Context: Observational studies have suggested an association between active smoking and the incidence of type 2 diabetes.
Objective: To conduct a systematic review with meta-analysis of studies assessing the association between active smoking and incidence of type 2 diabetes.
Data sources: A search of MEDLINE (1966 to May 2007) and EMBASE (1980 to May 2007) databases was supplemented by manual searches of bibliographies of key retrieved articles, reviews of abstracts from scientific meetings, and contact with experts.
Study selection: Studies were included if they reported risk of impaired fasting glucose, impaired glucose tolerance, or type 2 diabetes in relationship to smoking status at baseline; had a cohort design; and excluded persons with diabetes at baseline.
Data extraction and data synthesis: Two authors independently extracted the data, including the presence or absence of active smoking at baseline, the risk of diabetes, methods used to detect diabetes, and key criteria of study quality. Relative risks (RRs) were pooled using a random-effects model. Associations were tested in subgroups representing different patient characteristics and study quality criteria.
Results: The search yielded 25 prospective cohort studies (N = 1.2 million participants) that reported 45 844 incident cases of diabetes during a study follow-up period ranging from 5 to 30 years. Of the 25 studies, 24 reported adjusted RRs greater than 1 (range for all studies, 0.82-3.74). The pooled adjusted RR was 1.44 (95% confidence interval [CI], 1.31-1.58). Results were consistent and statistically significant in all subgroups. The risk of diabetes was greater for heavy smokers (> or =20 cigarettes/day; RR, 1.61; 95% CI, 1.43-1.80) than for lighter smokers (RR,1.29; 95% CI, 1.13-1.48) and lower for former smokers (RR, 1.23; 95% CI, 1.14-1.33) compared with active smokers, consistent with a dose-response phenomenon.
Conclusion: Active smoking is associated with an increased risk of type 2 diabetes. Future research should attempt to establish whether this association is causal and to clarify its mechanisms.