Tobacco product use and functionally-important respiratory symptoms among US adolescents/young adults

Abstract

Objective:

The relation between respiratory symptoms and the range of tobacco product use among U.S. adolescents/young adults is not yet clear. This cross-sectional analysis examines tobacco product use and respiratory symptoms in a nationally representative sample of 21,057 adolescents/young adults aged 12-24 years from Wave 4 (2016-17) of the Population Assessment of Tobacco and Health Study.

Methods:

Presence of functionally-important respiratory symptoms was defined by questions regarding wheezing and nighttime cough at a cutoff score associated with poorer functional health status. Past-30-day tobacco use was analyzed two ways: never-tobacco users (reference) vs. combustible users, noncombustible-only users, and former users; or frequency of use of cigarettes and/or e-cigarettes. Weighted Poisson regression adjusted for past-30-day marijuana use, secondhand smoke exposure, and asthma.

Results:

Functionally-important respiratory symptoms were present in 10.0% overall: 13.8% of combustible users, 9.0% of noncombustible users, 8.2% of noncurrent users and 9.7% of never users. Functionally-important respiratory symptoms were associated with combustible tobacco use (relative risk [RR]=1.52[95% CI 1.29, 1.80]), marijuana use (RR=1.54[1.34, 1.77]) and secondhand smoke exposure (RR=1.04[1.03, 1.05]). Higher cigarette smoking frequency was also associated with functionally-important respiratory symptoms for frequency categories >14 days/month (e.g. RR=1.93[1.50, 2.49] for 15-29 days/month). Frequency of e-cigarette use was not associated with functionally-important respiratory symptoms.

Conclusions:

During 2016-17, smoking cigarettes, marijuana use, and secondhand smoke exposure were cross-sectionally associated with functionally-important respiratory symptoms in adolescents/young adults. Risk increased with increased frequency of cigarette use but not e-cigarette use. Given changes to contemporary e-cigarettes and use, findings may not generalize to newer products.

Introduction

Tobacco use causes disease across all ages, with multiple studies demonstrating an association between adolescent cigarette smoking and asthma as well as respiratory symptoms including cough, phlegm, wheeze and shortness of breath.^1–3 Less is known about health risks from tobacco products other than cigarettes, and whether there are thresholds of use above which symptoms are experienced. E-cigarettes and other electronic vaping devices (hereafter e-cigarettes) have been readily adopted by adolescents, with prevalence of use surpassing combusted tobacco cigarettes and all other forms of tobacco among high school students. Pre-pandemic, in 2019 27.5% of high school students reported e-cigarette use in the past 30 days, compared to 5.8% reporting cigarette use in this period.^4,5 More recent assessments of e-cigarette use among high school students in 2021 have shown substantial decline, with 11.3% reporting current use,⁶ however these data were collected on-line to allow for remote student participation and are not directly comparable to previous in-person survey data. E-cigarettes represent a diverse product category regarding nicotine content, flavors and aerosol delivery that has only been sold since the mid 2000’s, with substantively different product characteristics (nicotine salt-based pod devices) entering the market in late 2015.⁷

There is biologic plausibility for respiratory compromise from e-cigarettes because they deliver ultrafine particulates at similar quantities to cigarettes, and these ultrafine particulates are known to cause inflammation and respiratory disease.⁸ Thus, it is important to try to ascertain e-cigarette-related health risks in population samples.⁹ Assessing the risk of respiratory impairment associated with e-cigarette use is challenging, however, given that e-cigarette users often also use other tobacco products and marijuana. To date, cross-sectional studies in adolescents have shown an association between e-cigarette use and higher risk of cough/phlegm;^10–12 these studies did not measure marijuana use and were inconsistent with regards to secondhand smoke exposure measurement. Another recent study assessed e-cigarette use among 12-17 year olds in the Population Assessment of Tobacco and Health (PATH) Study, finding no longitudinal relation between e-cigarette use or combustible tobacco use and a single item measuring wheeze.¹³ As wheeze may be due to airway hyperresponsiveness to stimuli, contemporaneous assessment may be more indicative of acute tobacco effects.

In order to better understand the relationships between multiple tobacco product use and respiratory symptoms among adolescents and young adults, data were analyzed from Wave 4 (2016-17) of the PATH Study. The PATH Study is a household-based, nationally representative, longitudinal cohort study of U.S. adults and adolescents on tobacco use patterns and associated health behaviors. Our focus was on the relation between tobacco product use and respiratory symptoms (wheezing/nighttime cough), while adjusting for potential confounders, including marijuana use, secondhand smoke exposure and concurrent diseases (asthma) that could impact tobacco use and respiratory status.

METHODS

Study design, setting and participants

The National Institute on Drug Abuse partners with the U.S. Food and Drug Administration’s Center for Tobacco Products to conduct the PATH Study under a contract with Westat. The PATH Study employed a stratified, address-based, area probability sampling design at Wave 1 (2013-14) that oversampled adult tobacco users, young adults (18-24 years), and African-American adults. A screener was used to select youths aged 12-17 (hereafter referred to as adolescents) and adults (18 years and older) from households for participation in the study, and audio computer-assisted self-interviews collected information on tobacco-use patterns and health behaviors.

Data from Wave 4 (2016-17) Restricted Use File were analyzed,¹⁴ and included 25,446 adolescents and young adults from 12-24 years of age without respiratory diseases other than asthma, of whom 21,054 subjects had complete data (see supplemental Figure 1 for a flow diagram) after missing data on sex and race/Hispanic ethnicity were imputed.¹⁵ Wave 4 sampling included a probability replenishment sample of adults and adolescents selected from the U.S. civilian noninstitutionalized population. A more detailed description of the sampling procedure for Wave 4 of the PATH Study, along with response rates, can be found in Supplement Section 2.

Full-sample and replicate weights were created that adjust for the complex sample design and nonresponse, enabling the PATH Study to obtain statistically valid estimates of U.S. civilians ages 12 years and older at the time of Wave 4, with associated measures of statistical precision. Further details regarding the PATH Study design and methods are published elsewhere;¹⁶ interview procedures, questionnaires, sampling, weighting, response rates, and data access are described in the PATH Study Restricted Use Files User Guide.¹⁷

All respondents ages 18 and older provided informed consent, with adolescent respondents ages 12 to 17 providing assent while a parent/legal guardian provided consent. The study was approved by the Westat Institutional Review Board.

Self-reported outcomes

Respiratory symptoms and functionally-important respiratory symptoms

Respiratory symptom questions included in the Wave 4 PATH Study were from the core wheezing and asthma module from the International Study of Asthma and Allergies in Children (ISAAC),¹⁸ which has been used in more than 100 countries and 2 million children to assess respiratory health. Questions included items related to wheezing and nighttime cough indicative of asthma and have high face validity and test-retest reliability.¹⁹ Items/responses, weighted prevalence of each item in Wave 4, and prevalence of similar questions from the 2011-12 National Health and Nutrition Examination Survey are all included in Supplemental Table 1; 39.6% of the PATH Study sample reported one or more of these symptoms. Responses from the ISAAC questions were summed to create a respiratory symptom index which ranged from 0 (no symptoms) to 9 (worst). Higher scores on the index were indicative of greater functional limitations (e.g., walking, fatigue) among adults.²¹ Given the high frequency of any reported respiratory symptoms among this adolescent/young adult sample, we sought to determine a cutoff score associated with impact on health and functioning e.g., self-reported overall health rating or health-related school absenteeism. The threshold validation for adolescents and young adults within this PATH Study is detailed in the Supplement and described briefly below.

Exposures

Adolescents and young adults reported their lifetime and current (i.e. past 30 day) use of cigarettes, traditional cigars, cigarillos, filtered cigars, pipe tobacco, hookah, snus pouches, other smokeless tobacco (loose snus, moist snuff, dip, spit, or chewing tobacco), and e-cigarettes. Pictures and descriptions were displayed for each product to ensure accuracy. Those who had never used any tobacco product and former experimental users (not currently using, never used any product “fairly regularly”, and had not used at least 100 cigarettes in their lifetime) were combined into a “never” reference category.

Tobacco exposure was analyzed in two ways. The first analysis categorized use of tobacco as four mutually exclusive product use groups: never use; noncurrent use (ever used tobacco, but not in the past 30 days); current (past 30-day) noncombustible only use, which included e-cigarettes; and current combustible use which included cigarettes, all types of cigars, pipes and hookah (regardless of noncombustible use). We considered breaking out dual users of cigarettes and e-cigarettes separately, but an unadjusted analysis indicated that the prevalence of functionally-important symptoms in that group was not significantly different from exclusive cigarette users.

The second analytic approach focused on the two most common subcategories of tobacco use, cigarette and e-cigarette use, assessing frequency of use regardless of how much product was used each day. For these analyses, frequency of use was based on number of days used in the past month; categories included never use, ever use but not current use, and current use on 1-14/30 days, on 15-29/30 days and daily.

Covariates

Wave 4 covariates (Supplemental Table 5) included variables that could be associated both with tobacco exposure and functionally-important respiratory symptoms, including age, sex, and race/ethnicity, asthma diagnosis (based on self-report of “have you been told by a doctor, nurse or other health professional that you have asthma”) and obesity (based on body mass index [BMI]). Asthma was further categorized by medication use in the past 12 months: no medications, rescue only medications (e.g. albuterol), and controller medications (e.g. inhaled corticosteroids). Other potentially smoke-related exposures that could confound the association between tobacco use and respiratory symptoms included secondhand smoke exposure and past-month marijuana use. Second-hand smoke (SHS) exposure was a continuous variable, with each 1-point increase indicative of an additional 5 hours of exposure per week; the variable was skewed with a mean of 1.6, a median of 0, 75^th percentile = 1, and 95^th percentile =15.

Statistical Analysis

Functionally-important respiratory symptoms scale validation

A respiratory symptoms index cut-off score was sought that would serve as a binary indicator of functionally-important respiratory symptoms, validated against functional status (see Supplement). Given limitations of questions spanning youth and adult survey instruments, relevant available measures were used to validate the index (Supplemental Table 2), including global ratings for physical health (assessed in both adolescents/young adults), frequency of missing school due to illness at Wave 2 (adolescent-only measure), and trouble walking 1 mile (young adult-only measure). As shown in Supplemental Figure 2, functional impairment increased as respiratory symptom scores increased. Unweighted receiver-operator curves (ROC) were used to determine a cutoff score while minimizing the false positive rate for its association with functional outcomes (Supplemental Table 3 shows values for several cut-off scores, with the selected cut-off being consistent with prevalence of asthma in the general population.)²⁰ The index was modestly associated with each functional outcome (area under the curve ranged from 0.54 to 0.63). A cut-off score of <3 vs. ≥3 was chosen, consistent with functionally-important respiratory symptoms among older adults in the PATH Study.²¹

A series of weighted logistic or multinomial regressions was then completed to fit each functional outcome on respiratory symptoms score ≥3, adjusted for age, gender, race/ethnicity, and BMI (Supplemental Table 4). In each regression, a score ≥3 was independently associated with the functional outcome. For example, a respiratory symptom score ≥3 was independently associated with fair/poor (compared to excellent) physical health among adolescents and young adults with combined AOR 3.74 (95% CI 3.13, 4.46) (AOR were similar for young adults (4.16 [3.10, 5.58]) and adolescents (3.17 [2.51, 4.00]) when analyzed separately.) Based on these analyses, a score ≥3 was concluded to indicate “functionally-important respiratory symptoms” among adolescents and young adults.

Analytical approach

All analyses were weighted using the Cohort 4, Wave 4 cross-sectional full-sample and replicate weights, except where noted in footnotes, to adjust for the complex sample design and loss to follow up. Variances were estimated using balanced repeated replication with Fay’s adjustment set to 0.3 to increase estimate stability.^22,23

Associations between tobacco product use and covariates were assessed using chi-square, t-tests or ANOVA, as appropriate. Multivariable weighted Poisson regression was used to obtain the adjusted risk of functionally-important respiratory symptoms by tobacco product use and, separately, by cigarette and e-cigarette use.

We also conducted a sensitivity analysis to assess the robustness of the results by repeating the analysis using multivariable linear regression to compare the respiratory symptoms index as a continuous measure.

All analyses were conducted using Stata survey data procedures, version 15.1 (StataCorp LLC, College Station, TX), and using the PATH Study Restricted Use Files.¹⁴ A complete set of aims, hypotheses and validation methods may be found in the Supplement.

Results:

Sample Description

Missing data

As shown in Supplemental Table 6, the analytic sample had a higher weighted proportion of 12-14 year olds (24.0%) than the non-analytic sample (18.6%, p<0.0001) and a lower proportion of 18-20 year olds (21.9% vs. 30.4%, p<0.0001). There were more “never” cigarette smokers in the analytic sample (71.6% vs. 60.7%, p<0.0001). The weighted percent of functionally important respiratory symptoms was similar in the analytic vs. non-analytic sample (10.0% vs. 11.2%, p=0.052).

Description of the sample

We use weighted percents to describe the sample (Table 2). The youths ranged in age from 12 to 24 years (24.0% 12-14, 23.6% 15-17, 21.9% 18-20, and 30.5% 21-24); 50.0% identified as female. The race/ethnicity categories were broadly representative of the US population: 54.2% White, 13.4% Black, and 22.1% Hispanic. Asthma diagnosis was reported in 18.3%, of whom 23.5% reported medication use. Functionally-important respiratory symptoms were present in 5.3% of youths without asthma, 21.8% of those with asthma but on no medications, 55.2% of those on rescue inhalers only, and 63.5% of those on controller medications.

Table 2.

Weighted cross-sectional associations between tobacco use and selected constructs and functionally-important respiratory symptoms for adolescents (age 12-17) and young adults (age 18-24) at Wave 4 of the Population Assessment of Tobacco and Health Study.^a

				Model 1		Model 2
Risk factor	Nb	%c	% with Functionally-Important Respiratory Symptomsc	Relative Riskd	95% CI	Relative Riskd	95% CI
Tobacco use statuse (Model 1 only)
Never use	12,864	56.5	9.7	Ref	Ref
Noncurrent use	4,413	23.5	8.2	0.96	[0.85,1.08]
Current noncombustible use only	787	3.9	9.0	0.87	[0.67,1.13]
Current combustible use	2,990	16.1	13.8	1.52	[1.29,1.80]
Cigarette use frequencyf (Model 2 only)
Never use	15,827	71.6	9.2			Ref	Ref
Noncurrent use	3,083	16.8	9.1			1.14	[0.98,1.32]
1-14 days	1,033	5.3	9.0			1.15	[0.91,1.46]
15-29 days	273	1.4	17.2			1.93	[1.50,2.49]
Daily	883	4.9	23.2			2.80	[2.25,3.47]
E-cigaretteg use frequencyh (Model 2 only)
Never use	14,943	68.2	9.3			Ref	Ref
Noncurrent use	4,596	24.0	11.1			0.99	[0.85,1.16]
1-14 days	1,019	5.0	13.2			1.02	[0.82,1.27]
15-29 days	195	1.0	10.7			0.73	[0.42,1.25]
Daily	301	1.8	13.1			1.25	[0.80,1.96]
Covariates
Age group
12-14	6,315	24.0	11.8	Ref	Ref	Ref	Ref
15-17	6,246	23.6	11.7	0.87	[0.79,0.96]	0.86	[0.78,0.95]
18-20	4,574	21.9	9.3	0.66	[0.57,0.76]	0.63	[0.55,0.73]
21-24	3,919	30.5	7.7	0.57	[0.50,0.65]	0.50	[0.43,0.59]
Sex
Female	10,523	50.0	10.8	Ref	Ref	Ref	Ref
Male	10,531	50.0	9.2	0.82	[0.75,0.90]	0.82	[0.75,0.90]
Race/Ethnicity
Whitei	10,054	54.2	10.7	Ref	Ref	Ref	Ref
Blacki	2,954	13.4	10.6	0.86	[0.76,0.96]	0.90	[0.81,1.01]
Otheri,j	2,027	10.3	9.6	0.89	[0.77,1.04]	0.91	[0.78,1.07]
Hispanic	6,019	22.1	7.8	0.74	[0.66,0.84]	0.77	[0.68,0.87]
Asthma status
No asthma	17,062	81.7	5.3	Ref	Ref	Ref	Ref
Asthma no meds	3,017	14.0	21.8	4.05	[3.65,4.49]	4.04	[3.65,4.48]
Rescue inhaler only	325	1.5	55.2	10.05	[8.86,11.41]	9.98	[8.81,11.32]
Controller meds	650	2.8	63.5	11.76	[10.48,13.19]	11.47	[10.19,12.91]
BMI
Underweight	882	4.4	8.6	1.01	[0.81,1.25]	0.99	[0.79,1.23]
Normal/Healthy	11,955	56.2	9.6	Ref	Ref	Ref	Ref
Overweight	4,176	20.5	9.9	1.02	[0.92,1.15]	1.02	[0.91,1.14]
Obese	4,041	18.9	11.6	1.07	[0.96,1.20]	1.05	[0.94,1.17]
Past-month marijuana usek
No	18,172	85.1	9.2	Ref	Ref	Ref	Ref
Yes	2,882	14.9	14.4	1.54	[1.34,1.77]	1.48	[1.30,1.69]
Secondhand smoke exposure (per each additional 5 hours/week)	21,054	n/a	n/a	1.04	[1.03,1.05]	1.03	[1.01,1.04]

^aN= 21,054 respondents with complete data at Wave 4.

^bUnweighted

^cWeighted to reflect the percent in the United States civilian noninstitutionalized population. Functionally-important respiratory symptoms were present in 10.0% (SE=0.2) overall.

^dAll relative risks adjust for the variables in the table.

^eNever use = never used tobacco; Noncurrent use = ever used tobacco, but not in the past 30 days; Noncombustible use only = exclusive noncombustible (e-cigarette, smokeless tobacco, or snus) use in the past 30 days; Combustible use = combustible (cigarette, pipe, hookah, or cigar) use in the past 30 days regardless of noncombustible use.

^fCigarette use frequency refers to the number of days of cigarette use within the past 30 days, regardless of other tobacco products used. Never = never used cigarettes; noncurrent = ever use of cigarettes, but no cigarette use within the past 30 days; other categories refer to the ranges of days cigarettes were used within the past 30 days.

^gAt W4, participants were asked about use of any e-products (e-cigarettes, e-cigars, e-pipes, and e-hookah), all of which are referred to as e-cigarettes here.

^hE-cigarette use frequency refers to the number of days of e-cigarette use within the past 30 days, regardless of other tobacco products used. Never = never used e-cigarettes; noncurrent = ever use of e-cigarettes, but no e-cigarette use within the past 30 days; other categories refer to the ranges of days e-cigarettes were used within the past 30 days.

ⁱNon-Hispanic

^jIncludes American Indian or Alaska Native, Asian Indian, Chinese, Filipino, Japanese, Korean, Vietnamese, Other Asian, Native Hawaiian, Guamanian or Chamorro, Samoan, Other Pacific Islander, and multiracial

^kMarijuana use variable does not distinguish between combustible and noncombustible use.

Supplemental Table 10 reports standard errors for all the weighted estimates presented in this table and adjusted RRs for all the variables in the model.

Tobacco use and age

Figure 1 shows unadjusted and weighted tobacco product use by age, demonstrating much higher rates of current combustible use among young adults. For example, the rate of current combustible use among those 15-17 years was 7.1% compared to 28.8% for 21-24 years. Cigarettes were the predominant combustible product used, comprising 78.6% of combustible use among 15-17 year olds and 74.1% in 21-24 year olds. The prevalence of noncombustible use was roughly the same among young adults and 15-17 year olds.

Figure 1.

Unadjusted weighted percent of tobacco use^a among four age groups of Population Assessment of Tobacco and Health Study Wave 4 adolescents (age 12-17) and young adults (age 18-24).^b

notes:

^a Never use = never used tobacco; Noncurrent use = ever used tobacco, but not in the past 30 days; Current noncombustible use only = exclusive noncombustible (e-cigarette^c, smokeless tobacco, or snus) use in the past 30 days; Current combustible use = combustible (cigarette, pipe, hookah, or cigar) use in the past 30 days regardless of noncombustible use. Error bars represent 95% confidence interval.

^b N = 21,054 adolescents and young adult respondents: 12-14 N = 6,315; 15-17 N = 6,246; 18-20 N = 4,574; 21-24 N = 3,919.

^c At W4, participants were asked about use of any e-products (e-cigarettes, e-cigars, e-pipes, and e-hookah), all of which are referred to as e-cigarettes here.

Tobacco use and covariates

Table 1 shows the relationships between sociodemographic, medical, and exposure characteristics and the four tobacco use categories. Males and whites were more likely to be current tobacco users than females and minorities. Combustible tobacco users had higher secondhand smoke exposure than any other category. Marijuana use was also more prevalent among past 30-day combustible users (48.6%) than current noncombustible (32.0%), noncurrent (19.4%), and never tobacco users (2.2%).

Table 1 –

Characteristics for adolescents (age 12-17) and young adults (age 18-24) at Wave 4 of the Population Assessment of Tobacco and Health Study,^a by tobacco use status^b (percentages and means are weighted to reflect U.S. population).

Tobacco Use Status
					Current (past 30-day)
	Never use (N = 12,864)		Noncurrent use (N = 4,413)		Noncombustible use only (N = 787)		Combustible use (N = 2,990)
	Nc	%	Nc	%	Nc	%	Nc	%	p-valued
Sociodemographics
Age
12-14	5,756	38.9	415	6.4	72	7.2	72	1.5	<0.0001
15-17	4,282	28.6	1,213	19.2	306	32.4	445	10.4
18-20	1,959	16.8	1,257	25.2	242	29.7	1,116	33.4
21-24	867	15.7	1,528	49.3	167	30.7	1,357	54.7
Sex
Female	6,565	52.0	2,346	52.9	280	34.9	1,332	42.3	<0.0001
Male	6,299	48.0	2,067	47.1	507	65.1	1,658	57.7
Race/Ethnicity
Whitee	5,939	52.4	2,028	52.6	517	71.2	1,570	58.6	<0.0001
Blacke	1,893	14.1	580	12.5	43	6.2	438	13.7
Othere,f	1,269	11.2	402	9.9	72	8.3	284	8.2
Hispanic	3,763	22.2	1,403	25.1	155	14.3	698	19.5
Relevant medical history
Asthma status
No asthma	10,411	81.4	3,579	82.0	627	80.0	2,445	82.7	0.001
Asthma no meds	1,787	13.7	648	13.9	126	16.0	456	14.5
Rescue inhaler only	203	1.5	78	1.8	9	1.2	35	1.1
Controller meds	463	3.3	108	2.3	25	2.8†	54	1.7
BMI
Underweight	599	5.0	137	3.1	26	4.7†	120	4.0	<0.0001
Normal	7,728	60.1	2,332	52.3	443	53.3	1,452	48.8
Overweight	2,325	18.2	967	23.0	160	21.6	724	24.7
Obese	2,212	16.6	977	21.6	158	20.4	694	22.5
Other likely smoke exposures
Past month marijuana useg
No	12,593	97.8	3,536	80.6	528	68.0	1,515	51.4	<0.0001
Yes	271	2.2	877	19.4	259	32.0	1,475	48.6
	Mean	SE	Mean	SE	Mean	SE	Mean	SE
Past week SHSh exposure	1.6	0.1	3.1	0.2	4.7	0.4	9.2	0.5	<0.0001

^aN = 21,054 adolescents and young adult respondents with complete data on all variables

^bNever use = never used tobacco; Noncurrent use = ever used tobacco, but not in the past 30 days; Noncombustible use only = exclusive noncombustible (e-cigaretteⁱ, smokeless tobacco, or snus) use in the past 30 days; Combustible use = combustible (cigarette, pipe, hookah, or cigar) use in the past 30 days regardless of noncombustible use.

^cUnweighted

^dChi-square test was used for categorical variables and ANOVA was used for continuous variables.

^eNon-Hispanic

^fIncludes American Indian or Alaska Native, Asian Indian, Chinese, Filipino, Japanese, Korean, Vietnamese, Other Asian, Native Hawaiian, Guamanian or Chamorro, Samoan, Other Pacific Islander, and multiracial

^gMarijuana use variable does not distinguish between combustible and noncombustible use.

^hSHS = secondhand smoke

ⁱAt W4, participants were asked about use of any e-products (e-cigarettes, e-cigars, e-pipes, and e-hookah), all of which are referred to as e-cigarettes here.

^†Estimate should be interpreted with caution because it has low statistical precision. It is based on a denominator sample size of less than 50, or the coefficient of variation of the estimate or its complement is larger than 30%.

Supplemental Table 7 reports standard errors for all the weighted estimates presented in this table.

Association between tobacco use and functionally-important respiratory symptoms

Functionally-important respiratory symptoms were present in 10.0% overall, 13.8% of current combustible users, 9% of current noncombustible users, 8.2% of noncurrent users and 9.7% of never users (p<0.0001). As shown in Table 2, functionally-important respiratory symptoms were more prevalent among females, younger age groups, those with higher BMI, those with past month marijuana use (14.4%), and those with asthma.

Model 1—categorical variables describing combustible and noncombustible use

Table 2 includes the two adjusted multivariable regression models assessing the associations between tobacco use and functionally-important respiratory symptoms. With respect to the four categories of use, Model 1 demonstrated that only current combustible use was independently associated with functionally-important respiratory symptoms (relative risk [RR] =1.52, [95% CI 1.29, 1.80]) compared to never users. Males, older age groups, and Hispanics were less likely to experience functionally-important respiratory symptoms. Asthmatics, those with past-month marijuana use (1.54 [1.34, 1.77]) and those with secondhand smoke exposure (1.04 [1.03, 1.05] for each additional 5 hours/week) were more likely to experience functionally-important respiratory symptoms.

Model 2: cigarette and e-cigarette frequency

With respect to cigarette and e-cigarette frequency, the prevalence of past 30-day cigarette and e-cigarette use was 11.6% and 7.8%, respectively. Figure 2 shows the unadjusted relationships, illustrating a dose-response between higher frequency of cigarette use and functionally-important respiratory symptoms (with prevalence increasing from 9.2% among never users to 23.2% among daily users), but little change with higher frequency of e-cigarette use. This also suggests an apparent threshold for the association between higher frequency of cigarette use and functionally-important symptoms, with little response up to 1-14 days per month of cigarette use. The linear trend for cigarettes was statistically significant (p<0.0001) whereas the trend for e-cigarettes was not (p >0.1).

Figure 2.

Relation^a between the frequency of cigarette^b and e-cigarette^c,d use and weighted percentage with functionally-important respiratory symptoms for adolescents (age 12-17) and young adults (age 18-24), unadjusted for covariates, at Wave 4 of the Population Assessment of Tobacco and Health (PATH) Study.

notes:

^a The * indicates that these weighted prevalence estimates are significantly different (p<0.05) from never users of that product. The linear trend for cigarettes was statistically significant (p<0.0001) whereas the trend for e-cigarettes was not (p >0.1)

^b Cigarette use frequency refers to cigarette use within the past 30 days, regardless of other tobacco products used. Never = never used cigarettes (N=15,827); Noncurrent = ever use of cigarettes, but no cigarette use within the past 30 days (N=3,038); Other categories refer to the ranges of days cigarettes were used within the past 30 days: 1-14 days (N=1,033); 15-29 days (N=273); Daily (N=883). Overall weighted P30D cigarette prevalence was 11.6% (SE=0.3)

^c At W4, participants were asked about use of any e-products (e-cigarettes, e-cigars, e-pipes, and e-hookah), all of which will be referred to as e-cigarettes here.

^d E-cigarette use frequency refers to e-cigarette use within the past 30 days, regardless of other tobacco products used. Never = never used e-cigarettes (N=14,943); noncurrent = ever use of e-cigarettes, but no e-cigarette use within the past 30 days (N=4,596); other categories refer to the ranges of days e-cigarettes were used within the past 30 days: 1-14 days (N=1,019); 15-29 days (N=195); Daily (N=301). Overall weighted P30D e-cigarette prevalence was 7.8 (SE=0.2).

In the second multivariable regression (Model 2, Table 2), cigarette frequency categories above 14 days/month were significantly associated with functionally-important respiratory symptoms in a dose-dependent manner: RR=1.15 [0.91, 1.46] for 1-14 days of cigarette use, 1.93 [1.50, 2.49] for 15-29 days of cigarette use, and 2.80 [2.25, 3.47] for daily cigarette use. As with the previous model, frequency of e-cigarette use was not associated with functionally-important respiratory symptoms.

Sensitivity analysis

The findings in Models 1 and 2 described above were unchanged when all analyses were rerun using the respiratory symptoms index as a continuous dependent variable and multivariable linear regression.

Discussion

In this population-based study of adolescents and young adults, use of combusted tobacco products was independently associated with functionally-important respiratory symptoms, with higher frequency of cigarette smoking associated with higher prevalence of these symptoms. In contrast, we did not find evidence associating any measure of e-cigarette use and functionally-important respiratory symptoms after accounting for cigarette and marijuana use, secondhand smoke exposure and asthma. These findings are similar to that of McConnell et al.,¹² who found no relation between e-cigarette use in adolescents and wheezing, although they did find increased risk of cough/phlegm. Unsurprisingly our findings mirror those assessing wheeze in an adolescent-only PATH sample; however, that study did not include cough or look at comparative effects of cigarette use or marijuana on respiratory symptoms.¹³ Our findings differ from others in the adult literature, including Li et al,²⁴ who found increased risk of wheezing and related respiratory symptoms in current e-cigarette users compared to never users in in the PATH Study adult sample; however, they did not account for past cigarette use or marijuana smoking.

In adolescents/young adults, the tobacco use history is limited. Moreover, cigarette smoking was uncommon among adolescents but increased four-fold in young adults; in contrast, e-cigarette use prevalence was similar among adolescents and young adults. More research can examine the role of early-onset e-cigarette use in young adult transitions to cigarette smoking, given strong evidence from longitudinal studies of adolescents associating e-cigarette use and subsequent onset of cigarette use.²⁵

With regard to marijuana use, there is high plausibility of marijuana contributing to respiratory symptoms given similarities in particulates and tar constituents.^26,27 Furthermore, numerous studies to date suggest a link between marijuana use and respiratory disorders.^28–30 Although there is no detail for marijuana exposure mode (ingested, smoked, vaped, etc.) in the PATH Study, smoking is the predominant exposure route by frequency in adolescent/young adult users.³¹ In spite of this measurement limitation, there was a clinically important association between marijuana use and functionally-important respiratory symptoms that clinicians and users should be aware of with legalization and widespread use of recreational marijuana across the U.S.

It is unsurprising that subjects with asthma are more likely to experience functionally-important respiratory symptoms, as wheezing is the cardinal symptom of this disease. Similarly, the link between secondhand smoke exposure and respiratory symptoms is well known.¹ Although high exposure to SHS was relatively rare, some 5% of respondents reported 75 hours or more per week of SHS exposure. Those at higher exposure had increased risk for reporting functionally-important respiratory symptoms compared to those not exposed to SHS (a relative risk of 1.04^¹⁵ or 1.8). Similar to other national studies that have found reduced odds of asthma in Hispanic adolescents,³⁰ this study found a lower prevalence of functionally-important respiratory symptoms among Hispanic adolescents. The prevalence of asthma within this cohort was higher than expected based on US prevalence data, likely due to self-reporting. This conjecture is supported by the low rate of medication use for asthma among this group. Subjects with asthma not on medications, however, did have markedly increased risk of functionally-important respiratory symptoms.

Limitations

These data are cross-sectional and thus temporal relationships cannot be determined. In addition, these data predate the significant increases in e-cigarette use from 2017, when 11.3% of high school students reported e-cigarette use to the peak seen in 2019 and subsequent decline.^4,5 Although lower rates were found in 2021 (11.3%), intensity of use was increased; 43.6% of e-cigarette users report using on 20 or more days/month.⁶ As previously mentioned, due to changes in the survey collection methodology, no direct comparison can be made of 2021 data with that of previous years. These data also predate the popularity of nicotine-salt pod-based devices such as JUUL that occurred among young people in 2018 to 2019, although these products were on the market during the time of this survey (2016-17). While a combusted tobacco cigarette is a relatively standardized product, there is substantial heterogeneity in the product characteristics of e-cigarette devices (e.g., temperature, voltage, vapor production, refill capacity), as well as in the characteristics of the solutions used within the devices (e.g., humectant type, flavors and nicotine levels). Of concern, studies have found that glycerin and propylene glycol aerosol alter inflammatory pathways in the lung,³² and direct cytotoxicity has been shown in vitro from flavoring chemicals.³³ These flavor chemicals are major ingredients of many e-liquids and may have higher concentrations than that of nicotine.³³ Further, while toxicologically-relevant constituents have been identified in sufficient quantities within e-cigarette products to cause disease, including increased susceptibility to pneumococcal ³⁴and viral^35,36 infections, there have been few human studies.

The outbreak of serious lung injury and respiratory compromise associated with e-cigarette use in the U.S. that was recognized in 2019 is cause for concern.^37,38 Recognizing that products in common use in 2016-17 differ from the product and e-juice characteristics in 2019, the present findings regarding e-cigarettes and respiratory symptoms may not be generalizable to newer products. Additional limitations include that all data are self-reported and may include recall bias. Further, tobacco use prevalence was quite low in adolescents under the age of 15, thus caution is warranted in extending these findings to younger adolescents. Asthma diagnosis is reported based on youth’s recollection that a health professional told them that they have asthma, which may be subject to recall bias or misclassification. There may also be other aspects of cognitive bias that may have influenced reporting, such as desire to alter symptom reporting they may attribute to product use or change in use, however this cannot be controlled. The significant strengths of this current study include a large, nationally representative sample enabling population estimates. In addition, survey items have been adapted from national and international surveys with high validity and reliability.

In summary, adolescents and young adults who use combustible tobacco, use marijuana or have secondhand smoke exposure were more likely to experience functionally-important respiratory symptoms, with risks increasing further with more frequent cigarette use. These modifiable exposures underscore the roles that tobacco and marijuana may play with respect to asthma symptoms in young people. We also emphasize concern about young adult transitions to cigarette smoking and the role e-cigarette use among adolescents may play in this transition. While there was no independent association between e-cigarette use and functionally-important respiratory symptoms, there are important limitations to this finding, with substantial product heterogeneity and timing prior to the marked increase in e-cigarette use from pod-based systems being introduced to the market. Additional epidemiologic and toxicologic studies assessing e-cigarette exposure and respiratory symptoms will be of high importance, along with longitudinal studies like the PATH Study that carefully assess tobacco use and health outcomes over time.

Article Summary:

In this study cigarette use, second-hand smoke exposure and marijuana use were associated with respiratory symptoms among youth, but e-cigarette use was not.

What’s Known on This Subject:

Tobacco use causes disease across all age groups, with studies linking adolescent cigarette smoking and respiratory symptoms including cough and wheeze. Less is known about health risks from other tobacco products and whether there are thresholds of use for symptoms.

What’s New:

In this nationally-representative study of 21,057 12-24 year olds, those who smoke cigarettes, use marijuana, and have secondhand smoke exposure were more likely to report functionally-important respiratory symptoms. Risk increased with increased frequency of cigarette use but not e-cigarette use.