Use of Electronic Cigarettes with Conventional Tobacco Is Associated with Decreased Sleep Quality in Women

To the Editor:

Electronic cigarettes (e-cigarettes) are highly popular, although their long-term health effects remain unknown. Clearly, conventional tobacco and nicotine adversely affect sleep, but it is unknown whether e-cigarettes will also have deleterious effects. Because the most popular e-cigarettes have higher levels of nicotine than tobacco cigarettes, and vapers tend to use them immediately before bedtime, we hypothesized that e-cigarette and dual use (e-cigarettes and cigarettes) would lead to worse sleep disruptions compared with conventional smoking.

Sleep disturbance is common both in nicotine-dependent young adults and in patients with tobacco-related lung disease (1, 2). In addition, sex differences have been demonstrated in prospective observational trials studying relapse, long-term abstinence, and sleep quality in conventional cigarette users (3–5). We aimed to assess whether sleep disturbance is associated with e-cigarette use by clinically validated sleep and cough questionnaires: Pittsburgh Sleep Quality Index (PSQI; a 19-item assessment of sleep quality during the prior month, with reasonable sensitivity and specificity for detection of sleep disorders) and Leicester Cough Questionnaire (LCQ; a 19-item assessment of the effect of cough severity on health-related quality of life, across physical, psychological, and social domains). Some of the results of these studies have been previously reported in the form of an abstract (6).

Methods

Two hundred seventy-four participants comprising nonsmokers, e-cigarette users, smokers, and dual users were recruited to complete a nationwide online survey through social media advertisements. The online questionnaire was designed with three sections (University of California, San Diego, inhalant use questionnaire; PSQI; and LCQ, with the approval of the University of California, San Diego, institutional review board).

The primary outcome was PSQI scores (range, 0–21, with lower scores denoting healthier sleep quality). Secondary outcomes included sleep latency, presence of cough in the last 30 days, and LCQ scores among those who reported presence of cough (range, 3–21, with higher scores indicating better quality of life). Descriptive analyses were performed on these outcomes by inhalant groups and by sex. ANOVA and chi-square tests were used for group comparisons of continuous variables and categorical variables, respectively.

Linear regression models were performed to study the association between inhalant groups and sleep quality and sleep latency, adjusting for age and sex. Similarly, logistic regression was used to evaluate presence of cough. An interaction term between inhalant groups and sex was included to assess whether effects differ between males and females. All statistical analyses were performed in R (http://cran.r-project.org), version 3.5.2.

Results

Of the 274 respondents included in this study, 139 self-identified as women and participants were ethnically diverse with age ranging from 16 to 74 years (Table 1). Descriptive analysis showed differences in PSQI scores between groups (with lower scores denoting healthier sleep quality). Dual users had the highest scores (8.77) when compared with e-cigarette users (6.87), conventional smokers (7.88), and nonsmokers (7.09, ANOVA P = 0.027; Table 2). There were significant group differences among females but not among males. Linear regression analysis showed a significant interaction effect between sex and dual-user group on sleep quality (P = 0.042). For females, dual use was associated with higher PSQI scores compared with nonsmokers (mean difference = 3.43; 95% confidence interval [CI], 1.63–5.22; P < 0.001), e-cigarette users (mean difference = 3.54; 95% CI, 1.41–5.67; P = 0.001), and conventional cigarette users (mean difference = 2.69; 95% CI, 0.27–5.11; P = 0.029). For males, dual use did not affect PSQI scores compared with nonsmokers (mean difference = 0.851; 95% CI, −0.867 to 2.568; P = 0.330), e-cigarette users (mean difference = 0.570; 95% CI, −1.097 to 2.236; P = 0.502), or conventional smokers (mean difference = −0.254; 95% CI, −2.832 to 2.324; P = 0.846). Sole e-cigarette use was not associated with worse PSQI scores compared with nonsmokers or conventional smokers for both females and males.

Table 1.

Demographics by Group Category

	Nonsmoker (n = 126)	Conventional (n = 25)	Electronic Cigarette User (n = 79)	Dual User (n = 44)	Overall (n = 274)	P Value
Sex, n (%)
Female	81 (64.29%)	15 (60.00%)	25 (31.65%)	18 (41.86%)	139 (50.81%)	<0.001
Male	45 (35.71%)	10 (40.00%)	54 (68.35%)	25 (58.14%)	134 (49.08%)
Total	126 (100%)	25 (100%)	79 (100%)	43 (100%)	273 (100%)
Ethnicity, n (%)
African American	11 (8.94%)	4 (16.67%)	2 (2.63%)	3 (6.82%)	20 (7.49%)	0.003
Caucasian	46 (37.40%)	13 (54.17%)	51 (67.11%)	25 (56.82%)	135 (50.56%)
East Asian	29 (23.58%)	3 (12.50%)	13 (17.11%)	5 (11.36%)	50 (18.73%)
Hispanic	28 (22.76%)	4 (16.67%)	6 (7.89%)	5 (11.36%)	43 (16.10%)
Other	9 (7.32%)	0 (0.00%)	4 (5.26%)	6 (13.64%)	19 (7.12%)
Total	123 (100%)	24 (100%)	76 (100%)	44 (100%)	267 (100%)
Age, yr, mean (SD)	29.4 (13.6)	28.5 (8.3)	31.8 (11.4)	30.5 (12.2)	30.2 (12.3)	0.519

Table 2.

Pittsburgh Sleep Quality Index Scores by Sex and Group

	n	Mean	SD	Minimum	Median	Maximum	ANOVA P Value
Sex
Female	139	8.12	3.97	1	8	18	<0.001
Male	134	6.57	3.15	1	6	15
Overall	273	7.36	3.67	1	7	18
Group
Nonsmoker	126	7.09	3.41	1	6	17	0.027
Conventional	25	7.88	3.84	2	8	17
Electronic cigarette	79	6.87	3.65	1	7	17
Dual user	44	8.77	4.01	1	10	18
Overall	274	7.37	3.66	1	7	18
Group (female)
Nonsmoker	81	7.61	3.60	1	7	17	0.006
Conventional	15	8.20	4.11	2	8	17
Electronic cigarette	25	7.60	4.58	1	7	17
Dual user	18	11.11	3.46	6	11	18
Overall	139	8.12	3.97	1	8	18
Group (male)
Nonsmoker	45	6.16	2.84	1	6	14	0.569
Conventional	10	7.40	3.57	3	7	13
Electronic cigarette	54	6.54	3.11	2	6.5	15
Dual user	25	7.04	3.61	1	6	15
Overall	134	6.57	3.15	1	6	15

Linear regression also revealed a significant interaction effect between sex and dual user group on sleep latency. Female dual users, but not males, were more likely to have increased sleep latency compared with nonsmokers (mean difference = 24.1 min; 95% CI, 12.9–35.2; P < 0.001) and e-cigarette users (mean difference = 17.3 min; 95% CI, 4.2–30.4; P = 0.01). Conventional cigarette use was also associated with increased sleep latency compared with nonsmokers in both males and females.

Report of chronic cough in the prior 30 days increased stepwise from nonsmokers (39.7%) to e-cigarette users (50.6%) to conventional cigarette users (60.0%) to dual users (63.6%). Logistic regression revealed that dual use was associated with increased odds of coughing in the last 30 days compared with nonsmokers (adjusted odds ratio = 2.51; 95% CI, 1.21–5.19; P = 0.013). E-cigarette use alone was not associated with increased odds of coughing. No interaction effect was observed between sex and inhalant groups for presence of cough.

In subjects reporting cough in the prior 30 days, linear regression demonstrated that dual use was associated with lower LCQ scores (i.e., worse severity of cough and diminished quality of life) compared with e-cigarette users (mean difference = −1.65; 95% CI, −3.00 to −0.30; P = 0.017) and nonsmokers (mean difference = −1.31; 95% CI: −2.63 to 0.01; P = 0.051).

Analyzing PSQI scores by presence of cough revealed that those with cough had worse sleep quality (mean = 7.79; SD, 3.43) when compared with those without a cough (mean = 6.97; SD, 3.84; t test P = 0.03). There was a modest correlation between PSQI scores and LCQ scores among subjects who reported cough (Spearman’s rho = −0.22).

Discussion

To our knowledge, this cross-sectional study is the first research to evaluate effects of e-cigarettes on sleep. As the most popular brand of e-cigarettes, JUUL, used by >50% of current e-cigarette vapers since late 2017, contains higher concentrations of nicotine compared with conventional cigarettes (up to 60 mg/ml), we hypothesized that e-cigarette users would have higher nicotine intake and poorer sleep outcomes compared with smokers. We found that dual use of e-cigarettes with conventional tobacco has the highest risk for causing sleep disruption. Mechanistically, this finding is logical if nicotine is the causal agent of sleep disruption, as dual users are more likely to consume greater concentrations of nicotine than either smokers or vapers (7). This notion may reveal the underlying mechanism for poorer sleep quality and increased odds and severity of cough in dual users.

In the context of tobacco use, dual use of e-cigarettes with conventional cigarettes has been shown to be the most common use pattern among consumers (8). Our data suggest that female, but not male, dual users are susceptible to sleep disturbances. We hypothesize that because women are more vulnerable to sleep disturbances (3), have increased risk of developing insomnia (9–11), and have more severe chronic nocturnal cough secondary to gastroesophageal reflux (resulting in sleep fragmentation) (12, 13), nicotine-induced acid reflux has more profound effects on women and leads to poorer sleep quality (14, 15). Women also have greater withdrawal-related distress and urge to smoke to relieve withdrawal distress when compared with men, which may also be leading to increased sleep disruption (5).

This study is limited, in that subjects chose to participate based on advertisements that were disseminated through social media. Thus, some participation bias is likely to have occurred. We were deliberately broad in our reach to avoid potential selection bias, reaching 134 men and 139 women across a broad range of ages and ethnicities, in more than 20 different states and 12 countries. Nonetheless, we cannot objectively determine the motivation for participating in our survey.

We conclude that dual use of e-cigarettes with conventional tobacco by women is worse for sleep quality than not smoking or sole use of only one inhalant. Future studies to assess objective sleep measures in a longitudinal manner, including but not limited to polysomnography and actigraphy, may give insight as to the specific differences among e-cigarette, cigarette, and dual use on sleep quality. In addition, evaluation of factors known to affect sleep quality, including body mass index, medications, depression, and anxiety, will be important to assess in future studies. Further efforts are needed to understand e-cigarette toxicity, identify underlying mechanisms, and define strategies to avoid nicotine dependence/addiction.