The association between smoking prevalence and eating disorders: a systematic review and meta-analysis

Abstract

BACKGROUND AND AIMS: Cigarette smoking is associated with severe mental illness, including schizophrenia and bipolar disorder, and with morbidity and mortality, but the association with anorexia (AN), bulimia nervosa (BN) and binge eating disorder (BED) is unclear. This meta-analysis compared the odds of smoking in eating disorders (ED) (ED = AN or BN or BED) versus healthy controls (HC) and calculated the prevalence of smokers in people with ED. METHODS: Three independent authors searched PubMed, MEDLINE and Scopus from database inception until 31 December 2015 for studies reporting data on life-time or current smoking prevalence in BED, BN and AN with or without control group. Meta-analyses were undertaken, calculating odds ratios (ORs) of life-time smoking in BED, BN, AN versus healthy controls (HCs) or prevalence of smoking in BED, BN and AN with 95% confidence intervals (CI). RESULTS: Thirty-one studies (ED = 8517, controls = 68 335) were meta-analysed. Compared with HCs, there were significantly more smokers among people with BN (life-time OR = 2.165) and BED (life-time OR = 1.792) but not AN (life-time OR = 0.927). BED was associated with smoking the most (life-time prevalence = 47.73%) followed by BN (life-time prevalence = 39.4%) and AN (life-time prevalence = 30.8%). In BN, life-time smoking prevalence was highest in Europe. In AN, higher age moderated both life-time and current smoking prevalence, and body mass index moderated higher life-time smoking prevalence. In BN, female sex moderated higher life-time smoking prevalence. CONCLUSIONS: People with binge eating disorder and bulimia nervosa are significantly more likely to be life-time smokers than healthy controls, which is not the case for anorexia nervosa.