Secondary Logo

Journal Logo

The Association of Barrett's Esophagus and Body Mass Index


Stein, Daniel J., M.D.; El-Serag, Hashem B., M.D.; Kuczynski, John, M.D.; Kramer, Jennifer R., Ph.D.; Sampliner, Richard E., M.D.

American Journal of Gastroenterology: September 2005 - Volume 100 - Issue - p S24
Supplement Abstracts Submitted for the 70th Annual Scientific Meeting of the American College of Gastroenterology: ESOPHAGUS

Section of Gastroenterology, Southern Arizona VA Health Care System, Tucson, AZ and Sections of Gastroenterology and Health Services Research, Houston Center for Quality of Care and Utilization Studies, Baylor College of Medicine, Houston, TX.

Purpose: Obesity is a recognized risk factor for esophageal adenocarcinoma. Its relationshiop to Barrett's esophagus (BE) is less well defined.

Aim: Examine the relationship between the presence and length of BE to weight, height, and body mass index (BMI).

Methods: A retrospective cross-sectional study of male patients undergoing upper endoscopy at SAVAHCS between 1998 and 2004 was conducted. BE was diagnosed by one endoscopist using standardized (ACG guidelines) endoscopic (columnar appearing distal esophagus) and histologic (intestinal metaplasia) criteria. The non-BE patients had normal esophageal mucosa at endoscopy. Chart review was performed for demographics, history of malignancy, death, and weight and height within 1 year of the endoscopy. Patients with BE and without BE were compared for demographic features, weight, height and BMI. Multivariate logistic regression analysis was conducted to examine the association of BMI and weight with BE, and analysis of covariance was used to examine the effect of BMI and weight on length of BE.

Results: 65 patients newly diagnosed with BE and 385 non-BE patients who did not have any recorded malignancy were compared. BE patients had a higher mean BMI (29.8 vs. 28, p = 0.03) and greater mean weight (206 lb vs. 190 lb, p = 0.005). There was no significant difference in height. Multivariate logistic regression adjusted for race and age showed an association between BE and BMI, both when categorized and continuous BMI: 25–30 (vs. <25) odds ratio 2.38, p = 0.03 and BMI >30 (vs. <25) odds ratio 2.45, p = 0.03. For each 10 pound increase in weight or 5-point increase in BMI there was a 10 and 35% increased risk of BE, respectively (p = 0.002 and p = 0.01). There was no significant correlation between the length of BE and BMI.

Conclusions: Obesity is associated with a greater than 2-fold increase in the risk of BE, providing another possible risk factor in the development of BE. BMI was not associated with BE length.

© The American College of Gastroenterology 2005. All Rights Reserved.