ABSTRACT
INTRODUCTION: Most descriptions of tobacco withdrawal have not changed in >30 years despite new research. This meta-analysis tested whether abstinence leads to decreased positive affect (PA) because abstinence-induced symptom changes are a core feature of the tobacco withdrawal syndrome. In addition, we examined whether reduced PA was due to withdrawal (ie, temporary decrease in a "U-shaped" curve) or offset (ie, return to baseline) effect. METHODS: Our main inclusion criterion was a prospective within-participant test of change in PA during abstinence conditions among people who smoke cigarettes daily who were not using a cessation medication. Our search of PubMed, PsycINFO, and personal libraries yielded a total of 32 tests with 2054 participants. RESULTS: There was a medium effect size indicating an overall decrease in PA following abstinence from cigarettes (Cohen's d = -0.40, 95% CI = -0.30 to -0.49). There was large heterogeneity (I2 = 70.7%). Most (79%) of the 24 trials that conducted significance tests reported that reduction in PA was significant. Seven tests were adequately designed to detect a withdrawal versus offset effect. Over half (57%) displayed a U-shaped curve for abstinence-induced change in PA indicative of a withdrawal symptom rather than offset effect. CONCLUSIONS: Abstinence from cigarettes is associated with a decrease in PA. Whether low PA should be added to withdrawal measures and diagnostic criteria requires replication of the time-course of change in PA and tests of whether abstinence-induced changes in PA and negative affect occur independently. IMPLICATIONS: Though there was substantial heterogeneity among trials, our findings suggest that (1) abstinence from cigarettes decreases positive affect and (2) this decrease may represent a withdrawal effect (vs. an offset effect). However, it is unclear whether abstinence-induced losses in positive affect are independent from increased negative affect.
