OBJECTIVE: To investigate which glycemic index is more strongly associated with colorectal neoplasia. METHOD: This cross-sectional study enrolled 2,776 participants in a comprehensive health management program which included measurement of fasting plasma glucose and HbA1c, along with screening colonoscopy. Primary outcome was colorectal adenoma with or without dysplasia. Risk factors for colorectal neoplasia were determined by the multivariate regression analysis, which evaluated the interrelationship among different glycemic indices in a hierarchical way. RESULTS: Colorectal neoplasms were found in 605 (21.79%) examinees, 68 (2.45%) of whom had high-risk tumors. Glycemic indices including diagnosis of diabetes mellitus, fasting plasma glucose, and HbA1c were all associated with colorectal tumors in the univariate analysis. However, HbA1c outperformed the other two markers as an independent risk factor (adjusted odds ratio, 1.22; 95% confidence interval, 1.10-1.36%) for colorectal neoplasia. Moreover, only HbA1c remained independently associated with colorectal tumor after patients with established diagnosis of diabetes (n = 132) were excluded. We also identified age, male gender, and smoking were independent risk factors for colorectal neoplasia. CONCLUSION: HbA1c as compared with fasting plasma glucose is more strongly and independently associated with colorectal neoplasia. Further research is warranted to elucidate the value of HbA1c in stratifying risk of colorectal cancer.
OBJECTIVE: To investigate which glycemic index is more strongly associated with colorectal neoplasia. METHOD: This cross-sectional study enrolled 2,776 participants in a comprehensive health management program which included measurement of fasting plasma glucose and HbA1c, along with screening colonoscopy. Primary outcome was colorectal adenoma with or without dysplasia. Risk factors for colorectal neoplasia were determined by the multivariate regression analysis, which evaluated the interrelationship among different glycemic indices in a hierarchical way. RESULTS:Colorectal neoplasms were found in 605 (21.79%) examinees, 68 (2.45%) of whom had high-risk tumors. Glycemic indices including diagnosis of diabetes mellitus, fasting plasma glucose, and HbA1c were all associated with colorectal tumors in the univariate analysis. However, HbA1c outperformed the other two markers as an independent risk factor (adjusted odds ratio, 1.22; 95% confidence interval, 1.10-1.36%) for colorectal neoplasia. Moreover, only HbA1c remained independently associated with colorectal tumor after patients with established diagnosis of diabetes (n = 132) were excluded. We also identified age, male gender, and smoking were independent risk factors for colorectal neoplasia. CONCLUSION: HbA1c as compared with fasting plasma glucose is more strongly and independently associated with colorectal neoplasia. Further research is warranted to elucidate the value of HbA1c in stratifying risk of colorectal cancer.