Background: To investigate the independent effects of smoking cessation on the risk for developing type 2 diabetes mellitus (DM) in the general population.
Design: An 8-year prospective study.
Methods: This prospective study started with baseline examinations in 1990 and 1992, and continued with follow-up examinations every 2 years up to 1998 and 2000. A total of 27 635 nondiabetic men, aged 35-44 years were classified as 5701 nonsmokers, 7477 ex-smokers and 14 457 sustained smokers, based on repeated self-reported questionnaires in 1992, 1994 and 1996. Baseline fasting serum glucose level and other risk factors were measured in 1990 and 1992. The outcome was newly developed DM, defined as a fasting glucose level > or =7.0 mmol/l in 1998 and 2000 (averaged).
Results: Over the 8 years, 1170 men (4.2%) developed DM. When compared with nonsmokers, the fully adjusted risk ratio of ex-smokers and sustained smokers for diabetes was 1.22 [95% confidence interval (CI), 0.96-1.55] and 1.60 (95% CI, 1.29-1.97), respectively. Among the ex-smokers, the risk for diabetes differed according to the quit-smoking period. Compared with nonsmokers, the fully adjusted risk ratio for DM in men who quit smoking before 1992, during 1992-1993 and during 1994-1995 was 0.95 (0.72-1.25), 1.44 (0.96-2.15) and 2.13 (1.51-3.00), respectively, after adjustment for age, baseline fasting serum glucose, weight change, baseline body mass index, family history of DM, alcohol consumption and exercise status.
Conclusions: Our results support cigarette smoking as an independent and modifiable risk factor for DM. Early smoking cessation could decrease the risk for developing DM to that of nonsmokers in the long term.