Nicotine or Expectancies? Using the Balanced-Placebo Design to Test Immediate Outcomes of Vaping

Abstract

Introduction:

Electronic (e-)cigarette use has increased in popularity, especially among those attempting to quit smoking. Previous studies evaluating the therapeutic efficacy of e-cigarettes have suggested that non-pharmacologic factors, such as expectancies about nicotine effects, may influence the experienced effects of e-cigarettes.

Method:

The independent and synergistic influences of drug and expectancies were parsed using a balanced-placebo design, whereby 128 participants (52 dual users) were provided an e-cigarette that either contained nicotine or non-nicotine solution, while told that it did or did not contain nicotine. We hypothesized main effects of nicotine content on physiological, objective outcomes (attention, appetite, aversion, respiratory tract sensations), and main effects of the instructions on more subjective, psychosocial outcomes (affect, reward, satisfaction). Sex was included as a moderator.

Results:

Results showed that nicotine increased sustained attention, and more so among females. Nicotine delivery was associated with aversion among females, but not males. Among those who were both told and did not receive nicotine, higher enjoyment of respiratory tract sensations was reported. Nicotine with complementary instructions produced the highest reward ratings; additionally, nicotine was rewarding to males but not females.

Discussion:

Findings demonstrated that both nicotine content and non-pharmacologic factors impact acute outcome effects of e-cigarettes, with moderation by sex in some cases. Results are relevant to the interpretation of clinical trials of e-cigarettes and suggest a more nuanced view of reinforcement pathways.

INTRODUCTION

As cigarette smoking rates decline, it is estimated that 2.8% of adults use electronic (e-) cigarettes regularly (Centers for Disease Control and Prevention, 2018). In fact, a majority of e-cigarette users, or “vapers,” are current or former smokers who report that their primary reason for using e-cigarettes is to quit or reduce smoking of traditional, combustible cigarettes (Pepper, Ribisl, Emery, & Brewer, 2014). The introduction of these products has generated debate among tobacco experts, ensuing a need for further research examining the effects and implications of e-cigarette use (Abrams et al., 2018; Glantz & Bareham, 2018, Fagerström, Etter, & Unger, 2015).

Thus far, only three large-scale clinical trials have been published directly assessing the therapeutic utility of e-cigarettes for smoking cessation (Bullen et al., 2013; Caponnetto et al., 2013; Hajek et al., 2019). In the first two studies, e-cigarettes containing nicotine were not found to be superior to non-nicotine (i.e. placebo) e-cigarettes in smoking abstinence or reduction, albeit with limited statistical power. However, a meta-analysis of the two studies suggested greater benefit of nicotine versus placebo e-cigarettes (Hartmann-Boyce, Bullen, Begh, Stead, & Hajek, 2016). Nevertheless, these findings suggest that e-cigarettes may influence smoking cessation via routes beyond nicotine delivery itself. Finally, Hajek et al., 2019 found e-cigarettes performed better than nicotine replacement therapy (NRT) on cessation rates and long-term abstinence rates among treatment-seeking smokers.

E-cigarettes provide a nicotine delivery mechanism that mimics the habitual and sensorimotor aspects of smoking. Qualitative and experimental studies of individuals using e-cigarettes report that these ritualistic reinforcing factors are an important component of the cessation process (Caponnetto, Polosa, Russo, Leotta, & Campagna, 2011; Van Heel, Van Gucht, Vanbrabant, & Baeyens, 2017). It is well understood that non-pharmacologic factors separate from and/or enhanced by the effects of nicotine also drive smoking behaviors (Robinson & Pritchard, 1992; Perkins, Karelitz, & Michael, 2017). These factors include expectancies, fundamental information processes that affect behavior generally (Goldman, 1999). Drug-related expectancies refer to anticipatory associations regarding the outcomes of drug use, as learned through biological, social, and experiential means (Goldman, Brown, & Christiansen, 1987). Expectancies are implicated in initiation as well as continuation and maintenance of drug use, and can inform the perceived effects of drug use via the “placebo effect” (Brandon, Juliano, & Copeland, 1999; Goldman, 1999).

Expectancies for cigarette smoking have been studied extensively using self-report and experimental means (Copeland, Brandon, & Quinn, 1995; Dar & Barrett, 2014). Known expectancies include the ability for cigarettes to increase positive affect and decrease negative affect, control appetite, alleviate cravings, stimulate, and enhance sensorimotor stimuli. Smokers will also endorse a number of adverse expectancies as well, such as cigarettes producing harmful health effects, uncomfortable physical sensations, and social stigma. Research on expectancies for e-cigarette use, however, is limited at present. Survey data indicates that e-cigarette users hold stronger positive expectancies regarding the taste and satisfaction achieved by vaping as compared to smoking, and weaker negative expectancies regarding addiction and health risks (Harrell, Marquinez et al., 2015). When compared to NRT (e.g. patch, gum), e-cigarettes are perceived to have fewer adverse side effects and to be superior at reducing cravings to smoke (Harrell, Marquinez et al., 2015; Hendricks et al., 2014). Furthermore, expectancies vary based on concurrent cigarette smoking, intention to quit, and sex (Copp, Collins, Dar, & Barrett, 2015; Harrell, Simmons et al., 2015; Piñeiro et al., 2016).

To separate the effects of drug delivery versus expectancies about the drug upon perceived drug effects, the balanced-placebo design (BPD) has been utilized (Hull & Bond, 1986; Marlatt, Demming, & Reid, 1973; Juliano & Brandon, 2002; Juliano, Fucito, & Harrell, 2011; Kelemen & Kaighobadi, 2007; Perkins et al., 2004, 2009; Gottlieb, Killen, Marlatt, & Taylor, 1987). In this paradigm, drug content (drug or no drug) is manipulated independently from instructional set (told drug or told no drug). Thus, a 2×2 factorial design allows for evaluations of independent main effects of drug pharmacology and expectancies, as well as their interactions. In general, the results from these studies demonstrate that drug delivery (i.e., alcohol or nicotine, versus placebo) appears to have primary influence over physiological or objective domains, such as cognition and physical symptoms, whereas drug expectancies may influence more emotionally salient or subjective domains, such as mood, satisfaction, and craving (Hull & Bond, 1986).

Recently, Palmer and Brandon (2018) utilized a BPD to evaluate the role of nicotine and expectancies on craving reduction outcomes of e-cigarette use. E-cigarette users were randomized to use e-cigarettes that contained either nicotine or non-nicotine solutions, and they were independently instructed that the e-cigarette contained nicotine or non-nicotine, resulting in four experimental conditions as illustrated in Table 1. Results showed that among dual users of both combustible cigarettes and e-cigarettes, cravings for cigarettes declined when the participants were told they were receiving nicotine, regardless of the actual drug content. In other words, expectancies about nicotine drove craving reduction effects to a greater degree than the drug itself. Additionally, among all participants, a drug X instruction interaction emerged on cravings to vape, suggesting a synergistic effect of both nicotine and the expectation of nicotine, in that participants in the True Positive condition showed the greatest reduction in cravings for e-cigarettes. These findings are consistent with the results of the aforementioned clinical trials that failed to show robust effects of nicotine per se on smoking cessation (Hartmann-Boyce et al., 2016).

Table 1.

Experimental design

Instruction	Drug Content
	Nicotine e-cigarette	Non-nicotine e-cigarette
Told Nicotine	True Positive	Placebo
Told Non-Nicotine	Anti-Placebo	True Negative

The present analysis examines secondary data collected during the Palmer and Brandon (2018) study. Using the BPD, the effects of nicotine and expectancies were parsed on additional outcome variables. These constructs were chosen based on previously identified expectancies for cigarette smoking: psychological reward, appetite control, sustained attention, enjoyment of respiratory tract sensations, aversive sensations, and affect regulation.

Main effects of drug were hypothesized to affect objective and physiological outcomes of nicotine use; that is, those receiving nicotine should report different physical responses (e.g. greater attention, lower appetite, lower aversion, and greater respiratory tract sensations) as compared to those not receiving nicotine. Additionally, it was hypothesized that main effects of instruction would impact subjective effects of nicotine use; that is, those told they received nicotine should endorse different psychological responses (e.g. higher positive psychological rewards, higher satisfaction, and lower negative affect) as compared to those told they did not receive nicotine. These predictions are in line with previous results on the influence of expectancies on perceived effects of tobacco, alcohol, and e-cigarette use (i.e., Copp et al., 2015; Dawkins, Turner, Hasna, & Soar, 2012; Gottlieb, Killen, Marlatt, & Taylor, 1987; Hull & Bond, 1986; Juliano & Brandon, 2002; Tate et al., 1994; Dawkins & Corcoran, 2014; Harrell & Juliano, 2012).

METHOD

Participants

Participants included 130 individuals who met the following eligibility criteria by telephone: 1) At least 18 years old; 2) daily e-cigarette users (use at least once per day for the past 30 days; must use nicotine solutions; must like tobacco, menthol, or fruit flavor); 3) smoking history of at least 100 lifetime cigarettes; 4) history of or current smoking rate of at least 1 cigarette per day for at least 30 days; 5) no current engagement in an e-cigarette cessation attempt; and 6) not currently pregnant, attempting to get pregnant, or nursing (by self-report). Generally, participants were recruited from flyers posted at local vape shops.

Experimental Procedure

Participants were asked to abstain from using e-cigarettes and combustible cigarettes for three hours prior to the session, and upon arrival, a breath carbon monoxide reading was administered. Participants were then randomized to both drug (nicotine) content and instructional set and factors, stratified on sex, flavor preference (tobacco, menthol, or fruit), and cigarette smoking status (current [defined by smoking >1 cigarette per week] or former). Drug content was double blind.

Demographic questionnaires, baseline measures, and the first set of dependent measures were administered. Next, participants engaged in an ad-lib vaping session using the provided e-cigarette, with nicotine content and instructional set as per randomization. Participants were instructed to take at least 10 puffs over the 10 minute session. Following the ad-lib session, the dependent measures were re-administered along with other assessments.

Apparatus

The e-cigarette used was a “second generation” eGo-style 3.6–4.2 Volt, 1100 mAh battery with a 2.8-Ohm, 510-style clearomizer with 1ml liquid capacity. The battery included an LCD display of number of puffs. The solution used was a custom made “research blend” (Avail Vapor, LLC) with target nicotine content of either 0 mg/ml or 12 mg/ml. Nicotine content was confirmed via mass spectrometry and liquid chromatography. A 50% vegetable glycerin (VG), 50% propylene glycol (PG) was utilized to intensify the throat sensations; hence, masking the non-nicotine solution. Flavor options included tobacco, menthol, or fruit, as determined by the participant during telephone screening. This particular configuration has demonstrated similar nicotine delivery as a combustible cigarette (Ramôa et al., 2015; Russell, Wilson, Patel, Feyerabend, & Cole, 1975; Yan & D’Ruiz, 2014).

Assessments

Baseline assessments.

Baseline data include demographic information, smoking history, and vaping history. Dependence on cigarettes and e-cigarettes was assessed via the Fagerström Test for Nicotine Dependence (FTND; Heatherton, Kozlowski, Frecker, & Fagerström, 1991; α = 0.56) and Penn State Electronic Cigarette Dependence Index (ECDI; Foulds et al., 2014; α = 0.70), respectively. Baseline e-cigarette expectancies were assessed using a modified version of the Smoking Consequences Questionnaire-Adult (SCQ-A; Copeland, Brandon, & Quinn, 1995; Harrell et al., 2014).

Dependent variables.

The following measures were assessed before and after the ad-lib vaping session, except where indicated.

Affect regulation is a well-researched motivational factor that contributes to cigarette smoking behavior (Brandon, 1994; Carmody, 1989). Thus, the Positive and Negative Affect Scale (PANAS; Watson, Clark, & Tellegen, 1988) was administered to evaluate current mood state pre- and post-ad-lib vaping session (α = 0.82 – 0.90). Metabolically, nicotine has a suppressive effect on appetite (Winders & Grunberg, 1990), and these expectancies are self-reported by smokers (Brandon et al., 1999). A three item Visual Analogue Scale (VAS; Flint, Raben, Blundell, & Astrup, 2000) was, therefore, used to assess changes in appetite emerging following the ad-lib vaping session (α = 0.85 – 0.89). The Modified Cigarette Evaluation Questionnaire (mCEQ; Cappelleri et al., 2007) was designed to reflect subjective immediate effects of cigarettes. As these effects reinforce smoking behaviors, this questionnaire was altered to assess perceived outcomes of vaping. Scales within this measure include Satisfaction (3 items; α = 0.87), Psychological Reward (5 items; α = 0.85), Aversion (2 items; α = 0.66), and Enjoyment of Respiratory Tract Sensations (1 item). Finally, nicotine’s stimulant effects have been shown to improve sustained attention following smoking (Heishman, Kleykamp, & Singleton, 2010). In the present study, short-term sustained attention was assessed after the ad lib vaping session using the Rapid Visual Information Processing Task (RVIP; Wesnes & Warburton, 1983), which has been previously found to also be sensitive to nicotine response expectancies (Harrell & Juliano, 2012; Juliano, Fucito, & Harrell, 2011). The task was administered via E-Prime (Psychology Software Tools, Inc.). Response sensitivity was calculated using the individual’s hit rate (hr; correct responses) and false alarm rate (far; incorrect responses) in the formula 0.5 + [(hr - far) + (hr - far)²] / 4*hr*(1 - far) (Harrell & Juliano, 2012; Sahgal, 1987). For the final analysis, 8 outlier participants were removed and a reciprocal transformation was performed to reduce negative skew and kurtosis. Unlike the other dependent measures, the RVIP was only administered after the ad lib vaping session, so change scores could not be calculated. Note that cravings to smoke and vape were also assessed, as previously reported (Palmer & Brandon, 2018).

Analyses

Factorial analyses of variance (ANOVAs; 2 [drug] X 2 [instructional set] X 2 [sex]) were used to test the main effects and interactions on outcome variables. Sex was included as a third factor due to established sex differences on expectancy and general effects of smoking and vaping (Copp et al., 2015; Perkins, 1996; Piñeiro et al., 2016). For outcome variables measured both pre- and post-ad-lib session, difference scores were calculated. Significant interactions effects were followed with post-hoc paired comparisons using t-tests to explore the nature of these effects. Because of the explaratory, early stage nature of these secondary analyses, we did not correct for Type I error.

RESULTS

Descriptive Statistics

After removing two participants, (one for instructional non-compliance, one for incorrect randomization) the final analyzed sample size was 128. Full descriptive characteristics of participants can be found elsewhere (Palmer & Brandon, 2018). The sample had a mean age of 36.4 (SD = 13.79), was 38% female, 83% Caucasian, 16% Hispanic, and 49% reported an annual income beyond $30,000. Fifty-two participants (41%) reported smoking cigarettes at least once per week, and were deemed “dual users.” Dual users endorsed smoking an average of 8.02 (SD = 8.57) cigarettes per day on an average of 5.01 (2.35) days per week. E-cigarette use differed between e-cigarette only users and dual users, with averages of 43.91 (SD = 59.60) and 26.66 (SD = 42.40) daily uses, respectively.

Physiological and Objective Outcome Variables

Enjoyment of Respiratory Tract Sensations.

Cell and marginal means for this mCEQ scale and all dependent variables are shown in Table 2. We did not find a main effect of drug (F [1, 120] = 0.64), instructional set (F [1, 120] = 0.34), or sex (F [1, 120] = 0.33). However, a drug X instructional set interaction emerged, F [1, 120] = 5.60, p < .05, η² = 0.05. Post-hoc analyses show an effect of instructional set only for those not receiving nicotine: True Negative (M = 4.12, SD = 1.93) vs Placebo (M = 3.26, SD = 1.44; t (58.96) = 2.04, p < .05, d = .51; Figure 1a).

Table 2.

Cell Means and Means of Manipulation Effects: Drug, Instruction, and Sex

Variable	Cell means: Drug X Instruction				Cell Means: Drug X Sex				Cell Means: Instruction X Sex
	True Positive	Placebo	Anti-Placebo	True Negative	Nicotine/Male	Nicotine/Female	No Nicotine/Male	No Nicotine/Female	Told Nic/Male	Told Nic/Female	Told nonnic/Male	Told nonnic/Female
Physiological and Objective Outcome Variables
Modified mCEQ-Enjoyment of Respiratory Tract Sensations	3.91	3.26a	3.13	4.12a	3.26	3.44	3.54	4.00	3.74	3.33	3.37	4.08
Modified mCEQ-Aversion	3.31	3.06	2.59	3.03	2.85	3.12a	3.44	2.35a	3.36	2.92	2.95	2.58
RVIP Sensitivity	1.65	1.59	1.59	1.51	1.60	1.65a	1.60	1.46a	1.64	1.57	1.55	1.54
VAS - Appetite	4.41	−3.03	−4.34	6.27	9.00	−13.96	2.89	21.13	1.36	−0.25	−1.83	5.96
Psychosocial and Subjective Outcome Variables
Modified mCEQ-Reward	19.81 a,b	14.03a	15.66b	17.45	19.49a	15.00	14.88a	17.43	18.23	14.92	16.07	17.42
PANAS-Positive	−1.00	−0.23	1.22	0.58	−0.11	0.48	−0.10	0.70	−0.32	−1.08	0.10	2.25
PANAS-Negative	1.63	2.32	1.88	2.12	1.95	1.46	1.80	2.96	2.05	1.87	1.70	2.50
Modified mCEQ-Satisfaction	14.94	12.84	12.25	13.27	13.62	16.56	12.20	14.61	13.56	14.46	12.24	13.67

	Marginal means: Drug content		Marginal means: Instructional set		Marginal means: Sex		F Values
Variable	Nicotine	No nicotine	Told Nicotine	Told No Nicotine	Male	Female	Drug X Instruction	Drug X Sex	Instruction X Sex
Physiological and Objective Outcome Variables
Modified mCEQEnjoyment of Respiratory Tract Sensations	3.70	3.52	3.59	3.63	3.55	3.71	5.60*	0.78	2.91
Modified mCEQAversion	2.95	3.05	3.19	2.82	3.15	2.75	1.08	3.96*	0.02
RVIP Sensitivity	1.62a	1.55 a	1.62	1.55	1.60	1.55	0.02	6.24*	0.39
VAS - Appetite	0.03	1.76	.075	1.05	−0.28	2.85	0.91	6.87*	0.27
Psychosocial and Subjective Outcome Variables
Modified mCEQPsychological Reward	17.74	15.80	16.97	16.57	17.13	16.17	8.34*	7.39*	3.29
PANAS- Positive	0.13	0.19	−0.61	0.89	−0.10	0.58	0.16	0.02	2.03
PANAS- Negative	1.76	2.22	1.98	2.00	1.87	2.19	0.07	1.66	0.67
Modified mCEQSatis faction	13.59	13.06	13.90	12.77	12.89	14.06	3.46	1.69	.112

Notes: Superscripts indicate significant differences in cell or marginal means, p < .05. An * indicates a significant interaction F-value, p < .05. Modified mCEQ = modified Cigarette Evaluation Questionnaire, modified for e-cigarettes. VAS = Visual Analogue Scale. RVIP = Rapid Visual Information Processing task (reciprocal transformation performed). PANAS = Positive and Negative Affect Scale. VAS and PANAS scores reported are difference scores from pre- to post-tests.

Figure 1. Drug X Instruction interactions.

Note: Superscripts indicate significant differences at p < .05. Error bars are standard error of the mean.

Aversion.

Whereas no main effects were found for drug (F [1, 120] = 0.06), instructional set (F [1, 120] = 1.25), or sex (F [1, 120] = 1.29), a drug X sex interaction emerged for the aversion factor of the modified mCEQ (F [1, 120] = 3.96, p < .05 η² = 0.03). Post-hoc t-tests showed the hypothesized effect of drug among females, but not males, in that females receiving nicotine had higher aversion scores (M = 3.12, SD = 1.67) than those not receiving nicotine (M = 2.35, SD = 0.71; t (33.01) = −2.12, p < .05, d = .60) (Figure 2a).

Figure 2. Drug X Sex interactions.

Note: Superscripts indicate significant differences at p < .05. Error bars are standard error of the mean.

Sustained Attention.

As hypothesized, a main effect of drug was found on sustained attention as measured by RVIP sensitivity (F [1, 111] = 6.35, p < .05, η² = 0.05) in that those who received nicotine had higher scores (M = 1.62, SD = 0.20) than those who did not (M = 1.55, SD = 0.20). However, a sex X drug interaction also emerged (F [1, 111] = 6.24, p < .05, η² = 0.05) such that the drug effect occurred only among females, with those receiving nicotine showing higher sensitivity scores (M = 1.65, SD = 0.21; t (39) = −2.71) than those who did not (M = 1.46, SD = 0.24, p < .05, d = .84) (Figure 2b). No main effect of instruction (F [1, 111] = 3.78) or sex (F [1, 111] = 2.12) was found.

Appetite.

No main effects of drug, instruction, or sex emerged for change in appetite as measured by the VAS (Fs [1, 120] = 0.71, 0.08, 0.16, respectively). A sex X drug interaction was observed, (F [1, 120] = 6.87, p < .05, η² = 0.05). This appeared as a cross-over interaction, with females reporting greatest appetite reduction when they did not receive nicotine, and males reporting greatest appetite reduction when they did receive nicotine. However, post-hoc paired comparisons failed to identify any significant differences between cells.

Psychosocial and Subjective Outcome Variables

Psychological Reward.

For the Psychological Reward factor of the mCEQ, no main effects of drug (F [1, 120] = 0.61), instruction (F [1, 120] = 0.08), or sex (F [1, 120] = 0.38) were found. A drug X instruction interaction was observed, F [1, 120] = 8.34, p < .01, η² = 0.07. Post hoc analyses indicate that the True Positive group had higher scores (M = 19.81, SD = 7.74) than the Placebo group (M = 14.03, SD = 7.09; t (61) = −3.09, p < .01, d = .78) and Anti-Placebo group (M = 15.66, SD = 7.92; t (62) = −2.13, p < .05, d = .53; Figure 1b). In addition, a drug X sex interaction was found (F [1, 120] = 7.39, p < .01, η² = 0.06) on psychological reward. Post-hoc t-tests revealed that males receiving nicotine had higher reward scores (M = 19.49, SD = 8.13) than those not receiving nicotine (M = 14.88, SD = 6.68; t (78) = −2.78, p < .01, d = .62), but no differences were found between drug condition in females (Figure 2c).

Other Scales.

No significant main effects of drug, instruction, or sex were observed on change scores for the positive and negative affect subscales of the PANAS. No further interactions emerged. Similarly, no significant results emerged for the Satisfaction subscale of the mCEQ.

DISCUSSION

The emergence of e-cigarettes has engendered substantial controversy, sustained by the paucity of available evidence regarding the effects of e-cigarette use. Thus, it is imperative that contemporary tobacco research assesses multiple factors that may contribute to the initiation and maintenance of e-cigarette use. Utilizing a fully-crossed BPD, the independent and synergistic influences of both nicotine delivery and expectancies were parsed on outcomes postulated to be related to maintenance of e-cigarette use. Based on nicotine and alcohol literature, it was hypothesized that physiological and objective outcomes would be most impacted by nicotine content, whereas more subjective effects would be influenced by expectancies for nicotine.

In line with these hypotheses, nicotine delivery produced improved sustained attention as measured by the RVIP. However, this effect was moderated by sex such that it was found among females but not males. Nicotine appears to produce a range of short-term beneficial effects upon cognitive processing, including attention (Ashare, Falcone, & Lerman, 2014; Evans & Drobes, 2009), whereas nicotine withdrawal produces acute attentional deficits (Hendricks, Ditre, Drobes, & Brandon, 2006). The main effect is consistent with this general finding. Although sex effects have been found in some studies on cognitive effects of nicotine delivery or withdrawal, the findings are inconsistent, and this is an area that requires additional research (Ashare et al., 2014).

Interactions between drug content and instructional set were observed for Psychological Reward and Enjoyment of Respiratory Tract Sensations. For the former, the True Negative condition produced the greatest enjoyment of respiratory tract sensations; for the latter, the True Positive condition produced the greatest reported reward. This effect was significantly moderated by sex such that nicotine, compared to no-nicotine, was rewarding to males but not females. Together with the finding that nicotine delivery was associated with aversion among females but not males, these results are consistent with previous reports suggesting that nicotine has greater subjective and reinforcing effects upon males (e.g., Perkins, 1996). It is interesting that participants reported the greatest respiratory track enjoyment in the True Negative condition. This suggests that both nicotine and nicotine-related expectancies are associated with less enjoyment of this type. The positive perceived effects of smoke in the upper airways have been found to be independent of cigarette nicotine content (Westman, Behm, & Rose, 1996), indicating that in the present study, nicotine or expected nicotine, may have interfered with this aspect of participants’ enjoyment of vaping.

Results from this study have several limitations, and should be interpreted accordingly. To begin, inconsistent nicotine delivery has been observed across e-cigarettes in general (e.g., Eissenberg, 2010; Farsalinos et al., 2014). Although the nicotine dosage and second-generation e-cigarette were selected to deliver adequate nicotine, blood nicotine levels would be required to quantify actual nicotine delivery. Additionally, measures utilized in the present study that were adapted for e-cigarettes were not validated as such. Given the early stage of this research, multiple tests were conducted without correction, which could inflate Type I error rate. Finally, the fidelity of manipulations in the BPD are always imperfect. As presented in Palmer and Brandon (2018), when asked to guess nicotine content following the ad-lib session, not all participants guessed consistently with their instructional assignment, a common challenge observed in placebo and balanced-placebo research (Dar & Barrett, 2014).

Although nicotine is far from the most harmful ingredient in combustible cigarettes, its presence in e-cigarettes has raised concerns with respect to both addiction liability and toxicity (World Health Organization, 2014). The findings from the present study suggest that outcomes of e-cigarette use result from interactions of both nicotine pharmacology and cognitive outcome expectancies, with additional moderation by individual differences, such as sex. Therefore, the focus on nicotine in e-cigarettes should be expanded to include a more nuanced understanding of non-pharmacological effects of e-cigarette use, and their role in dependence on these products. Eventually, this line of research could inform both intervention development and public policy with respect to e-cigarettes’ utility for smoking cessation as well as the minimization of their addiction liability.

Highlights.

• Outcomes of e-cigarette use (cravings) are driven by expectancies as well as nicotine

• To parse these effects on additional outcomes, we conducted a balanced-placebo design

• Nicotine, expectancies, and sex interacted on a variety of outcome variables

• Results provide insight into the addictive and clinical potential of e-cigarettes