Aims: To investigate the association between smoking and leptin, and to discuss their influence on diabetes in a large-scale study of Japanese men.
Methods: A cross-sectional study was carried out in 2002. The subjects were 2836 men aged 35-66. Smoking history was investigated in a self-administered questionnaire. Blood leptin, glucose and insulin were measured.
Results: Significant differences in leptin levels and homeostasis model assessment of insulin resistance (HOMA-IR) related to smoking status were observed (P=0.001 and P=0.008, respectively). The multivariate-adjusted geometric means of leptin in current, past and never smokers were 3.88, 4.08 and 4.12 ng/ml, respectively, while the means of HOMA-IR were 1.64, 1.61 and 1.49, respectively. The age-, body mass index-, and other lifestyle-adjusted prevalences of diabetes in current and never smokers were 9.2 and 4.7%, respectively. That of current smokers was significantly higher than in never smokers (P<0.001). The dose-dependent association found between the intensity of smoking and leptin levels in current smokers was statistically significant (P=0.030).
Conclusions: The present finding may explain in part an association among smoking, leptin levels and diabetes. Smoking is one of the important modifiable risk factors for the prevention of diabetes.