Background & aims: Obesity and diabetes mellitus are associated with an increased incidence of proximal colon cancer. Colonic adenoma that has been reported to be associated with elevated serum insulin levels and subsets of hyperplastic polyps might serve as a precursor of colon cancer. In this study, we sought to determine segment-specific associations between serum insulin levels and the prevalence of adenoma and hyperplastic polyps in the proximal and distal colon.
Methods: We studied 343 consecutive patients who underwent colonoscopy in our hospital. All medical information, including fasting serum insulin, was obtained at colonoscopy. We performed multinomial logistic regression models by using the outcome categories of none (reference), proximal-only, distal-only, and both-segment lesions for the presence of adenoma/hyperplastic polyp with serum insulin, age, gender, lifestyle characteristics, and the presence of other types of lesions as predictors. Odds ratios (ORs) and 95% confidence intervals (CIs) are provided for a 5-muU/mL increase in serum insulin.
Results: Overall, serum insulin levels were significantly associated with adenoma (OR, 1.5; 95% CI, 1.1-2.0; P = .005) and borderline associated with hyperplastic polyps (OR, 1.3; 95% CI, 1.0-1.7; P = .075). In multinomial logistic regression models, elevated serum insulin levels were significantly associated with proximal-only adenoma (OR, 1.8; 95% CI, 1.2-2.5; P = .002), both-side hyperplastic polyp (OR, 1.7; 95% CI, 1.1-2.5; P = .015), and proximal-only hyperplastic polyp (OR, 1.5; 95% CI, 1.0-2.1; P = .048) and borderline associated with distal-only adenoma (OR, 1.5; 95% CI, 1.0-2.1; P =.059) but not with distal-only hyperplastic polyp.
Conclusions: Serum insulin levels directly correlate with the presence of adenoma and hyperplastic polyps in the proximal colon and might also less strongly correlate with the presence of distal adenoma.