TITLE PAGE

Environmental Tobacco Smoke Exposure among Electronic Cigarette Users

Hai V. Nguyen, PhDa*

Aziz Sheikh, MDb,c

a Memorial University of Newfoundland, Canada

b Director, Asthma UK Centre for Applied Research, Usher Institute of Population Health

Sciences and Informatics, University of Edinburgh, UK

c Visiting Professor of Medicine, Harvard Medical School, Boston MA, USA

Corresponding author:

Hai V. Nguyen, Ph.D

School of Pharmacy

Memorial University of Newfoundland

Canada

Phone number: 1 (709) 777 6536

Email: hvnguyen@mun.ca

Word Count: 3,325

Number of Tables: 3

Number of Figures: 0

Abstract (233 words)

IntroductionExposure to environmental tobacco smoke (ETS) from combustible tobacco products causes various diseases and makes quitting smoking more difficult. However, little is known about exposure of e-cigarette users to ETS from combustible tobacco products. This study aimed to investigate e-cigarette users’ exposure to ETS from tobacco smokers.

MethodsThe association between ETS exposure frequency and different types of smokers including e-cigarette users was examined using ordered logistic regression analysis and nationally representative survey data on 28,765 individuals who were interviewed in the Canadian Tobacco, Alcohol and Drugs Surveys conducted during 2013 and 2015. Survey respondents were classified into one of five smoker types: smokers of tobacco only, dual users of tobacco and e-cigarettes, users of e-cigarette only, former smokers and never smokers. The analyses were conducted using the entire sample and by age group. ResultsYoung to mid-age (15-54) dual users of both regular cigarettes and e-cigarettes have higher ETS exposure than even tobacco smokers. Young to mid-age single users of e-cigarettes are less exposed to ETS than tobacco smokers, but still have higher ETS than never smokers. At older age (55+), both dual and single e-cigarette users face similar risks of ETS exposure as tobacco smokers. ConclusionsE-cigarette users are at high risk of ETS exposure. Policies that target the behaviour of e-cigarette users as well as the environments surrounding them to address their high ETS exposure risk would be beneficial.

Key words: e-cigarettes, environmental tobacco smoke exposure.

2

1. Introduction

Use of electronic cigarettes or e-cigarettes has increased rapidly, both among young

people (Arrazola et al., 2015; Reid, Rynard, Czoli, & Hammond, 2015) and adults (Caraballo,

Jamal, Nguyen, Kuiper, & Arrazola, 2016; King, Patel, Nguyen, & Dube, 2015). E-cigarettes

are advocated as a less harmful way than combustible tobacco to deliver nicotine without

creating harmful environmental tobacco smoke (ETS) for non-smokers and as a smoking-

cessation aid for tobacco smokers (Glynn, 2014; Goniewicz et al., 2014). Notably, two recent

reports from Public Health England and the Royal College of Physicians suggested that e-

cigarettes are far safer than traditional tobacco and called for smokers to switch to e-cigarettes

(McNeill et al., 2014; Public Health England, 2015; Royal College of Physicians, 2016).

However, there is not yet a scientific consensus on the long-term health effects of e-cigarettes

(Callahan-Lyon, 2014; Dinakar & O’Connor, 2016; Hajek, Etter, Benowitz, Eissenberg, &

McRobbie, 2014). A recent review raised concerns about the long-term health effects of e-

cigarettes and discouraged its use (Grana, Benowitz, & Glantz, 2014). Further, evidence

concerning e-cigarette’s role as a smoking-cessation aid is mixed (Dinakar & O’Connor, 2016;

Lindson‐Hawley et al., 2016; Malas et al., 2016).

Overlooked in the debate on the benefits of e-cigarettes is the extent of e-cigarette

users’ exposure to ETS generated by tobacco smokers. This is, however, an important issue

because ETS exposure can affect the extent of e-cigarettes’ health benefits and its effectiveness

as a smoking-cessation aid. The effects of ETS on non-smokers are nearly as large as that of

smoking on smokers (Centers for Disease Control and Prevention, 2017). Specifically, ETS

exposure causes various heart and lung diseases, low birth weight, sudden infant death

syndrome and respiratory distress syndrome (Royal College of Physicians, 2010), as well as

makes it hard for smokers to quit (Eng et al., 2014; Prugger et al., 2014). This is because 3

biologically, ETS exposure can increase nicotine acetylcholine receptors in the brain, and thus

nicotine dependence (Brody et al., 2011). It may also be difficult to quit smoking around other

smokers because of peer pressure (Powell, Tauras, & Ross, 2005), imitation of smoking

behaviour (Harakeh, Engels, Van Baaren, & Scholte, 2007) and smoking cues (Erblich &

Montgomery, 2012).

In this study, we investigated exposure of e-cigarette users to ETS from tobacco

smokers. Our analysis first quantified the extent of ETS exposure by e-cigarette users and

compared it with that of tobacco smokers as well as former smokers and never-smokers. We

distinguished between two types of e-cigarette users: dual users who used both e-cigarettes and

regular cigarettes versus single users who used only e-cigarettes. We then examined predictors

of ETS exposure focusing on the role of e-cigarette use. We hypothesized that e-cigarette

users’ exposure to ETS depends on whether they are single or dual users. Since smokers tend to

connect primarily to other smokers, and smokers and non-smokers cluster into discrete

subgroups that share similar characteristics (Christakis & Fowler, 2008; Poland et al., 2000),

we anticipated that single e-cigarette users likely interact mostly with their peer single e-

cigarette users and have low ETS exposure, while dual users likely interact mostly with their

fellow tobacco smokers and have substantial SHS exposure. We also hypothesized that e-

cigarette users’ exposure to ETS varied by age as younger smokers have different smoking

patterns and different levels of social interactions than older ones (Clark & Loheac, 2007;

Miething, Rostila, Edling, & Rydgren, 2016).

4

2. Methods

2.1 Data and variables

We pooled individual-level data from the two most recent cycles of Canadian Tobacco,

Alcohol and Drugs Survey (CTADS) 2013 and 2015 conducted by Statistics Canada. The

CTADS biennially interviewed a nationally representative sample of nearly 15,000 Canadians

aged 15 years and older on their use of tobacco, alcohol and illicit drugs. All survey

participants were asked the following question about their ETS exposure to smoke from

combustible cigarettes: ‘Overall, [excluding your own smoking] in the past month how often

were you exposed to second-hand smoke?’ This is the standard question used by Statistics

Canada to extract information about exposure to environmental tobacco smoke from

combustible cigarettes. Based on responses to this question, we constructed our study outcome

of ETS exposure frequency. It was an ordered discrete variable coded as 1 if the person

reported being never exposed, 2 if exposed at least once in the past month, 3 if exposed at least

once a week, 4 if exposed almost every day, and 5 if exposed every day.

The survey contained conventional classification of tobacco smoking status, i.e. current

smokers, former smokers and never smokers (Statistics Canada, 2012). In addition, all

respondents were asked about their use of e-cigarettes. Specifically, they were asked whether

they ever used e-cigarette in their life, and among those who reported ever using e-cigarettes,

whether they used it over past 30 days. Based on their responses, we created two indicators of

e-cigarette use: ever use and past 30-day use. The ‘ever use of e-cigarettes’ thus included both

past 30-day users and those who ever used but did not use past 30 days. Combining the

information on whether a person smoked regular cigarettes and/or e-cigarettes, we classified

survey respondents into five groups: smokers of tobacco only, dual users of tobacco and e-

cigarettes, single e-cigarette users (users of e-cigarettes only), former smokers (those who are

5

former smokers of tobacco and never used any e-cigarettes before) and never smokers (those

who never used tobacco or e-cigarettes). The classification below was based on the ‘ever use’

definition of e-cigarette use and was used to derive the main results for this paper. As ‘ever’

users may be experimenters and do not necessarily become regular users, we also used ‘past

30-day use of e-cigarettes’ to define e-cigarette use. The results using this e-cigarette use

definition were similar and reported in the Online Supplemental.

2.2 Descriptive analysis

We reported the means of demographic characterises for the entire sample and by

smoking status. We also estimated the prevalence of various smoking statuses by different age

groups. To test whether there was a statistically significant difference in composition of

smoking status across different age groups, we used the Chi-square test of independence

between two categorical variables: smoking status (5 categories: tobacco smokers, single e-

cigarette users, dual users, former smokers and never smokers) and age group (3 categories:

age<25, age 25-44, and age 45 and above). The null hypothesis was that the two variables are

independent.

2.3 Regression model

The association between ETS exposure frequency and different types of smokers was

studied using ordered logistic regression analysis. The covariates of primary interest were

indicators of the five groups (i.e. dual users, single e-cigarette users, former smokers and

never-smokers, with tobacco smokers as reference category). The regression models included

age (continuous variable), sex (male/female), household size (continuous variable for the

number of persons living in the respondent’s household), marital status (indicator for single or

6

married), urban status (indicator for whether the respondent resides in an urban or rural area),

and province fixed effects (dummy variables for each province to capture time-invariant

province-specific factors that affect the outcomes). We also controlled for presence of

smoking at home (defined as having at least one cigarette smoked at home a day) which is a

source of ETS, illicit drug use (indicator for whether the respondent used any opioids) which

correlates with smoking, employment status (indicator for whether the respondent worked at a

job or business in the past week) to account for the fact that most workplaces are smoke free

nowadays which affects ETS exposure, and drinking status (indicator for being a current

drinker, former drinker or abstainer) as drinking and smoking are often correlated affecting

ETS exposure (Dee, 1999)). The analyses were conducted for the entire sample and repeated

for age subgroups (less than 25, between 25 and 54, and 55 or above). All analyses were

conducted using Stata 15 (StataCorp, 2017).

3. Results

3.1 Descriptive analysis

Table 1 shows the prevalence of the five groups. In column 1, nearly one-fifth of the

sample reported using either regular cigarettes or electronic cigarettes or both, with a

breakdown of tobacco smokers (7.7%), dual users (6.0%), and single e-cigarette users (4.9%).

Never-smokers accounted for 57.2% and former smokers for 24.2%. The next three columns 2-

4 display the prevalence by age group. For the group aged less than 25 years, the prevalence of

only e-cigarette use and of dual use each was three times as high as that of tobacco use (13.2%

and 10.7% vs. 3.7%). This indicated that e-cigarettes have become very popular among the

youth group. For the group aged 25-54, the pattern was reversed: tobacco smokers accounted

for the highest percentage (8.8%), followed by dual users (6.7%) and single e-cigarette users

7

(4.7%). Among people aged 55 and above, the prevalence of dual use was higher than that of

single use (2.8% vs. 1.4%) but substantially lower than that of regular cigarette use (2.8% vs.

7.8%).

Table 2 displays ETS exposure and demographic characteristics of the whole sample

and the five groups. 13.1% of sample was exposed to ETS every day or almost every day, and

up to 1/3 of the sample exposed at least once a week. Among the 5 groups, dual users had the

highest ETS exposure with 45.3% reporting exposure almost every day or more, and more than

two-third of this group (70.8%) was exposed at least once a week. Tobacco smokers are the

second highest exposed group, with 34.2% exposed almost every day or more and 56.2%

exposed at least once a week. Former smokers and never-smokers were least exposed with

10.5% and 7.4% respectively reporting ETS exposure almost every day or more. Compared

with tobacco smokers, dual users were younger (37 vs. 47 years), more likely to be current

drinkers (85.7% vs. 78.9%) and to use illicit drugs (20.7% vs. 15.7%). Single e-cigarette users

were the youngest with an average age of 32. Nearly one-third of tobacco smokers and one-

third of dual users reported the presence of smoking in their home while 5% of single e-

cigarette users and 2% of former smokers and never-smokers reported the same.

3.2 Regression estimates

Table 3 reports the regression estimates for the relationship between ETS exposure and

different types of smokers. To explore possible confounding effects of age, drinking and illicit

drug use suggested by Table 2, we first estimated a baseline regression that excluded these

three covariates. The results reported in column 1 indicate that dual users had a higher risk of

ETS exposure than tobacco smokers (OR = 1.78; 95% CI = 1.35 – 2.36; p<0.01). Meanwhile,

8

Table 1. Smoking status prevalence, all sample and by age group, CTADS 2013 & 2015 (pooled)

All sample N Age<25 N Age 25-54 N Age 55+ N p-value

(1) (2) (3) (4)

Tobacco smoker (%)

7.7 2,142 3.7 439 8.8 888 7.8 815 <0.001

(7.4 - 8.0) (3.3 - 4.0) (8.2 - 9.4) (7.3 - 8.4)

Dual user (%) 6.0 2,279 10.7 1,217 6.7 743 2.8 319(5.7 - 6.3) (10.1 - 11.2) (6.2 - 7.3) (2.4 - 3.1)

Single e-cigarette user (%)

4.9 2,203 13.2 1,519 4.7 473 1.4 139

(4.7 - 5.2) (12.6 - 13.9) (4.2 - 5.1) (1.2 - 1.7)

Former smoker (%)

24.2 5,958 1.5 168 20.1 1,878 40.8 3,912

(23.7 - 24.7) (1.3 - 1.7) (19.2 - 20.9) (39.8 - 41.8)

Never smoker (%) 57.2 16,183 70.9 7,483 59.7 4,668 47.2 4,032(56.6 - 57.8) (70.1 - 71.8) (58.7 - 60.8) (46.1 - 48.2)

Total N 28,765 10,898 8,650 9,217Note: Weighted percentages for indicator variables are presented; 95% confidence intervals are in parentheses. Smoking classification is based on e-cigarette use defined as "ever use of e-cigarettes"

Table 2. Descriptive statistics (ETS exposure level and demographic characteristics), all sample and by smoking status; CTADS 2013 & 2015

Variable All sample (N=28,609)

Tobacco smoker

(N=2,110)

Dual user (N=2,256)

Single e-cigarette

user (N=2,200)

Former smoker

(N=5,923)

Never smoker

(N=16,120)

ETS Exposure frequency Everyday

(%)6.9 22.0 30.7 8.7 5.6 2.8

(6.6 - 7.2) (20.3 - 23.9) (28.8 - 32.6) (7.5 - 9.8) (5.0 - 6.2) (2.6 - 3.1)

Almost 6.2 12.2 14.6 10.8 4.9 4.6everyday

(%)(5.9 - 6.5) (10.8 - 13.6) (13.2 - 16.1) (9.5 - 12.1) (4.4 - 5.5) (4.3 - 5.0)

>=Once a 18.4 22.0 25.5 28.3 18.1 16.6week (%) (18.0 - 18.9) (20.2 - 23.7) (23.6 - 27.2) (26.4 - 30.2) (17.1 - 19.1) (16.0 - 17.1)

>=Once a 26.9 15.8 12.6 29.0 23.5 31.2month (%) (26.4 - 27.4) (14.2 - 17.3) (11.2 - 14.0) (27.1 - 30.8) (22.4 - 24.6) (30.4 - 31.9)

Never (%) 41.6 28.0 16.6 23.2 47.9 44.8(41.0 - 42.1) (26.1 - 29.9) (15.1 - 18.2) (21.5 - 25.1) (46.7 - 49.2) (44.0 - 45.6)

Age (years) 45.9 47.2 36.9 31.9 56.6 43.3(45.7 – 46.1) (46.6 - 47.9) (36.3 – 37.5) (31.3 - 32.5) (56.2 – 56.9) (43.0 – 43.6)

Household size

2.9 2.7 2.8 3.1 2.5 3.0

(persons) (2.9 – 2.9) (2.6 – 2.7) (2.8 – 2.9) (3.0 – 3.1) (2.5 – 2.5) (3.0 – 3.1)

Single (%) 23.5 24.7 37.4 42.1 11.2 25.4(23.0 - 23.9) (22.8 - 26.5) (35.4 - 39.4) (40.0 - 44.2) (10.4 - 12.0) (24.8 - 26.1)

Male (%) 49.1 57.6 55.2 57.8 54.9 44.2(48.6 - 49.7) (55.5 - 59.7) (53.2 - 57.3) (55.7 - 59.8) (53.6 - 56.2) (43.4 - 45.0)

Urban status (%)

80.3 77.5 80.1 82.0 77.4 81.8

(79.9 - 80.8) (75.8 - 79.3) (78.4 - 81.7) (80.4 - 83.6) (76.3 - 78.5) (81.2 - 82.4)

Employ (%)

51.8 53.5 60.1 60.8 44.1 53.2

(51.2 - 52.4) (51.4 - 55.6) (58.1 - 62.1) (58.7 - 62.8) (42.8 - 45.4) (52.4 - 54.0)

Home smoking

5.7 29.4 28.9 5.1 2.2 1.6

(%) (5.5 - 6.0) (27.5 - 31.4) (27.1 - 30.8) (4.2 - 6.1) (1.9 - 2.6) (1.4 - 1.8)

Drinking status

Current drinker

76.7 78.9 85.7 89.3 81.7 72.2

(%) (76.2 - 77.1) (77.1 - 80.6) (84.3 - 87.2) (88.0 - 90.6) (80.7 - 82.7) (71.5 - 72.9)

Former drinker

14.1 18.6 13.0 7.9 16.2 13.2

(%) (13.7 - 14.5) (17.0 - 20.3) (11.7 - 14.4) (6.7 - 9.0) (15.3 - 17.2) (12.7 - 13.7)

Non-drinker (%)

9.3 2.5 1.2 2.8 2.1 14.6

(8.9 - 9.6) (1.8 - 3.2) (0.8 - 1.7) (2.1 - 3.5) (1.7 - 2.4) (14.0 - 15.1)

Illegal drug use

13.6 15.7 20.7 18.0 15.7 11.3

(%) (13.2 - 14.0) (14.2 - 17.3) (19.0 - 22.3) (16.4 - 19.6) (14.8 - 16.6) (10.8 - 11.7)Note: Weighted percentages for indicator variables and weighted means for continuous variables are presented; 95% confidence intervals are in parentheses;; Smoking classification is based on e-cigarette use defined as "ever use of e-cigarettes"; “Single” refers to those who were single or never married; “Household size” refers to number of persons in the respondent’s household; “Urban status” refers to whether the respondent resides in an urban area.

single e-cigarette users were exposed to ETS less than tobacco smokers (OR = 0.72; 95% CI =

0.64 – 0.80; p<0.01) but still more than former smokers and never-smokers. When the

regression was augmented to include age, drinking and illicit drug use (column 2), the

magnitude of the dual user and single e-cigarette user coefficients decreased. This suggests that

these three factors confounded the relationship between ETS exposure and e-cigarette use.

Notably, after controlling for these factors, dual users still had a higher risk of ETS exposure

(OR = 1.39; 95% CI = 1.10 – 1.75; p=0.01), suggesting that some other factors were also

driving the high ETS exposure of dual users. Single e-cigarette users still had lower ETS

exposure risk than tobacco smokers (OR = 0.50; 95% CI = 0.44 – 0.57; p<0.01).

Table 3. Predictors of ETS exposure frequency, all sample and by age group; Ordered logistic models; CTADS 2013 & 2015; Smoking classification is based on e-cigarette use defined as "ever use of e-cigarettes"

Baseline(N=28,609)

Augmented (N=28,516)

Age<25 (N=10,853)

Age 25-44 (N=8,575)

Age 45+ (N=9,088)

(1) (2) (3) (4) (5)

Type of smokerDual user 1.782** 1.385** 1.282** 1.421* 1.269

(1.345 - 2.361) (1.099 - 1.745) (1.083 - 1.519) (1.056 - 1.911) (0.943 - 1.706)Single e-cigarette

user 0.716** 0.504** 0.420** 0.455** 0.841

(0.642 - 0.799) (0.444 - 0.572) (0.375 - 0.471) (0.329 - 0.629) (0.507 - 1.394)Never smoker 0.331** 0.323** 0.247** 0.292** 0.433**

(0.307 - 0.356) (0.305 - 0.341) (0.206 - 0.295) (0.260 - 0.328) (0.376 - 0.498)Former smoker 0.416** 0.497** 0.331** 0.433** 0.684**

(0.367 - 0.471) (0.439 - 0.562) (0.167 - 0.656) (0.368 - 0.509) (0.569 - 0.821)Home smoking 3.163** 3.412** 5.138** 2.957** 3.472**

(2.767 - 3.616) (2.975 - 3.912) (4.616 - 5.719) (2.540 - 3.442) (2.374 - 5.078)Employ 1.776** 1.340** 1.217** 1.406** 1.254*

(1.662 - 1.897) (1.255 - 1.431) (1.071 - 1.382) (1.166 - 1.696) (1.046 - 1.504)Male 1.330** 1.339** 1.083* 1.420** 1.338**

11

(1.222 - 1.448) (1.228 - 1.459) (1.012 - 1.158) (1.241 - 1.626) (1.236 - 1.449)Single 1.904** 1.203** 1.044 1.157* 1.308**

(1.777 - 2.040) (1.139 - 1.271) (0.856 - 1.272) (1.014 - 1.319) (1.068 - 1.602)Household size 1.115** 0.938** 0.963 0.905** 0.986

(1.094 - 1.137) (0.900 - 0.978) (0.914 - 1.014) (0.844 - 0.970) (0.862 - 1.128)Urban status 0.947 0.973 1.173* 0.874* 1.078**

(0.878 - 1.021) (0.918 - 1.031) (1.025 - 1.342) (0.772 - 0.990) (1.030 - 1.128)Age 0.967** 0.989 0.974** 0.966**

(0.962 - 0.973) (0.963 - 1.017) (0.969 - 0.979) (0.959 - 0.972)Drinking status

Current drinker 1.714** 1.737** 2.194** 1.299*(1.503 - 1.954) (1.438 - 2.099) (1.681 - 2.863) (1.063 - 1.588)

Former drinker 1.570** 1.595** 2.098** 1.152**(1.359 - 1.814) (1.172 - 2.171) (1.439 - 3.060) (1.044 - 1.270)

Illegal drug use 1.364** 1.222** 1.421** 1.354*(1.108 - 1.679) (1.144 - 1.305) (1.125 - 1.795) (1.041 - 1.761)

Note: Odds ratios from ordered logistic models are reported; 95% confidence intervals are in parentheses; Five-point scale for ordered ETS exposure frequency outcome: 1 (never exposed), 2 (exposed at least once a month), 3 (exposed at least once a week), 4 (exposed almost every day), and 5(exposed every day). All models include province fixed effects. Reference categories are female, non-drinker, tobacco smoker, married, non-user of illicit drugs and unemployed. Robust standard errors throughout are clustered at the province level and estimates are weighted. Smoking classification is based on e-cigarette use defined as "ever use of e-cigarettes"; Significance levels are **p<0.01; *p<0.05.

We next explored the relationship between ETS exposure and e-cigarette use by age

group. The results are reported in columns 3-5 of Table 3. Compared with the reference group

of tobacco smokers, young to mid-age dual users face higher risks of ETS than tobacco

smokers (OR = 1.28; 95% CI = 1.08 – 1.52; p<0.01 for the age group less than 25, and OR =

1.42; 95% CI = 1.06 – 1.91; p=0.02 for the age group 25-54). For dual users aged 55 and

above, the odds ratio is greater than 1 but not statistically significant (OR = 1.27; 95% CI =

0.94 – 1.71; p=0.11). Meanwhile, the pattern of ETS exposure risk for young to mid-age single

e-cigarette users was reversed: compared with tobacco smokers, ETS exposure risk was lower

for the youth aged less than 25 (OR = 0.42; 95% CI = 0.38 – 0.47; p<0.01) and for adults aged

25-54 (OR = 0.46; 95% CI = 0.33 – 0.63; p<0.01). As with older age dual users, older age

single e-cigarette users faced similar risks of ETS exposure as tobacco smokers (OR = 0.84;

CI=0.51– 1.39; p=0.50).

12

Regression estimates for other covariates, also reported in Table 3, indicated that

presence of smoking at home raised the risk of ETS exposure significantly. Males were at

higher risk of ETS exposure than females while respondents who were single had a higher risk

of ETS exposure than married respondents. Risk of ETS exposure also increased with the size

of household. Those who used illicit drugs were more exposed to ETS than those who did not.

Similarly, both current and former drinkers were more exposed to ETS than non-drinkers.

Despite the fact that most workplaces have adopted smoking restrictions during the study

period of 2013-2015 (Statistics Canada), those who were employed had a higher ETS exposure

risk than the unemployed. It is possible that employed people have higher level of social

interactions that exposed them to more ETS than those who are not. This also suggests that

most of ETS exposure for employed people must have occurred outside their workplaces.

4. Discussion

Our study provides the first comparison of self-reported ETS exposure among different

types of tobacco and e-cigarette users. We found that ETS exposure risk was higher for dual

users than for those only smoking tobacco. This finding was driven by a high ETS exposure of

young to mid-age dual users. In contrast, young to mid-age single e-cigarette users faced a

lower risk of ETS exposure than tobacco smokers.

Why did young to mid-age dual users (age 15-54) face a higher risk of ETS exposure

than older dual users (age 55 and above)? It is possible that younger dual users had a more

socially active lifestyle and interacted more often with tobacco smokers than older ones. This

hypothesis is consistent with the finding that drinking alcohol was associated with higher ETS

exposure. Meanwhile, it is likely that single e-cigarette users interacted mostly with other

single e-cigarette users in their social network (Christakis & Fowler, 2008) and as a result,

13

were not exposed to ETS as frequently as tobacco smokers. The lower ETS exposure for single

e-cigarette users (compared with what their ETS exposure would be if they take up tobacco

smoking) can be seen as another benefit of using e-cigarettes. This, however, does not

necessarily mean that early use of e-cigarettes is not a gateway to initiation of regular cigarette

smoking (Leventhal et al., 2015; Rigotti, 2015).

Our findings have a number of policy implications. First, to maximize potential benefits

associated with e-cigarette use (relative to regular cigarette smoking), policymakers should

target young to mid-age dual users who had higher level of ETS exposure than tobacco

smokers. Understanding the smoking behaviour of e-cigarette users (e.g. where and when they

smoke e-cigarettes) and the environments around them (e.g. whom they often smoke with)

would provide useful information to help design interventions for reducing their ETS exposure.

Second, given that ETS exposure can act as a barrier to smoking cessation, high ETS exposure

of dual users might explain the mixed findings in the literature on the role of e-cigarettes in

helping its users to quit smoking. Tobacco control efforts should promote awareness of the

risk of ETS exposure among e-cigarette users. Further, as e-cigarette smokers are increasingly

prohibited from smoking in public places and thus have to smoke e-cigarettes in the same areas

where combustible cigarettes are allowed, separate designated smoking areas for e-cigarette

users might be needed to prevent them being exposed to ETS from combustible cigarette

smokers. Third, our finding of high ETS exposure of e-cigarette users draws attention to a

potential impact of the current policy trend of banning e-cigarette use in public places. If

banning e-cigarettes in public places induces e-cigarette users to switch their e-cigarette use to

the same places as tobacco smokers, this may expose e-cigarette users to harmful ETS from

tobacco smokers. There has been evidence of displacement effect of public smoking bans. For

example, it was reported that public-place smoking laws significantly increased non-smokers'

14

exposure to ETS at building entrances (Carpenter, Postolek, & Warman, 2011). While this

potential unintended consequence is unlikely to outweigh the obvious and significant benefits

of bans including protecting non-smokers from e-cigarette aerosol and de-normalizing smoking

behaviour, it is nevertheless should not be ignored. Fourth, our results indicate that the risk of

ETS exposure substantially increased with the presence of smoking at home. This highlights

the need to implement measures to curb home smoking.

It is worth noting that we used the term ‘Environmental Tobacco Smoke (ETS)’ instead

of ‘Second-Hand Smoke (SHS)’ to refer to passive smoking in this study (Protano & Vitali,

2011). ETS includes not only SHS (the combination of smoke emitted from the burning end of

a cigarette or other tobacco products and smoke exhaled by the smoke) but also the thirdhand

smoke (the residue from tobacco smoke that persists on the clothing and hair of smokers, on

environmental surfaces and in dust long after a cigarette has been extinguished). As such, the

term ETS captures passive smoking more accurately than the term SHS.

Our study has a number of limitations. First, our finding of the association between

ETS exposure and e-cigarette use was based on cross-sectional data and does not mean that

smoking e-cigarettes causes higher ETS exposure. The high ETS exposure experienced by

young to mid-age dual users is likely to have been driven by self-selection issues. That is,

people who self-selected into using e-cigarettes have certain characteristics that make them

more likely to be exposed to ETS rather than the use of e-cigarettes itself. We have shown that

age, drinking and illicit drug use status of dual users may contribute to their high ETS

exposure. But the fact that dual users still have higher ETS exposure even after controlling for

these factors suggests that there may be other factors that drive ETS exposure, such as e-

cigarette users’ social networks. Unfortunately, there is no information in the survey on e-

cigarette users’ social network to explore this issue further. This is a topic for future research.

15

Second, we used an alternative definition of e-cigarette use based on “past 30-day use” to

account for experimental use of e-cigarettes. While this measure excludes users who only

experimented with e-cigarettes in the distant past, these may still include respondents who

experimented with e-cigarettes in the past 30 days. Another limitation of our study is that there

was no objective data on respondents’ actual exposure to ETS. Next, the survey we used for

this study contains no information on second hand exposure to e-cigarette vapor. Future

surveys and research should explicitly address the issue of exposure to e-cigarette vapor. This

will be important as second-hand exposure to vapor from e-cigarettes is not harmless,

particularly among infants and children (Protano, Manigrasso, Avino, Sernia, & Vitali, 2016;

Protano, Manigrasso, Avino, & Vitali, 2017). Cotinine levels in non-smokers exposed to e-

cigarette vapor have been found to be similar to those exposed to ETS from combustible

cigarettes (Ballbè et al., 2014; Flouris et al., 2013). Finally, there is no information on specific

places of ETS (e.g., workplace, bars, building entrance, bus stop, park, etc.) and our findings

could not shed light on specific locations where interventions should be targeted.

5. Conclusions

Studying ETS exposure of e-cigarette users provides a new perspective with which to

assess and inform potential benefits and risks of using e-cigarettes. Our analysis demonstrated

that health benefits associated with using e-cigarettes should not be taken for granted when e-

cigarette users are at high risk of ETS exposure. Policies that target the behaviour of e-cigarette

users as well as the environments surrounding them to address their high ETS exposure risk

would be beneficial.

16

References

Arrazola, R. A., Singh, T., Corey, C. G., Husten, C. G., Neff, L. J., Apelberg, B. J., … Centers

for Disease Control and Prevention (CDC). (2015). Tobacco use among middle and

high school students - United States, 2011-2014. MMWR. Morbidity and Mortality

Weekly Report, 64(14), 381–385.

Ballbè, M., Martínez-Sánchez, J. M., Sureda, X., Fu, M., Pérez-Ortuño, R., Pascual, J. A., …

Fernández, E. (2014). Cigarettes vs. e-cigarettes: Passive exposure at home measured

by means of airborne marker and biomarkers. Environmental Research, 135, 76–80.

https://doi.org/10.1016/j.envres.2014.09.005

Brody, A. L., Mandelkern, M. A., London, E. D., Khan, A., Kozman, D., Costello, M. R., …

Mukhin, A. G. (2011). Effect of secondhand smoke on occupancy of nicotinic

acetylcholine receptors in brain. Archives of General Psychiatry, 68, 953–960.

Callahan-Lyon, P. (2014). Electronic cigarettes: human health effects. Tobacco Control, 23

Suppl 2, ii36-40. https://doi.org/10.1136/tobaccocontrol-2013-051470

Caraballo, R. S., Jamal, A., Nguyen, K. H., Kuiper, N. M., & Arrazola, R. A. (2016).

Electronic Nicotine Delivery System Use Among U.S. Adults, 2014. American Journal

of Preventive Medicine, 50(2), 226–229. https://doi.org/10.1016/j.amepre.2015.09.013

Carpenter, C., Postolek, S., & Warman, C. R. (2011). Public-place smoking laws and exposure

to environmental tobacco smoke (ETS) in public places. Am Econ J Econ Policy, 3(3),

35–61.

Centers for Disease Control and Prevention. (2017). Fact Sheet: Health Effects of Secondhand

Smoke. 2017.

Christakis, N. A., & Fowler, J. H. (2008). The collective dynamics of smoking in a large social

network. New England Journal of Medicine, 358, 2249–2258.

17

Clark, A. E., & Loheac, Y. (2007). “It wasn’t me, it was them!” social influence in risky

behavior by adolescents. Journal of Health Economics, 26, 763–784.

https://doi.org/10.1016/j.jhealeco.2006.11.005

Dee, T. S. (1999). The complementarity of teen smoking and drinking. Journal of Health

Economics, 18, 769–793.

Dinakar, C., & O’Connor, G. T. (2016). The health effects of electronic cigarettes. New

England Journal of Medicine, 375, 1372–1381.

Eng, L., Su, J., Qiu, X., Palepu, P. R., Hon, H., Fadhel, E., … Kashigar, A. (2014). Second-

hand smoke as a predictor of smoking cessation among lung cancer survivors. Journal

of Clinical Oncology, 32, 564–570.

Erblich, J., & Montgomery, G. H. (2012). Cue-induced cigarette cravings and smoking

cessation: the role of expectancies. Nicotine Tob Res, 14, 809–815.

https://doi.org/10.1093/ntr/ntr288

Flouris, A. D., Chorti, M. S., Poulianiti, K. P., Jamurtas, A. Z., Kostikas, K., Tzatzarakis, M.

N., … Koutedakis, Y. (2013). Acute impact of active and passive electronic cigarette

smoking on serum cotinine and lung function. Inhalation Toxicology, 25(2), 91–101.

https://doi.org/10.3109/08958378.2012.758197

Glynn, T. J. (2014). E-cigarettes and the future of tobacco control. CA: A Cancer Journal for

Clinicians, 64, 164–168. https://doi.org/10.3322/caac.21226

Goniewicz, M. L., Knysak, J., Gawron, M., Kosmider, L., Sobczak, A., Kurek, J., … Benowitz,

N. (2014). Levels of selected carcinogens and toxicants in vapour from electronic

cigarettes. Tobacco Control, 23, 133–139. https://doi.org/10.1136/tobaccocontrol-2012-

050859

18

Grana, R., Benowitz, N., & Glantz, S. A. (2014). E-cigarettes: a scientific review. Circulation,

129, 1972–1986. https://doi.org/10.1161/circulationaha.114.007667

Hajek, P., Etter, J. F., Benowitz, N., Eissenberg, T., & McRobbie, H. (2014). Electronic

cigarettes: review of use, content, safety, effects on smokers and potential for harm and

benefit. Addiction, 109, 1801–1810. https://doi.org/10.1111/add.12659

Harakeh, Z., Engels, R. C., Van Baaren, R. B., & Scholte, R. H. (2007). Imitation of cigarette

smoking: an experimental study on smoking in a naturalistic setting. Drug and Alcohol

Dependence, 86, 199–206. https://doi.org/10.1016/j.drugalcdep.2006.06.006

King, B. A., Patel, R., Nguyen, K. H., & Dube, S. R. (2015). Trends in awareness and use of

electronic cigarettes among US adults, 2010-2013. Nicotine & Tobacco Research:

Official Journal of the Society for Research on Nicotine and Tobacco, 17(2), 219–227.

https://doi.org/10.1093/ntr/ntu191

Leventhal, A. M., Strong, D. R., Kirkpatrick, M. G., Unger, J. B., Sussman, S., Riggs, N. R., …

Audrain-McGovern, J. (2015). Association of electronic cigarette use with initiation of

combustible tobacco product smoking in early adolescence. JAMA, 314, 700–707.

https://doi.org/10.1001/jama.2015.8950

McNeill, A., Etter, J. F., Farsalinos, K., Hajek, P., le Houezec, J., & McRobbie, H. (2014). A

critique of a World Health Organization-commissioned report and associated paper on

electronic cigarettes. Addiction, 109, 2128–2134. https://doi.org/10.1111/add.12730

Miething, A., Rostila, M., Edling, C., & Rydgren, J. (2016). The Influence of Social Network

Characteristics on Peer Clustering in Smoking: A Two-Wave Panel Study of 19- and

23-Year-Old Swedes. PloS One, 11, e0164611.

https://doi.org/10.1371/journal.pone.0164611

19

Poland, B., Cohen, J., Ashley, M., Adlaf, E., Ferrence, R., Pederson, L., … Raphael, D. (2000).

Heterogeneity among smokers and non-smokers in attitudes and behaviour regarding

smoking and smoking restrictions. Tobacco Control, 9, 364–371.

https://doi.org/10.1136/tc.9.4.364

Powell, L. M., Tauras, J. A., & Ross, H. (2005). The importance of peer effects, cigarette

prices and tobacco control policies for youth smoking behavior. Journal of Health

Economics, 24, 950–968. https://doi.org/10.1016/j.jhealeco.2005.02.002

Protano, C., Manigrasso, M., Avino, P., Sernia, S., & Vitali, M. (2016). Second-hand smoke

exposure generated by new electronic devices (IQOS® and e-cigs) and traditional

cigarettes: submicron particle behaviour in human respiratory system. Annali Di Igiene:

Medicina Preventiva E Di Comunita, 28(2), 109–112.

Protano, C., Manigrasso, M., Avino, P., & Vitali, M. (2017). Second-hand smoke generated by

combustion and electronic smoking devices used in real scenarios: Ultrafine particle

pollution and age-related dose assessment. Environment International, 107, 190–195.

https://doi.org/10.1016/j.envint.2017.07.014

Protano, Carmela, & Vitali, M. (2011). The New Danger of Thirdhand Smoke: Why Passive

Smoking Does Not Stop at Secondhand Smoke. Environmental Health Perspectives,

119(10), a422. https://doi.org/10.1289/ehp.1103956

Prugger, C., Wellmann, J., Heidrich, J., De Bacquer, D., Perier, M. C., Empana, J. P., … Keil,

U. (2014). Passive smoking and smoking cessation among patients with coronary heart

disease across Europe: results from the EUROASPIRE III survey. European Heart

Journal, 35, 590–598. https://doi.org/10.1093/eurheartj/eht538

Public Health England. (2015). Publications - E-cigarettes: a new foundation for evidence-

based policy and practice. Retrieved June 19, 2018, from

20

https://www.heartland.org/publications-resources/publications/e-cigarettes-a-new-

foundation-for-evidence-based-policy-and-practice

Reid, J. L., Rynard, V. L., Czoli, C. D., & Hammond, D. (2015). Who is using e-cigarettes in

Canada? Nationally representative data on the prevalence of e-cigarette use among

Canadians. Preventive Medicine, 81, 180–183.

https://doi.org/10.1016/j.ypmed.2015.08.019

Rigotti, N. A. (2015). e-Cigarette Use and Subsequent Tobacco Use by Adolescents: New

Evidence About a Potential Risk of e-Cigarettes. In Jama (Vol. 314, pp. 673–674).

United States.

Royal College of Physicians. (2010). Passive smoking and children. Retrieved June 19, 2018,

from https://shop.rcplondon.ac.uk/products/passive-smoking-and-children

Royal College of Physicians. (2016, April 28). Nicotine without smoke: Tobacco harm

reduction. Retrieved June 19, 2018, from

https://www.rcplondon.ac.uk/projects/outputs/nicotine-without-smoke-tobacco-harm-

reduction-0

StataCorp. (2017). Stata Statistical Software: Release 15. College Station, TX: StataCorp LLC.

Statistics Canada. (2012). Microdata User Guide, Canadian Tobacco Use Monitoring Survey

Cycle 1. Retrieved from http://www23.statcan.gc.ca/imdb/p2SV.pl?

Function=getSurvey&Id=64426#a4

Statistics Canada. (n.d.). Canadian Tobacco Use Monitoring Survey (CTUMS) 2012.

21