Screening for Barrett's esophagus in colonoscopy patients with and without heartburn

The population prevalence of Barrett's esophagus (BE) is uncertain. Our aim was to describe the prevalence of BE in a volunteer population.

Upper endoscopy (EGD) was performed in 961 persons with no prior history of EGD who were scheduled for colonoscopy. Symptom questionnaires were completed prior to endoscopy. Biopsy specimens were taken from the gastric cardia and any columnar mucosa extending > or =5 mm into the tubular esophagus and from the stomach for H. pylori infection in the last 812 patients.

The study sample was biased toward persons undergoing colonoscopy, males, and persons with upper GI symptoms. The prevalence of BE was 65 of 961 (6.8%) patients, including 12 (1.2%) with long-segment BE (LSBE). Among 556 subjects who had never had heartburn, the prevalences of BE and LSBE were 5.6% and 0.36%, respectively. Among 384 subjects with a history of any heartburn, the prevalences of BE and LSBE were 8.3% and 2.6%, respectively. In a univariate analysis, LSBE was more common in those with any heartburn vs. those with no heartburn (P = 0.01), but the sample size was insufficient to allow multivariate analysis of predictors of LSBE. In a multivariate analysis, BE was associated with increasing age (P = 0.02), white race (P = 0.03), and negative H. pylori status (P = 0.04). Overall, BE was not associated with heartburn, although heartburn was more common in persons with LSBE or circumferential short segments.

LSBE is very uncommon in patients who have no history of heartburn. SSBE is relatively common in persons age > or =40 years with no prior endoscopy, irrespective of heartburn history.