Negative affect among daily smokers: A systematic review and meta-analysis

• Published in: Journal of affective disorders 2020-09, Vol.274, p.553-567
• Main Authors: Akbari, Mehdi, Hasani, Jafar, Seydavi, Mohammad
• Format: Article
• Language: English
• Subjects: Adolescent; Adult; Cross-Sectional Studies; Daily smokers; Female; Humans; Male; Meta-analysis; Negative affect; Prospective Studies; Smokers; Smoking; Smoking - epidemiology; Systematic review; Tobacco Smoking
• ISSN: 0165-0327; 1573-2517
• DOI: 10.1016/j.jad.2020.05.063

•This paper synthesizes cross-sectional and longitudinal studies on negative affect (NA) and smoking.•The effect size (ES) between NA and smoking was 0.11 (95% CI 0.071–0.15).•The ES in longitudinal studies was more significant compared to cross-sectional studies.•The ES was higher among adolescents than adults.•Based on gender proportion, ES seems higher among males.The ES was higher in light-to-moderate smokers than in moderate-to-heavy smokers. Negative Affect (NA), as a personality trait is a tendency towards experiencing a more negative emotion. The body of research suggests that NA encourages smoking relapse and smoking as a reason for NA reduction, though. The likelihood of this connection does not seem to be bright yet. The present study critically reviews researches to synthesize the existing literature to determine the strength of this linkage. Key-word related research was systematically searched in PubMed, PsychINFO, Science Direct and Google Scholar for studies conducted from 1980 to 2019, followed by, the assessment and selection of retrieved studies based on defined inclusion criteria. A random-effects meta-analysis model was used to examine the prospective relationship between NA and smoking. Meta-regression was also used to dig for possible explanations of heterogeneity. Furthermore a multi-moderators model and sub-group analyses examined the moderating factors. Forty effect-sizes comprising 12 cross-sectional studies, 28 longitudinal studies and 24,913 participants were included in the meta-analysis. The forest plot of the pooled correlation effect size in the random model indicates a significant effect size of the relationship between NA and smoking (r = 0.11; 95%CI 0.071–0.15, P = 0.001) in the meta-analysis with high heterogeneity (Q = 473.916; df=39; P = 0.001; I2=91.77%). Also, the pooled effect size was obtained as 0.143 (95%CI 0.071–0.214) for light-to-moderate and 0.112 (95%CI 0.057–0.166) for moderate-to-heavy smokers, with the effect size ranging from 0.061 to 0.195 which was significant among all subtypes, though this trend seem higher among adolescents, males, and longitudinal studies than in adults, females, and cross-sectional studies. The review was limited to English articles, and the heterogeneity of the studies were high. These results support the notion that NA was positively and weakly linked to smoking and this linkage is stronger in light-to-moderate smokers, males, and adolescents. Theoretical and clinical implicati