Body Composition and Barrett’s Esophagus
Article Outline
See “Abdominal obesity and body mass index as risk factors for Barrett’s esophagus” by Corley DA, Kubo A, Levin TR, Block G, Habel L, Zhao W, Leighton P, Quesenberry C, Rumore G, Buffler PA, on page 34.
In the current issue of Gastroenterology, Corley et al1 present an interesting and valuable investigation assessing the role of various body measures in relation to risk of developing Barrett’s esophagus, a metaplasia of the distal esophageal mucosa caused by long-standing gastroesophageal reflux (Figure 1). The critical aspect of Barrett’s esophagus is the substantially increased risk of the development of esophageal adenocarcinoma, a cancer with generally poor prognosis. Although overweight and obesity are established risk factors for esophageal adenocarcinoma,2, 3, 4 the relation between these factors, as well as other variables representing body composition and the risk of developing Barrett’s esophagus, are more sparsely studied.
Figure 1.
An endoscopic view of the distal esophagus with a mucosa representing Barrett’s esophagus distal to an irregularly shaped squamo-columnar line (Z-line). Photo courtesy of Dr. Edgar Jaramillo, Karolinska University Hospital.
Corley et al1 conducted a large, case-control study in the United States, evaluating cases with Barrett’s esophagus and 2 control groups: population-based control participants and patients with recorded gastroesophageal reflux disease (GERD) and no Barrett’s esophagus. Because Barrett’s is without any specific symptoms, it is difficult to study true incident cases of Barrett’s esophagus, and thus avoiding the inclusion of prevalent cases is a reasonable approach. The principal way to identify Barrett’s esophagus is through endoscopy conducted for various reasons, for example, reflux symptoms or other upper gastrointestinal symptoms or disorders. The investigators included only newly diagnosed cases of Barrett’s esophagus. The use of 2 control groups, representing population-based participants as well as persons with GERD, adds to the validity of the study, but the limited participation frequencies are potential threats to the validity. The low participation rate is generally an increasing problem in population-based epidemiologic research, and this source of error might have contributed to biased selection of cases or controls to participate. However, it is unlikely that such error would materially influence the overall conclusions of the investigation; body composition should not be a major reason for nonparticipation. The exposures under study in the investigation conducted by Corley et al included measurements of several body variables rather than self-reported body mass index (BMI) alone. This more detailed and objective assessment is a step forward in the epidemiologic research addressing the influence of body measures in relation to risk of premalignant (Barrett’s esophagus) or malignant esophageal diseases.
The main finding of the Corley study1 was that abdominal diameter was an independent risk factor of Barrett’s esophagus, that is, the association remained after adjustment for BMI, whereas the association between BMI and Barrett’s disappeared after adjustment of abdominal diameter. Thus, the study indicates that abdominal fat is the key factor for the link between obesity and Barrett’s esophagus. This finding gains support from a recent study conducted by El-Serag et al5 that used computed tomography as the objective evaluation of fat distribution among persons with or without Barrett’s esophagus. This study indicated that abdominal fat was an independent risk factor for Barrett’s esophagus, although BMI did not remain associated with this condition after adjustment for abdominal fat. The results of the Corley study are also supported by a case-control study of risk factors for Barrett’s esophagus conducted by Edelstein et al.6 The Edelstein study also included objectively assessed anthropometric measurements, and all measures representing central adiposity were more strongly related to risk of Barrett’s esophagus than BMI.
The lack of an independent association between BMI and risk of Barrett’s esophagus might at a first glance seem surprising in view of the substantial evidence of a strong and dose-dependent association between increasing BMI and risk of esophageal adenocarcinoma.2, 3, 4 These studies did not include other body measures, however, and therefore did not adjust their results for abdominal diameter. It is fully possible, and in view of the recent studies of Barrett’s esophagus rather likely, that the link between abdominal diameter or abdominal fat distribution is stronger than BMI also with regard to risk of esophageal adenocarcinoma. BMI might thus only be a marker of an increased likelihood of more abdominal fat. This hypothesis remains to be proven, however.
A key question is, of course, how abdominal diameter can cause transformation of the esophageal mucosa to the premalignant condition of Barrett’s esophagus? It would be tempting to assume that the mechanism behind the association between abdominal diameter and Barrett’s esophagus would be through GERD, particularly because recent large studies, including a recent meta-analysis, have provided firm evidence of a causal and dose-dependent relation between BMI and GERD.4, 7, 8 Yet, the association between abdominal diameter and Barrett’s esophagus was only slightly attenuated after adjustment for GERD symptoms in the Corley study, suggesting a limited role of GERD in the causal pathway. Similarly, the results of the Edelstein study revealed no change in the associations between waist-to-hip ratio and Barrett’s esophagus after adjustment for heartburn.6 Moreover, the biological mechanism behind the association between BMI and risk of esophageal adenocarcinoma remains uncertain, because the available large-scale epidemiologic studies have failed to identify support for GERD symptoms acting as the key factor for this association.2, 3 Several alternative or complementary mechanisms have been suggested, but these are currently only speculative. Therefore, an intensified research addressing the mechanisms behind the association between anthropometric factors and these conditions are warranted.
Although a more thorough evaluation of body composition in relation to risk of esophageal adenocarcinoma is warranted, this might not be feasible. Although Barrett’s esophagus per se does not lead to changes in body composition, and thus the body measures are not subjective to biased estimates, the cardinal symptom of esophageal adenocarcinoma is substantial weight reduction. Currently, the only realistic epidemiologic approach to evaluate body measures in relation to risk of developing this comparatively rare cancer seems to be the case-control design, with an inherent retrospective assessment of the exposures. To be able to assess more detailed measurements of body composition as exposures for this cancer would require very large, prospective cohort studies with repeated evaluation of body measures and decades of follow-up.
It is likely that intra-abdominal fat, more than BMI alone, is a marker of risk of morbidity of the esophagus, namely, GERD, Barrett’s esophagus, and adenocarcinoma of the esophagus. It would be interesting to address the potential difference in the relation between body measures and risk of GERD between women and men, particularly because women are less prone to develop intra-abdominal fat compared with men. However, some but not all studies found that BMI is a stronger risk factor for GERD among women than men7; this is not compatible with the hypothesis of intra-abdominal fat as the key factor for morbidity of the esophagus. A more thorough evaluation of fat distribution with regard to this issue is, however, warranted, particularly because interaction by known or unknown factors might, for example, play an important role. One such potential factor of interaction that might strongly influence the risk of malignant transformation of the esophagus among persons with obesity and GERD is estrogen exposure.7 If estrogen protects against such transformation, it might explain the male predominance in Barrett’s esophagus and esophageal adenocarcinoma; estrogen receptors have been identified in Barrett’s esophagus and in adenocarcinoma of the esophagus.9
In summary, Corley et al1 have contributed to an increased knowledge of the epidemiology of Barrett’s esophagus with regard to body measures, and highlighted the influence of abdominal fat on the risk of this premalignant condition. Moreover, they have revealed that a more detailed and objective evaluation of such exposures is important in the evaluation of the relation between body measures and development of premalignant and malignant conditions of the upper gastrointestinal tract. Inferences to esophageal adenocarcinoma might also be justified. Abdominal fat distribution rather than BMI only might be a more appropriate measure of the link between body measures in relation to Barrett’s esophagus and esophageal adenocarcinoma.
References
- . Abdominal obesity and body mass index as risk factors for Barrett’s esophagus. Gastroenterology. 2007;133:34–41
- . Body mass index and risk of adenocarcinoma of the esophagus and gastric cardia. J Natl Cancer Inst. 1998;90:150–155
- . Association between body mass and adenocarcinoma of the esophagus and gastric cardia. Ann Intern Med. 1999;130:883–890
- . Meta-analysis: obesity and the risk for gastroesophageal reflux disease and its complications. Ann Intern Med. 2005;143:199–211
- . Abdominal obesity and the risk of Barrett’s esophagus. Am J Gastroenterol. 2005;100:2151–2156
- . Central adiposity and risk of Barrett’s esophagus. Gastroenterology. 2007;133;(In press).
- . Obesity and estrogen as risk factors for symptomatic gastroesophageal reflux. JAMA. 2003;290:66–72
- . Body-mass index and symptoms of gastroesophageal reflux in women. N Engl J Med. 2006;354:2340–2348
- . Estrogen receptor-beta is expressed in Barrett’s metaplasia and associated adenocarcinoma of the esophagus. Anticancer Res. 2002;22:1459–1461
PII: S0016-5085(07)01023-2
doi:10.1053/j.gastro.2007.05.039
© 2007 AGA Institute. Published by Elsevier Inc. All rights reserved.
Refers to article:
- Abdominal Obesity and Body Mass Index as Risk Factors for Barrett’s Esophagus , 26 April 2007
