Electronic Cigarette Use and Chronic Respiratory Symptoms among U.S. Adults

To the Editor:

E-cigarette use is rising in the United States, especially among adolescents and young adults (1). Although limited evidence suggests that e-cigarettes may contribute to a higher respiratory symptom burden, the role of age and concurrent tobacco smoking in this relationship remains unclear (2–8). We examined the association of e-cigarette use with chronic respiratory symptoms, specifically focusing on young adults and never tobacco smokers—groups traditionally at low risk for respiratory symptoms.

Methods

We analyzed data on noninstitutionalized U.S. adults (≥18 years old) from the 2017 Behavioral Risk Factor Surveillance System (BRFSS), a national health-related telephone panel survey conducted annually by the CDC. In 2017, 11 states collected data on both respiratory health and e-cigarette use. We included individuals who responded to at least one respiratory symptom question plus the e-cigarette use question.

E-cigarette use was defined as responding “every day” or “some days” to “Do you now use e-cigarettes or other electronic vaping products every day, some days, or not at all”? We considered individuals who responded “not at all” as unexposed. We excluded individuals who responded “don’t know” or refused to answer.

Our primary composite outcome was chronic respiratory symptoms, defined as responding “yes” to at least one question assessing daily cough, sputum production, or breathlessness during the past 3 months. We considered missing responses to individual questions as negative responses, provided that the respondents answered at least one question. Secondary outcomes included cough, sputum production, or breathlessness separately. For these secondary outcomes, we excluded participants who refused the respective question.

We used multivariable log-binomial regression to estimate adjusted prevalence ratios for the association of e-cigarette use with respiratory symptoms. We applied BRFSS weights (9) to minimize potential bias from differential survey response rates and selection probabilities. Given our clinical question, we stratified the models by tobacco smoking status (current, recent-former [quit ≤ 1 yr], remote-former [quit > 1 yr], and never) and age group (18–34, 35–54, and 55+ yr). The final models adjusted for sex, obesity (body mass index ≥30 vs. <30 kg/m²), and cardiac or respiratory disease. We prespecified sensitivity analyses excluding individuals at high risk for respiratory symptoms (current inhaled-marijuana users and participants who reported chronic obstructive pulmonary disease, asthma, or heart disease).

Results

Of 87,067 individuals who answered both respiratory health and e-cigarette questions, 2,992 (weighted prevalence 4.6%) reported current e-cigarette use. Compared with nonusers, e-cigarette users were younger, more often male, and more often reported prior respiratory diagnoses. Nearly half (49.5%) currently smoked tobacco (Table 1). Overall, chronic respiratory symptoms were more prevalent among current tobacco smokers than among never-smokers (53.7% vs. 27.2%), and among persons ≥55 years of age than among those 18–34 years of age (43.5% vs. 27.8%).

Table 1.

Characteristics of Respiratory Question Respondents by E-Cigarette Use, Behavioral Risk Factor Surveillance System, 2017^*

Characteristics	E-Cigarette Never-Users† (%) (n = 84,075)	E-Cigarette Users† (%) (n = 2,992)
Sex, M	47.6	57.1
Age, yr
18–34	26.1	49.0
35–54	32.2	31.4
≥55	41.7	19.6
Body mass index ≥ 30 kg/m2	28.5	26.4
Tobacco smoking status
Never-smoker	59.7	20.3
Remote-former smoker	21.3	18.6
Recent-former smoker	2.7	10.8
Current smoker	15.3	49.5
Inhaled marijuana use	1.8	7.7
Diagnosed comorbidity
Coronary heart disease	7.1	5.9
Chronic obstructive pulmonary disease	7.4	12.2
Asthma	12.6	18.8

^*Includes individuals from Arizona, Florida, Georgia, Minnesota, Nevada, Tennessee, West Virginia, Iowa, Kansas, Nebraska, and Washington, D.C. who answered at least one respiratory symptom question and the question assessing e-cigarette use.

^†Weighted prevalence using standard Behavioral Risk Factor Surveillance System weights.

After adjustment for sex, obesity, and cardiopulmonary disease, e-cigarette use was not associated with respiratory symptoms among current or recent-former smokers across age groups, or among never-smokers and remote-former smokers ≥55 years of age (Table 2). In contrast, among younger never-smokers and remote-former smokers, e-cigarette use was associated with a higher prevalence of respiratory symptoms (never-smokers age 18–34 yr [prevalence ratio (PR), 1.36; 95% confidence interval (CI), 1.08–1.70]; remote-former smokers age 18–34 yr [PR, 1.27; 95% CI, 0.72–2.22]).

Table 2.

Adjusted Prevalence and Prevalence Ratios for Any Respiratory Symptoms with E-Cigarette Use, Stratified by Age Group and Smoking Status

Subgroup	Respondents (n)	Adjusted Prevalence of Symptoms* (%)		Adjusted Prevalence Ratio* (95% Confidence Interval)
		E-Cigarette Never-Users	E-Cigarette Users
Never-smokers
18–35 yr	16,656	22.1	30.0	1.36 (1.08–1.70)
36–54 yr	16,673	22.6	26.3	1.16 (0.67–2.01)
≥55 yr	17,570	35.8	35.8	1.00 (0.69–1.46)
Remote-former smokers†
18–35 yr	1,706	24.5	31.0	1.27 (0.72–2.22)
36–54 yr	4,850	26.5	36.2	1.37 (1.18–1.59)
≥55 yr	12,033	47.0	51.7	1.10 (0.97–1.25)
Recent-former smokers‡
18–35 yr	1,053	30.0	26.7	0.89 (0.57–1.40)
36–54 yr	932	33.2	33.3	1.00 (0.70–1.44)
≥55 yr	775	54.6	50.1	0.92 (0.78–1.08)
Current smokers
18–35 yr	4,257	45.3	52.5	1.16 (0.99–1.36)
36–54 yr	5,591	52.7	54.4	1.03 (0.91–1.17)
≥55 yr	4,971	61.0	60.0	0.98 (0.89–1.09)

^*Adjusted for sex, obesity, and any diagnosis of coronary heart disease, chronic obstructive pulmonary disease, or asthma. Estimates also apply standard Behavioral Risk Factor Surveillance System weights.

^†Quit > 1 year.

^‡Quit ≤ 1 year.

Across age groups, e-cigarette use was not associated with individual symptoms among current or recent-former smokers. Among remote-former smokers ≥55 years of age, we found a lower prevalence of sputum production with e-cigarette use (PR, 0.42; 95% CI, 0.26–0.67). Conversely, among young never-smokers, the association of e-cigarette use with chronic respiratory symptoms was primarily driven by a higher prevalence of cough (age 18–34 yr [PR, 1.60; 95% CI, 1.11–2.31]).

Similar results were obtained after we excluded persons who reported current inhaled-marijuana use or had received a diagnosis of cardiopulmonary disease (data not shown).

Discussion

Among a representative sample of U.S. adults, e-cigarette use was associated with a higher prevalence of chronic respiratory symptoms among young never-smokers and remote-former smokers. Our findings suggest that the association of chronic respiratory symptoms with e-cigarette use is present only among individuals with a lower background risk. The persistence of our results after we excluded inhaled-marijuana users and respondents with diagnosed cardiopulmonary disease argues against alternate explanations for the observed associations.

Our study is the first to examine the association of e-cigarette use with respiratory symptoms among U.S. adults and captures the largest sample to date of e-cigarette users who never smoked tobacco. Our results agree with those of the largest prior study to account for smoking status, in which the prevalence of wheezing associated with e-cigarette use was higher among never-smokers only (3). A study of Swedish never-smokers also suggested a higher prevalence of respiratory symptoms among e-cigarette users (odds ratio, 1.46; 95% CI, 0.93–2.29) (2).

The strengths of our study include the robust methodology of BRFSS, with sampling representative of the U.S. population (10). However, the cross-sectional design limits our ability to infer causality. Some current and recent-former smokers may have started using e-cigarettes because of chronic respiratory symptoms, obscuring any potential detrimental effect in these subgroups. Similarly, the lower prevalence of sputum production with e-cigarette use among remote-former smokers ≥55 years of age could reflect reverse causality. Nonetheless, it is unlikely that young never-smokers with respiratory symptoms would have used e-cigarettes more than individuals without respiratory symptoms.

The study’s reliance on self-report and the relative time difference between respiratory symptom (“during the past 3 months”) and e-cigarette use (“now”) questions further limit causal inference. The small number of e-cigarette users in certain subgroups may have reduced our ability to detect associations. BRFSS did not query about the intensity or duration of e-cigarette use, which would help clarify the magnitude and mechanisms of associated risk. Finally, we could not account for inhaled medication use; however, the persistence of our findings after we excluded individuals with chronic obstructive pulmonary disease or asthma suggests that this did not influence our results.

In summary, e-cigarette use was associated with a higher prevalence of chronic respiratory symptoms among young never-smokers and remote-former tobacco smokers. It remains unclear whether this higher prevalence of chronic respiratory symptoms predicts future respiratory impairment (8). With the rising prevalence of e-cigarette use among young adults, our findings underscore the importance of counseling patients about their potential harm. Longitudinal studies examining the long-term risks of e-cigarette use among traditionally low-risk populations are needed.