The study included a randomised between-subjects design. Participants were randomised to either passive exposure to combustible cigarette use (n=30; 13 female) or e-cigarette use (n=30; 13 female). Study candidates were recruited by online advertisements for a 2 h study described as ‘assessing mood response following exposure to common tasks and social interactions’ (see online supplementary appendix for details). This general description was chosen to reduce expectancy of cigarette and e-cigarette exposure. Inclusion criteria were age between 18 and 35 years, daily smoking of 5–18 cigarettes per day, not having any major medical or psychiatric disorders excluding nicotine dependence, and not currently trying to quit smoking. Candidates (N=69) arrived between 10:00 and 15:00 h, completed informed consent and underwent screening that included interviews and surveys on background characteristics, the Fagerström Test for Nicotine Dependence (FTND)17 (link) and a modified non-patient version of the Structured Clinical Interview for DSM-IV.18 Abstinence from alcohol and recreational drugs was required for at least 24 h prior to the study, and an alcohol breath test was used to verify current sobriety (0.000 mg%). Smoking abstinence for at least 2 h prior to arrival was also required and verified by an expired air carbon monoxide reading of ≤15 ppm.
Eligible participants (60/69; 88%) were informed that the 1 h study session would immediately follow the screening and would include completing computerised surveys before and after engaging in two randomly assigned tasks for 5 min, separated by a 10 min break. The tasks were described as engaging with a second participant in conversation, viewing pictures, eating food, drinking a beverage or smoking. In fact, task selection at both intervals was predetermined for the study participant to engage in conversation with another participant and choose the topic from the list provided, that is, talk about the weather, pets, places to eat, movies and television, local landmarks or vacations. In addition, the other participant was a study confederate predetermined to drink water for his/her first task (ie, control cue) and then to smoke either an e-cigarette that is visually similar to a regular cigarette (NJOY King) or a combustible cigarette (American Spirit or Benson and Hedges) (ie, active cues) for the second task. After each cue, participants completed the digit symbol substitution task19 to maintain concentration and mask the focus of the study. Upon study completion, each participant was debriefed and paid US$30. The study was approved by the University of Chicago Institutional Review Board.
Videotapes of the exchange between the confederates (one female and one male, ages 21 and 24 years, respectively) and the participants were later scored by two independent raters to ascertain the quality of the interactions using the Two-Dimensional Social Interaction Scale.20 (link) Results showed no differences in participant–confederate interactions in the e-cigarette versus combustible cigarette cue groups (ps≥0.16).
Measures were given at baseline (time 0), following the control cue (15 min) and following the randomised active cue (35 and 50 min). The main dependent measures were two visual analogue scale (VAS) items for Desire to smoke an electronic cigarette and Desire to smoke a regular cigarette (your preferred brand) anchored from ‘not at all’ (0) to ‘most ever’ (100)21 (link) and the Brief Questionnaire of Smoking Urges (BQSU)22 (link) with 10 items rated from 1 (strongly disagree) to 7 (strongly agree) and summed for a total score.22 (link) Additional VAS items were included to mask the focus on smoking (see online supplementary appendix for details). Data were analysed by 2 group (active cue type)×4 time (0, 15, 35, 50 min) analyses of variance. Significant main effects or interactions were analysed by simple effects tests.