Abstract
BACKGROUND: Serum total bilirubin (TB) was used as a predictor of diabetes mellitus (DM) and this study was undertaken to investigate the relationship between serum TB and DM. METHODS: In this cross-sectional study, a total of 3,867 subjects older than 65 years were recruited from East China. The anthropometric data, lifestyle and past history were collected. The fasting blood glucose, total cholesterol (TC), triglycerides (TGs), TB and alanine aminotransferase (ALT) were detected. The prevalence of DM was calculated for every serum TB quartile. Logistic regression analysis was employed to evaluate the relationship between serum TB and DM. RESULTS: Serum TB was significantly higher in non-DM than DM patients (P=0.001). Serum TB was negatively related to the prevalence of DM (P=0.004). Logistic regression analysis revealed that serum TB was an independent prognostic factor of DM [odds ratio (OR): 0.876, 95% confidence interval (CI): 0.807-0.951]. CONCLUSIONS: Our results showed serum TB in physiological range is an independent risk factor of DM in old people; the prevalence of DM in old people has a linear relationship with serum TB quartiles; the prevalence of DM reduces with the increasing of serum TB within physiological range.