Abstract
BACKGROUND AND AIMS: Mechanisms of varenicline preloading in promoting smoking abstinence have not been evaluated. Based on an extinction of reinforcement framework, we tested the hypothesis that pre-quit reductions in smoking rate mediate the effect of extended preloading on abstinence. We also tested alternative indicators of change in smoking reinforcement, as well as smoking aversion, nausea and abstinence self-efficacy as candidate mediators. DESIGN, PARTICIPANTS AND INTERVENTION: Randomized, double-blind, placebo-controlled trial (NCT03262662) comparing extended (4-week varenicline) to standard (3 weeks of placebo, 1-week varenicline) preloading, preceding 11 weeks of open-label varenicline, in 320 adults (56% female). The primary outcome was self-reported continuous smoking abstinence during the last 4 weeks of treatment, with cotinine bio-verification at end of treatment (EOT). SETTING: University at Buffalo, State University of New York, USA (part of the trial was conducted at participants' homes due to the COVID-19 pandemic). MEASUREMENTS: Candidate mediators, including smoking rate and subjective effects of smoking (reward, satisfaction, aversion), self-reported craving, withdrawal, nausea and abstinence self-efficacy, were assessed daily during the pre-quit period with ecological momentary assessment. At two laboratory visits participants completed a choice task to assess smoking reinforcement. FINDINGS: There was a statistically significant indirect effect of extended preloading on greater EOT abstinence rates through pre-quit declines in smoking rate [a*b = 0.284, 95% confidence interval (0.072,0.616)] and percent reduction in smoking across the pre-quit period [a*b = 0.225, (0.080,0.437)]. There were also statistically significant indirect effects through reductions in pre-quit craving [a*b = 0.150, (0.01,0.420)] and increases in pre-quit self-efficacy [a*b = 0.157, (0.038,0.375)]. Sex-specific analyses suggested these mediated effects were consistently limited to females. No other candidate mediators yielded statistically significant indirect effects. CONCLUSIONS: Extended varenicline preloading mediated smoking abstinence through reduced pre-quit smoking and craving among female smokers seeking to quit; increased pre-quit abstinence self-efficacy was also a significant mediator.