Cigarette smoking and diabetes mellitus.

OBJECTIVE: To assess whether an independent relationship between cigarette smoking and type 2 diabetes exists in both men and women selected from a French population, and to assess the effects of active smoking and smoking cessation on the prevalence of diabetes.
METHODS: A population-based cross sectional study in 28,409 volunteers.
RESULTS: After adjustment for age, BMI, waist-hip ratio (WHR) and alcohol, the risk of diabetes mellitus (estimated by the odds ratio) was 1.49 (1.13-1.96, P=0.004) and 1.31 (1.01-1.17, P=0.03) for current and former smoker men, respectively, as compared to non-smoker men. The risk was even higher in men aged 40 to 69. No association was found with the duration of smoking cessation. In women, the risk of diabetes associated with current smoking was much less significant [HR: 1.46 (0.92-2.22, P=0.09)], even in women aged 40 to 69 [HR: 1.60 (1.00-2.58, P=0.05)]. No relationship was found for past smoking in women. In non-diabetic men, the adjusted fasting glucose was similar in current and in non smokers, but it was higher in current smokers aged 40 to 69 (99.2 +/- 0.27 vs 98.7 +/- 0.14 mg/dl, P=0.05). It was higher in former than in non smokers (97.4 +/- 0.20 vs 96.0 +/- 0.10 mg/dl, P=0.0001), regardless of age. In non-diabetic women, the adjusted fasting glucose was lower in current than in non smokers (90.7 +/- 0.20 vs 91.4 +/- 0.12 mg/dl, P=0.0001), even in women aged 40 to 69 (93.0 +/- 0.35 vs 93.7 +/- 0.18 mg/dl, P=0.03). It was similar in former and in non smokers, regardless of the age.
CONCLUSIONS: Current and past smoking are associated with a risk of diabetes mellitus essentially in men, but much less in women, and the relationship between fasting glucose and smoking appears different in men and women. No dose-relationship between the number of cigarettes smoked and diabetes mellitus was found. Smoking cessation is not associated with a reduced risk of diabetes.