A Pilot Study of E-Cigarette Naïve Cigarette Smokers and the Effects on Craving After Acute Exposure to E-Cigarettes in the Laboratory

Abstract

Background and Objective.

Recent surveys confirm continued increases in the use of electronic-cigarettes (e-cigarettes) in adolescents and adults. Users often state that e-cigarettes reduce tobacco craving and withdrawal symptoms in addition to their smoking. Data from laboratory studies and clinical trials have confirmed these statements, though there are inconsistencies in the outcomes. In this pilot study, we set out to evaluate the effects of e-cigarettes, as compared to the participants’ own cigarettes, on baseline craving and smoking severity.

Methods.

Using a within-subjects, placebo-controlled study design, 15 tobacco-dependent, e-cigarette naïve participants sustained abstinence overnight. They completed distinct phases of this protocol during 4 separate study sessions. Participants were randomized to an e-cigarette device containing one of 3 doses of nicotine (0, 18, or 36 mg/ml) or their own cigarette. Each study visit was ~3 hours long and separated by at least 7 days. Visits included assessments of craving and smoking severity.

Results.

The data showed that after 10 puffs in both the Own cigarette and e-cigarette conditions, breath CO levels increased significantly in the former but not the latter. QSU and Choices to Smoke scores were not statistically different across groups after two distinct bouts of 10 puffs each. Additionally, E-cigarette Perceptions Questionnaire responses were not significantly different according to dose.

Conclusion and Scientific Significance.

This experiment provides data demonstrating that e-cigarettes did not reduce craving or smoking severity in e-cigarette naïve users. However, since this was a pilot study, the conclusions that can be drawn are limited.

Introduction

Self-reported combustible cigarette use has been on the decline among U.S. adults and adolescents, while the use and awareness of e-cigarettes has increased significantly. 15.4% of adults aged ≥ 18 years have tried an e-cigarette and 3.2% frequently used e-cigarettes as of 2016.¹ Additionally, nationwide sales of e-cigarettes jumped 14.4% in 2014-2015 alone and continue to rise.²

The popularity of e-cigarettes is likely related to their ability to closely mimic both the physical and behavioral components of combustible cigarette smoking. E-cigarettes are distinct from other nicotine replacement therapies in that they provide a similar ritual to combustible cigarette smoking behavior consisting of the hand-to-mouth movement, visible smoke exhaled, and sensory stimulation.³ This may explain why these products may be especially effective in reducing the use of combustible cigarettes. However, the extent to which e-cigarette devices resemble and feel like combustible cigarettes may greatly vary.

Outcomes from laboratory and outpatient studies have shown that e-cigarette users are able to reliably obtain nicotine levels in venous blood comparable to combustible cigarette users.⁴ Yet, despite such advances in knowledge of the effects of e-cigarettes among users, many questions remain unanswered — especially in regard to naïve users. Eissenberg specifically investigated the effect of acute e-cigarette administration on craving suppression and subjective effects in e-cigarette naïve smokers. His results demonstrated that the tested e-cigarette brands failed both to suppress craving as effectively as combustible cigarettes and to substantially increase plasma nicotine levels.⁵ Bullen et al. evaluated smoking desire and withdrawal symptom suppression in e-cigarette naïve smokers randomized to an inhalator, 0 or 16 mg e-cigarette dose, or participants’ own brand of cigarette. Following overnight abstinence, there were significant reductions in desire to smoke after an hour of smoking the 16 mg e-cigarette, however these reductions were not as great as those seen following the combustible cigarettes.⁶ Additionally, a recent study of e-cigarette naïve users by Hiler et al. found that e-cigarette exposure significantly reduced tobacco craving and withdrawal. Nonetheless, these effects were significantly smaller in magnitude when compared to experienced users.⁷ Despite these outcomes, the Hiler study did not compare effects to the participants’ own combustible cigarettes as a control, and the cohort was only ~30 years of age. The study findings, therefore, cannot be extrapolated to older smokers whose responses to e-cigarettes may differ due to their extensive smoking history.

We posit that it is important to consider how e-cigarettes impact craving while comparing outcomes to an Own cigarette as well as investigating these outcomes in smokers ~50 years of age. Therefore, this pilot study was set forth to compare the effects of an Own cigarette versus that of an e-cigarette on craving and Choices to Smoke in e-cigarette naïve participants.

Methods

Participants.

Participants consisted of non-treatment seeking combustible cigarette smokers (N = 15) who reported having never tried an e-cigarette. They were recruited to this study from the Houston area via newspaper, internet advertisements, flyers, and referral. All potential participants were pre-screened by telephone to ensure that they met the basic requirements of the study. Those deemed eligible were invited for an in-person interview. During the interview the participants received an explanation of the study purpose and requirements, in addition to the chance to review, inquire about, and sign the informed consent. The consenting process involved an explicit description as to smoking e-cigarettes at which point participants were asked about potential concerns and willingness to participate. Afterwards, staff assured that participants met DSM criteria for nicotine dependence as determined by the Mini-International Neuropsychiatric Interview. This pilot study was approved by the Baylor College of Medicine and Michael E. DeBakey VA Medical Center institutional review boards and participants were compensated for their time.

The inclusion criteria comprised of not seeking abstinence-focused therapy treatment concurrently with the study, reporting smoking of ≥ 10 combustible cigarettes per day for the last year, providing a breath carbon monoxide (CO) ≥ 10 ppm (participants agreed to refrain from smoking for 24 hours prior), having a FTND ≥ 5, and self-reporting no previous use (naïve) of any form of e-cigarette. Exclusion criteria included dependence on other non-tobacco substances, self-reporting that preferred route of nicotine use was any other than smoking, diagnosis of any Axis I disorder, pregnancy or nursing, or having any other illness, condition, or use medications which would preclude safe and/or successful completion of the study.

Study Cigarettes, E-Cigarettes, and E-Liquid.

The combustible cigarettes used in this pilot study were the participants’ own brand. The e-cigarettes used were eGo devices with a 3.3-Volt e-cigarette battery attached to a 1.5-ohm dual-coil cartomizer (Smoktech, Shenzhen, China). This device was chosen since eGo systems are consistently among the most popular and appealing e-cigarette devices.^{8, 9} Additionally, recent research by Lopez et al. showed that the specific eGo device used was capable of delivering blood plasma nicotine at levels approximating those produced by a regular combustible cigarette.⁴ The study’s e-liquid was Virginia Pure tobacco flavored, containing 0, 18, or 36 mg/ml nicotine loaded with 1 ml of a 70% propylene glycol/30% vegetable glycerin (Avail Liquids, Richmond, VA). Nicotine levels were independently assessed and confirmed. The 18 and 36 mg doses were chosen based on the Lopez study, which demonstrated that these were the only doses resulting in a reliable increase in nicotine plasma concentrations.⁴

Procedure.

Participants completed four distinct sessions separated by at least 7 days. Prior to each session, participants engaged in overnight abstinence from their combustible cigarette, as confirmed by presenting a breath CO level ≤ 4 ppm prior to each session.¹⁰ Participants who did not meet this criterion had their session rescheduled. Across the four sessions, the 0, 18, or 36 mg/ml e-cigarette doses, or the participants’ own combustible cigarettes were tested. The order of sessions was randomized for each subject. E-cigarettes were presented in a double-blind manner.

Sessions were conducted in a specially designated negative pressure room at the MEDVAMC. Participants were trained how to use the e-cigarette device using a standardized smoking procedure mimicking ad-lib smoking.^{11, 12} In each session, participants completed two 10-puff e-cigarette bouts with a 30 second interval in between each individual puff. Puff bouts were separated by ~60 minutes.

Assessments

Fagerström Test of Nicotine Dependence (FTND).

The FTND was administered to measure the level of nicotine dependence. The FTND consists of 6 questions with a mixture of multiple-choice questions (scored 0–3) and yes/no questions (scored 0 or 1). The total score range was from 0–10 with 10 being indicating the highest level of dependence.

Questionnaire of Smoking Urges (QSU).

The QSU was administered to assess craving and smoking urges. The resulting score was used to assess feelings and thoughts towards the desire to smoke, in addition to any relief of negative symptoms. The assessment was comprised of 10 questions and used a Likert scale from 1 to 7. Craving was evaluated after each distinct bout of 10 puffs each. QSU factor scores were also calculated, but outcomes are not shown since differences between groups did not reach statistical significance.

Dual-Component Combustible Cigarette Self-Administration Session (“Choices to Smoke”).

This task modeled latency to relapse and the reinforcing strength of combustible cigarettes.^13–15 Participants were presented with a lighter, an ashtray, and their preferred brand of combustible cigarettes. Participants were informed that they were free to smoke at any time but would earn $1 for every 5 minutes that they did not smoke for up to 50 minutes total. Once the participants either decided to smoke or the 50 minutes had fully elapsed, the participant had one hour to smoke freely. Participants chose how many combustible cigarettes that they would smoke from an $8 tab, and each cigarette cost $1. The primary dependent variables were total number of combustible cigarettes purchased and latency to choose to smoke. For the latter, data are not shown since differences between groups did not reach statistical significance.

E-Cigarette Perceptions Questionnaire.

This questionnaire consisted of 9 questions evaluating the effectiveness and liking of each specific e-cigarette dose. Questions 1-8 were scored using a Likert scale from 1-7 (Not At All to Very Much). Question 9 consisted of two check boxes focusing on whether the participants would prefer to smoke that e-cigarette dose again or their own combustible cigarette.

Virtual Reality Procedure.

VR testing was conducted as described in Thompson-Lake et al. and included active cues such as a smoking paraphernalia room and a party room, as well as a neutral room with no smoking cues.¹⁶ Given that these findings were negative (e.g. no differences among groups), the results are not shown here, but can be made available upon request.

Sample Size Calculation.

With regard to sample size, we referred to the power analysis conducted by Spindle et al. in which ANOVAs were used to examine subjective effects data (among other measures) in e-cigarette users. Their power analysis revealed that 15 participants were required to provide greater than 80% power, assuming a medium effect size and repeated measures correlations greater than 0.80.¹⁷ Another study that involved 4 e-cigarette doses and the identical device and e-liquid used in the current study confirmed this relative sample size to be sufficient. The sample size they utilized was 16 individuals.¹⁸ On this basis, we felt confident that a sample size of 15 would suffice for this study.

Data Analyses.

Outcomes were evaluated as change from baseline (for breath CO and QSU) or absolute values (for Choices to Smoke) and analyzed using one-way analysis of variance (ANOVA). In all cases p < .05 was considered statistically significant. All analyses were conducted using Statview™ version 5.0. Effect sizes were calculated using Cohen’s d, where 0.2 = small, 0.5 = medium, and 0.8 = large.¹⁹

Results

Demographics and Smoking Characteristics.

A total of 15 patients (10M:5F) with an average age of 50.6 ± 7.6 years participated. They were daily combustible cigarette users who had self-reported no prior e-cigarette use. Participants identified as Caucasian (26%), Black (66%), or Mixed (7%), as well as Hispanic (13%) or Non-Hispanic (87%). On average, they had smoked for 32.1 ± 7.1 years, consuming 19.3 ±13.1 combustible cigarettes per day, and had an average FTND score of 6.6 ± 1.1.

Breath CO.

A subtraction of CO levels after 10 puffs from baseline reveled a significant main effect (F_3,55 = 19.5, p < .0001). The post-hoc comparison was significant for CO in the Own smoking condition only (p < .0001).

QSU.

A subtraction of mean QSU scores after 10 puffs from baseline did not reveal a significant main effect (F_3,56 = .56, p = .65) (Figure 1a). Effects sizes were small for the comparisons between Own and each e-cigarette dose (Cohen’s d = 0.16 – 0.23).

Figure 1.

QSU Scores Were Not Significant After Puff Bouts.

Similarly, a subtraction of mean QSU scores after a second bout of 10 puffs from baseline reveled no significant main effect (F_3,56 = 1.43, p = .24) (Figure 1b). Effects sizes were medium for the comparisons between Own and each e-cigarette dose (Cohen’s d = 0.60 – 0.79).

Choices to Smoke.

There was no significant main effect (F_3,56 = 1.85, p = .15) on Choices to Smoke (Figure 2). Effects sizes were medium to high for the comparisons between Own and each e-cigarette dose (Cohen’s d = 0.79 – 0.86).

Figure 2.

Choices to Smoke Showed No Significant Effect

E-Cigarette Perceptions Questionnaire.

There were no significant main effects detected across groups for any of the individual questions (all p’s > 0.1) (Table 1).

Table 1.

E-Cigarette Perceptions Questionnaire Showed No Significant Effect

E-Cigarette Perceptions Questionnaire 0-7 Scale (Not At All to Very Much)	E-Cig Dose
	0 mg	18 mg	36 mg
Does this E-Cig dose reduce withdrawal symptoms?	3.6 ± 1.5	4.1 ± 1.6	3.8 ± 1.3
Does this E-Cig dose match the feeling of smoking own cigs?	2.9 ± 1.7	3.0 ± 1.6	2.7 ± 1.6
Would you be likely to smoke only this E-Cig dose?	3.6 ± 1.7	3.5 ± 1.7	3.1 ± 1.8
Do you think this E-Cig dose would help you quit smoking?	4.2 ± 1.4	4.3 ± 1.2	4.1 ± 1.6
Do you think this E-Cig dose is safer than own cigs?	4.7 ± 1.3	4.4 ± 1.5	3.9 ± 1.6
Would you be likely to buy this dose of E-cig?	3.9 ± 1.4	3.7 ± 1.4	3.3 ± 1.5
Would you recommend this dose of E-Cig to a cig smoker?	4.0 ± 1.5	4.3 ± 0.9	3.6 ± 1.7
How rewarding (satisfying) is this E-Cig dose compared to own?	3.1 ± 1.9	3.0 ± 1.8	2.7 ± 1.7
Which would you rather smoke -This E-cig dose or Own cig (ratio)	3:11	4:11	4:11

Conclusion

In the current pilot study, e-cigarettes did not reduce neither craving nor Choices to Smoke in e-cigarette naïve users. The differences between groups were generally of small to medium effect size. As expected, smoking an Own cigarette increased breath CO, but this was not seen for any e-cigarette condition. This replicates findings previously reported by Vansickel et al..¹¹

Smoking an Own cigarette reduced craving, as measured by the QSU, though this did not reach statistical significance. These effects were not observed after any e-cigarette dose. The lack of effect on craving, however, is more nuanced. Specifically, initial versus subsequent ratings (differential reductions in craving) did show effects. The outcomes on craving in this report are supported by recent work with combustible cigarette smokers showing that craving did not change after varying doses of nicotine (similar to that used here).²⁰ Notwithstanding, the data in this report do not replicate those reported by Bullen et al. and Hiler et al., who showed that e-cigarettes suppressed nicotine abstinence symptoms in e-cigarette naïve individuals.^{6, 7} Upon completion of this research, another report was published on e-cigarette naïve individuals showing an asymmetric substitution pattern in which e-liquid (24 mg/ml) served as a partial substitute for combustible cigarettes but not the reverse. Additionally, the participants in that study valued e-liquid positively and purchased it frequently both as a substitute for and independently of combustible cigarettes.²¹

Of interest, data from the current study showed the fewest Choices to Smoke in the Own cigarette condition, but this did not reach statistical significance. It is possible that on days that the participants were exposed to e-cigarettes they were inclined, on average, to make more Choices to Smoke in the e-cigarette conditions. This presumes that the e-cigarette doses were not as reinforcing as the Own cigarettes and that the participants needed more e-cigarette puffs to achieve the same effect.

One way to interpret the negative findings reported here is that the participants may have perceived this study as an experiment rather than an alternative to not smoking. The participants, therefore, may have been partial in their assessment and less accepting of the e-cigarettes. Another interpretation is that the older average age of participants enrolled (~50 years), considered in the context of their long history of smoking combustible cigarettes (~32 years), may have contributed to this cohort being less likely to feel any effects produced by acute e-cigarette exposure. However intriguing, the issue of age may simply be speculation as we do not have enough of an age range or sample size to evaluate it as a covariate. Additionally, these individuals were non-treatment seeking which, as compared to a treatment seeking population, may have contributed more towards perceiving e-cigarettes as ineffective.

It is possible that increasing the nicotine concentration could have increased differences among groups. JUUL e-cigarettes, a popular vaping device on the market, offers e-liquid containing 40 mg per cartridge — equivalent to 59 mg/ml.²² Given that the cohort for this study smoked about a pack of cigarettes per day on average, these participants may not have been satisfied with the comparatively lower e-cigarette doses of 18 and 36 mg. Utilizing a higher dose such as that of the JUUL, however, would likely not have been approved by the BCM IRB without peer-reviewed scientific data demonstrating its safety. Increasing the number of puffs taken would also not likely have changed these outcomes. The puff count we used was consistent with similar studies previously mentioned.^{5, 11} On the other hand, experienced e-cigarette smokers tend to puff intermittently throughout the day at a greater puff volume and in less clustered sessions.²³ It is unlikely, given the naivety of our participants to these devices, that increasing the amount of puffs in each session would substantiate differences among groups.

This was a pilot study so the conclusions that can be drawn are limited. Other limitations of this study included an overrepresentation of males and Black participants. These results, therefore, may not be generalizable to females and other races considering previous research showing gender and racial differences in nicotine intake and metabolism.^{24, 25, 26} Black smokers also report a higher usage of menthol cigarettes than traditional cigarettes, so using tobacco flavored e-cigarettes may have hindered our results.²⁷ Additionally, the sample size may have limited our ability to detect effects in the Own cigarette condition as well. Another limitation was the use of a non-blinded Own cigarette condition as the participants may have preferred their own brand simply because it was well-known to them.

Finally, the e-cigarette naïve population is critical in terms of understanding adoptability and effectiveness of e-cigarettes as a nicotine replacement therapy. This population remains of interest when evaluating the safety and effectiveness of these devices, and these outcomes may differ considerably in adolescent and young adults as compared to older individuals.