T1052 Acid-Related Upper Endoscopy Findings in Subjects with Diabetes Versus Non-Diabetics

From the few studies in the literature, it has been suggested that diabetics may be less likely to experience symptoms of gastroesophageal reflux disease but more likely to have acidrelated mucosal damage. To better quantify risk of acid-related damage among diabetics and non-diabetics undergoing esophagogastroduodenoscopy (EGD), we conducted a prospective study using the Clinical Outcomes Research Initiative (CORI) National Endoscopy Database. Methods: Data were collected at 10CORI sites on all patients undergoing EGD fromDecember 2006 to November 2007 unless evaluation was for a varices indication or feeding tube (PEG) placement. Within 30 days after EGD, eligible patients were contacted by telephone and questioned on their medical history, symptoms, medications and demographic data. Diabetics provided further details about their condition. Participants were considered to have an acidrelated finding if at least one of the following findings was entered on the CORI report: Barrett's esophagus; esophageal inflammation (unless etiology was not acid-related); healed ulcer; duodenal, gastric or esophageal ulcer; stricture; and mucosal abnormality with the sub-categorization of erosion or ulcer. Results: Of 2,090 eligible patients, 1,569 (75.1%) were included in the final analysis. 256 (16%) had diabetes, the majority being Type II (95%). Diabetics were significantly more likely to be male (65% vs 55%; p=.01), older (mean age 63 vs 58 years; p<.0001) and have a higher body mass index (mean 31 vs 28; p<.0001) than non-diabetics. The prevalence of acid reflux (self-reported) was similar among diabetics (70%) and non-diabetics (71%), however diabetics were significantly less likely to report frequent heartburn symptoms (at least 2-3 times per week) compared with nondiabetics (21% vs 30%; p=.01). Proton pump inhibitor (PPI) use did not differ between groups (56% in diabetics vs 55% in non-diabetics; p=.69), whereas diabetics were less likely to report non-steroidal anti-inflammatory drug use (11% vs 16%; p=.03). In unadjusted analyses, diabetics had similar risk for acid-related finding as non-diabetics (OR 1.09; 95% CI: 0.83, 1.42) which persisted after adjusting for gender, age, acid reflux, acid indication for EGD and PPI use (OR 1.04; 95% CI:0.79, 1.39). Conclusion: Diabetics were overrepresented in the cohort (16%) relative to the estimated prevalence of treated diabetics in the United States (7%), suggesting they may be more likely than non-diabetics to undergo EGD. However, no difference in risk of acid-related finding was observed between diabetics and non-diabetics, even after adjustment for potential confounders in multivariate models.