114
A CASE-CONTROL STUDY TO ASSESS THE RISK OF COLONIC NEOPLASMS IN DIABETIC AND NON-DIABETIC PATIENTS UNDERGOING COLONOSCOPY
N Anand, C Ottaway, E Irvine
St. Michael’s Hospital, Toronto, Ontario
Background: Risk factors for colorectal adenoma or carcinoma include age >50 years, prior abdominal or pelvic radiation, smoking, inherited polyposis syndromes - familial adenomatous polyposis (FAP) or hereditary non-polyposis colon cancer (HNPCC), Crohn’s disease or ulcerative colitis and a positive family history of colorectal cancer. Several epidemiologic and animal studies suggest that hyperglycemia or hyperinsulinemia may increase the risk of developing colon cancer by stimulating insulin and insulin like growth factor (IGF - 1) receptors on colonic epithelial cells.
Aim of Study: To assess whether patients with diabetes are more likely to have colorectal neoplasia at colonoscopy compared with non-diabetics.
Methods: Using ICD-10 codes, all patients who had colonoscopy between January 1, 2004 and December 31, 2006 were identified. Linkage analysis was performed to identify those with or without diabetes. A case was identified as diabetic by linkage analysis, colonoscopy report or other electronic records. Controls were the first 265 consecutive subjects who did not have diabetes. Standardized demographic and medical data were extracted from charts; for patients with any polyp or cancer, the pathology report was examined to diagnose adenoma or carcinoma. Odds ratio and 95% CI were calculated to test for increased risk.
Results: Of ~ 9,000 subjects identified 366 were examined and 305 charts were eligible (29 had no colonoscopy report and 32 were flexible sigmoidoscopies). Forty patients were identified as diabetic. The average age was 63.62 yrs. ± 12.11 for diabetics and 57.74 yrs. ± 11.92 for non-diabetics. There were 6.6 controls for each case. One patient with diabetes and two without diabetes had recurrent neoplasms and were excluded from the analysis. There were 3 new neoplasms in the diabetic group and 3 in the non-diabetic group. The odds ratio of patients with diabetes having new adenomas and/or adenocarcinoma was 7.44 (95% CI 1.44 to 38.35) compared with patients with no diabetes.
Conclusions: Patients with diabetes appear to be at an increased risk of incident neoplasms compared with non-diabetics. They may require earlier or more frequent screening than average risk subjects.
Results: