Significant decline in the prevalence of Barrett's esophagus among patients with gastroesophageal reflux disease.

Barrett's esophagus (BE), a premalignant condition for the development of esophageal adenocarcinoma (EAC), is a consequence of chronic gastroesophageal reflux disease (GERD). Although the incidence of EAC is increasing, a similar trend for BE is not clear. We aimed to evaluate the prevalence of newly diagnosed BE over time in a cohort of patients presenting with GERD symptoms. Information was prospectively collected between 1998 and 2015 for patients presenting to the endoscopy unit at a tertiary referral center for their index upper endoscopy for evaluation of GERD symptoms. Patients were asked to complete a validated GERD questionnaire that documents the onset of GERD symptoms (heartburn and acid regurgitation) and grades the frequency and severity of symptoms experienced. Demographic information, body mass index (BMI), and use of aspirin, nonsteroidal anti-inflammatory drugs, acid suppression therapy if any, smoking, family history, and endoscopic findings: erosive esophagitis, BE, and hiatal hernia were recorded. Patients evaluated during 1998-2003 (control) were compared with those presented in subsequent years (3-year cohorts) using chi-square test, and a multivariable logistic regression model was used to evaluate independent predictors. A total of 1109 patients were included in the analysis: mean age 56.9 years (standard deviation [SD] 12.8), 83% Caucasian, 93% male, and mean BMI 29.8 (SD 5.5). Overall, 226 (20.3%) patients were diagnosed with BE, with a mean BE length of 2.1 cm (SD 2.6). There was a significant decrease in the prevalence of BE over time from 24.3% in 1998-2003 to 13.5% in 2013-2015 (P = 0.002). During the same time period, a significant increasing trend in proton pump inhibitor (PPI) (41.7%; 1998-2003 vs. 80.2%; 2013-2015) (P < 0.001) and aspirin (ASA) use (23.7%; 1998-2003 vs. 25.9%; 2013-2015) (P = 0.034) was noted. There was also a significant reduction in cigarette smoking. In a multivariable logistic regression model for predicting the presence of newly diagnosed BE, there was a significant effect of timeframe even after adjusting for confounding variables. The results of our study indicate that there has been a steady and significant decline in the prevalence of BE in GERD patients over the last 2 decades. During this same time period, there has been an accompanying increase in the use of PPI, aspirin therapy, and a reduction in smoking, all modifiable risk factors potentially contributing to the decreasing prevalence of BE. Whether this decreasing prevalence of BE will lead to a reduction in EAC remains to be seen.