Many smokers say that they have the intention to quit and many make actual attempts, but the vast majority of these attempts is unsuccessful. Typically, no more than 3%–5% of smokers trying to quit without any assistance are still abstinent 6–12 months later [1
]. Those smokers who do seek evidence-based assistance in their quit attempt are only two to three times more likely to become and remain abstinent in the long-term [2
]. Assistance includes medically approved smoking cessation aids and professional quit-smoking assistance, such as quit-smoking medication, short-term pharmaceutical Nicotine-Replacement Therapy (NRT) and/or behavioral counseling.
Tobacco Harm Reduction (THR) may prove a viable alternative for those smokers who are not able or not willing to cease all tobacco and/or nicotine consumption. THR refers to any actions taken to reduce the substantial health risks associated with tobacco smoking, such as educating people about the risks of different sources of nicotine, and providing smokers with the opportunity to switch from tobacco smoking to the use of low-risk nicotine products [3
] including smokeless tobacco (e.g., Swedish snus), long-term pharmaceutical NRT, or more recently, e-cigs. A unique characteristic of e-cigs is that, while acknowledging the importance of efficient nicotine delivery, they also aim to address the behavioral and sensory aspects of smoking. At least at face value, e-cigs appear to provide good mimicry of many of the crucial Conditional Stimuli (CSs) related to the act of smoking tobacco cigarettes (for example, hand-to-mouth action, visual cues, sensorimotor cues of inhaling and exhaling, throat hit, tobacco flavor) [3
Taking into account that one pack-year of smoking equals over 70,000 puffs, a typical 40 year-old smoker will have experienced 1–2 million trials in which nicotine administration and its rewarding effects has been contingently paired with a well-specified set of proximal CSs (hand-to-mouth action, visual cues, sensorimotor cues of inhaling/exhaling, throat hit, flavor/smell) and a less well-specified set of more distal CSs (location, time-of-the-day, preceding or concurrent activities, social context, internal/emotional cues). Exposure to these cues (when smoking a new tobacco cigarette) or to perceptually highly similar cues (for example, when using an e-cig) can thus be expected to tap into and activate an already well-established associative memory network consisting of smoking/nicotine-associated cues linked to previously experienced nicotine effects. The critical involvement of such non-pharmacological CSs associated with smoking in the reinforcing and craving reduction effects of nicotine has been documented in both animal and human research [7
Animal research suggests that the primary reinforcing effects of nicotine are relatively weak compared to other substances, such as cocaine or heroin [7
]. There is strong evidence that animal nicotine self-administration not only depends on the primary rewarding effects of response-contingent nicotine delivery but also on the conditioned rewarding effects of non-nicotine cues (e.g., visual or auditory stimuli) that are associated with nicotine delivery [14
]. Moreover, nicotine has also been shown to act as a (non-contingent) reinforcement enhancer of other unconditioned or conditioned (nicotine-associated) reinforcers, synergizing in a multiplicative rather than an additive fashion with nicotine in the acquisition and maintenance of nicotine self-administration [15
In the same vein, human data on nicotine craving reduction and smoking satisfaction point to the critical role of non-pharmacological CSs [10
]. For example, Rose and colleagues [12
] designed an experimental procedure allowing assessment of the relative contribution of the smoking ritual and of nicotine administration on cigarette craving reduction after a 12-h overnight abstinence period. Smokers received intravenous injections of either nicotine or saline (control), and, orthogonal to this manipulation, they were or were not allowed to simultaneously smoke a denicotinized cigarette. The resulting four experimental conditions were compared to a condition in which participants smoked their usual brand of nicotine-containing cigarettes. Cigarette craving was assessed prior to and 30 and 60 min after the experimental manipulation. The relieving effects solely due to the intravenous nicotine administrations were observed to be substantially lower than the satiation resulting from smoking a usual brand cigarette. The sensorimotor stimuli alone (provided by smoking the denicotinized cigarette) were observed to result in a similarly small but reliable relief of the craving to smoke. Only when intravenous nicotine and sensorimotor components were combined (nicotine injections plus smoking denicotinized cigarette), a relief level similar to the craving reduction resulting from smoking a usual brand cigarette was achieved.
In this condition of the study by Rose and colleagues [12
], the smokers were exposed to most of the stimuli associated with regular smoking—participants smoked an actual cigarette, be it denicotinized—while simultaneously obtaining nicotine, which probably created an optimal situation for craving reduction. As argued above, vaping an e-cig creates a sensorimotor experience that is substantially similar (but not identical) to that resulting from smoking: it exposes the user to many (but not all) of the critical stimuli associated with the ritual of smoking while intermittently providing her with doses of nicotine. Vaping essentially emulates the condition of the study by Rose and colleagues [12
] in which participants were allowed to smoke while they received doses of nicotine intravenously. Some studies have found that the nicotine delivery resulting from vaping may be less efficient (less and more gradual delivery of nicotine to the blood and brain) than nicotine delivery resulting from tobacco smoking [16
]. However, in other studies using more advanced (high-wattage) devices, higher nicotine concentrations in the e-liquid and experienced users, or a combination of those, nicotine delivery profiles approaching or even surpassing those of combustible tobacco cigarettes have been observed [18
]. Despite this heterogeneity in nicotine delivery profiles, and despite the fact that the non-nicotine cues provided by vaping are similar but not identical to those of smoking combustibles, multiple experimental studies have convincingly and unanimously shown that vaping can indeed lead to craving reduction.
] allowed participants to take 10 puffs of an e-cig with 16 or 18 mg/mL nicotine, to smoke (10 puffs) a combustible tobacco cigarette, or to sham smoke (holding an unlit tobacco cigarette) after a 12-h abstinence period. The reduction in desire to smoke was reliable for the e-cigs with nicotine but weaker than for tobacco cigarette smoking, and absent after sham smoking. Bullen and colleagues [16
] found that the desire to smoke after overnight abstinence was significantly reduced after five minutes of vaping an e-cig with 16 mg/mL nicotine, but also with an e-cig containing 0 mg/mL nicotine. However, the reduction found in the 16 mg/mL nicotine condition was significantly stronger than in the no-nicotine condition. The same conclusions were reached by Dawkins and colleagues [22
] who presented their participants with either an e-cig containing 0 mg/mL or 18 mg/mL nicotine to vape for five minutes. An important difference is that the abstinence period here was limited to one hour, which indicates the robustness of the reduction findings over different durations of smoking abstinence. Dawkins and Corcoran [23
] recruited experienced vapers and asked them to abstain overnight from all tobacco/nicotine products. In the morning, they were presented with a first generation e-cig containing 18 mg/mL nicotine, from which they were allowed to take 10 puffs. As a result, the nicotine craving and urge to smoke were significantly reduced in these experienced vapers. In a group of smokers without any previous experience with vaping and using a more advanced open-system e-cig with e-liquid containing 18 mg/mL of nicotine, Adriaens and colleagues found that five minutes of vaping resulted in immediate craving reduction after four hours of abstinence that was equally strong as the craving reduction obtained by smoking a tobacco cigarette [24
]. In summary, earlier studies clearly established the craving reduction effects of vaping in controlled laboratory trials; these effects were sometimes also present when using e-cigs containing no nicotine, be it to a lesser degree. This suggests that nicotine cannot be solely accountable for the observed craving reduction. The question then becomes which particular aspects of the smoking ritual (CSs) are responsible for this craving reduction.
Dawkins and colleagues [25
] presented participants with either a red or white cig-a-like nicotine-containing (18 mg/mL) “tobacco” flavored e-cig after a 10-h abstinence period. Their results suggest that the visual appearance of an e-cig has an effect on cigarette craving reduction. They found that the more the e-cig resembled a regular cigarette (white color), the stronger the craving reduction. However, this effect was only found in people without prior experience with e-cigs [25
]. As discussed above, the study of Rose and colleagues [12
] suggested that handling a (tobacco) cigarette, including the act of inhaling and exhaling, also contributes to craving reduction; it remains to be seen, however, whether or not this finding can be replicated in the context of vaping. Finally, there is some indirect experimental evidence suggestive of the involvement of aroma as important non-nicotine cues: Audrain-McGovern and colleagues [26
] demonstrated that flavored e-cigs (“green apple” or “chocolate” vs. unflavored) resulted in a subjectively more rewarding vaping experience, whereas Goldenson and colleagues [27
] showed that nicotine-free and nicotine-containing (6 mg/mL) e-cigs produced greater appeal when containing sweet flavors (e.g., peach, blackberry) than when containing non-sweet flavors (e.g., tobacco, menthol) or no flavor, but the authors of both experimental studies did not assess whether or not this had any impact on the craving reduction potential of vaping. Altogether, these lab studies suggest that sensorimotor stimuli, as well as visual and also flavor cues, may contribute to the effectiveness of the e-cig in terms of craving reduction [12
The aim of this experimental study was to examine the impact of four variables pertaining to the use of e-cigs on craving for tobacco cigarettes and for e-cigs after an overnight abstinence period. The four variables were the nicotine level (0% vs. 3.6% nicotine), the sensorimotor component (e-cig handled by participants vs. held by a unipod), the visual aspect (no visual restrictions vs. blindfold), and the aroma (tobacco flavor vs. apple flavor) of the e-cig. These variables were manipulated in an orthogonal manner, resulting in a between-group 2 × 2 × 2 × 2 factorial design (16 experimental conditions). The study consisted of two lab sessions (session 1: intake plus baseline measurements; session 2: experimental vaping session) that were separated by a (approximately) one-week period during which participants, who were all naïve about e-cigs, could familiarize themselves with vaping.
The effects of nicotine and of the non-nicotine factors were investigated as exploratory. No specific hypotheses were put forward concerning the (relative) strength of possible main effects or with respect to the existence or importance of interactions between the manipulated variables. In general, however, we expected (a) an overall craving reduction following use of the e-cigs; (b) which would show up in the nicotine (3.6% nicotine) as well as in the no nicotine (0% nicotine) conditions but (c) which was hypothesized to be weaker in the no nicotine conditions than in the nicotine conditions. Furthermore, we predicted (d) a facilitating influence from the presence of handling cues (manual handling better than unipod handling), visual cues (unrestricted vision better than blindfolded), and aroma (tobacco flavor better than apple flavor).
In the present study, it was investigated which factors contributed to cigarette and e-cig craving reduction after a 12-h abstinence period. More specifically, the relative contribution of nicotine and non-nicotine (conditioned) factors were taken under scrutiny. Apart from the nicotine concentration of the e-cigs, we manipulated three non-pharmacological smoking/vaping-associated CSs for which there was some preliminary experimental evidence of their potential involvement in craving reduction [12
]: Handling (Manual/Unipod), Visual Cues (visual/blindfold) and Aroma (apple/tobacco). The induction of craving, by means of an abstinence period prior to the test session, was successful. The CO level did significantly decline from the intake to the beginning of the test session. This indicates a period of abstinence from smoking in which (e-) cigarette craving is expected to build up.
The first experimental finding is that a decrease in craving was obtained regardless of whether nicotine was present or not. We observed reliable craving reduction when the e-cig contained no nicotine, but the decrease was significantly stronger when nicotine was present. With this finding, we replicated earlier findings by Bullen and colleagues and Dawkins and colleagues [16
]. This superiority of nicotine-containing e-cigs showed up in both craving measures (TCQ and VAS) and both with respect to cigarette craving and e-cig craving. What this study does not speak to is the specific mode of action of nicotine with respect to craving reduction. The most obvious route is, of course, the direct pharmacological central nervous system effects of nicotine (nicotine-acetylcholine-glutamate- gamma-aminobutyric acid (GABA)-dopamine neural circuitry promoting nicotine reward, dependence, and withdrawal) [37
]. Another mechanism may be the airway sensory impact of nicotine (“throat hit”), which has been shown to contribute to the conditioned rewarding effects (pleasantness, desirability) of (tobacco) cigarette puffs independently from the dopaminergic central nervous system effects [38
], as well as to the potential of e-cigs to evoke a self-reported readiness to switch from tobacco smoking to e-cigs in smokers [39
]. Our finding confirms the important role of nicotine in craving reduction, but, at the same time, it indicates that other, conditioned factors may play a role as well [14
A second experimental finding resulted from the exploratory analysis of the VAS measures. The presence or absence of handling cues proved to interact with the presence or absence of nicotine with respect to the amount of craving reduction, both with respect to cigarette craving and e-cig craving. When the cues related to the standard hand-to-mouth action were present, the craving reduction was obtained irrespective of the presence or absence of nicotine. Only when a unipod was used so that the habitual hand-to-mouth sensorimotor stimuli were eliminated were the craving reducing effects of nicotine observed. A plausible explanation for this observation is that it is due to a floor effect with respect to the maximum observable craving reduction in this artificial experimental lab situation. Moreover, it should be interpreted with caution, as it only showed up in the VAS data, not in the TCQ. One possible explanation of the latter is the fact that the TCQ assesses more and different aspects of craving (e.g., the ability to limit one’s smoking, when a cigarette is available). For example, it may be possible that the hand-to-mouth cues influence the extent to which the participants feel like smoking, but do not influence all aspects of the craving construct as measured by the TCQ (namely, “emotionality”, “expectancy”, “compulsivity”, and “purposefulness”).
The third experimental finding resulting from the exploratory analyses was an effect of the visual cues on e-cig craving reduction. When participants were blindfolded, craving reduction was obtained irrespective of the presence or absence of nicotine. Only when participants had unrestricted vision did the craving reducing effects of nicotine come into play. This pattern of results is at the same time counterintuitive, hard to explain given previous research [26
], and does not easily fit within the framework of classical conditioning and the role of conditioned stimuli. However, it must be stressed that this was only observed with respect to e-cigarette craving, not cigarette craving, and only showed up in the VAS data, not the TCQ data.
Finally, despite evidence from previous research [27
], the aroma of the e-cig did not seem to affect the cigarette or e-cig craving.
Despite its strengths, this study also has a number of limitations. A first potential weakness is the fact that it is difficult to check whether the participants indeed did not smoke or vape for 12 h prior to the test session. A control system was built into the procedures, namely the measurement of CO, but this does not provide conclusive evidence. CO measures only give an indication of (the abstinence of) smoking regular cigarettes, not of vaping e-cigs, which does not influence the CO level [24
]. The possible variance in the length of the actual periods of abstinence of both smoking and vaping can influence the baseline level of craving and possibly also the further course of post-vaping craving levels, which is one of our main dependent variables. This is especially problematic in the case where the actual abstinence periods may have been different between different experimental conditions.
A second potential weakness concerns our operationalization of the critical conditioned stimuli. We worked from the hypothesis that the characteristics of the e-cig resemble the conditioned stimuli of regular tobacco smoking. During the experiment, the participants noted, however, that the tobacco flavor of the e-cigs did substantially differ from the flavor of real tobacco, and especially of smoked tobacco. The question then becomes whether the flavors were sufficiently alike to generalize the conditioning effects from the tobacco flavor in combustible cigarettes to the tobacco flavor in e-cigs. This also implies that, in this study, the presentation of the tobacco flavor cues was only partially realized, which may at least in part explain the lack of effects of the aroma manipulation. It is possible that results would be different with another e-cig tobacco aroma, but with current vaping technology, it is very unlikely that the aroma of smoked tobacco would be replicated to perfection. With regard to the handling characteristics, a similar remark can be made. The e-cig is and was reported by the participants to be remarkably longer and especially heavier than a tobacco cigarette. Given this fact, the handling cues of the e-cigs also probably only partially mimicked the handling cues of a regular cigarette. Again, this can influence the degree of generalization of the conditioned handling aspects from smoking a tobacco cigarette to using an e-cig. Despite the fact that the cues were not perfectly simulated, the e-cig was a non-refillable “cig-a-like”, which offers a better approximation of the handling cues than the refillable second, third, or fourth generation e-cigs with larger and heavier batteries (i.e., box-mods). On the other hand, it has to be noted that some aspects of the handling, namely the “mouth action” and the stimuli resulting from the puffing (inhaling/exhaling), remained present even in the Unipod condition. This way, it is possible that some crucial aspects of the handling were still present in the condition that was intended to remove these stimuli. In this sense, our manipulation of the sensorimotor stimuli only allowed for a less than ideal assessment of the role of handling cues on craving reduction. Note that some studies did take away the handling completely by administering the nicotine intravenously [12
A third potential weakness concerns the power of the present study. Investigating 16 conditions with 81 participants will have certain drawbacks, such as the fact that small effects cannot be found or that the found effects are potentially overestimated. Especially in the three-way interactions, sufficient power is an issue that should be taken into account.
A fourth potential weakness is the generalizability of the findings. Generalizing the findings to more prolonged/more frequent vaping in everyday situations should be done with great care since this study (a) took place in an artificial laboratory setting requiring repeated explicit craving reports and (b) implemented only a single five minute opportunity to vape in order to reduce craving with a one-hour follow-up.
A final potential weakness concerns the duration of the period that participants could get acquainted with using an e-cig before the actual lab session. The literature is not consistent about the amount of time that is needed to provide the participants with sufficient experience with the e-cig [24
]. Sufficient experience may be important, however, not only to attain optimal craving reduction but also to be able to demonstrate sufficient sensitivity to some subtle manipulations of the smoking/vaping-related CSs.