Preventive Medicine Reports 30 (2022) 102016

Contents lists available at ScienceDirect

Preventive Medicine Reports

journal homepage: www.elsevier.com/locate/pmedr

Tobacco use profiles by respiratory disorder status for adults in the wave
1-wave 4 population assessment of tobacco and health (PATH) study
Jamie Cordova a, b, Ruth M. Pfeiffer c, Kelvin Choi d, Rachel Grana Mayne a, Laura Baker e,
Jacqueline Bachand f, Kristen Constantine g, Sean Altekruse h, Carolyn Reyes-Guzman a, *
a
Tobacco Control Research Branch, Behavioral Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, National Institutes of
Health, 9609 Medical Center Drive, MSC 9776, Bethesda, MD 20892, United States
b
Noninfectious Disease Programs, National Foundation for the Centers for Disease Control and Prevention, P.O. Box 117300, Atlanta, GA 30368, United States
c
Biostatistics Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, 9609 Medical Center Drive, MSC 9776,
Bethesda, MD 20892, United States
d
Division of Intramural Research, National Institute of Minority Health and Health Disparities, National Institutes of Health, 6707 Democracy Blvd., Suite 800, Bethesda,
MD 20892, United States
e
The Bizzell Group, LLC, 4500 Forbes Blvd., Suite 400, Lanham, MD 20706, United States
f
US Department of Housing and Urban Development, Office of Policy Development and Research, 451 7th St. SW, Washington, D.C. 20410, United States
g
Center for Medicare & Medicaid Innovation, U.S. Centers for Medicare & Medicaid Services. 7500 Security Boulevard, Baltimore, MD 21244, United States
h
National Heart, Lung, and Blood Institute, National Institutes of Health, 31 Center Drive, Bethesda, MD 20892, United States

A R T I C L E I N F O A B S T R A C T

Keywords: Limited evidence exists on the association between electronic nicotine delivery systems (ENDS) and chronic
Cigarette smoking respiratory disorders. This study examines the association of combustible tobacco and ENDS use with chronic
Electronic nicotine delivery systems respiratory disorders among US adults. Public-use data from the Population Assessment of Tobacco and Health
E-cigarettes
(PATH) Study Wave 1 (2013–2014), Wave 2 (2014–2015), Wave 3 (2015–2016), and Wave 4 (2016–2018) were
Vaping
pooled. Analyses focused on adults with W1–W4 respiratory disorder data and current tobacco use at W4, as well
Tobacco co-use
Asthma as youth entering the adult cohort at W2 through W4 (N = 26,072). We fit weighted multivariable logistic
COPD regression models for each respiratory outcome (asthma, COPD, bronchitis) using W4 longitudinal weights.
Respiratory disorders Cigarette smokers (adjusted odds ratio [AOR] = 0.8, 95 % CI 0.7–0.9) were less likely to report an asthma
Chronic respiratory diseases diagnosis (p = 0.013). In contrast, ENDS users (AOR = 6.5, 95 % CI 3.7–11.5), cigarette smokers (AOR = 6.1, 95
Bronchitis % CI 4.0–9.1), dual users of cigarettes and ENDS (AOR = 5.4, 95 % CI 3.4–8.7), current users of non-cigarette
combustible, smokeless, and polytobacco products (AOR = 4.4, 95 % CI 3.1–6.4), and former users of any
product (AOR = 3.0, 95 % CI 1.9–4.7) had significantly elevated odds of reporting a diagnosis of COPD (p <
0.001). Similar patterns to COPD were observed for bronchitis (p < 0.001). Current and former tobacco use,
including ENDS, were significantly associated with prevalence of self-reported COPD and bronchitis after con­
trolling for demographic and psychosocial confounders.

1. Introduction Health Consequences of Smoking, 2014). However, national surveil­

lance data show that many individuals with COPD and asthma in the US
Combustible cigarette smoking has been linked to multiple respira­ currently smoke cigarettes. Among adults of all ages, 38 % of those with
tory diseases and is known to exacerbate chronic respiratory disorders COPD were current smokers in 2013 (Wheaton et al., 2019), and nearly
(Forey et al., 2011; Tamimi et al., 2012). The US Surgeon General has 18 % of those with asthma were current smokers spanning from 2014 to
concluded that there is sufficient evidence to infer that smoking causes 2017 (Deshpande et al., 2020).
poor asthma control and asthma exacerbation in adults and is the Electronic nicotine delivery systems (ENDS, which include e-ciga­
dominant cause of chronic obstructive pulmonary disorder (COPD) (The rettes) have been on the US market since 2007 and have sometimes been

* Corresponding author at: National Institutes of Health, National Cancer Institute, Division of Cancer Control & Population Sciences, Tobacco Control Research
Branch, 9609 Medical Center Drive, Room 3E564, MSC 9761, Bethesda, MD, United States.
E-mail address: carolyn.reyes-guzman@nih.gov (C. Reyes-Guzman).

https://doi.org/10.1016/j.pmedr.2022.102016
Received 18 January 2022; Received in revised form 6 October 2022; Accepted 9 October 2022
Available online 12 October 2022
2211-3355/Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
J. Cordova et al. Preventive Medicine Reports 30 (2022) 102016

marketed as a safer alternative to combustible cigarette smoking (Grana 2.2. Respiratory disorder status
and Ling, 2014; Klein et al., 2016). Over the past decade, the prevalence
of ENDS use has increased, especially among youth (Wang et al., 2018) Respiratory disorder status was defined as either a lifetime diagnosis
and young adults aged 18–24 (Dai and Leventhal, 2019), who perceive at W1 or a past 12-month diagnosis at W2, W3, or W4, of any of the
e-cigarettes as less harmful and less addictive than cigarettes (E-Ciga­ following conditions: asthma, bronchitis, or COPD. At W1, respondents
rette Use Among Youth and Young Adults, 2016). Additionally, current were asked to self-report an ever, or lifetime, diagnosis of a respiratory
adult e-cigarette users often cite a desire to quit or reduce cigarette disorder: “Has a doctor, nurse or other health professional ever told you
smoking, and believe that e-cigarettes pose less harm than cigarettes, as that you had any of the following lung or respiratory conditions?” At
some of the main reasons for using them (Berg, 2016; Coleman et al., subsequent waves, respondents were asked to self-report a diagnosis of a
2017). In particular, individuals who smoke and have respiratory dis­ respiratory disorder within the past 12 months: “In the past 12 months,
orders may be attracted by these perceived benefits of e-cigarettes. has a doctor, nurse, or other health professional told you that you had
Studies have found that current smokers with asthma have higher odds any of the following lung or respiratory conditions?” Additionally, we
of ever and current e-cigarette use compared to smokers without asthma included new baseline participants who reported a lifetime diagnosis of
(Deshpande et al., 2020; Kruse et al., 2017), and smokers with COPD any condition.
have higher odds of ever e-cigarette use compared to those without
COPD (Kruse et al., 2017). 2.3. Tobacco use profiles
Although ENDS liquids and aerosols typically contain lower levels of
carcinogens than combustible cigarette smoke (Eaton et al., 2018), they Current tobacco use was determined by “someday” or “every day”
contain numerous toxicants that can irritate or impair lung tissue, use of at least one of the following products at W4: cigarettes, ENDS,
including some substances not found in combustible cigarettes (Clapp hookah, smokeless tobacco, snus or snuff, traditional cigars, filtered
and Jaspers, 2017). E-cigarette users have reported adverse effects such cigars, cigarillos, or pipes. Former users were defined as individuals who
as cough, dry or irritated mouth/throat, and shortness of breath (King met the lifetime threshold for established use of the product of interest,
et al., 2019). Cross-sectional studies of adolescents (Choi and Bernat, but who reported not currently using the product at the time of assess­
2016; McConnell et al., 2017), and adults19 (Li et al., 2019), have found ment or within the past 12 months. For cigarettes, the threshold for
higher rates of wheezing, coughing, asthma attacks, and difficulty established use was smoking 100 cigarettes in lifetime and currently
breathing among current e-cigarette users than among non-users. smoking at least “fairly regularly.” For all other products, the threshold
Further, a 2019 review of the respiratory effects of e-cigarette use for established use was ever use of the product “fairly regularly.”
concluded that e-liquids and aerosols have adverse biologic effects on We subsequently created a set of tobacco use profiles to represent
organ and cellular health in humans, animals, and in vitro, but the different tobacco use histories and patterns. We used existing derived
impact on long-term health outcomes is unclear (Gotts et al., 2019). variables from the PATH W4 dataset to construct the tobacco use pro­
Few epidemiological studies have examined the relationship be­ files. These were: (1) “ENDS only” (N = 621: current exclusive ENDS
tween ENDS use and prevalence of chronic respiratory diseases. In this users), using the PATH derived variables for current use of ENDS and no
study, we examined the association of combustible tobacco and ENDS current use of all other products; (2) “cigarettes only” (N = 5,776:
use with multiple chronic respiratory disorders in a sample of US adults current exclusive combustible cigarette smokers), using variables for
weighted to represent the US adult population. current use of combustible cigarettes and no current use of all other
products; (3) “dual cigarettes and ENDS” (N = 806: current concurrent
2. Methods users of only cigarettes and ENDS), using variables for current use of
ENDS and cigarettes, and no current use of all other products; (4)
2.1. Data source and analytic population “current users of non-cigarette combustible, smokeless, and polytobacco
products” (N = 3,786: any combination of single or poly use of
The Population Assessment of Tobacco and Health (PATH) study is a combustible cigarettes, ENDS, traditional cigars, filtered cigars, ciga­
US representative longitudinal cohort of approximately 49,000 adults rillos, pipes, hookah, smokeless tobacco, and snus (with the exception of
(18 years and older) and youth (12–17 years). The study collects exclusive cigarette and ENDS use), using derived variables for each
epidemiologic data on tobacco use patterns and associated health out­ product; (5) “former users of any product” (N = 4,872: former estab­
comes; data for Wave 1 were collected from 2013 to 2014, Wave 2 from lished users of combustible cigarettes, ENDS, traditional cigars, filtered
2014 to 2015, Wave 3 from 2015 to 2016, and Wave 4 from 2016 to cigars, cigarillos, pipes, hookah, smokeless tobacco, and snus, using the
2018, which we analyzed in 2020. To recruit a sample representative of derived former established use variables for each product; and (6) “non-
the civilian, non-institutionalized US population, PATH employed a current, non-former, or never users of any product” (N = 10,211), the
stratified address-based, area-probability sample design with over­ reference group, included participants who did not report current or
sampling of tobacco users, young adults (18–24 years), and African former use of any product or who were never users of any product.
Americans. Non-response adjusted population and replicate weights Groups 1–4 are illustrated in Fig. 1 “Scenario I.”
were created to account for oversampling of certain groups and possible
deficiencies in the sampling frame. Further information regarding the 2.4. Covariates
PATH Study design and methods are available at https://www.icpsr.
umich.edu. Our analysis was based on deidentified data and was We examined sociodemographic characteristics assessed at W4
deemed exempt from Institutional Review Board. including age group (18–24, 23–34, 35–54, and ≥ 55 years); education
We pooled PATH adult questionnaire data from Wave 1 (W1, N = (some high school or less, high school graduate or GED, some college or
32,320), Wave 2 (W2, N = 28,362), Wave 3 (W3, N = 28,148), and Wave associate’s degree, and college degree or greater); race/ethnicity (Non-
4 (W4, N = 33,822 to increase sample size and statistical power due to Hispanic (NH) White, NH Black/African American, NH Other (including
the small number of respondents in select tobacco user groups with multi-racial), and Hispanic or Latino); annual household income
respect to respiratory disorders. Analyses were restricted to adults, as (<$10,000, $10,000–$24,999, $25,000–$49,999, $50,000–$99,999,
well as youth who aged into the adult cohort during W2 to W4, who and ≥$100,000); and marital status (married, widowed, divorced,
reported respiratory disorder data across all four waves and tobacco use separated, and never married).
data assessed at W4 (N = 26,072). Behavioral health symptomatology was measured using the GAIN-SS
screener. Per GAIN-SS guidelines, we calculated composite scores for
internalizing behaviors (e.g., depression, anxiety), externalizing

2
J. Cordova et al. Preventive Medicine Reports 30 (2022) 102016

were still significant after a Bonferroni adjustment for multiple testing.

We used a p-value of 0.05/30 = 0.0017 where 30 was the total number
of tests conducted (6 models with 5 parameters).

3. Results

The final analytic sample contained 26,072 adults, a majority of

whom were female (52.0 %), aged 35 years and older (70.4 %), married
(51.8 %), NH White (65.1 %), had at least some college education (61.1
%), and reported an annual household income of at least $25,000 (64.8
%), as shown in Table 1. The overall prevalence of ever having been
diagnosed with a respiratory disorder was 14.2 % for asthma, 4.5 % for
COPD and 6.3 % for bronchitis, while the prevalence of high probability
of a behavioral health disorder was: 22.8 % for an internalizing
behavioral disorder, 9.0 % for an externalizing behavioral disorder, and
4.4 % for a substance use disorder. Additionally, 1.3 % were exclusive
ENDS users; 13.6 % were exclusive cigarette smokers; 1.7 % were dual
cigarettes and ENDS users; 8.7 % were current users of non-cigarette
combustible, smokeless, and polytobacco products; 24.0 % were
former users of any product; and 50.8 % were non-current, non-former,
or never users of any product.
Table 2 summarizes results from minimally and fully adjusted
models for asthma, COPD, and bronchitis. In minimally adjusted models,
current users of non-cigarette combustible, smokeless, and polytobacco
products (AOR = 1.3, 95 % CI 1.1–1.5) and former users of any product
(AOR = 1.2, 95 % CI 1.1–1.4) had increased odds of reporting an ever
asthma diagnosis as compared to non-current, non-former, or never
users of any product (p = 0.002), while there was no significant asso­
ciation with asthma for exclusive ENDS users, exclusive cigarette
smokers, and dual cigarettes and ENDS users. Conversely, in fully
adjusted models, exclusive cigarette smokers (AOR = 0.8, 95 % CI
0.7–0.9) were less likely to report an asthma diagnosis as compared to
non-current, non-former, or never users of any product (p = 0.013),
while there was no significant association with asthma for exclusive
ENDS users, dual cigarettes and ENDS users, current users of non-
cigarette combustible, smokeless, and polytobacco products, and
former users of any product.
In fully adjusted models, the odds of COPD were much higher for
Fig. 1. Tobacco use profiles, PATH Study, 2013–2018. exclusive ENDS users (AOR = 6.5, 95 % CI 3.7–11.5), exclusive cigarette
smokers (AOR = 6.1, 95 % CI 4.0–9.1), dual users of cigarettes and ENDS
(AOR = 5.4, 95 % CI 3.4–8.7), current users of non-cigarette combus­
behaviors (e.g., hyperactivity, impulsivity), and substance use, and
tible, smokeless, and polytobacco products (AOR = 4.4, 95 % CI
classified individuals into two levels of severity: low/moderate versus
3.1–6.4), and former users of any product (AOR = 3.0, 95 % CI 1.9–4.7),
high probability of a diagnosis.
as compared to non-current, non-former, or never users of any product
(p < 0.001). Compared to the reference group, exclusive cigarette
2.5. Analyses smokers (AOR = 1.7, 95 % CI 1.3–2.2), exclusive dual cigarette and
ENDS users (AOR = 2.3, 95 % CI 1.6–3.5), current users of non-cigarette
We examined weighted distributions of tobacco use, sociodemo­ combustible, smokeless, and polytobacco products (AOR = 1.8, 95 % CI
graphic, and behavioral characteristics. Chi-squared tests assessed dif­ 1.4–2.4), and former users of any product (AOR = 1.6, 95 % CI 1.2–2.1),
ferences between respondents with or without asthma, COPD, or had nearly twice the odds of reporting a diagnosis of bronchitis (p <
bronchitis. To examine associations of tobacco use profiles with each 0.001). Supplemental Table 1 presents additional results from fully
chronic respiratory outcome, we fit weighted logistic regression models adjusted models for each respiratory disorder outcome. Of note, a par­
with two different sets of adjustment variables. First, minimally adjusted ticipant’s indication of an internalizing behavior disorder also increased
models included only the aforementioned sociodemographic predictors. the odds of reporting an asthma (AOR = 1.3, 95 % CI 1.1–1.4, p = 0.001)
Then, we fit fully adjusted models for sociodemographic, internalizing or bronchitis (AOR = 1.7, 95 % CI 1.4–2.1, p < 0.001) diagnosis.
behavior, externalizing behavior, substance use disorders, and other As part of a sensitivity analysis, we created a second set of tobacco
respiratory disorder covariates (i.e., models for COPD included asthma use profiles to assess whether altering the profiles would impact our
and bronchitis as adjustment variables). In minimally and fully adjusted findings. As illustrated in Fig. 1 “Scenario II,” we grouped individuals
models, no evidence of collinearity was found when assessing the vari­ differently from “Scenario I” to examine patterns of single and poly to­
ance inflation factor (VIF). Variances were estimated using the balanced bacco use as: (1) “any ENDS, single or poly” (N = 2,148), defined as
repeated replications (BRR) method with Fay’s adjustment = 0.3 to in­ exclusive use of ENDS or co-use of ENDS with any combustible or non-
crease estimate stability (Judkins, 1990). All analyses were conducted in combustible product; (2) “Any cigarettes, single or poly (no ENDS)” (N
SAS 9.4 and SUDAAN 11.0 using PATH W4 all-wave (longitudinal) = 7,095), including exclusive cigarettes or co-use with any other
sample weights to account for data on respiratory conditions from combustible or non-combustible product, excluding ENDS; (3) “Any
W1–4. All tests were two-sided with significance level set at 5 %. other product without cigarettes or ENDS” (N = 1,746); (4) “former
However, when interpreting the results, we focused on findings that users of any product” (N = 4,872), as previously defined; and (5) “non-

3
J. Cordova et al. Preventive Medicine Reports 30 (2022) 102016

Table 1 Table 1 (continued )

Weighted Demographic Characteristics and Tobacco Use Profiles at Wave 4. Unweighted Weighted %
Pooled Adult Questionnaire Data from Waves 1–4 of the PATH Study, N (95 % CI)
2013–2018 (N = 26,072).
Substance use
Unweighted Weighted % Low/moderate 23,965 94.4 (94.0, 94.7)
N (95 % CI) High 1,761 4.4 (4.1, 4.7)
Gender Missing 346 1.2 (1.1, 1.4)
Male 12,615 48.0 (47.8, 48.2)
Female 13,456 52.0 (51.8, 52.1) COPD c
Missing 1 0.0 (0.0, 0.1) Yes 1,220 4.5 (4.1, 4.9)
No 24,844 95.5 (95.1, 95.9)
Age Missing 8 0.0 (0.0, 0.1)
18–24 8,021 12.1 (11.9, 12.3)
25–34 5,463 17.4 (16.8, 18.1) Bronchitisc
35–54 6,885 32.7 (32.0, 33.4) Yes 1,737 6.3 (5.9, 6.7)
55+ 5,700 37.7 (37.2, 38.3) No 24,327 93.7 (93.3, 94.1)
Missing 3 0.0 (0.0, 0.1) Missing 8 0.0 (0.0, 0.1)

Marital status Asthmac

Married 9,534 51.8 (50.7, 52.8) Yes 4,245 14.2 (13.6, 14.7)
Widowed, divorced, separated, never married 16,414 47.8 (46.8, 48.8) No 21,819 85.8 (85.3, 86.3)
Missing 124 0.5 (0.4, 0.6) Missing 8 0.0 (0.0, 0.1)

Note:
Race
Unweighted frequencies and weighted percentages (95% CI) are presented.
White, Non-Hispanic 15,034 65.1 (64.9, 65.2)
a.) Accounts for ever use of ENDS, combustible cigarettes, traditional cigars,
Black or African American, Non-Hispanic 3,941 11.7 (11.6, 11.8)
Other, Non-Hispanic (including multiracial) 2,007 7.8 (7.6, 7.9) filtered cigars, cigarillos, pipes, hookah, smokeless tobacco, snus.
Hispanic or Latino 5,089 15.5 (15.4, 15.5) b.) Current tobacco use assessed at W4:
Missing 1 0.0 (0.0, 0.1) ENDS only: current exclusive ENDS users.
Cigarettes only: current exclusive combustible cigarette smokers.
Dual cigarettes and ENDS: current concurrent users of only combustible ciga­
Highest education completed
Some high school or less 3,272 10.9 (10.4, 11.4) rettes and ENDS.
High school graduate or GED 7,761 27.7 (27.1, 28.3) Current users of non-cigarette combustible, smokeless, and polytobacco prod­
Some college or associate’s degree 9,263 31.8 (31.2, 32.3) ucts: any combination of single or poly use of combustible cigarettes, ENDS,
College degree or greater 5,677 29.3 (28.9, 29.8) traditional cigars, filtered cigars, cigarillos, pipes, hookah, smokeless tobacco,
Missing 99 0.3 (0.3, 0.5) and snus (with the exception of exclusive cigarette and ENDS use).
Former users of any product: former established users of combustible cigarettes,
Annual household income ENDS, traditional cigars, filtered cigars, cigarillos, pipes, hookah, smokeless
<$10,000 4,095 10.5 (9.9, 11.1) tobacco, and snus.
$10,000–$24,999 5,180 17.5 (16.8, 18.2) Non-current, non-former, or never users of any product: participants who did not
$25,000–$49,999 5,537 20.8 (20.0, 21.6) report current or former use of any product or who were never users of any
$50,000–$99,999 5,623 24.6 (23.8, 25.4) product.
≥$100,000 4,005 19.4 (18.5, 20.3) c.) Respiratory disorder (asthma, COPD, bronchitis) status accounts for W1
Missing 1,632 7.3 (6.7, 7.8)
lifetime diagnosis, W2 past 12-month diagnosis, W2 new baseline lifetime
diagnosis, W3 past 12-month diagnosis, W3 new baseline lifetime diagnosis, W4
Ever use of any tobacco product a past 12-month diagnosis, W4 new baseline lifetime diagnosis.
Yes 21,360 73.8 (72.6, 75.0) CI = confidence interval. COPD = chronic obstructive pulmonary disease. ENDS
No 4,485 24.6 (23.5, 25.6) = electronic nicotine delivery system. PATH = Population Assessment of To­
Missing 227 1.6 (1.4, 1.9)
bacco and Health.

Tobacco use profiles (scenario I) b

current, non-former, or never users of any product” (N = 10,211), as
ENDS only 621 1.3 (1.2, 1.5)
Cigarettes only 5,776 13.6 (13.1, 14.1) previously defined. Results were similar to those of the original tobacco
Dual cigarettes and ENDS 806 1.7 (1.6, 1.8) use profiles and thus, were not discussed, but were presented in Sup­
Current users of non-cigarette combustible, 3,786 8.7 (8.3, 9.1) plemental Table 2.
smokeless, and polytobacco products
Former users of any product 4,872 24.0 (23.0, 24.9)
Non-current, non-former, or never users of 10,211 50.8 (49.5, 52.0) 4. Discussion
any product
Missing 0 0.0 (0.0, 0.0) In a sample of approximately 26,000 US adults weighted to represent
the US adult population, current and former use of combustible and non-
Internalizing behavior combustible tobacco products, including ENDS, was significantly asso­
Low/moderate 17,936 76.3 (75.5, 77.0) ciated with higher odds of reporting a prior (prevalent) diagnosis of
High 7,926 22.8 (22.0, 23.5)
Missing 210 0.9 (0.8, 1.1)
either asthma, bronchitis or COPD. For asthma, different covariate ad­
justments strongly impacted the estimates, but no effects were signifi­
cant after adjusting for multiple testing, although a significant
Externalizing behavior
Low/moderate 22,187 89.9 (89.5, 90.4) association remained for bronchitis and COPD.
High 3,593 9.0 (8.6, 9.4) Exclusive cigarette smokers and dual users had a much lower risk of
Missing 292 1.1 (0.9, 1.3) reporting a prior bronchitis diagnosis compared with a prior COPD
diagnosis, while ENDS users had a non-statistically significant lower
risk. In another longitudinal study of older adults with or at-risk for
COPD, ever use of e-cigarettes was associated with chronic bronchitis

4
J. Cordova et al. Preventive Medicine Reports 30 (2022) 102016

Table 2
Associations of Tobacco Use (Scenario I) with Respiratory Disorder Status. Select Multivariable Logistic Regression Results with Pooled Adult Questionnaire Data from
Waves 1–4 of the PATH Study, 2013–2018.
Tobacco use (N) Asthma a COPD a
Bronchitis a

b c b c b
Minimal (N = Full (N = Minimal (N = Full (N = 23,788) Minimal (N = Full c (N = 23,788)
24,303) 23,788) 24,303) 24,303)

Minimal Full p- p- p-value p-value p-value p-value

value value

Tobacco use profiles 0.002 0.013 <0.001 <0.001 <0.001 <0.001

(scenario I) d
ENDS only 575 564 0.9 0.8 5.1 6.5 1.2 0.8
(0.7, (0.6, (2.7, (3.7, (0.6, (0.5,
1.2) 1.0) 9.4) 11.5) 2.2) 1.6)
Cigarettes only 5,432 5,274 1.1 0.8 5.9 6.1 2.4 1.7
(0.9, (0.7, (4.1, (4.0, (1.9, (1.3,
1.2) 0.9) 8.5) 9.1) 3.0) 2.2)
Dual cigarettes and 763 740 1.1 0.8 6.9 5.4 3.4 2.3
ENDS (0.9, (0.6, (4.6, (3.4, (2.4, (1.6,
1.4) 1.0) 10.6) 8.7) 4.7) 3.5)
Current users of non- 3,576 3,481 1.3 1.0 4.8 4.4 2.6 1.8
cigarette combustible, (1.1, (0.9, (3.3, (3.1, (2.0, (1.4,
smokeless, and 1.5) 1.2) 7.0) 6.4) 3.4) 2.4)
polytobacco products
Former users of any 4,601 4,509 1.2 1.1 3.4 3.0 2.1 1.6
product (1.1, (0.9, (2.3, (1.9, (1.6, (1.2,
1.4) 1.2) 5.0) 4.7) 2.6) 2.1)
Non-current, non- 9,356 9,220 Ref. Ref. Ref. Ref. Ref. Ref.
former, or never users
of any product

Notes:
Unweighted frequencies and adjusted odds ratios (AORs, 95 % CIs) for asthma, COPD, and bronchitis are presented. Bolded estimates are statistically significant (p <
0.05).
a.) The outcome variable, respiratory disorder (asthma, COPD, bronchitis) status, accounts for W1 lifetime diagnosis, W2 past 12-month diagnosis, W2 new baseline
lifetime diagnosis, W3 past 12-month diagnosis, W3 new baseline lifetime diagnosis, W4 past 12-month diagnosis, W4 new baseline lifetime diagnosis.
b.) Minimal model adjusts for gender, age, marital status, race/ethnicity, education, and annual household income.
c.) Full model adjusts for gender, age, marital status, race/ethnicity, education, annual household income, internalizing behavior, externalizing behavior, substance
use, and respiratory disorders (i.e., model for asthma adjusts for COPD and bronchitis).
d.) Current tobacco use assessed at W4:
ENDS only: current exclusive ENDS users.
Cigarettes only: current exclusive combustible cigarette smokers.
Dual cigarettes and ENDS: current concurrent users of only combustible cigarettes and ENDS.
Current users of non-cigarette combustible, smokeless, and polytobacco products: any combination of single or poly use of combustible cigarettes, ENDS, traditional
cigars, filtered cigars, cigarillos, pipes, hookah, smokeless tobacco, and snus (with the exception of exclusive cigarette and ENDS use).
Former users of any product: former established users of combustible cigarettes, ENDS, traditional cigars, filtered cigars, cigarillos, pipes, hookah, smokeless tobacco,
and snus.
Non-current, non-former, or never users of any product: participants who did not report current or former use of any product or who were never users of any product.
AOR = adjusted odds ratio. CI = confidence interval. COPD = chronic obstructive pulmonary disease. ENDS = electronic nicotine delivery system.

and COPD exacerbations after adjusting for current cigarette smoking, studies, demonstrating that ENDS use was significantly associated with a
while more rapid decline in lung function was not associated with e- higher prevalence of respiratory symptoms (Hedman et al., 2018; Wills
cigarette use after controlling for current smoking (Bowler et al., 2017). et al., 2019; Wills et al., 2021). Individuals with behavioral health
Also, for both bronchitis and COPD, the risk of reporting either prior conditions are more likely to use tobacco products than those without
condition was lower for former users of any product compared to current such conditions, and tobacco product use is associated with worsened
users, but higher than among non-users. Although the effects of e-ciga­ symptoms and reduced effectiveness of behavioral health treatment
rette use on COPD remission remain unclear, the benefits of complete (Prochaska et al., 2017). Given that individuals with these conditions
tobacco product cessation in slowing the progression of COPD and use tobacco products at disproportionately higher rates than the general
reducing the risk of respiratory complications are well documented population, it was necessary to adjust for behavioral health symptom­
(Coronini-Cronberg et al., 2011; Bai et al., 2017). atology in the fully adjusted models to control for potential confounding
The findings for asthma should be interpreted with caution, as after effects. The main drivers of the reversal of the association, however,
applying a multiple-comparison correction (Bonferroni correction), the were the inclusion of the other respiratory conditions as covariates (i.e.,
results were no longer statistically significant. Yet, prior to adjusting for not the respiratory outcome of interest), which we included because of
multiple testing, we observed that individuals who exclusively smoked the correlation across the three respiratory conditions that we examined.
cigarettes were significantly less likely to report a diagnosis of asthma in For instance, asthma and COPD tend to be intercorrelated, and a subset
the fully adjusted model, a reversal in the direction of association from of individuals experience exacerbated symptoms of both disorders,
the minimally adjusted model which only considered sociodemographic known as asthma-COPD overlap syndrome (ACOS) (Alshabanat et al.,
characteristics. Although the lower associations of cigarettes-only on 2015). We observed overlap of respiratory disorders in our sample:
asthma observed in the fully adjusted model may seem counterintuitive, among 4,245 respondents with asthma, 503 (13.2 %) also reported a
we offer some possible explanations. First, when we did not adjust the COPD diagnosis and 743 (19.0 %) reported a diagnosis of bronchitis.
model for composite scores for mental health and other respiratory Further, our minimally adjusted models (without other respiratory
disorder covariates, our findings for asthma were consistent with other covariates) yielded qualitatively similar conclusions. It is possible that

5
J. Cordova et al. Preventive Medicine Reports 30 (2022) 102016

conclusions from other studies that have examined ENDS and asthma or asthma-related quality of life (Del Giacco et al., 2016; Li et al., 2020).
asthma-like respiratory conditions, such as wheezing (Li et al., 2019;
Yao et al., 2017), differed due to non-adjustment for competing pul­ 5. Conclusion
monary outcomes. Second, nicotine has been shown to inhibit immune
response and weaken the immune system overall. Individuals with im­ Our study found a strong association between dual cigarette and
mune disorders, such as asthma, may be averse to initiating or sustaining ENDS use on the prevalence of COPD and bronchitis. Further research
use of combustible tobacco products due to the immunosuppressive ef­ examining tobacco product use over time can further clarify the
fects of nicotine (Alkhattabi et al., 2018; Piao et al., 2009). This is a observed inverse association between tobacco use and asthma. In
potential explanation for the lower associations of exclusive cigarette particular, a greater understanding of the timing of smoking behavior
use with asthma observed in the fully adjusted model, in addition to the change and asthma (which has a shorter latency period for development
possibility of selection bias. and clearing of symptoms, compared to COPD and bronchitis) and a
The main strength of this study was the examination of a series of more precise assessment of the temporality of the effect of tobacco use
tobacco use profiles among US adults that captures the effect of on respiratory conditions is warranted. Additionally, longitudinal
combustible cigarettes alone, ENDS alone, dual cigarette and ENDS studies can examine whether adult smokers who develop any of the
users, as well as combinations of current use of non-cigarette combus­ respiratory disorders we studied are more likely to quit smoking or to
tible, smokeless, and polytobacco products, and former use of any transition to ENDS use (exclusively or dually with cigarettes), compared
product. Cigarettes were categorized separately from other combustible to adult smokers without respiratory disorders. Healthcare providers are
tobacco products (i.e., cigars, pipes and hookah) because cigarettes uniquely positioned to discuss tobacco use with their patients as part of
remain the most common single-use tobacco product (Cornelius et al., respiratory disorder treatment; they are particularly well-suited to
2020). This tobacco use categorization allowed us to examine the po­ emphasize to their patients that complete cessation of all tobacco
tential respiratory health consequences among US adults who used products is required to improve lung functioning and asthma symptoms.
various tobacco products after diagnosis of a respiratory condition. The
detailed observational data on tobacco use and respiratory outcomes Funding
also demonstrates another study strength. Our key limitation in this
study was that tobacco use was assessed for the month prior to the This work was supported in part by the National Cancer Institute at
questionnaire, while history of respiratory illnesses was ascertained for the National Institutes of Health (contract number
the prior year (W2-4) or ever/never (W1). Hence, it is possible that HHSN261201700004I). Dr. Choi’s effort was supported in part by the
participants who were diagnosed with a respiratory disorder and were National Institute on Minority Health and Health Disparities at the Na­
current smokers at the time of diagnosis could have changed their tional Institutes of Health. Dr. Altekruse’s effort was supported in part by
behavior and therefore self-reported as non-smokers at the time of the the National Heart, Lung, and Blood Institute at the National Institutes of
survey. The cross-sectional design, with tobacco use status measured at Health.
W4, precludes inference on causal relationships between tobacco use
and respiratory disorders. Another limitation to our study is our inability Declaration of Competing Interest
to meaningfully disentangle the effects of “Group 4” (current user of
non-cigarette combustible, smokeless, and polytobacco products) in The authors declare that they have no known competing financial
relation to respiratory health, given the combination of single and poly interests or personal relationships that could have appeared to influence
use of combusted and non-combusted products (excluding exclusive the work reported in this paper.
cigarettes and ENDS). However, we were compelled to collapse these
subgroups as we were unable to examine respiratory health effects Data availability
among such a small number of participants. Furthermore, another lim­
itation is the inclusion of former cigarette smokers within the “ENDS Data is publicly available at https://www.icpsr.umich.edu/web/
only” tobacco user group, which may introduce residual confounding NAHDAP/studies/36231
effects when observing the association between current exclusive ENDS
use and COPD. We were unable to disentangle former cigarette smokers
Acknowledgments
from the “ENDS only” tobacco user group due to an extremely small
number of respondents that reported current exclusive ENDS-only with
Role of Funder: The National Institutes of Health had no role in the
no history of former combustible cigarette use. In the minimally
design and conduct of the study; collection, management, analysis, and
adjusted models, of 575 ENDS only users, 366 respondents reported
interpretation of the data; preparation, review, or approval of the
former combustible cigarette use. In the fully adjusted models, of 564
manuscript; and decision to submit the manuscript for publication.
ENDs only users, 357 respondents reported former combustible cigarette
Disclaimers: The views and opinions expressed in this paper are
use. Our study also does not address cigarettes or ENDS use with respect
those of the authors only and do not necessarily represent the views,
to acute respiratory effects of exposure to these products as was seen
official policy or position of the U.S. Department of Health and Human
during the outbreak of e-cigarette or vaping product associated lung
Services or any of its affiliated institutions or agencies, the U.S.
injury (EVALI) in Fall 2019.
Department of Housing and Urban Development, and the U.S. Centers
In summary, our goals were to better understand the relationships
for Medicare and Medicaid Services.
between use of several types of tobacco products, particularly cigarettes
and ENDS, on the common respiratory conditions of bronchitis, COPD,
References
and asthma, and examine whether dual use of cigarettes and ENDS had a
differential effect on these conditions compared to the single use of Alkhattabi, N., Todd, I., Negm, O., Tighe, P.J., Fairclough, L.C., 2018. Tobacco smoke
products. We observed a sixfold risk of reporting a prior COPD diagnosis and nicotine suppress expression of activating signaling molecules in human
for the individual use of either ENDS or cigarettes, consistent with other dendritic cells. Toxicol. Lett. 15 (299), 40–46.
Alshabanat, A., Zafari, Z., Albanyan, O., Dairi, M., FitzGerald, J.M., Short, P., 2015.
research on ENDS (Wills et al., 2019), and a fivefold risk associated with Asthma and COPD overlap syndrome (ACOS): a systematic review and meta analysis.
the dual use of these products. Additionally, the significant estimates PLoS ONE 10 (9), e0136065.
observed for internalizing behaviors for asthma and bronchitis also Bai, J.W., Chen, X.X., Liu, S., Yu, L., Xu, J.F., 2017. Smoking cessation affects the natural
history of COPD. Int. J. Chron. Obstruct. Pulmon. Dis. 12, 3323–3328.
agree with other research showing that anxiety and depression are Berg, C.J., 2016. Preferred flavors and reasons for e-cigarette use and discontinued use
associated with a greater risk of asthma, and poorer asthma control and among never, current, and former smokers. Int. J. Public Health 61 (2), 225–236.

6
J. Cordova et al. Preventive Medicine Reports 30 (2022) 102016

Bowler, R.P., Hansel, N.N., Jacobson, S., Graham Barr, R., Make, B.J., Han, M.K., King, J.L., Reboussin, B.A., Wiseman, K.D., Ribisl, K.M., Seidenberg, A.B., Wagoner, K.G.,
O’Neal, W.K., Oelsner, E.C., Casaburi, R., Barjaktarevic, I., Cooper, C., Foreman, M., Wolfson, M., Sutfin, E.L., 2019. Adverse symptoms users attribute to e-cigarettes:
Wise, R.A., DeMeo, D.L., Silverman, E.K., Bailey, W., Harrington, K.F., Woodruff, P. Results from a national survey of US adults. Drug Alcohol. Depend. 196, 9–13.
G., Drummond, M.B., 2017. Electronic Cigarette Use in US the cadults at risk for or Klein, E.G., Berman, M., Hemmerich, N., Carlson, C., Htut, S., Slater, M., 2016. Online E-
wiOPD: analysis from two observational cohorts. J. Gen. Intern. Med. 32 (12), cigarette marketing claims: a systematic content and legal analysis. Tob Regul Sci. 2
1315–1322. (3), 252–262.
Choi, K., Bernat, D., 2016. E-cigarette use among florida youth with and without asthma. Kruse, G.R., Kalkhoran, S., Rigotti, N.A., 2017. Use of electronic cigarettes among U.S.
Am. J. Prev. Med. 51 (4), 446–453. adults with medical comorbidities. Am. J. Prev. Med. 52 (6), 798–804.
Clapp, P.W., Jaspers, I., 2017. Electronic cigarettes: their constituents and potential links Li, Y., Jiang, Q., Ji, Y., Cao, C., 2020. Anxiety and depression may associate with poorer
to asthma. Curr. Allergy Asthma Rep. 17 (11), 79. control and quality of life in adults with asthma. Allergy. Feb 8;all.14189.
Coleman, B.N., Rostron, B., Johnson, S.E., Ambrose, B.K., Pearson, J., Stanton, C.A., Li, D., Sundar, I.K., McIntosh, S., Ossip, D.J., Goniewicz, M.L., O’Connor, R.J., et al.,
Wang, B., Delnevo, C., Bansal-Travers, M., Kimmel, H.L., Goniewicz, M.L., 2019. Association of smoking and electronic cigarette use with wheezing and related
Niaura, R., Abrams, D., Conway, K.P., Borek, N., Compton, W.M., Hyland, A., 2017. respiratory symptoms in adults: cross-sectional results from the Population
Electronic cigarette use among US adults in the population assessment of tobacco Assessment of Tobacco and Health (PATH) study, wave 2. Tob Control. Feb 13;
and health (PATH) Study, 2013–2014. Tob. Control. 26 (e2), e117–e126. tobaccocontrol-2018-054694.
Cornelius, M.E., Wang, T.W., Jamal, A., Loretan, C.G., Neff, L.J., 2020. Tobacco Product McConnell, R., Barrington-Trimis, J.L., Wang, K., Urman, R., Hong, H., Unger, J.,
Use Among Adults — United States, 2019. MMWR Morb. Mortal. Wkly Rep. 69 (46), Samet, J., Leventhal, A., Berhane, K., 2017. Electronic cigarette use and respiratory
1736–1742. symptoms in adolescents. Am. J. Respir. Crit. Care Med. 195 (8), 1043–1049.
Coronini-Cronberg, S., Heffernan, C., Robinson, M., 2011. Effective smoking cessation Piao, W.H., Campagnolo, D., Dayao, C., Lukas, R.J., Wu, J., Shi, F.D., 2009. Nicotine and
interventions for COPD patients: a review of the evidence. JRSM Short Rep. 2 (10), inflammatory neurological disorders. Acta Pharmacol. Sin. 30 (6), 715–722.
1–12. Prochaska, J.J., Das, S., Young-Wolff, K.C., 2017. Smoking, mental illness, and public
Dai, H., Leventhal, A.M., 2019. Prevalence of e-cigarette use among adults in the United health. Annu. Rev. Public Health 20 (38), 165–185.
States, 2014–2018. JAMA 322 (18), 1824. Tamimi, A., Serdarevic, D., Hanania, N.A., 2012. The effects of cigarette smoke on
Del Giacco, S.R., Cappai, A., Gambula, L., Cabras, S., Perra, S., Manconi, P.E., et al., airway inflammation in asthma and COPD: Therapeutic implications. Respir. Med.
2016. The asthma-anxiety connection. Respir. Med. 120, 44–53. 106 (3), 319–328.
Deshpande, M., Bromann, S., Arnoldi, J., 2020. Electronic cigarette use among adults The Health Consequences of Smoking—50 Years of Progress: A Report of the Surgeon
with asthma: 2014–2017 national health interview survey. Res Soc Adm Pharm. 16 General. Atlanta (GA): Centers for Disease Control and Prevention (US); 2014.
(2), 202–207. Wang, T.W., Gentzke, A., Sharapova, S., Cullen, K.A., Ambrose, B.K., Jamal, A., 2018.
Eaton, D.L., Kwan, L.Y., Stratton, K., editors., 2018. Public health consequences of E- Tobacco product use among middle and high school students — United States,
cigarettes. National Academies Press (US), Washington (DC). 2011–2017. MMWR Morb. Mortal. Wkly Rep. 67 (22), 629–633.
E-Cigarette Use Among Youth and Young Adults: A Report of the Surgeon General. Wheaton, A.G., Liu, Y., Croft, J.B., VanFrank, B., Croxton, T.L., Punturieri, A., Postow, L.,
Atlanta (GA): Centers for Disease Control and Prevention (US); 2016. Greenlund, K.J., 2019. Chronic Obstructive Pulmonary Disease and Smoking Status
Forey, B.A., Thornton, A.J., Lee, P.N., 2011. Systematic review with meta-analysis of the — United States, 2017. MMWR Morb. Mortal. Weekly Rep. 68 (24), 533–538.
epidemiological evidence relating smoking to COPD, chronic bronchitis and Wills, T.A., Pagano, I., Williams, R.J., Tam, E.K., 2019. E-cigarette use and respiratory
emphysema. BMC Pulm. Med. 11 (1), 36. disorder in an adult sample. Drug Alcohol Depend. 194, 363–370.
Gotts, J.E., Jordt, S.E., McConnell, R., Tarran, R., 2019. What are the respiratory effects Wills, T.A., Soneji, S.S., Choi, K., Jaspers, I., Tam, E.K., 2021. E-cigarette use and
of e-cigarettes? BMJ 30, l5275. respiratory disorders: an integrative review of converging evidence from
Grana, R.A., Ling, P.M., 2014. Smoking revolution. Am. J. Prev. Med. 46 (4), 395–403. epidemiological and laboratory studies. Eur. Respir. J. 57 (1), 1901815.
Hedman, L., Backman, H., Stridsman, C., Bosson, J.A., Lundbäck, M., Lindberg, A., Yao T, Max W, Sung HY, Glantz SA, Goldberg RL, Wang JB, et al. Relationship between
Rönmark, E., Ekerljung, L., 2018. Association of electronic cigarette use with spending on electronic cigarettes, 30-day use, and disease symptoms among current
smoking habits, demographic factors, and respiratory symptoms. JAMA Netw Open. adult cigarette smokers in the U.S. van Zyl-Smit R, editor. PLOS ONE. 2017 Nov 7;12
1 (3), e180789. (11):e0187399.
Judkins, D.R., 1990. Fay’s method for variance estimation. J Off Stat. 6 (3), 223–239.