February 04, 2013
2 min read
Save

Age, abdominal obesity, tobacco use improved predictive model for Barrett’s esophagus

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

A model incorporating age, waist–hip ratio and smoking in addition to GERD symptoms more effectively predicted Barrett’s esophagus in older men than a model using GERD alone in a recent study.

In a cross-sectional study, researchers performed upper endoscopy in 822 men aged 50 to 79 years who underwent screening for colorectal cancer, in order to detect new cases of Barrett’s esophagus (BE). Patients’ weight, height and waist–hip ratio were measured and information on the severity and duration of GERD symptoms was collected.

BE was diagnosed in 70 participants, including 15.5% of those who reported weekly symptoms of GERD and 7.1% of those who did not. Independent associations were observed between BE and weekly GERD symptoms (adjusted OR=2.33; 95% CI, 1.34-4.05), advanced age (aOR=1.53; 95% CI, 1.05-2.25 per 10 years) and tobacco use (aOR=1.09; 95% CI, 1.04-1.14 per every 10 pack-years). A near-significant association was made for waist–hip ratio (WHR) (aOR=1.44; 95% CI, 0.898-2.32 per 0.10 increment; P=.13 for trend).

Investigators developed the Michigan Barrett’s Esophagus pREdiction Tool (M-BERET), a model incorporating frequency of GERD symptoms (weekly or more frequent vs. less frequent or no symptoms), pack-years of tobacco use, age and WHR. When compared via AUROC analysis with a model only incorporating frequency and duration of GERD symptoms, M-BERET had a greater AUC (0.72 vs. 0.61; P<.001).

Comparing the models via NRI index, with prevalence thresholds of less than 7% (low), 7-15% (intermediate) and more than 15% (elevated), indicated that M-BERET was 45% more accurate among those with BE, while GERD-only was 20% more accurate among those without BE (NRI index 25%; P=.007). Sensitivity analysis with altered thresholds resulted in NRI indices of 19% (10% and 20% thresholds; P<.01) and 29% (5% and 20% thresholds; P<.001).

Joel H. Rubenstein, MD, MSc

Joel H. Rubenstein

“This study highlights that risk factors in addition to GERD should be considered for identifying people at risk for esophageal adenocarcinoma,” researcher Joel H. Rubenstein, MD, MSc, assistant professor in the gastroenterology division at University of Michigan Medical School, told Healio.com. A model that uses age, smoking, and abdominal obesity in addition to GERD symptoms does a better job identifying men with Barrett’s esophagus than a model based only on GERD symptoms.”

Disclosure: See the study for a full list of relevant disclosures.