AR, significantly reducing the odds by 56.9% and 52.8%, respectively. In contrast, the presence of asthma or atopic dermatitis significantly increased the odds of AR by more than twofold (OR, 3.075 and 3.725, respectively; P<0.001 for both). An investigation into the association between current smoking status and the incidence of AR revealed a significant trend of decreased AR occurrence among CC users. According to the data, CC users had a 27.7% lower risk of developing AR compared to non-smokers (control group), with an OR of 0.723 (95% CI, 0.643–0.812; P<0.001). By contrast, EC users had a 35.8% higher risk of AR than non-smokers, as indicated by an OR of 1.358 (95% CI, 1.160–1.590; P<0.001). Thus, the use of ECs may increase the risk of developing AR, highlighting that the type of smoking can have differential effects on the likelihood of AR incidence. Cotinine levels were not statistically significant in predicting AR risk across most categories, except for levels exceeding 500 ng/mL, which were associated with a decrease in AR occurrence (OR, 0.792; P<0.001) (Table 2).
Univariable logistic regression analysis according to the presence of AR
Impact of smoking status on AR prevalence
In the adjusted multivariable logistic regression analysis designed to assess the impact of smoking status on AR prevalence, clear patterns were observed. By accounting for confounding factors such as age, sex, area of residence, education, income, alcohol consumption, and comorbid health conditions, the relationship between CC use and AR became more evident. The adjusted OR for CC users indicated a reduced likelihood of AR, with an OR of 0.797 (95% CI, 0.702–0.905; P=0.001), representing a 20% decrease in the odds of AR occurrence compared to non-smokers. This finding suggests that, consistent with results from unadjusted models, CC use may be inversely associated with AR when demographic and health-related factors are considered. In contrast, the relationship between EC use and AR initially appeared as an increased risk in unadjusted models (OR, 1.36; 95% CI, 1.160–1.590; P=0.001). However, this association was diminished after adjustment, with an OR of 1.05 (95% CI, 0.89–1.24; P=0.578), suggesting that the observed increased risk of AR among EC users might be influenced by other covariates (Table 3).
Multivariable logistic regression analysis according to the presence of AR
Cotinine levels and AR prevalence: analyzing the dose–response relationship
The relationship between cotinine levels and the prevalence of AR exhibited a dose-response trend. The reference group, with a cotinine level ≤10 ng/mL, was compared to other cotinine subgroups. In the initial, unadjusted analysis, participants with cotinine levels ranging from 11 to 30 ng/mL had an OR of 1.455 (95% CI, 0.917–2.307), and those with levels from 31 to 500 ng/mL had an OR of 1.154 (95% CI, 0.967–1.377). Neither group reached statistical significance, with P-values of 0.111 and 0.113, respectively. However, a significant decrease in the odds of AR occurrence was observed in individuals with a cotinine level >500 ng/mL, demonstrating an OR of 0.792 (95% CI, 0.711–0.882) with a P-value of <0.001. To more precisely evaluate the association between cotinine levels and the prevalence of AR, an analysis was conducted by categorizing cotinine levels in increments of 100 units (ng/mL). The results indicated that for every 100-ng/mL increase in cotinine concentration, the likelihood of AR occurrence decreased by 1.3% (OR, 0.987; 95% CI, 0.980–0.995; P=0.001). These findings suggest a significant inverse correlation between urinary cotinine levels and AR prevalence (Table 4).
Prevalence of allergic rhinitis according to the urine cotinine level
Comparative analysis of cotinine levels and AR odds among CC and EC users
We compared the ORs for different levels of cotinine among individuals with AR, and those using CC or EC. For AR controls, there were no significant differences in cotinine levels, with ORs close to 1 and P-values above 0.05. In contrast, CC users showed a significant increase in odds for the cotinine level 11–30 ng/mL subgroup (OR, 3.47; 95% CI, 1.48–5.99; P=0.007), while those with cotinine levels >500 ng/mL experienced a significant decrease (OR, 0.65; 95% CI, 0.437–0.968; P=0.034). EC users, however, exhibited a different pattern. There was no significant change in odds for the lower cotinine level subgroup (cotinine level 11–30 ng/mL) with an OR of 0.764 (95% CI, 0.084–6.944; P=0.811). For those with midrange cotinine levels (cotinine level 31–500 ng/mL), there was an observed increase in odds with an OR of 1.675 (95% CI, 0.81–3.463), though this increase did not reach statistical significance (P=0.164). The OR for EC users with cotinine levels >500 ng/mL was 1.06 (95% CI, 0.554–2.028), with a P-value of 0.860, indicating no significant association.
When analyzing the relationship between urine cotinine levels and the odds of AR occurrence among different types of smokers, it is essential to consider potential confounders that could influence the outcomes. In the unadjusted model, we explored the direct associations between cotinine levels and the likelihood of AR occurrence without accounting for other variables that might impact this relationship. While this approach offers a preliminary insight, it fails to consider additional factors like age, sex, socioeconomic status, environmental exposures, or genetic predispositions that could skew the results. Therefore, we also analyzed the adjusted statistical results.
In the adjusted multivariable logistic regression analysis, the relationship between CC users and AR showed a slightly different pattern compared to the unadjusted analysis. In the subgroup with cotinine levels ranging from 11–30 ng/mL, the OR was 3.359 (95% CI, 1.266–8.909) with a P-value of 0.015, mirroring the unadjusted results. However, in the 31–500 ng/mL cotinine level group, the OR increased to 1.521 (95% CI, 0.942–2.456), though this increase was not statistically significant (P=0.087). In contrast, for those with cotinine levels above 500 ng/mL, the OR rose from 0.65 in the unadjusted analysis to 1.114, but this change was also not statistically significant (P=0.628). For EC users, there was a slight increase in the OR across each cotinine concentration group; however, none of the results reached statistical significance, with P-values of 0.850, 0.110, and 0.491, respectively (Table 5).
Multivariable logistic regression analysis according to the urine cotinine level by current smoking status and type
Analyzing the impact of elevated cotinine levels on the management of AR
In this analysis, we further elucidated the association between varying cotinine levels and AR management. The unadjusted analysis suggested a trend toward less effective AR management with increasing cotinine concentrations. Specifically, participants with cotinine levels between 11 ng/mL and 30 ng/mL showed a nonsignificant decrease in AR management (OR, 0.956; 95% CI, 0.351–2.601; P=0.929). Similarly, those with levels between 31 ng/mL and 500 ng/mL did not demonstrate a significant association (OR, 0.809; 95% CI, 0.551–1.187; P=0.278). However, individuals with cotinine levels greater than 500 ng/mL exhibited a statistically significant reduction in the likelihood of AR management (OR, 0.781; 95% CI, 0.635–0.962; P=0.020). This inverse relationship at the highest cotinine levels underscores a potential negative correlation with AR management. After adjusting for confounding factors, the trend toward decreased effectiveness in AR management persisted, although it did not reach statistical significance in any cotinine group. The adjusted ORs were 0.937 (95% CI, 0.341–2.571; P=0.899) for the 11–30 ng/mL group, 0.791 (95% CI, 0.535–1.17; P=0.241) for the 31–500 ng/mL group, and 0.863 (95% CI, 0.69–1.081; P=0.200) for those with levels over 500 ng/mL. These findings suggest that any apparent effect of cotinine on AR management is likely attributable to other confounding variables, rather than a direct impact of cotinine levels themselves (Table 6).
Cotinine levels according to the management of AR
DISCUSSION
Using data from the KNHANES VII (2018) and VIII (2019–2021), we examined the correlation between AR prevalence and management among Korean CC and EC smokers. Additionally, we assessed the correlations of sociodemographic indicators, lifestyle factors, and clinical indices with smoking status and AR prevalence. Multivariable logistic regression analysis revealed a significant association between CC use and a lower prevalence of AR, particularly in CC users with cotinine levels of 500 or higher. The OR for this subgroup was 0.797 (95% CI, 0.702–0.905; P=0.001), indicating that these individuals had approximately 20% lower odds of experiencing AR. Additionally, there was a trend toward a decreasing OR value in AR management, suggesting a potential link between higher cotinine levels in CC users and reduced AR prevalence.
Our study’s analysis of demographic, socioeconomic, and lifestyle factors revealed that the majority of participants were non-smokers (81.2%), with the remainder consisting of CC (75.13%) and EC (24.87%) users. EC users, who had a mean age of ≤37.6 years, predominantly lived in urban areas (88.47%). This urban residency may be linked to easier access to ECs. Additionally, EC users generally had higher socioeconomic statuses, including income and education levels, compared to their CC-using and non-smoking counterparts. This group also tends to consume alcohol more frequently. The sex distribution among smokers and non-smokers was relatively consistent, showing no significant differences between CC and EC users. It is necessary to highlight the integrity of our smoker versus non-smoker classification, which is further supported by objective urine cotinine levels. Notably, about 7% of participants who self-identified as smokers in surveys were classified as non-smokers based on their cotinine levels. This underscores the importance of using objective biomarkers to accurately determine smoking status in studies investigating the health impacts of smoking.
The association between CC use and AR prevalence is a compelling aspect of our research findings. Statistical analyses support the hypothesis that heavy CC smoking may be negatively correlated with AR prevalence. Our data revealed that among individuals who smoked CCs, those with a cotinine level greater than 500 ng/mL had a significantly reduced risk of developing AR—approximately 20% lower—compared to non-smokers (OR, 0.8; 95% CI, 0.7–0.91; P<0.001), a result that remained robust in both univariate and multivariate logistic regression analyses. However, in the group with cotinine levels exceeding 500 ng/mL, the management of AR did not show statistical significance. In these analyses, potential confounding variables such as demographic factors, socioeconomic status, and other lifestyle choices were rigorously adjusted for, thereby reinforcing the observed association. This finding suggests a negative correlation between heavy CC smoking and both the development and management of AR.
Nicotine, a primary compound in tobacco products, plays a complex role in modulating the immune system, particularly in relation to allergic conditions such as AR. While it specifically reduces Th2 cell responses, which are crucial for triggering allergic reactions, it does not affect the functions of goblet and muscle cells that are responsible for mucus secretion and airway narrowing. This selective suppression could potentially reduce the severity of allergic reactions without alleviating the respiratory symptoms associated with smoking [5,8]. Active smoking aggravates irritation of the nasal mucosa, potentially worsening symptoms like congestion and a runny nose [11]. However, it may also decrease allergen sensitization due to the immunomodulatory properties of nicotine, indicating a dual effect on AR. To better understand how smoking and nicotine might influence the prevalence of AR, we reviewed various studies on the immunomodulatory effects of nicotine. The research indicates that nicotine helps suppress Th2 immune responses, which could reduce allergic sensitivity. Wang et al. [6] reported that nicotine’s anti-inflammatory effects are mediated through the α7 nicotinic acetylcholine receptor, potentially reducing Th2 cell activation involved in allergic reactions. Similarly, Mishra et al. discovered that nicotine diminishes the production of Th2 cytokines, such as IL-4, IL-5, and IL-13, by inhibiting pulmonary Th2 responses, which could help reduce allergen sensitization in smokers [8]. This mechanism might explain the lower prevalence of AR observed among cigarette users. Further studies corroborate these findings. Razani-Boroujerdi et al. [7] demonstrated that nicotine suppresses chronic inflammation but also reduces immune responses to viral infections, suggesting that certain immune responses may be suppressed in immune-mediated diseases like AR. Pisinger and Dossing [10] emphasized the role of nicotine in modulating specific immune responses, potentially leading to reduced allergic responses through Th2 cytokine suppression. Furthermore, Gomez et al. [5] identified differences in biomarkers for upper airway inflammatory diseases, such as AR, between smokers and non-smokers. Smokers exhibited lower levels of IL-33 than non-smokers, indicating reduced inflammatory markers related to allergic responses. These findings suggest that changes in the immune system induced by smoking may be linked to a decreased prevalence of AR [5]. In a comparison of ECs to CCs, Marques et al. [12] identified potential risks associated with the aerosol compounds and additives in ECs, which may have distinct effects on immune responses and respiratory health compared to CCs. Unlike nicotine, other chemical components in EC aerosols could uniquely impact the immune system.
ECs, which deliver nicotine without the harmful byproducts of tobacco combustion found in CC, present a different set of potential effects on respiratory health and allergic diseases [13-15]. The absence of combustion byproducts may reduce some risks associated with traditional smoking, yet the variety of ingredients and flavors in ECs introduces new factors to consider regarding their impact on the immune system and respiratory conditions [12,16]. The interaction between smoking habits and allergic diseases is complex, as evidenced by the differing effects of CCs and ECs on AR. This complexity is highlighted by the contrasting impacts of CCs and ECs on AR, underscoring the intricate nature of these relationships. The precise mechanisms, which may involve changes in the immune system’s sensitivity to allergens due to smoke constituents, require further investigation. Initial studies on the effects of ECs on AR might suggest a straightforward relationship; however, this becomes more complicated when adjusting for confounding variables. This indicates a need to clarify the specific impacts of EC aerosol components and additives.
A recent report by Rha et al. [17] revealed that EC use was significantly associated with a higher prevalence of AR in the Korean adult population. The study was a nationwide cross-sectional analysis that evaluated the relationship between EC use and the prevalence of chronic rhinosinusitis (CRS) and AR. Although both studies analyzed data from the same KNHANES and examined AR and CRS, the focus of this study was solely on AR, providing a more detailed analysis of the increased risk among CC and EC users. The findings indicated that EC users had a significantly higher prevalence of AR, and this study also reports that EC use increases the risk of AR by 35.8%. However, this association was only observed in the univariable logistic regression analysis, with no significant findings in the multivariable analysis. Additionally, no statistical analysis was conducted on CC users in their study. In contrast, this study included both CC and EC users, showing that while EC users had an increased risk of AR, CC users with urine cotinine levels of 500 ng/mL demonstrated a negative correlation with AR prevalence. This suggests differing impacts between EC and CC users. Furthermore, this study is unique in that it used urine cotinine levels to directly measure nicotine exposure. While the previous study primarily relied on self-reported survey data, this study quantitatively measured nicotine exposure through urine cotinine levels, providing more accurate data for the results. Nicotine is absorbed into the body through cigarette smoke and is metabolized into cotinine, which remains in the body longer than nicotine. This allows for the quantitative measurement of smoking status and consumption [18-20]. Since the results of this study were based on participants’ questionnaire responses, the self-reported smoking amount may not accurately reflect the actual smoking amount. Additionally, the half-life of cotinine, approximately 16–20 hours, limits its usefulness as a short-term indicator. Nevertheless, this biochemical marker can quantitatively assess the presence and extent of smoking, thereby overcoming the limitations associated with self-reported data. In these results, nicotine exposure was quantified through the analysis of urine cotinine levels, showing no significant difference between EC (1,325.3±847.13 ng/mL) and CC users (1,277.4±826.79 ng/mL). The lack of a significant difference in nicotine intake, as measured by urine cotinine levels between CC and EC users, suggests comparable levels of nicotine exposure despite the different sources.
Moreover, in a related study by Gomez et al. [5], comparisons between smokers and non-smokers among AR patients showed no significant differences in demographic characteristics or allergen sensitivity between the two groups. Serum and nasal lavage immunoglobulin E (IgE) levels were also statistically similar, with values of 2.94 kU/L in smokers versus 3.92 kU/L in non-smokers. However, there was a significant difference in salivary cotinine levels, with smokers exhibiting much higher levels. Notably, the study observed a significant decrease in IL-33 levels in smokers, both systemically and locally, indicating that smoking may impair natural defense mechanisms, including IgE-mediated inflammatory responses. Despite these reductions in inflammatory markers, there were no significant differences in the quality of life between smokers and non-smokers, as assessed by the Mini Rhinoconjunctivitis Quality of Life Questionnaire. These findings suggest that smokers might not fully recognize the adverse effects of smoking on their condition, which could complicate efforts to quit smoking [5].
The study also explored the relationships between AR and various demographic and clinical factors. There were notable correlations between age, BMI, and AR. Specifically, each 1-year increase in age was associated with a 3% decrease in the odds of having AR (OR, 0.97; 95% CI, 0.968–0.972; P<0.001). Similarly, each 1-kg/m2 increase in BMI was linked to a 2.3% reduction in the odds of AR occurrence (OR, 0.977; 95% CI, 0.967–0.988; P<0.001). We further investigated differences in mean BMI values among the three groups using the Tukey honest significant difference test, which revealed significant differences between the group means. Additionally, there was a significant association with sex; women had a 43.3% higher likelihood of having AR than men (OR, 1.433; 95% CI, 1.328–1.546; P<0.001). These findings are consistent with global trends and are not unique to Korea. Sensitivity to allergens such as pollen and dust mites has been shown to decrease with age. Consequently, as individuals age, their reactivity to allergens tends to diminish, potentially contributing to the lower prevalence of AR in older populations [21,22].
This study examined the correlation between smoking and AR using data from the KNHANES. However, several limitations must be considered. Firstly, the determination of allergic diseases in KNHANES was based on self-reported surveys, which inherently carry the risk of recall bias and response bias. Participants may either exaggerate or downplay their symptoms. Additionally, KNHANES employs a stratified multi-stage cluster sampling method, dividing the country into large cities, small and medium-sized cities, and rural areas. Efforts to minimize selection bias included using supplementary samples and targeting non-responsive households. Despite these efforts, challenges such as accessibility in remote areas and the non-participation of certain groups could still influence the data. Selection bias was addressed by pre-selecting supplementary samples to replace non-responsive households and making additional survey efforts in areas with low accessibility. Another significant limitation is that this study did not account for the detailed treatment regimens or specific management strategies employed for AR. These factors may play a crucial role in understanding the full scope of the disease and its management. Therefore, it is difficult to assess how variations in treatment modalities across individuals or regions may have influenced the prevalence or severity of AR.
Our findings suggest a potential AR inhibitory effect in CC users with high urine cotinine levels. However, it is important to emphasize that smoking in any form carries significant health risks and should not be considered a viable method for AR inhibition. Further research into the mechanisms by which smoking suppresses AR could be beneficial. Additionally, this study was based on survey data and urine cotinine levels from a specific population, which may limit its generalizability to other groups. Future research is necessary to verify these findings and to further investigate the mechanisms behind the observed relationships.
HIGHLIGHTS
▪ The prevalence of allergic rhinitis (AR) exhibited a 3% annual decrease with increasing age and a 2.3% decrease per unit increase in body mass index.
▪ Women exhibited a 43.3% higher prevalence of AR than men; additionally, urban residency and alcohol consumption more than once per month were significantly associated with an increased risk of AR.
▪ The presence of asthma or atopic dermatitis was linked to more than a twofold increase in the likelihood of developing AR.
▪ Urinary cotinine levels showed a significant inverse correlation with the prevalence and management of AR, decreasing the likelihood of AR occurrence by 1.3% for every 100-ng/mL increase in urinary cotinine concentration.
▪ In a multivariable logistic regression analysis of a national cross-sectional study, conventional cigarette use (indicated by urine cotinine levels >500 ng/mL) was associated with a lower prevalence of AR.
Notes
No potential conflict of interest relevant to this article was reported.
ACKNOWLEDGMENTS
This research was supported by the National Research Foundation (NRF) funded by the Korean government (MSIT) (RS-2024-00441029, 2020R1C1C1004572), ICT Creative Consilience Program through the Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korea government (MSIT) (IITP-2025-RS-2020-II201819, 34%), and a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant no. RS-2022-KH129266).
AUTHOR CONTRIBUTIONS
Conceptualization: JMS, THK. Methodology: YJ. Validation: YJ. Investigation: JMS, YJ, JK, JL. Data curation: JMS, YJ, JK, JL. Supervision: THK. Project administration: THK. Funding acquisition: THK. Writing-original draft: JMS, THK. Writing-review & editing: THK. All authors read and agreed to the published version of the manuscript.