Sa1076 Abdominal Diameter Index Is a Stronger Predictor of Barrett's Esophagus Than BMI or Waist-to-Hip Ratio

Abstract

Background: Abdominal obesity is associated with the development of gastroesophageal reflux disease (GERD) and, subsequently, Barrett's esophagus (BE). Increased body mass index (BMI) and waist-to-hip ratio (WHR) have individually been associated with BE; however, other anthropometric measurements exist and may be more accurate. Abdominal diameter index (ADI, sagittal abdominal diameter divided by thigh circumference) was previously shown to be a more accurate predictor of incident cardiovascular disease compared to other body measurements. Our aim was to examine whether ADI was a more accurate predictor of prevalent BE compared to other anthropometric measurements. Methods: We conducted a case-control study of patients presenting to our institution from October 2013November 2014. Our study population was consecutive Caucasian men with a known history of BE confirmed by endoscopy and histology. We recruited Caucasian male control patients who underwent endoscopy for any reason and who did not have evidence of BE by history or endoscopy. Prior to endoscopy or outpatient visit both groups completed a questionnaire about demographics, smoking status, and medication use and underwent a series of body measurements including height, weight, waist circumference, hip circumference, thigh circumference, and sagittal abdominal diameter using standardized measuring tools. BMI, WHR, and ADI were calculated, and the data was analyzed using SPSS 22.0. Results: A total of 31 BE patients and 42 control patients were recruited. The BE cohort were older (mean age 62.5 vs. 53.2 yrs, p=0.009) and had a higher rate of hiatal hernia (74.2% vs. 19.0%, p<0.001) and proton pump inhibitor use (90.3% vs. 69.0%, p=0.03). The mean ADI for patients with BE was 0.65 ± 0.07 and without BE was 0.59 ± 0.08 (p= 0.01). In univariate analysis, an ADI≥0.60 vs. <0.60 conferred an increased risk of BE (OR= 3.8, 95% CI=1.42-10.10). When controlling for age, history of tobacco use, and BMI, an ADI≥0.60 remained a significant independent risk factor for BE (OR=3.0, 95% CI=1.078.55). Of note, BMI was not a significant predictor of BE. Similarly, WHR was not associated with BE in either univariate or multivariate analysis. The predictive value of ADI was analyzed using a receiver-operator characteristic (ROC) curve and was a more powerful predictor of BE than WHR or BMI (AUROC=0.72 vs. 0.60 vs. 0.52, respectively). Using a cut-point ADI value of 0.60, ADI had a sensitivity of 77.4% and a specificity of 64.3% for the presence of BE. Conclusion: ADI appears to be a more powerful predictor of the presence of BE than BMI and WHR. ADI may be a better measure of central obesity than WHR. In the future, physicians may be able to use ADI to help risk-stratify those patients that should undergo screening for Barrett's esophagus.