Use of E-Cigarettes and Self-Reported Lung Disease Among US Adults

Abstract

Objectives

Initially marketed for smoking cessation, electronic cigarettes (e-cigarettes) are commonly regarded as safer than combustible cigarettes because they usually contain less nicotine and do not use combustion. However, few studies have examined the health effects of e-cigarettes. The objective of this study was to examine whether e-cigarette use had a differential effect on the prevalence of lung disease among current, former, and never tobacco users.

Methods

We analyzed data from respondents aged ≥18 (n = 45 908) who responded to questions about e-cigarette use and lung disease in the 2016 Behavioral Risk Factor Surveillance System (BRFSS) survey. We calculated crude odd ratios (ORs) and ORs adjusted by 15 sociodemographic and health behavior factors: age, sex, race/ethnicity, annual household income, health insurance, personal physician, health status, body mass index, education, marital status, exercise, alcohol use, tobacco smoking, tobacco chewing, and metropolitan status.

Results

We found a significant association between e-cigarette use and lung disease, which was significantly modified by tobacco use. Among never tobacco users, the adjusted odds of reporting lung disease were 4.36 (95% CI, 1.76-10.77) times higher among everyday e-cigarette users than among never e-cigarette users. Among current tobacco users, the adjusted odds of reporting lung disease were 1.47 (95% CI, 1.13-1.92) times higher among everyday e-cigarette users than among never e-cigarette users.

Conclusions

People who have never smoked combustible cigarettes should refrain from starting e-cigarettes, because e-cigarettes carry a significant risk of lung disease independent of tobacco smoking. Additional prospective research into the harmful effects of e-cigarettes would help to further elucidate this link.

Electronic cigarettes (e-cigarettes) are battery-powered devices that deliver aerosols containing nicotine, chemicals, and flavors such as fruit, mint, or candy for users to inhale. These devices include other electronic vaping products, such as electronic hookahs, vape pens, and electronic cigars (e-cigars). E-cigarettes function by heating up a liquid solution until it is aerosolized and can be vaped.¹ Since the introduction of e-cigarettes into the US market in 2007, their popularity and use have markedly increased among adults.^2,3 In 2016, the prevalence of e-cigarette use among US adults aged ≥18 was 4.5%.⁴ Among US adults aged 18-24, current e-cigarette use increased by 46.2% from 2017 (2.8%) to 2018 (3.2%).⁵ The e-cigarette industry generated an estimated revenue of $3.4 billion in 2019 in the United States, which was an increase from $1.3 billion in 2018.⁶

Although the causality between combustible cigarettes or tobacco smoking and lung disease is well established, the relative novelty of e-cigarettes has made it difficult to study the safety and effects of e-cigarette use. This potential association warrants further investigation.⁷ In August 2019, the New York State Department of Health released a health advisory discouraging people from using vaping products.⁸ Thirty-four cases of severe pulmonary illness requiring hospitalization among patients aged 15-46 have been reported since September 5, 2019, and each patient had a vaping history of weeks to months before experiencing symptoms.⁸ As of November 2019, the Centers for Disease Control and Prevention had received reports of 2172 product-associated lung injury from e-cigarettes or vaping from 49 states, the District of Columbia, and 2 US territories.⁹ These reports of lung injury suggest that, despite their relatively short time in the market, e-cigarettes have the potential to cause lung damage with use.⁹

The objective of our study was to examine whether e-cigarette use has a differential effect on the prevalence of lung disease among current, former, and never tobacco users. We hypothesized that continuous use of e-cigarettes would lead to continuous harm on lung function and could potentially proliferate into lung disease.

Methods

We conducted a secondary analysis of data from the 2016 Behavioral Risk Factor Surveillance System (BRFSS) survey¹⁰ to assess the association between e-cigarette use and lung disease among US adults. The BRFSS is a national survey of health conducted in all 50 US states and territories. Using random-digit-dialing techniques through landline telephones and cell phones and in-house interviews, the BRFSS administers annual questionnaires to noninstitutionalized US residents aged ≥18 to collect information on health-related risk behaviors, chronic health conditions, and the use of preventive services. For our study, we examined data on adults aged ≥18 who responded to questions about e-cigarette use and diagnosis of lung disease in the BRFSS survey. Respondents were asked to describe their e-cigarette smoker status as everyday user, some-day user, former user, or never user. They were also asked whether they had ever been told that they had chronic obstructive pulmonary disease (COPD), emphysema, or chronic bronchitis.

Of the 486 303 US adults aged ≥18 who responded to the 2016 BRFSS questionnaire, 6697 did not have data on age, 18 552 did not respond to the question about e-cigarette smoker status, and 1956 did not answer whether they had ever been diagnosed with COPD, emphysema, or chronic bronchitis; all were excluded. The remaining 459 098 respondents were included in the study.

The independent variable was e-cigarette smoker status, which included e-cigarettes and other electronic vaping products (electronic hookah, vape pens, e-cigars, and others) as defined by the BRFSS. We categorized e-cigarette smoking status based on the respondent’s history and frequency of e-cigarette use. The variable included the following modifiers: current user, everyday; current user, some days; former user; and never user. The dependent variable was presence of COPD, emphysema, or chronic bronchitis, as defined by a yes/no response to the following question: “(Ever told) by a doctor that you have COPD, emphysema, or chronic bronchitis?”

We assessed the following possible confounders: race/ethnicity (non-Hispanic White, non-Hispanic Black, non-Hispanic other race/multiracial, and Hispanic), age (18-24, 25-44, 45-64, ≥65), sex (male, female), annual household income (<$15 000, $15 000-$24 999, $25 000-$34 999, $35 000-$49 999, ≥$50 000), health insurance (any kind of health care coverage, including health insurance, prepaid plans such as health maintenance organizations, or government plans such as Medicare or Indian Health Service vs none), access to a personal physician (having ≥1 personal physician or health care provider vs none), health status (self-reported as good/better or fair/poor), body mass index (defined as underweight [<18.5 kg/m²], normal weight [18.5-24.9 kg/m²], overweight [25.0-29.9 kg/m²], and obese [≥30.0 kg/m²]), education (<high school graduate, high school graduate, attended college/technical school, and college graduate/technical school graduate), marital status (married or in a relationshipvs. not married [ie, divorced, widowed, separated, or never married]), engagement in physical activity (self-reporting physical activity or exercise during the past 30 days other than for a job), alcohol use (self-reported as heavy drinker or not a heavy drinker), metropolitan status (self-reported as urban, suburban, or rural), cigarette use (defined as smoking ≥100 cigarettes in their lifetime and categorized as current smoker, former smoker, and never smoker), and tobacco chewing (defined as chewing tobacco, snuff, and snus). Heavy drinker in alcohol use was defined for men as having >14 drinks per week and for women as having >7 drinks per week. Current smoker was defined as someone who has smoked in the past 30 days, former smoker as someone who has smoked ≥100 cigarettes in their lifetime, and never smoker as someone who has never smoked. The Florida International University Office of Research Integrity determined that this study was exempt from institutional review board review.

Statistical Analysis

We conducted a descriptive analysis to assess the characteristics of our study sample. We conducted a bivariate analysis using the Pearson χ² test to determine potential confounders for our sample. We used binary logistic regression to calculate crude odds ratios (ORs) and adjusted ORs (aORs) with 95% CIs to assess the independent association between e-cigarette use and lung disease. We adjusted for potential confounders. We examined the correlation matrix of all potential confounders and main exposure to check for collinearity in the adjusted model. We examined whether the use of e-cigarettes had a differential effect on the prevalence of lung disease among current, former, and never tobacco users. We conducted a sensitivity analysis including only respondents aged ≥45. We considered P < .05 to be significant. We used Stata version 15.0 to account for the BRFSS survey design and for all statistical analyses.¹¹

Results

Of 459 098 US adults in the study sample, 56.7% were female, 77.5% self-identified as non-Hispanic White, 8.1% as non-Hispanic Black, 8.1% as Hispanic, and 6.3% as non-Hispanic other race/multiracial.

All 15 selected characteristics from the BRFSS were significantly associated with the use of e-cigarettes (Table 1). The percentage of current everyday users was significantly higher among respondents who were aged 25-44 (vs other age categories; 46.3%, P < .001), male (vs female; 64.7%, P < .001), non-Hispanic White (vs other races/ethnicities; 79.1%, P < .001), not married (vs married; 53.3%, P < .001), former smokers (vs current or never smokers; 56.1%, P < .001), and not heavy drinkers (vs heavy drinkers; 88.9%, P < .001); had an annual household income >$35 000 (vs other income levels; 95.4%, P < .001), health insurance (vs no health insurance; 86.8%, P < .001), access to a personal physician (vs no access; 70.4%, P < .001), and a normal weight (vs other weight category; 34.0%, P < .001); reported good or better health status (vs fair or poor health status; 79.2%, P < .001) and physical activity (vs no physical activity; 77.6%, P = .02); attended college/technical school (vs other education level; 40.1%, P < .001); never chewed tobacco (vs current tobacco user; 93.7%, P < .001); and lived in urban areas (vs suburban or rural areas; 42.3%, P < .001).

Table 1.

Self-reported demographic characteristics of US adults aged ≥18 (N = 459 098), by e-cigarette use,^a 2016 Behavioral Risk Factor Surveillance System^b

Characteristic	Current everyday user, % (n = 4957)c	Current some-day user, % (n = 10 089)c	Former user, % (n = 58 286)c	Never user, % (n = 385 766)c	P valued
Age, y					<.001
18-24	21.4	28.4	23.7	9.7
25-44	46.3	41.0	44.8	30.6
45-64	27.0	26.1	25.7	35.4
≥65	5.3	4.5	5.8	24.3
Sex					<.001
Male	64.7	57.9	55.6	46.5
Female	35.3	42.1	44.4	53.5
Race/ethnicity					<.001
Non-Hispanic White	79.1	68.2	66.4	63.0
Non-Hispanic Black	5.2	10.6	11.1	11.8
Non-Hispanic other race/multiracial	7.6	9.3	7.6	7.9
Hispanic	8.2	12.0	15.0	17.2
Annual household income, $					<.001
<15 000	9.8	14.9	13.1	10.3
15 000-24 999	16.5	21.0	19.5	16.5
25 000-34 999	11.9	12.5	11.2	10.1
35 000-49 999	47.7	14.6	14.6	13.2
≥50 000	47.7	37.0	41.5	49.9
Health insurancee					<.001
Yes	86.8	83.3	85.4	89.6
No	13.2	16.7	14.6	10.4
Access to a personal physician					<.001
Yes	70.4	68.6	69.1	80.8
No	29.6	31.4	30.9	19.2
Health status					<.001
Good/better	79.2	77.6	80.5	82.7
Fair/poor	20.8	22.4	19.5	17.3
Body mass index, kg/m2					<.001
Underweight, <18.5	2.6	2.9	2.6	1.8
Normal weight, 18.5-24.9	34.0	38.4	36.5	32.1
Overweight, 25.0-29.9	32.9	31.5	33.1	35.9
Obese, ≥30.0	30.5	27.1	27.7	30.2
Education					<.001
<High school graduate	11.4	16.1	13.8	13.6
High school graduate	36.2	35.7	32.7	26.7
Attended college/technical school	40.1	34.9	36.3	29.8
College graduate/technical school graduate	12.3	13.3	17.2	29.9
Marital statusf					<.001
Married or in a relationship	46.7	38.3	42.2	59.2
Not married	53.3	61.7	57.8	40.8
Exerciseg					.02
Physical activity	77.6	76.2	76.4	75.5
None	22.4	23.8	23.6	24.5
Alcohol useh					<.001
Not a heavy drinker	88.9	87.2	88.3	95.1
Heavy drinker	11.1	12.8	11.7	4.9
Tobacco smokingi					<.001
Current smoker	34.2	62.1	46.5	7.6
Former smoker	56.1	15.2	22.5	24.6
Never smoker	9.7	22.7	31.1	67.8
Tobacco chewingj					<.001
Current user	6.3	8.2	6.3	2.8
Never user	93.7	91.8	93.7	97.2
Metropolitan statusk					<.001
Urban	42.3	38.1	38.5	41.9
Suburban	36.5	38.8	39.9	39.8
Rural	21.1	23.2	21.6	18.3

Abbreviations: BRFSS, Behavioral Risk Factor Surveillance System; e-cigarette, electronic cigarette.

^aBased on responses to the following BRFSS survey question: “Do you now use e-cigarettes or other electronic ‘vaping’ products every day, some days, or not at all?” E-cigarette use is defined as use of an e-cigarette or other electronic “vaping” products, including electronic hookahs, vape pens, and electronic cigars.

^bData source: Centers for Disease Control and Prevention.¹⁰

^cNumbers are unweighted and percentages are weighted based on the complex sampling methodology of the BRFSS, which took into consideration unequal probability of selection, differential nonresponse, and potential deficiencies in sampling frame.

^dUsing the Pearson χ² test, with P < .05 considered significant.

^eHealth insurance is defined as having any kind of health care coverage, including health insurance, prepaid plans such as health maintenance organizations, or government plans such as Medicare or Indian Health Service.

^fNot married includes divorced, widowed, separated, or never married.

^gPhysical activity is defined as doing physical activity or exercise during the past 30 days other than for a regular job.

^hHeavy drinker is defined for men as having >14 drinks per week and for women as having >7 drinks per week.

ⁱCurrent smokers are defined as those who currently smoke, either every day or some days. Former smokers are defined as those who used to smoke but do not smoke currently. Never smokers are defined as those who never smoked tobacco.

^jChewing tobacco includes chewing tobacco, snuff, and snus.

^kSelf-reported metropolitan status, with urban defined as in the center of a metropolitan statistical area (MSA), suburban as inside a suburban county of the MSA, and rural as not in an MSA.

The percentage of respondents who reported lung disease was significantly higher among respondents who were aged ≥65 (vs <65; 12.4%, P < .001), female (vs male; 7.3%, P < .001), non-Hispanic White (vs other races/ethnicities; 7.5%, P < .001), underweight (vs other weight categories; 10.3%, P < .001), not married (vs married; 8.0%, P < .001), and current tobacco smokers (vs former or never smokers; 14.4%, P < .001); had health insurance (vs no health insurance; 6.7%, P < .001), access to a physician (vs no access; 7.4%, P < .001), and <high school degree (vs ≥high school degree; 11.7%, P < .001); reported fair or poor health (vs good or better health; 19.3%, P < .001), an annual household income <$15 000 (vs ≥$15 000; 13.6%, P < .001), and no physical activity (vs physical activity; 12.1%, P < .001); and lived in rural areas (vs suburban or urban areas; 10.6%, P < .001; Table 2). The prevalence of lung disease was significantly higher among respondents who reported using e-cigarettes every day, some days, or formerly compared with never users. The unadjusted odds of having lung disease among everyday, some-day, and former e-cigarette users were about twice that of respondents who had never used e-cigarettes.

Table 2.

Self-reported characteristics of US adults aged ≥18 (N = 459 098), by presence of lung disease,^a and the unadjusted associations between e-cigarette use^b and potential confounders with lung disease, 2016 Behavioral Risk Factor Surveillance System^c

Characteristic	Lung disease, % (n = 38 917)	No lung disease, % (n = 420 181)	P valued	Odds ratio (95% CI)
E-cigarette use			<.001
Current everyday use	9.7	90.3		1.83 (1.59-2.10)
Current some-day use	12.0	88.0		2.33 (2.07-2.62)
Former user	10.1	89.9		1.92 (1.82-2.03)
Never user	5.5	94.5		1 [Reference]
Age, y			<.001
18-24	2.0	98.0		1 [Reference]
25-44	3.2	96.8		1.62 (1.40-1.88)
45-64	8.1	91.9		4.29 (3.73-4.93)
≥65	12.4	87.6		6.83 (5.94-7.85)
Sex			<.001
Male	5.8	94.2		1 [Reference]
Female	7.3	92.7		1.30 (1.24-1.36)
Race/ethnicity			<.001
Non-Hispanic White	7.5	92.5		1 [Reference]
Non-Hispanic Black	6.8	93.2		0.90 (0.83-0.97)
Non-Hispanic other race/multiracial	5.0	95.0		0.65 (0.57-0.75)
Hispanic	3.6	96.4		0.46 (0.42-0.50)
Annual household income, $			<.001
<15 000	13.6	86.4		4.46 (4.14-4.80)
15 000-24 999	10.0	90.0		3.14 (2.94-3.36)
25 000-34 999	7.7	92.3		2.36 (2.17-2.57)
35 000-49 999	6.3	93.7		1.92 (1.76-2.08)
≥50 000	3.4	96.6		1 [Reference]
Health insurancee			<.001
Yes	6.7	93.3		1 [Reference]
No	5.2	94.8		0.76 (0.70-0.83)
Access to a personal physician			<.001
Yes	7.4	92.6		2.11 (1.96-2.27)
No	3.6	96.4		1 [Reference]
Health status			<.001
Good/better	3.8	96.2		1 [Reference]
Fair/poor	19.3	80.7		6.09 (5.82-6.38)
Body mass index, kg/m2			<.001
Underweight, <18.5	10.3	89.7		2.03 (1.76-2.34)
Normal weight, 18.5-24.9	5.4	94.6		1 [Reference]
Overweight, 25.0-29.9	5.7	94.3		1.07 (1.01-1.14)
Obese, ≥30.0	6.6	91.1		1.71 (1.62-1.81)
Education			<.001
<High school graduate	11.7	88.3		4.43 (4.13-4.75)
High school graduate	7.7	92.3		2.81 (2.64-2.98)
Attended college/technical school	6.4	93.6		2.30 (2.16-2.45)
College graduate/technical school graduate	2.9	97.1		1 [Reference]
Marital statusf			<.001
Married or in a relationship	5.4	94.6		1 [Reference]
Not married	8.0	92.0		1.53 (1.46-1.60)
Exerciseg			<.001
Physical activity	4.8	95.2		1 [Reference]
None	12.1	87.9		2.72 (2.60-2.85)
Alcohol useh			.45
Not a heavy drinker	6.6	93.5	1 [Reference]
Heavy drinker	6.8	93.2	1.04 (0.94-1.15)
Tobacco smokingi			<.001
Current smokers	14.4	85.6		5.71 (5.39-6.05)
Former smokers	10.3	89.7		3.87 (3.65-4.10)
Never smokers	2.9	97.1		1 [Reference]
Tobacco chewingj			.46
Current user	6.8	93.2	1.04 (0.93-1.17)
Never user	6.6	93.5	1 [Reference]
Metropolitan statusk			<.001
Urban	8.3	91.7		—l
Suburban	8.7	91.3
Rural	10.6	89.4

Abbreviations: BRFSS, Behavioral Risk Factor Surveillance System; e-cigarette, electronic cigarette.

^aLung disease includes chronic obstructive pulmonary disease, emphysema, and chronic bronchitis.

^bBased on responses to the following BRFSS survey question: “Do you now use e-cigarettes or other electronic ‘vaping’ products every day, some days, or not at all?” E-cigarette use is defined as use of an e-cigarette or other electronic “vaping” products, including electronic hookahs, vape pens, and electronic cigars.

^cData source: Centers for Disease Control and Prevention.¹⁰

^dUsing the Pearson χ² test, with P < .05 considered significant.

^eHealth insurance is defined as having any kind of health care coverage, including health insurance, prepaid plans such as health maintenance organizations, or government plans such as Medicare or Indian Health Service.

^fNot married includes divorced, widowed, separated, or never married.

^gPhysical activity is defined as doing physical activity or exercise during the past 30 days other than for a regular job.

^hHeavy drinker is defined for men as having >14 drinks per week and for women as having >7 drinks per week.

ⁱCurrent smokers are defined as those who currently smoke, either every day or some days. Former smokers are defined as those who used to smoke but do not smoke currently. Never smokers are defined as those who never smoked tobacco.

^jChewing tobacco includes chewing tobacco, snuff, and snus.

^kSelf-reported metropolitan status, with urban defined as in the center of a metropolitan statistical area (MSA), suburban as inside a suburban county of the MSA, and rural as not in an MSA.

^lNot deemed clinically relevant.

The role of combustible cigarette smoking as an effect modifier created a significant association between e-cigarette use and the presence of lung disease when we used an interaction term. When baseline characteristics and e-cigarette use were stratified by tobacco smoking status and adjusted for confounders, the odds of having lung disease among current everyday e-cigarette users who were never smokers was 4.36 (95% CI, 1.76-10.77) times higher than among never e-cigarette users (Table 3). Among former combustible cigarette smokers, the odds of having lung disease were 1.46 (95% CI, 1.23-1.88) times higher among current everyday e-cigarette users than among never e-cigarette users. Among current combustible cigarette smokers, the odds of having lung disease were 1.47 (95% CI, 1.13-1.92) times higher among current everyday e-cigarette users than among never e-cigarette users.

Table 3.

Adjusted model stratified by smoking status among US adults aged ≥18 (N = 459 098) based on effect modification log, adjusted for current, former, and never combustible cigarette smokers,^a 2016 Behavioral Risk Factor Surveillance System^b

Characteristic	OR (95% CI)	Current combustible cigarette smokers, aOR (95% CI)	Former combustible cigarette smokers, aOR (95% CI)	Never combustible cigarette smokers, aOR (95% CI)
E-cigarette usec
Current everyday user	1.83 (1.59-2.10)	1.47 (1.13-1.92)	1.46 (1.23-1.88)	4.36 (1.76-10.77)
Current some-day user	2.33 (2.07-2.62)	1.82 (1.56-2.14)	2.05 (1.42-2.94)	1.27 (0.77-2.08)
Former user	1.92 (1.82-2.03)	1.65 (1.48-1.84)	2.05 (1.78-2.37)	1.58 (1.24-2.02)
Never user	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]
Age, y
18-24	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]
25-44	1.62 (1.40-1.88)	2.08 (1.55-2.80)	1.15 (0.65-2.03)	1.47 (1.12-1.94)
45-64	4.29 (3.73-4.93)	4.68 (3.52-6.23)	3.30 (1.89-5.75)	2.56 (1.96-3.36)
≥65	6.83 (5.94-7.85)	8.57 (6.33-11.60)	6.14 (3.51-10.74)	3.64 (2.79-4.74)
Sex
Male	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]
Female	1.30 (1.24-1.36)	1.28 (1.16-1.41)	1.19 (1.09-1.30)	1.39 (1.24-1.57)
Race/ethnicity
Non-Hispanic White	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]
Non-Hispanic Black	0.90 (0.83-0.97)	0.65 (0.54-0.78)	0.67 (0.56-0.79)	1.07 (0.90-1.27)
Non-Hispanic other race/multiracial	0.65 (0.57-0.75)	0.83 (0.69-1.00)	1.08 (0.78-1.52)	0.78 (0.61-1.00)
Hispanic	0.46 (0.42-0.50)	0.36 (0.29-0.46)	0.42 (0.34-0.52)	0.63 (0.52-0.77)
Annual household income, $
<15 000	4.46 (4.14-4.80)	2.14 (1.82-2.52)	1.94 (1.63-2.31)	1.98 (1.63-2.42)
15 000-24 999	3.14 (2.94-3.36)	1.92 (1.66-2.23)	1.61 (1.38-1.86)	1.65 (1.39-1.95)
25 000-34 999	2.36 (2.17-2.57)	1.40 (1.17-1.68)	1.41 (1.21-1.65)	1.42 (1.19-1.71)
35 000-49 999	1.92 (1.76-2.08)	1.33 (1.12-1.59)	1.32 (1.14-1.52)	1.36 (1.12-1.64)
≥50 000	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]
Health insuranced
Yes	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]
No	0.76 (0.70-0.83)	0.76 (0.64-0.90)	0.87 (0.69-1.12)	1.00 (0.79-1.25)
Access to a personal physician
Yes	2.11 (1.96-2.27)	1.51 (1.31-1.74)	1.43 (1.19-1.71)	1.29 (1.08-1.55)
No	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]
Health status
Good/better	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]
Fair/poor	6.09 (5.82-6.38)	3.23 (2.90-3.60)	3.54 (3.22-3.88)	3.11 (2.76-3.50)
Body mass index, kg/m2
Underweight, <18.5	2.03 (1.76-2.34)	1.51 (1.21-1.90)	1.72 (1.31-2.24)	1.51 (1.05-2.17)
Normal weight, 18.5-24.9	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]
Overweight, 25.0-29.9	1.07 (1.01-1.14)	0.90 (0.80-1.02)	0.97 (0.86-1.09)	1.17 (1.00-1.37)
Obese, ≥30.0	1.71 (1.62-1.81)	1.15 (1.02-1.30)	1.23 (1.10-1.38)	1.71 (1.46-2.00)
Education
<High school graduate	4.43 (4.13-4.75)	1.89 (1.60-2.22)	1.76 (1.51-2.04)	1.20 (0.95-1.51)
High school graduate	2.81 (2.64-2.98)	1.31 (1.13-1.51)	1.48 (1.31-1.66)	1.30 (1.12-1.51)
Attended college/technical school	2.30 (2.16-2.45)	1.35 (1.16-1.56)	1.37 (1.21-1.55)	1.48 (1.29-1.70)
College graduate/technical school graduate	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]
Marital statuse
Married or in a relationship	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]
Not married	1.53 (1.46-1.60)	1.03 (0.93-1.14)	1.16 (1.05-1.29)	1.21 (1.07-1.36)
Exercisef
Physical activity	1 [Reference]	1 [Reference]	1 [Reference]	1 [Reference]
None	2.72 (2.60-2.85)	1.32 (1.19-1.45)	1.42 (1.29-1.57)	1.32 (1.17-1.49)

Abbreviations: aOR, adjusted odds ratio; BRFSS, Behavioral Risk Factor Surveillance System; e-cigarette, electronic cigarette; OR, odds ratio.

^aCurrent smokers are defined as those who currently smoke, either every day or some days. Former smokers are defined as those who used to smoke but do not smoke currently. Never smokers are defined as those who never smoked tobacco.

^bData source: Centers for Disease Control and Prevention.¹⁰

^cBased on responses to the following BRFSS survey question: “Do you now use e-cigarettes or other electronic ‘vaping’ products every day, some days, or not at all?” E-cigarette use is defined as use of an e-cigarette or other electronic “vaping” products, including electronic hookahs, vape pens, and electronic cigars.

^dHealth insurance is defined as having any kind of health care coverage, including health insurance, prepaid plans such as health maintenance organizations, or government plans such as Medicare or Indian Health Service.

^eNot married includes divorced, widowed, separated, or never married.

^fPhysical activity is defined as doing physical activity or exercise during the past 30 days other than for a regular job.

In our sensitivity analysis, adjusted for confounders among respondents aged ≥45, the odds of having lung disease among never combustible cigarette smokers were 5.17 times higher among current everyday e-cigarette users than among never e-cigarette users. Among former combustible cigarette smokers, the odds of having lung disease was 1.75 times higher among current everyday e-cigarette users than among never e-cigarette users. Among current tobacco smokers, the odds of having lung disease were 1.66 times higher among current everyday e-cigarette users than among never e-cigarette users.

Discussion

We found a significant association between e-cigarette use and lung disease among adults even after controlling for confounders and stratifying for our effect modifier, tobacco smoking. Current everyday users of e-cigarettes who had never smoked combustible cigarettes had significantly higher odds of having lung disease than people who had never smoked e-cigarettes or combustible cigarettes.

BRFSS respondents aged 18-44 reported lung disease despite being younger than the age at which COPD is typically manifested, usually from age 40 to 50. However, according to a secondary analysis of the Southern California Children’s Health Study in 2017, increased odds of developing bronchitis symptoms were significantly associated with e-cigarette use among adolescents.¹²

Similar to our study, a study that explored the relationship between e-cigarette use and COPD progression found that the use of e-cigarettes was significantly associated with an increase in chronic bronchitis and an exacerbation of COPD, even after adjusting for confounders such as age, race/ethnicity, sex, pack-years (defined as the number of cigarettes smoked per day multiplied by the number of years an individual has smoked), and current smoking status.¹³

A laboratory-based intervention study found that short-term e-cigarette use for 5 minutes caused a decrease in fractional exhaled nitric oxide, an increase in airway impedance, and an increase in peripheral lung airway resistance among adults.¹⁴ The exhaled nitric oxide test is an important tool for evaluating lung function, because nitric oxide mediates oxidative stress, eosinophilic inflammation, and bronchial hyperreactivity. With these results from the intervention study, we hypothesized that continuous use of e-cigarettes would lead to continuous harm on lung function and proliferate into lung disease.

A study conducted in 2017 found that heating propylene glycol and vegetable glycerin, 2 components of e-cigarette liquids, at temperatures greater than 215°C produced toxic carbonyls.⁷ These chemicals are linked to the pathogenesis of asthma.⁷ In addition, the review found another study that demonstrated that long-term inhalation of propylene glycol and vegetable glycerin resulted in a decrease in how much air a person could exhale in a breath during the first second (forced expiratory volume 1 [FEV1]) and a decrease in forced vital capacity (FVC), the total amount of air exhaled during the FEV test.¹⁵ These studies demonstrated that compounds contained in e-cigarette formulas are harmful to respiration and may explain the increased association of e-cigarettes with lung disease among people who had no history of tobacco smoking in our study.

Other studies showed that chemicals in e-cigarettes lead to decreased macrophage and neutrophil activity, increased inflammatory cytokine inflammation and airway hypersensitivity, and increased airway resistance.^6,16 Another study found that both combustible cigarettes and e-cigarette liquid containing flavorings and nicotine had similar cytotoxic effects on lung epithelia and keratinocytes.¹⁷ Thus, literature supports the biological plausibility that e-cigarette use is associated with the presence of lung disease.

Using a study design that was similar to our study design, research that examined the 2016 Hawaii BRFSS found an independent association between e-cigarette use and chronic respiratory disorders, including asthma and COPD.¹⁸ The main difference between that study and our study is that the latter examined data in only 1 US state, Hawaii, with a sample size of 8087, whereas our population size was larger and included all US states and territories. A similar secondary analysis of the Population Assessment of Tobacco and Health cohort study, conducted during 2013-2014 with 32 320 adult participants, showed that e-cigarette users had 43% higher odds of reporting a diagnosis of COPD than did nonusers of e-cigarettes.¹⁹

Our study found a smaller association between e-cigarette use and lung disease among respondents who had a history of tobacco smoking than among respondents who had never smoked tobacco. This observation may be explained by the ceiling effect, wherein people who smoke tobacco already have so much lung damage that the addition of e-cigarette use does not significantly increase the association of e-cigarette use with lung disease. A retrospective–prospective study assessing the health effects of combustible cigarette smokers with COPD who switched to e-cigarettes found no significant changes in postbronchodilator FEV1 and FVC from baseline between people who had switched from combustible cigarettes to e-cigarettes and people who still smoked only combustible cigarettes.²⁰ Therefore, in our study, the presence of lung disease among e-cigarette users who had a history of tobacco smoking was mostly likely due to the effects of combustible cigarette smoking rather than e-cigarette use.

Strengths

Our study had several strengths. First, our sample size was large and should be similar and generalizable to the adult population of the United States. We excluded only a few respondents who did not answer questions related to our exposure or outcome, which was necessary to minimize bias. Second, we adjusted for 15 confounders that may have had an effect on the association between e-cigarette use and presence of lung disease, which was more extensive than other studies we found in our literature review. Third, we stratified our analysis by tobacco smoking as an effect modifier on the association between e-cigarette use and lung disease. By doing this, we were able to better represent the association between our exposure and outcome. Finally, we conducted a sensitivity analysis of only adults aged ≥45, and the results supported our conclusion from our analysis of adults aged ≥18, thereby demonstrating the presence of an association between e-cigarette use and lung disease.

Limitations

Our study also had several limitations. First, our data analysis used a cross-sectional approach that assessed both our exposure, e-cigarette use, and outcome, presence of lung disease, simultaneously. Therefore, we were not able to determine the temporal relationship between the exposure and the outcome, and we could not draw conclusions on causality.

Second, we were limited to the predetermined questions and answer choices. In the question assessing e-cigarette user status, the BRFSS assessed only whether users smoked every day, some days, formerly, or never, but it did not quantify the dose and frequency of use. Thus, substantial individual variation in e-cigarette use may have skewed our association in either direction. The definition of e-cigarettes included electronic hookahs, vape pens, e-cigars, and others, which prevented us from determining which device could have been associated with lung disease. The lack of specificity in the answer choices prevented a clear analysis of how living environment may influence e-cigarette use.

Third, our study was unable to quantify the duration and frequency of e-cigarette use in each respondent and its relationship to lung disease. From the time of this BRFSS questionnaire, e-cigarette users would have only been able to use e-cigarettes for a maximum of 12 years. Compared with combustible cigarettes, which require chronic and heavy use to cause COPD, e-cigarettes may be too new to be associated with any type of lung disease. However, in a preliminary study published in September 2019 by the Illinois Department of Public Health, 53 patients were admitted at 1 hospital from June through August 2019 who were found to have bilateral pulmonary infiltrates and who had reported use of e-cigarette devices in the 90 days before onset of symptoms.²¹ These patients had no other medical explanation for the pulmonary infiltrates and were in good health before presenting to the hospital.¹⁹ This evidence suggests that the association between e-cigarette use and lung disease may require a shorter timeline than that required to determine the association between combustible cigarettes and lung disease. If e-cigarettes are already causing pathophysiological changes in the lungs, it is not unreasonable to believe that these lung changes can eventually develop into COPD.

Conclusion

Our study found that e-cigarette use could be significantly associated with lung disease. The stronger association between current everyday e-cigarette users and lung disease among respondents who had never smoked tobacco cigarettes suggests that e-cigarettes have a significant independent association with lung disease regardless of tobacco smoking history. We conclude that people who have never smoked tobacco should not start using e-cigarettes. Given the results of our study, future research is needed to explore a potential causal relationship between e-cigarette use and lung disease. As e-cigarettes become more popular, longitudinal observational studies will be needed to determine both the pathophysiology and long-term effects of e-cigarettes. Public health initiatives should be launched to educate the public about the potential harmful effects of e-cigarette use.

ORCID iD

Trisha Nguyen, BS https://orcid.org/0000-0001-9347-398X