Smoking as an independent determinant of Barrett's esophagus and, to a lesser degree, of reflux esophagitis.

PURPOSE

To evaluate the role of smoking in Barrett's esophagus (BE) and erosive esophagitis (E) compared to endoscopic controls with no BE or E. Smoking is considered a cause of both BE and E, but results on this topic are quite controversial.

METHODS

Patients with BE (339), E (462) and controls (619: 280 with GERD (gastroesophageal reflux disease)-negative and 339 with GERD-positive anamnesis) were recruited in 12 Italian endoscopy units. Data were obtained from structured questionnaires.

RESULTS

Among former smokers, a remarkable upward linear trend was found in BE for all smoking-related predictors. In particular, having smoked for more than 32 years increased the risk more than two times (OR 2.44, 95 % CL 1.33-4.45). When the analysis was performed in the subgroup of subjects with GERD-negative anamnesis, the risk of late quitters (<9 years) passed from OR 2.11 (95 % CL 1.19-3.72) to OR 4.42 (95 % CL 1.52-12.8). A noticeably positive dose-response relationship with duration was seen also among current smokers. As regards E, no straightforward evidence of association was detected, but for an increased risk of late quitters (OR 1.84, 95 % CL 1.14-2.98) in former smokers and for early age at starting (OR 3.63, 95 % CL 1.19-11.1) in GERD-negative current smokers.

CONCLUSIONS

Smoking seems to be an independent determinant of BE and, to a lesser degree, of E. The elevation in risk is independent from GERD and is already present in light cigarette smokers. Smoking cessation may reduce, but not remove this risk.