Background and aims: The high level of circulating insulin in diabetic patients and the potential proliferative properties of insulin as a result of its cross interaction with insulin-growth factor-I (IGF-I) receptors suggest a proto-oncogenic role on the colonocyte. The aim of our study was to determine whether poor control of diabetes mellitus (DM) is associated with increased prevalence of colonic adenomatous polyps (APs), especially those that are advanced .

Methods: A retrospective review of all patients with type-2 DM diagnosed with APs from 1996 to 2006 was performed. Hemoglobin A1c (HbA1c) levels were evaluated as an index of glycemic control over the year that preceded the diagnosis of APs. A total of 33 factors were assessed in the patients grouped as well controlled (HbA1c < 7.5%) and poorly controlled (HbA1c > or = 7.5%). Factors associated with advanced APs in each group were examined using univariate analysis (UA). Significant variables by UA were included in a stepwise logistic regression analysis to determine independent predictors of aggressive clinical behavior by the polyps. All values are presented as means +/- SE, and statistical significance was determined at P < or = 0.05.

Results: Approximately 652 patients with DM type-2 and APs were identified. UA demonstrated that patients with poorly controlled DM-2 had a significantly increased incidence of right-sided APs (P = 0.001), a greater number of APs (P < 0.005), more advanced APs (P < 0.005), a younger age of presentation (P = 0.001), a history of smoking (P = 0.05), and greater use of exogenous insulin (P = 0.01). Logistic regression, as measured by HbA1c, demonstrated that poorly controlled DM-2 independently predicted a greater prevalence of right-sided AP, a more advanced lesion at the time of presentation, a greater number of polyps, and greater use of exogenous insulin.

Conclusions: Poor glycemic control in patients with DM-2 independently predicts an aggressive clinical course for patients with APs. Small differences in HbA1c elevation and, by inference, small differences in circulating insulin levels may lead to large variations in the behavior of APs.