Does acute tobacco smoking prevent cue-induced craving?

Introduction

Nicotine replacement therapy (NRT) administration and expectancy (the belief that nicotine has been consumed) are generally ineffective in preventing episodic peaks in cigarette craving triggered through exposure to smoking-associated stimuli (i.e. cue-induced craving; Ferguson and Shiffman, 2009). We recently reported that neither the independent nor combined effects of nicotine administration, via acute NRT use, or nicotine expectancy prevented craving following exposure to smoking-associated stimuli (Schlagintweit et al., 2014). Research examining the impact of varenicline, an α4β2 nicotinic receptor partial agonist, on cue-induced craving have produced more ambiguous findings. For example, Franklin et al. (2011) reported diminished ventral striatum and medial orbitofrontal cortex responses to smoking cues following varenicline treatment, but this study failed to observe significant medication by time interactions involving cue-induced craving. In a second study, Brandon et al. (2011) reported relatively diminished cue-induced craving following varenicline treatment, but because the effect was moderated by perceived drug condition, it is unclear to what extent these findings are due to drug effects or expectancy effects. The persistence of cue-induced craving following smoking cessation aid use may represent an important factor contributing to cessation failure, as cue-induced craving has been found to play an important role in relapse into smoking (Ferguson and Shiffman, 2009).

To our knowledge, the extent to which cue-induced cravings persist following actual smoking behavior has not been directly examined. Assessing the effectiveness of denicotinized cigarettes, which contain much less nicotine than conventional cigarettes, in preventing cue-induced craving may be of particular interest, as switching from conventional to denicotinized tobacco smoking has been found to facilitate smoking cessation (Dermody and Donny, 2014; Hatsukami et al., 2010; Walker et al., 2012). In double-blind, placebo-controlled, laboratory-based studies, denicotinized tobacco has been found to be more effective than NRT in delaying subsequent smoking behavior (Barrett, 2010), as well as to curb craving and withdrawal in a manner comparable to conventional cigarettes (Dallery et al., 2003; Donny et al., 2007). Denicotinized and conventional tobacco smoking have similar sensori-motor characteristics (i.e. similar appearance, smell, taste and administration rituals; Rose et al., 2000). It is possible that the replacement of such conditioned stimuli might reduce the salience of other environmental smoking associated stimuli. In addition, tobacco smoke contains a number of non-nicotine psychoactive components, such as anabasine and acetaldehyde (Caine et al., 2014; Hoffman and Evans, 2013), as well as the monoamine oxidase inhibitors harman and norharman (e.g. Herraiz, 2004). These constituents might also provide adequate satiation in the presence of smoking-associated cues. Moreover, smokers have also been found to have much greater expectations that cigarettes will help ameliorate negative affect and craving relative to NRT (Juliano and Brandon, 2004), and it is possible that such findings translate to denicotinized tobacco. On the other hand, nicotine consumed through conventional tobacco smoking might be more effective in preventing cue-induced craving than denicotinized tobacco due to a rapid delivery of nicotine (Hukkanen et al., 2005), to interactions of nicotine with various non-nicotine tobacco entities (Clemens et al., 2009; Harris et al., 2010), to interactions of sensorimotor aspects of smoking with nicotine (Rose et al., 2000), or to psychological processes such as the belief that nicotine had been consumed (see Dar and Barrett, 2014).

An increasing body of research documents sex differences in cue-induced craving and reactivity to nicotine and non-nicotine aspects of tobacco smoking (Barrett, 2010; Field and Duka, 2004; Perkins et al., 1999; Rose, 2006; Schlagintweit et al., 2014). Females appear to demonstrate greater reactivity to smoking-related stimuli than males, even when exposed to smoking cues following acute NRT consumption (Field and Duka, 2004; Schlagintweit et al., 2014). Perhaps females may also show greater cue reactivity than males following denicotinized and nicotine-containing tobacco consumption. Alternatively, Barrett (2010) reported that denicotinized tobacco is more effective in suppressing tonic subjective craving in females than males. Thus, one may also expect that denicotinized tobacco may be more effective in suppressing cue-induced craving in females relative to males. To our knowledge, sex differences in cue reactivity following nicotine-containing and denicotinized tobacco consumption have not been assessed.

The present study aimed to (a) evaluate the independent and combined impact of acute nicotine administration and expectancy components of tobacco smoking on cue-induced craving and (b) identify potential sex differences in cue reactivity following nicotine-containing and denicotinized tobacco smoking. The study used a balanced placebo design, which manipulated participant expectancies about the nicotine content of nicotine-containing and denicotinized cigarettes. Following cigarette smoking, participants were exposed to smoking-associated stimuli, and subjective craving was assessed prior to and after cigarette consumption and stimulus exposure. Heart rate was also assessed prior to and after cigarette consumption and cue exposure, in order to verify differences in the physiological effects of nicotine-containing and denicotinized tobacco consumption, as nicotine-containing tobacco smoking has been found to lead to greater increases in heart rate than denicotinized tobacco smoking (Benowitz, 1986). It is hoped that findings from this study will help to identify pharmacological or psychological factors that might help prevent increases in craving that result from the exposure to smoking-related stimuli, as well as to clarify the mechanisms underlying the therapeutic benefits of denicotinized tobacco.

Material and methods

Participants

Thirty-five (18 male) healthy, medication-free dependent smokers (Fagerström Test for Cigarette Dependence (FTCD) ⩾ 3; Fagerström, 2012) participated in the current study; however, five of these participants (one male) failed to complete both study sessions. Thus, the final study sample consisted of 30 (17 male) participants with mean FTCD scores of 5.10 (standard deviation (SD)=1.60). Participants self-reported an absence of a lifetime history of psychiatric illness, smoked an average of 12.75 (SD=5.98) cigarettes per day, ranged in age from 20 to 58 years (mean=26.37, SD=7.28) and had been daily smokers for at least one year. Table 1 presents additional demographic characteristics of the participants.

OPEN IN VIEWER

Table 1.

Mean (standard deviation) demographic information for participants who completed both study sessions (n=30). No significant differences in demographic information were observed across study groups (p>0.05).

	Denicotinized cigarette (n=15)		Nicotine-containing cigarette (n=15)
	Female (n=7)	Male (n=8)	Female (n=6)	Male (n=9)
Age in years	23.57 (3.31)	30.63 (11.56)	23.67 (3.01)	26.56 (5.57)
Age of first cigarette	12.71 (0.95)	14.75 (3.20)	15.67 (2.50)	13.67 (4.39)
Cigarettes per day	12.36 (6.41)	14.88 (6.68)	12.33 (6.61)	11.44 (5.06)
Years as a daily smoker	6.79 (4.28)	13.13 (13.16)	6.33 (3.61)	7.01 (5.96)
FTCD	5.00 (1.41)	5.25 (1.58)	5.33 (2.07)	4.89 (1.69)

FTCD: Fagerström Test for Cigarette Dependence.

Results

Findings from all participants who completed both sessions

Cigarette craving

Tobacco consumption

The main effects of Time were observed for both factor 1 (intention to smoke; F(1,27.80)=58.76, p⩽0.001) and factor 2 (withdrawal-relief; F(1,30.69)=25.35, p⩽0.001) craving. Craving was found to be reduced following tobacco consumption (factor 1: Mean (M)=20.01, standard error (SE)=1.21; factor 2: M=13.80, SE=1.31) compared to baseline (factor 1: M=29.50, SE=1.21; factor 2: M=19.31, SE=1.31).

Receive by Time interactions were also observed for factor 1 (F(1,27.80)=11.65, p=0.002) and factor 2 (F(1,30.69)=4.44, p<0.05) craving. Following tobacco consumption, craving was reduced in the receive nicotine-containing cigarette condition (factor 1: M=14.28, SE=1.73; factor 2: M=10.96, SE=1.87) relative to the receive denicotinized cigarette condition (factor 1: M=25.74, SE=1.70, p⩽0.001; factor 2: M=16.63, SE=1.84, p<0.05). Craving was also reduced after tobacco consumption relative to baseline in the receive nicotine-containing cigarette (factor 1, T1: M=28.00, SE=1.73, p⩽0.001; factor 2, T1: M=18.78, SE=1.87, p⩽0.001) and receive denicotinized cigarette conditions (factor 1, T1: M=31.00, SE=1.70, p=0.005; factor 2, T1: M=19.84, SE=1.84, p<0.05).

Finally, a significant Instruction by Time interaction was observed for factor 1 craving (F(1,33.83)=4.69, p<0.05). Following tobacco consumption, craving was reduced in the told nicotine-containing cigarette condition (M=18.44, SE=1.32) relative to the told denicotinized cigarette condition (M=21.58, SE=1.32, p⩽0.001). Craving was also reduced after tobacco consumption relative to baseline in the told nicotine-containing cigarette (T1: M=29.21, SE=1.32, p⩽0.001) and told denicotinized cigarette conditions (T1: M=29.80, SE=1.32, p⩽0.001).

Cue reactivity

The main effects of Time were observed for both factor 1 (F(1,26.00)=28.08, p⩽0.001) and factor 2 (F(1,26.00)=15.72, p=0.001) craving. Craving was greater after the smoking cue (factor 1: M=24.93, SE=1.12; factor 2: M=15.60, SE=1.24) relative to the neutral cue (factor 1: M=21.24, SE=1.28; factor 2: M=13.61, SE=1.16). A near significant Sex by Instruction by Time interaction was also observed for factor 1 craving (F(1,26.00)=4.16, p=0.052, see Figure 1). Craving was elevated following the smoking cue relative to the neutral cue in females in the told nicotine-containing cigarette condition (T3: M=20.18, SE=2.07; T4: M=26.10, SE=1.98, p⩽0.001), as well as in males in the told nicotine-containing cigarette (T3: M=20.26, SE=1.80; T4: M=23.13, SE=1.73, p<0.05) and told denicotinized cigarette (T3: M=21.32, SE=1.85; T4: M=25.10, SE=1.64, p<0.01) conditions. Craving did not differ between the smoking and neutral cues in females who were told that they had received a denicotinized cigarette (p>0.05).

OPEN IN VIEWER

Figure 1.

Estimated marginal mean values (+/- standard error) for QSU-B factor 1 craving. Craving was elevated following the smoking cue relative to the neutral cue in females and males who were told that they had received a nicotine-containing cigarette and in males who were told that they had received a denicotinized cigarette, but not in females who were told the same.

Discussion

While cigarette consumption was found to reduce acute cigarette craving, exposure to smoking-associated stimuli increased craving regardless of expectancy and/or nicotine effects of recent smoking. The current findings are consistent with prior observations that smoking cues increase cigarette craving following as few as 30 minutes of abstinence (McBride et al., 2006), and suggest that smoking-associated stimuli may play a critical role in the maintenance of smoking behavior regardless of recency of smoking, nicotine administration or expectancy. The important role of cue-induced craving in maintaining smoking behavior has also been demonstrated in naturalistic studies, where exposure to smoking-associated stimuli leads to a greater likelihood of smoking (Shiffman et al., 2004, 2014). While there is some debate regarding the association between smoking cue exposure in the laboratory versus the natural environment (see Shiffman et al., 2015), findings attest to the critical role of cue-induced craving in maintained smoking behavior.

The persistence of cue-induced craving following denicotinized tobacco consumption suggests that denicotinized tobacco’s effectiveness as a smoking cessation aid is likely not a result of curbed cue-induced craving. In fact, nicotine-containing tobacco was also relatively ineffective in preventing cue-induced craving. These findings, along with those of previous studies examining cue-induced craving following acute NRT use (e.g. Schlagintweit et al., 2014) and prolonged varenicline use (Brandon et al., 2011; Franklin et al., 2011), suggest that smoking salient stimuli may continue to pose a risk for relapse despite the use of varied psychological and pharmacological interventions. Given that cue-induced cravings have been found to persist for several months following cessation, despite reduced background craving and withdrawal (Balter et al., 2015; Bedi et al., 2011), and that individuals who respond less strongly to smoking cues are more likely to successfully quit (Ferguson and Shiffman, 2009), identifying an effective means to combat cue-induced craving may contribute to a substantial increase in cessation success. At present, however, smoking cessation strategies may benefit from advising individuals attempting cessation to avoid smoking-related cues.

While no robust sex differences were observed in the current study, there was a trend toward blunted cue reactivity in females who were told that they had received a denicotinized cigarette. These findings are consistent with previous work suggesting that females may be more reactive to non-pharmacological manipulations, including expectancy and cue exposure, than males (Barrett, 2010; Field and Duka, 2004; Perkins et al., 1999; Rose, 2006; Schlagintweit et al., 2014); however, the present findings should be interpreted with caution due to the realtively small sample size and marginal statistical significance.

Nicotine administration and expectancy components of acute tobacco smoking were independently associated with an immediate reduction in intention to smoke and an increase in ratings of subjective liking of the cigarettes, while nicotine administration alone was associated with reduced withdrawal-related craving. Consistent with these findings, Darredeau and Barrett (2010) demonstrated that nicotine content instructions associated with inhaler use led to reduced intention to smoke but not withdrawal-related craving, while Darredeau et al. (2013) found that nicotine-containing, but not denicotinized, cigarette smoking was associated with reduced withdrawal-related craving, while expectancy manipulations had no impact. Taken together, these findings suggest that both psychological and pharmacological manipulations may impact the rewarding aspects of smoking behavior, yet pharmacological manipulations may be necessary to impact withdrawal and negative reinforcement.

Whether these effects are nicotine-specific or a result of interactions between non-nicotine tobacco constituents and nicotine remains unknown. Intravenously administered nicotine has been found to reduce withdrawal symptoms (Rose et al., 2000), yet the extent to which these effects were due to expectancy versus nicotine pharmacology remains unclear, as expectancy was not manipulated independently of nicotine pharmacology. Other research suggests that nicotine-containing tobacco may be more effective than nicotine administered via NRT in reducing withdrawal-related craving (Barrett, 2010), which is consistent with the possibility that non-nicotine tobacco constituents may contribute to withdrawal-relieving properties of smoking (Harris et al., 2010; Hoffman and Evans, 2013). Additional research is needed to clarify the various pharmacological and non-pharmacological mechanisms that underlie craving and withdrawal reduction associated with nicotine-containing and denicotinized tobacco use.

The findings of this study should be considered in light of the following limitations. The nicotine content of the nicotine-containing cigarettes used in the study was lower than that of most marketed brands, which may have limited findings of nicotine effects. However, the observed nicotine effects and increased heart rate following nicotine-containing cigarette administration suggest a pharmacologically active dose. In addition, the same brand of nicotine-containing cigarettes have been previously shown to reduce withdrawal-related craving and increase subjective satisfaction and stimulation compared to denicotinized cigarettes, suggesting that they have sufficient nicotine to yield prototypical smoking effects (Barrett, 2010; Darredeau et al., 2013). It is possible, however, that consumption of participants’ preferred brand of cigarettes may have been more effective in preventing cue-induced craving. Second, the design of this study allowed for the examination of the effects of nicotine as well as of nicotine content information, but not for a direct comparison between these, since this would have required a comparison of within-subject and between-subject effects. While the present study had greater power to detect expectancy effects than pharmacology effects, it is noteworthy that several nicotine-specific effects were observed. In addition, previous balanced placebo studies using similar designs and sample sizes have identified significant effects of both within and between subject factors (Darredeau and Barrett, 2010; Hughes et al., 1989), suggesting that the current study is adequately powered. Finally, the presentation of the neutral cue prior to the smoking cue raises the possibility of order effects. The decision not to counterbalance the cues was made in an effort to avoid carryover effects (Sayette et al., 2010) and because the cues were presented in close proximity to each other (within 5 minutes) it is unlikely that any differences observed can be attributed to the mere passage of time.

Conclusions

In conclusion, findings demonstrate that while both nicotine content and expectancy contribute to the liking of cigarettes and reduced intention to smoke, nicotine content appears to have a unique influence on relief from withdrawal-related craving. Neither nicotine content nor expectancy were observed to prevent craving following exposure to smoking-associated stimuli, suggesting that cue-induced craving is resistant to acute pharmacological and psychological manipulations. Overall, findings suggest that cue-induced craving persists regardless of acute nicotine-containing or denicotinized tobacco smoking, suggesting that another mechanism is responsible for the therapeutic benefits ascribed to denicotinized tobacco use.