2.2 Participants

Twelve adults (5 men; mean age of 36.3; range 21 to 64 years) took part of the study. One participant was categorized as “heavy smoker” (25> cigarettes per day), seven as “moderate smokers” (15–24 cigarettes per day) and four as “light smokers” (<15 cigarettes per day; Wilson et al., 1999). They were all in the contemplator stage, as assessed by the Prochaska and DiClemente questionnaire (Prochaska and DiClemente, 1983), indicating they wanted to quit smoking. Participants had no history of neurological or psychiatric disorders other than addiction for nicotine smoking and had normal physical and neurological exams. One participant (subject 2) was taking lisinopril for high blood pressure. Participants were also assessed on nicotine dependence (Fagerström Test for Nicotine Dependence; Fagerstrom et al., 1990) impulsivity (Barratt Impulsiveness Scale; Patton et al., 1995), mood (Beck Depression Inventory; Beck et al., 1996), sleep (Pittsburgh Sleep Quality Index; Buysse et al., 1988), food habits (General Craving Food Questionnaire-Trait; Cepeda-Benito et al., 2000), and handedness (Edinburgh Handedness Inventory). None of the participants dropped out of the study. Demographic and clinical measures are presented in Table 1.

Participants were screened for noninvasive brain stimulation contraindications (Wasserman, 1998; Rossi et al., 2009). All were naive to brain stimulation, the decision-making tasks, and were not informed about the experimental variables related to the cognitive tasks tested. Written informed consent was obtained from participants prior to inclusion in the study, which was approved by the Institutional Review Board of Beth Israel Deaconess Medical Center. All stimulation sessions were conducted with strict adherence to current safety guidelines and recommendations at the Harvard-Thorndike Clinical Research Center and the Berenson-Allen Center for Noninvasive Brain Stimulation (Beth Israel Deaconess Medical Center, Harvard Medical School). A standardized questionnaire on side effects of stimulation (modified after Keel et al., 2001) and mood (as previously used in Fregni et al., 2008) were administrated following each stimulation session. All side effects reported by participants are listed in Table 2.

2.3 Smoking intake with the reported number of cigarettes smoked and carbon monoxide measurement

We measured the number of cigarettes smoked using a daily calendar (Boggio et al., 2010). Subjects were required to complete the calendar at the end of every day during each of the two arms (nine days for each arm). We measured levels of carbon monoxide before stimulation (Day 1) and immediately after stimulation (Day 5) using a carbon monoxide monitor (piCO+ Smokerlyzer; Bedfont, Scientific Ltd., Kent, UK).

2.4 Nicotine craving

Level of nicotine craving was characterized with a cue-provoked paradigm. We used videos displaying individuals smoking employed previously (Fregni et al. 2007; Boggio et al., 2010). In order to rate level of craving, we used the standardized Questionnaire of Smoking Urges from Tiffany and Drobes (1991). This questionnaire consists of four different subscales: Desire to smoke, Anticipation of positive outcome, Relief from negative affect, and Intention to smoke. This paradigm was administrated before stimulation (Day 1) and immediately after the last stimulation session (Day 5).

2.5 Ultimatum Game

Subjects performed two versions of the Ultimatum Game. The two versions were identical with the only difference of the type of reward. One version consisted of the original task in which subjects were rewarded with money (Guth et al., 1982), and in the modified version, subjects were rewarded with cigarettes. In this task, a proposer offers the subject to split a certain amount of money ($10). The subject can either accept or reject the offer. If he accepts the offer, the money is split as proposed, but if he rejects, none of them receive the money/cigarette. Most individuals reject low offers ($1 and $2) to punish the proposer as they consider the offer unfair and this punishment drive overwhelms self-centered motives (e.g., Sanfey et al., 2003; Knoch et al., 2006b). Although smokers show the same punishment behavior for money, they however accept unfair offers of cigarettes, showing greater self-centered motives in that context (Takahashi, 2007). The main outcome in this study was the acceptance rate. Data were analyzed for unfair offers ($1 and $2; 1 and 2 cigarettes), midfair offers ($3 and 3 cigarettes), and fair offers ($4 and $5; 4 and 5 cigarettes) for both the money and cigarettes version of the Ultimatum Game. Rewards were not real. There were 150 trials for each version. For each tDCS arm, subjects performed the two versions of the Ultimatum Game before stimulation (Day 1) and immediately after the last stimulation session (Day 5).

2.6 Risk Task

Subjects performed two versions of the Risk Task. The two versions were identical with the only difference being the type of reward. One version consisted of the original Risk Task in which subjects were rewarded with points (Rogers et al., 1999; Fecteau et al., 2007a). A modified version was also administrated, in which cigarettes constituted the reward. In both tasks, subjects were presented with six boxes and each box could either be pink or blue. The ratio of pink and blue boxes varies from trial to trial and can be 5:1, 4:2, or 3:3. Subjects have to pick the color of the box that hides the winning token. For each trial, the ratio of pink to blue boxes (level of risk) effectively determines the probability of findings the winning token and thus the level of risk of the choice. The winning color and levels of risk and reward changed from one trial to another, disrupting any possible winning stratagem or tactic. Subjects were thus not aware of any winning strategy to maximize their earnings and were not instructed in a way or another to adopt such a strategy. Subjects were rewarded with an amount of points or cigarettes for correctly guessing the color of the box hiding the winning token and punished by loosing points for picking the incorrect color. The amount of reward (or penalty) associated with any scenario varies (balance of reward) and is indicated on the screen. Importantly, the conflict inherent in risk taking is reflected by the fact that the largest reward is always associated with the least likely of the two outcomes (i.e., the most risky option). Also, it has been shown that addicts take more risk at the Risk Task as compared to nonaddicts (Rogers et al., 1999). The outcome measure in this study was the choice of low risk versus high risk options. The choice was calculated as the number of times subjects choose the low risk option on all trials excluding the neutral conditions (i.e., equal number of blue and pink boxes). Rewards were not real. There were 96 trials for each version. For each tDCS arm, subjects performed the two versions of the Risk Task at two time points: before tDCS (Day 1) and immediately after the last tDCS session (Day 5).

2.7 Transcranial Direct Current Stimulation

Each subject received the two types of stimulation conditions (active and sham) in a counterbalanced order. Studies on smokers have reported greater beneficial effects on smoking behaviors associated with anodal stimulation over the right DLPFCs coupled with cathodal stimulation over the left DLPFC. We therefore focused on this electrode arrangement (referred here as ‘right anodal/left cathodal’). This arrangement is presumed to enhance excitability in the right DLPFC and decrease it in the left DLPFC. A second stimulation condition was a sham stimulation control with the same electrodes arrangement as in the active stimulation (anodal over the left and cathodal over the right DLPFC). Direct current was induced by two saline-soaked surface sponge electrodes (35 cm²) and delivered by a batterydriven, constant current stimulator (neuroConn, Germany). For active stimulation, participants received a constant current of 2mA intensity during 30 minutes. The anode electrode was placed over F4 (international EEG 10/20 system) and the cathode electrode over F3. For sham stimulation, the electrodes were placed at the same position as for active stimulation (F4 and F3), but the stimulator was turned on to 2 mA only for the first 30 seconds and at the last 30 seconds of the 30-minute period so participants felt the itching sensation associated with ramping up and down the current, but received no active current otherwise. This method of sham stimulation has been shown to be reliable, at least for a 1 mA current (Gandiga et al., 2006), though may not totally blind participants when 2 mA is applied. Nonetheless, when explicitly asked at the end of the study, none of the fourteen participants believed to have undergone sham or active stimulation with a confidence level greater than 50%, suggesting successful blinding of the stimulation conditions.

3.2 Nicotine craving

We tested whether stimulation had an effect cue-induced smoking craving using Wilcoxon test between active and sham stimulation conditions (change from baseline at Day 1 to post-stimulation at Day 5) for each of the four subscales of the standardized Questionnaire of Smoking Urges (Tiffany and Drobes, 1991). There was a significant difference for the Desire to smoke scale (p=.02), but not for the remaining scales: Anticipation of positive outcome (p=.37), Intention to smoke (p=.33), and Relief from negative affect (p=.64).

3.3 Ultimatum Game

In order to test whether stimulation had an effect on performance at the Ultimatum Game, the percent of change of acceptance rate of offers between baseline at Day 1 and post-stimulation at Day 5 was entered as dependent variable with stimulation (active, sham) and type of offer (unfair, mid-fair, fair) as independent fixed variables for each version of the Ultimatum Game. We analyzed the graphic representation of performance data from the two experiences (monetary reward and cigarettes) to ensure data fitted the normal distribution, and proceeded with two-way ANOVAs. For the Ultimatum Game with cigarettes as reward, there was an effect of stimulation (F(1,11)=6.932, p=.02), but no effect of type of offer (F(2,22)=.812; p=.46). Please refer to Figure 2 for an illustration and Table 4 for raw data. For the Ultimatum Game with money as reward, there were no effects of stimulation (p=.10) or type of offer (p=.20).

Open in a separate window

Figure 2

Group mean percent changes in the choice of accepting offers of cigarettes at the Ultimatum Game for each type of offer and stimulation condition (active and sham). Bars represent Standard Error of Mean; * p<0.05. Active stimulation is represented with grey bars; sham stimulation is represented with black bars.

4.1 Limitations

There are some limitations in this work. First, the present results comes a small sample size. There is a need to replicate these findings with a larger sample of subjects. Second, we failed to find a significant change in level of carbon monoxide. Subjects were thus smoking a significant lesser number of cigarettes but did not equally reduce the level of carbon monoxide (although one might consider there was a trend: p =.06), suggesting that they might have inhaled more each cigarette when they received active as compared to sham stimulation. Another limitation of this work is that we measured the level of certainty of the administered stimulation condition (active, sham) with a visual analog scale, but we did not measure accuracy per se, which might have resulted in different judgment on blindness of the stimulation condition. Finally, it should be noted that the rewarding cigarettes were not actually given to the subjects after the completion of the tasks. We made this decision because first we believe that providing smokers who wished to quit with real cigarettes was not ethical. Second, real and virtual rewards appear to be processed in a similar way. For instance, it has been shown that real and virtual reward induced similar behavioral responses (Bowman and Turnbull, 2003) and brain activation patterns (Miyapuram et al., 2012). Future work should further characterize smokers as slow and fast metabolizers using H3/cotinine ratios as this might influence the findings. Future work should also include neuroimaging measures to identify the neural substrates that are modulated by tDCS and associated with decreased in the number of cigarettes smoked. Here, we cannot state whether both electrodes or either the anodal or cathode electrode caused the observed effects. We can only say that these effects were found when comparing active to sham stimulation by applying the anode over F4 and cathode over F3.

We would like to thank our participants and Dr. Jean Leblond from the CIRRIS for statistical advice.

Funding

This work was supported by grants from the Canada Research Chair and Canadian Institutes of Health Research to SF and by the Harvard-Thorndike Clinical Research Center at Beth Israel Deaconess Medical Center integrated in the Harvard Clinical and Translational Science Center, supported by grants M01-RR- 01066 and UL1 RR025758 from the National Center for Research Resources; by grants from the National Institutes of Health (K24RR018875) to APL.