410 Barrett’s Esophagus and Obstructive Sleep Apnea: Is There a Link?

Abstract

INTRODUCTION: Barrett’s esophagus (BE) is a precursor lesion for esophageal adenocarcinoma. While some studies suggest relationship between obstructive sleep apnea (OSA) and GERD. Studies linking BE to obstructive sleep apnea (OSA) remains limited. Here, we hypothesized that OSA is independently associated with presence of Barrett’s esophagus. METHODS: We used the Nationwide Inpatient Sample (NIS) database from the year 2016 for this study. We identified patient with obstructive sleep apnea and Barrett’s esophagus using ICD-10-CM codes. Patients less than 18 years old were excluded from the study. Clinical factors associated with Barrett’s esophagus (age >50, white race, GERD, male gender, hiatal hernia, obesity and smoking) were also extracted from the database. We performed multivariate logistic regression to control for confounding factors associated with BE. All statistical analyses were performed using SAS 9.4. RESULTS: Out of 6,028,636 patients, 6.4% patients had OSA in year 2016. Prevalence of Barrett’s esophagus in OSA was 0.75%. We divided patients in two groups based on diagnosis of OSA. The mean age of the OSA population was 63.2 year, 40.8% were male, and 72% were white. Proportion of patients with age more than 50 year was high in OSA group (84.7%). Smoking and hiatal hernia were equally distributed among two groups. OSA group had higher prevalence of GERD, obesity and BE as shown in Table 1. Age more than 50, male gender, and white race were associated with almost 2-fold increase likelihood of having BE as shown in table 2. Patients with hiatal hernia are 7 times more likely to have BE (OR: 7.4, P < 0.0001). For unclear reason, obesity was associated with marginally lower risk of BE (OR: 0.935, P < 0.0001). Patients with OSA had 30% more chance of having BE compared to patient without OSA (OR: 1.293, 95% CI: 1.270 – 1.317, P < 0.0001). AUC (Area Under Curve) for this model was 0.81. CONCLUSION: OSA is an independently associated with higher likelihood of Barrett’s esophagus even after adjusting for potential confounders. Further studies can be pursued to reinforce the finding so that OSA can be included as a factor for screening for BE.