Tobacco Use and Respiratory Symptoms Among Adults: Findings From the Longitudinal Population Assessment of Tobacco and Health (PATH) Study 2014-2016

Abstract

Introduction: We examined the relationship between current tobacco use and functionally important respiratory symptoms.

Methods: Longitudinal cohort study of 16 295 US adults without COPD in Waves 2-3 (W2-3, 2014-2016) of the Population Assessment of Tobacco and Health Study. Exposure-Ten mutually exclusive categories of tobacco use including single product, multiple product, former, and never use (reference). Outcome-Seven questions assessing wheezing/cough were summed to create a respiratory symptom index; cutoffs of ≥2 and ≥3 were associated with functional limitations and poorer health. Multivariable regressions examined both cutoffs cross-sectionally and change over approximately 12 months, adjusting for confounders.

Results: All tobacco use categories featuring cigarettes (>2/3's of users) were associated with higher risk (vs. never users) for functionally important respiratory symptoms at W2, for example, at symptom severity ≥ 3, risk ratio for exclusive cigarette use was 2.34 [95% CI, 1.92, 2.85] and for worsening symptoms at W3 was 2.80 [2.08, 3.76]. There was largely no increased symptom risk for exclusive use of cigars, smokeless tobacco, hookah, or e-cigarettes (adjustment for pack-years and marijuana attenuated the cross-sectional e-cigarette association from 1.53(95% CI 0.98, 2.40) to 1.05 (0.67, 1.63); RRs for these products were also significantly lower compared to exclusive use of cigarettes. The longitudinal e-cigarette-respiratory symptom association was sensitive to the respiratory index cutoff level; exclusive e-cigarette use was associated with worsening symptoms at an index cutoff ≥ 2 (RR = 1.63 [1.02, 2.59]) and with symptom improvement at an index cutoff of ≥ 3 (RR = 1.64 [1.04, 2.58]).

Conclusions: Past and current cigarette smoking drove functionally important respiratory symptoms, while exclusive use of other tobacco products was largely not associated. However, the relationship between e-cigarette use and symptoms was sensitive to adjustment for pack-years and symptom severity.

Implications: How noncigarette tobacco products affect respiratory symptoms is not clear; some studies implicate e-cigarettes. We examined functionally important respiratory symptoms (wheezing/nighttime cough) among US adults without COPD. The majority of adult tobacco users smoke cigarettes and have higher risk of respiratory symptoms and worsening of symptoms, regardless of other products used with them. Exclusive use of other tobacco products (e-cigarettes, cigars, smokeless, hookah) was largely not associated with functionally important respiratory symptoms and risks associated with their use was significantly lower than for cigarettes. The association for e-cigarettes was greatly attenuated by adjustment for cigarette pack-years and sensitive to how symptoms were defined.

Figure 1.

Unweighted lowess smoothed curves^a illustrating the relation between the proportion with functionally important respiratory symptoms^b and intensity of cigarette smoking at Wave 2 of the Population Assessment of Tobacco and Health (PATH) Study for four groups of tobacco users among whom cigarette smoking played a prominent role in their tobacco use.^{c,d. a}Unweighted lowess smoothed curves show the proportion of the four tobacco use groups that contain cigarette users with functionally important respiratory symptoms with the quantity of monthly cigarettes consumed. Cigarettes per month was calculated by multiplying how many cigarettes a respondent used each day by the number of days they used the products in the past 30 days. This variable was modified by taking outliers above the 95th percentile and recoding them to the 95th percentile value based on descriptive analysis of data for cases in the analysis sample who used more than 0 cigarettes per month. In the PATH Study, only ever cigarette users were asked about their cigarette consumption, so respondents who were never cigarette smokers were then recoded as having consumed 0 cigarettes per month. ^bCut-off level for what is considered functionally important is a respiratory index value of ≥3. ^cWeighted proportion (SE) with P30D cigarette use in each polycombustible group: Multiple combustible product use = 90.0% (1.0). Multiple combustible and noncombustible product use = 80.4% (1.5). ^dWeighted mean (SE) monthly cigarettes consumed for each group. Exclusive use of cigarettes = 313 (5.5). Multiple combustible product use = 224 (9.1). Cigarettes + e-cigarettes = 296 (9.6). Multiple combustible and noncombustible product use = 216 (9.2).

Figure 2.

Relation between the intensity^a of tobacco product use and weighted percentage with functionally important respiratory symptoms for exclusive users of cigarettes or e-cigarettes, at Wave 2 of the Population Assessment of Tobacco and Health (PATH) Study,^b after adjusting for pack years of cigarette smoking. ^aCigarettes and e-cigarettes per day are calculated by multiplying how many product units a respondent used each day by the number of days they used the products in the past 30 days and then dividing by 30. These variables were modified by taking outliers above the 95th percentile and recoding them to the 95th percentile value based on descriptive analysis of data for cases in the analysis sample who used more than 0 cigarettes or e-cigarettes per month. For e-cigarettes, participants were asked about daily use depending on the type of e-cigarette they regularly used. The five applicable types were disposable e-cigarette, nonrefillable cartridge e-cigarette, refillable cartridge e-cigarette, refillable tank system e-cigarette, or unknown e-cigarette. Depending on their type, respondents were asked on average, how many (1) e-cigarettes; (2) e-cigarette cartridges; or (3) milliliters of e-liquid they now use each day. ^bN = 9402 adult respondents without chronic obstructive pulmonary disease or other nonasthma respiratory disease and with PATH Study longitudinal (all-waves) weights and complete data on all study variables, plus Wave 2 cigarette and e-cigarette frequency/intensity variables. Unweighted Ns for each mutually exclusive use group: Never tobacco = 5888, Non-daily exclusive cigarette use = 978, Daily exclusive cigarette use Q1 (1–9 per day) = 597, Daily exclusive cigarette use Q2 (10–13 per day) = 543, Daily exclusive cigarette use Q3 (14–20 per day) = 936, Daily exclusive cigarette use Q4 (Greater than 1 pack per day) = 175, Non-daily exclusive e-cigarette use = 174, Daily exclusive e-cigarette use Q1 (1 per day) = 35, Daily exclusive e-cigarette use Q2 (2 per day) = 34, Daily exclusive e-cigarette use Q3 (3-4 per day) = 17, Daily exclusive e-cigarette use Q4 (Greater than 4 e-cigarettes per day) = 25. Quartile values were determined based on descriptive analysis of data for cases in the analysis sample who were exclusive, everyday users of either cigarettes or e-cigarettes. Error bars represent 95% confidence interval.