Smoking Cessation, Weight Gain, and Risk of Type 2 Diabetes Mellitus Among Postmenopausal Women

Cigarette smoking is associated with an increased risk of type 2 diabetes mellitus.¹ However, smoking cessation is often accompanied by weight gain, which may explain the increased risk of diabetes that has been observed in several studies.^2-6 Two studies with data on weight came to different conclusions about whether the increased risk of diabetes after smoking cessation is primarily attributable to postcessation weight gain.^3,4 We used data from the Women's Health Initiative,⁷ a large prospective study with detailed information on smoking status, weight changes, and potential confounders, to assess the relationship between smoking cessation, weight gain, and subsequent diabetes risk. We examined diabetes risk by smoking status, including new quitters who smoked at baseline but no longer smoked at the 3-year follow-up visit.

A total of 115 092 women without known diabetes were followed up from year 3 to diabetes diagnosis, date of death, loss to follow-up, or September 30, 2010, whichever occurred first. The definition of incident diabetes was a positive answer to questions regarding “newly prescribed treatment for diabetes with pills or insulin shots” or using “diet and/or exercise for diabetes” on any of the semiannual or annual follow-up questionnaires. Self-reported diabetes in the Women's Health Initiative has been validated by medication inventories and laboratory data as a reliable indicator of diagnosed diabetes.⁸ A total of 11 056 incident diabetes cases occurred during an average of 8½ years after the 3-year follow-up visit.

The main exposure included never smokers at both baseline and 3-year follow-up visit, former smokers at both baseline and follow-up visit, continuing smokers at both baseline and follow-up visit, and new quitters who smoked at baseline but were abstinent at the follow-up visit. A small proportion of women (0.6%) whose smoking status changed from never or former smokers at baseline to current smokers in year 3 were excluded. Weight was measured at both baseline and year 3 in 107 471 women, 10 380 of whom developed diabetes.

Cox proportional hazards regression models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) of diabetes risk by smoking status overall and stratified by weight gain (categorized as <5 kg or ≥5 kg). In multivariable models, we adjusted for potential baseline confounders, including age at enrollment, race/ethnicity, education, body mass index, waist circumference, physical activity, alcohol intake, treatment for hypertension and high cholesterol levels, and participation in different Women's Health Initiative study cohorts (observational study or clinical trials and different treatment assignments for all 4 clinical trials).

Compared with never smokers, the HR for incident diabetes was 1.00 (95% CI, 0.96-1.04) for former smokers, 1.20 (95% CI, 1.09-1.31) for continuing smokers, and 1.43 (95% CI, 1.26-1.63) for new quitters after confounders were adjusted for. During the first 3 years of follow-up, a weight gain greater than 5 kg occurred in 10.2% of never smokers, 11.6% of former smokers, 12.3% of continuing smokers, and 30.5% of new quitters (median weight gain, 2.9 kg vs 0.3-0.5 kg in all other groups of smoking status). After weight gain was adjusted for (Table), the risk of diabetes remained elevated among continuing smokers (HR, 1.20; 95% CI, 1.10 1.32) and attenuated slightly but remained significantly elevated among new quitters (HR, 1.36; 95% CI, 1.19-1.54). Among former smokers, the diabetes risk decreased significantly as the time since quitting increased, and the risk was similar to that of never smokers after a cessation period of 10 years. We examined whether the increased risk of diabetes in new quitters was explained by weight gain in the analysis stratified by weight gain (Table). Among former smokers, the risk of diabetes was not greater compared with never smokers with similar weight gain. Among continuing smokers, the risk of diabetes was significantly increased compared with never smokers, both in women who gained 5 kg or more and in those who did not (P for interaction, 0.4). Among new quitters, the risk of diabetes among women who gained less than 5 kg was similar to the risk in continuing smokers, although it was not significantly elevated in this relatively small subgroup. However, among new quitters who gained 5 kg or more, there was a 67% excess risk of diabetes compared with nonsmokers (P for interaction, .02).

Our prospective analysis over 8½ years of follow-up shows that the increased risk of diabetes associated with quitting is confined to a subgroup that gains at least 5 kg. Our data in former smokers suggest that diabetes risk is likely to return to that in never smokers after about 10 years, independent of more recent weight gain. Since weight gain after smoking cessation can be prevented by regular moderate physical activity and dietary modification,⁹ smokers should not be deterred from quitting by concerns about an increase in the risk of diabetes.

Correspondence: Dr Luo, Department of Community Medicine, Mary Babb Randolph Cancer Center, West Virginia University, PO Box 9190, Morgantown, WV 26506 (Jiluo@hsc.wvu.edu).

Author Contributions:Study concept and design: Margolis and Luo. Acquisition of data: Rossouw, Ockene, Margolis, and Luo. Analysis and interpretation of data: Rossouw, Tong, Giovino, Lee, Chen, Ockene, Qi, Margolis, and Luo. Drafting of the manuscript: Rossouw and Luo. Critical revision of the manuscript for important intellectual content: Rossouw, Tong, Giovino, Lee, Chen, Ockene, Qi, Margolis, and Luo. Statistical analysis: Qi and Luo. Obtained funding: Rossouw, Ockene, and Margolis. Study supervision: Rossouw and Margolis.