Hiatal Hernia and the Risk of Barrett's Esophagus

Juliana Andrici; Martin Tio; Michael R Cox; Guy D Eslick

J Gastroenterol Hepatol. 2013;28(3):415-431.

Abstract and Introduction

Abstract

Background and Aim: Barrett's esophagus has been associated with the presence of hiatal hernia; however, to date no meta-analysis of the relationship has been performed. We aimed to conduct a systematic review and meta-analysis, providing a quantitative estimate of the increased risk of Barrett's esophagus associated with hiatal hernia.

Methods: A search was conducted through four electronic databases (Medline, PubMed, Embase, and Current Contents Connect) to 4 April 2012, for observational studies of Barrett's esophagus patients. We calculated pooled odds ratios and 95% confidence intervals using a random effects model for the association of hiatal hernia with any length Barrett's esophagus, as well as with short segment Barrett's esophagus and long segment Barrett's esophagus. 33 studies comprising 4390 Barrett's esophagus patients were eligible for the meta-analysis.

Results: Hiatal hernia was associated with an increased risk of Barrett's esophagus of any length (odds ratio 3.94; 95% confidence interval 3.02–5.13). Heterogeneity was present (I² = 82.03%, P < 0.001), and the Egger test for publication bias was significant (P = 0.0005). The short segment Barrett's esophagus subgroup analysis likewise showed an increased risk (odds ratio 2.87; 95% confidence interval 1.75–4.70). The strongest association was between hiatal hernia and long segment Barrett's esophagus (odds ratio 12.67; 95% confidence interval 8.33–19.25). The increased risk was present even after adjusting for reflux and body mass index.

Conclusions: The presence of hiatal hernia was associated with an increased risk of Barrett's esophagus, even after adjusting for clinically significant confounders. The strongest association was found between hiatal hernia and long segment Barrett's esophagus.

Introduction

Barrett's Esophagus (BE) is a condition in which the normal squamous esophageal lining is replaced by specialized or intestinal columnar epithelium.^[1,2] According to current guidelines in North America,^[3] BE is diagnosed when columnar mucosa is identified above the gastroesophageal junction on endoscopy and is confirmed to contain specialized intestinal epithelia (characterized by the presence of goblet cells) on histological examination. Since 1994, BE has been classified as either short segment BE (SSBE) or long segment BE (LSBE) according to the length of the metaplastic change observed on endoscopic examination.^[4] If the intestinal metaplasia extends less than 3 cm above the gastroesophageal junction, it is termed SSBE, and if it extends 3 cm or more, it is termed LSBE. The prevalence of BE in the general US population is uncertain, due to the fact that many individuals are asymptomatic, and because diagnosis requires endoscopy. A study from Sweden found a prevalence of 1.6% in a random sample of 3000 individuals from the general population.^[5] Among patients with gastroesophageal reflux disease (GERD), the prevalence of BE has been reported to be between 3% and 15%.^[6,7] The clinical significance of BE is its association with an increased risk in developing esophageal adenocarcinoma.^[8,9] Although esophageal adenocarcinoma is a relatively uncommon disease, its incidence has been increasing in the US and other Western countries in recent decades.^[10–15]

Risk factors for BE include white race, male sex, older age, obesity^[16] and persistent gastroesophageal reflux.^[17] Hiatal hernia has also been associated with BE. Hiatal hernia refers to the prolapse of elements of the abdominal cavity, most commonly parts of the stomach, through the esophageal hiatus of the diaphragm and into the thoracic cavity. The most common type is Type I, or sliding hernia, in which the lower esophageal sphincter and a portion of the gastric cardia herniate upwards due to a widening of the muscular hiatal aperture and circumferential laxity of the phrenoesophageal membrane.^[18–20] The presence of hiatal hernia results in anatomical impairment of the esophagogastric junction, which leads to reflux of gastric material into the esophagus. This includes gastric products such as hydrochloric acid and pepsin, as well as pancreatic enzymes and bile.^[17,19] It is hypothesized that chronic exposure to these substances is a contributing factor to the development of BE.

Although individual studies have shown a higher prevalence of hiatal hernia in BE patients compared with non-BE GERD patients,^[21–23] to date no meta-analysis of the relationship between BE and hiatal hernia has been performed. The purpose of this study was to conduct a meta-analysis combining the results of studies reporting the prevalence of hiatal hernia in BE subjects, and thus provide a quantitative estimate of the increased risk of BE associated with hiatal hernia. We hypothesized that hiatal hernia is associated with an increased risk BE, and that we would see a stronger association with LSBE than with SSBE.

1 2 3 4

Next Section

Abstract and Introduction
Methods
Results
Discussion
References

Table 1. Barrett's esophagus studies included in the systematic review
Authors	Year	Study type	Country	Cases	Controls	Adjusted variables	Number of cases	Total size	Length of BE
Abrams et al.²⁹ (2008) 1a	2008	Cross-sectional	USA	Patients undergoing upper endoscopy with histologically confirmed BE	Patients undergoing upper endoscopy without BE	Age, gender, race, indication for endoscopy	9	2100	LSBE
Abrams et al.²⁹ (2008) 1b	2008	Cross-sectional	USA	Patients undergoing upper endoscopy with histologically confirmed BE	Patients undergoing upper endoscopy without BE	Age, gender, race, indication for endoscopy	92	2100	Any Length BE
Amano et al.³⁰ (2006)	2006	Cross-sectional	Japan	Consecutive patients undergoing EGD with histologically confirmed BE	Consecutive patients undergoing EGD for their annual medical check-up or for the investigation of their GI symptoms	Age, gender, BMI, green tea consumption, fatty food consumption	106	1668	Any length BE
Avidan et al.²² (2002)	2002	Case control	USA	Symptomatic GERD patients with histologically confirmed BE	Patients with symptomatic GERD, but with histologically proven absence of BE and endoscopic absence of erosive esophagitis, esophageal ulcer, or peptic stricture	None	256	485	Any Length BE
Avidan et al.³¹ (2001) 1a	2001	Case control	USA	Patients undergoing upper endoscopy with histologically confirmed BE	Patients without symptoms of GERD and with normal esophageal mucosa during EGD	Age, gender, Caucasian ethnicity, smoking, alcohol consumption, prior gastric surgery	650	4063	SSBE
Avidan et al.³¹ (2001) 1b	2001	Case control	USA	Patients undergoing upper endoscopy with histologically confirmed BE	Patients without symptoms of GERD and with normal esophageal mucosa during EGD	Age, gender, Caucasian ethnicity, smoking, alcohol consumption, prior gastric surgery	366	4063	LSBE
Banki et al.³² (2005)	2005	Cross-sectional	USA	BE patients	Patients with abnormal 24-h pH tests without BE	None	186	506	Any Length BE
Byrne et al.³³ (1999) 1a	1999	Cross-sectional	UK	Symptomatic patients with histologically confirmed BE attending clinic for open-access endoscopy	Symptomatic patients without BE attending clinic for open-access endoscopy	None	15	225	SSBE
Byrne et al.³³ (1999) 1c	1999	Cross-sectional	UK	Symptomatic patients with histologically confirmed BE attending clinic for open-access endoscopy	Symptomatic patients without BE attending clinic for open-access endoscopy	None	8	225	LSBE
Byrne et al.³³ (1999) 1b	1999	Cross-sectional	UK	Symptomatic patients with histologically confirmed BE attending clinic for open-access endoscopy	Symptomatic patients without BE attending clinic for open-access endoscopy	None	23	225	Any Length BE
Cameron²¹ (1999) 1a	1999	Case control	USA	Patients referred for upper GI endoscopy for clinical indications, patients with histologically confirmed BE	Consecutive patients referred for endoscopy either with symptoms of reflux or for non-reflux symptoms, without histologically confirmed BE	None	18	167	SSBE
Cameron	1999	Case control	USA	Patients referred for upper GI endoscopy for clinical indications, patients with histologically confirmed BE	Consecutive patients referred for endoscopy either with symptoms of reflux or for non-reflux symptoms, without histologically confirmed BE	None	46	167	LSBE
Cameron²¹ (1999) 1c	1999	Case control	USA	Patients referred for upper GI endoscopy for clinical indications, patients with histologically confirmed BE	Consecutive patients referred for endoscopy either with symptoms of reflux or for non-reflux symptoms, without histologically confirmed BE	None	64	167	Any Length BE
Campos et al.³⁴ (2001) 1a	2001	Cross-sectional	USA	Patients with GERD documented by abnormal acid exposure on 24-h esophageal pH monitoring, with histological evidence of BE	Patients with GERD documented by abnormal acid exposure on 24-h esophageal pH monitoring, without histological evidence of BE	None	67	502	SSBE
Campos et al.³⁴ (2001) 1b	2001	Cross-sectional	USA	Patients with GERD documented by abnormal acid exposure on 24-h esophageal pH monitoring, with histological evidence of BE	Patients with GERD documented by abnormal acid exposure on 24-h esophageal pH monitoring, without histological evidence of BE	Defective LES, gender, number of reflux episodes	107	502	LSBE
Campos et al.³⁴ (2001) 1c	2001	Cross-sectional	USA	Patients with GERD documented by abnormal acid exposure on 24-h esophageal pH monitoring, with histological evidence of BE	Patients with GERD documented by abnormal acid exposure on 24-h esophageal pH monitoring, without histological evidence of BE	Abnormal bilirubin exposure, defective LES, gender, defective distal contraction amplitude, number of reflux episodes longer than 5 min, duration of GERD symptoms	174	502	Any Length BE
Chacaltana et al.³⁵ (2009) 1a	2009	Case control	Peru	Patients participating in a gastric cancer screening campaign, with histologically confirmed BE	Patients participating in a gastric cancer screening campaign, without symptoms of GERD and no esophagitis or BE on endoscopy	Tobacco consumption, alcohol consumption, use of NSAIDs	11	975	Any Length BE
Chacaltana et al.³⁵ (2009) 1b	2009	Case control	Peru	Patients participating in a gastric cancer screening campaign, with histologically confirmed BE	Patients participating in a gastric cancer screening campaign, with diagnosed GERD	Tobacco consumption, alcohol consumption, use of NSAIDs	11	975	Any Length BE
Conio et al.³⁶ (2002) 1a	2002	Case control	Italy	Patients referred for upper gastrointestinal endoscopy, with histologically confirmed BE	Patients admitted to hospitals in the same catchment areas as the cases for acute, non-neoplastic, non-GI conditions	None	109	457	SSBE
Conio et al.³⁶ (2002) 1b	2002	Case control	Italy	Patients referred for upper gastrointestinal endoscopy, with histologically confirmed BE	Patients admitted to hospitals in the same catchment areas as the cases for acute, non-neoplastic, non-GI conditions	None	40	457	LSBE
Conio et al.³⁶ (2002) 1c	2002	Case control	Italy	Patients referred for upper gastrointestinal endoscopy, with histologically confirmed BE	Patients admitted to hospitals in the same catchment areas as the cases for acute, non-neoplastic, non-GI conditions	Center, gender and age	149	457	Any Length BE
Conio et al.³⁶ (2002) 1d	2002	Case control	Italy	Patients referred for upper gastrointestinal endoscopy, with histologically confirmed BE	GERD controls	Center, gender and age	149	457	Any Length BE
Dickman et al.²³ (2010)	2010	Cross-sectional	Israel	Consecutive endoscopy cases with histologically confirmed BO	Consecutive endoscopy cases without histologically confirmed BO	Gender	64	166	Any Length BE
Grassi et al.³⁷ (2006)	2006	Cross-sectional	Italy	Outpatients referred for endoscopy, with histologically confirmed BE	Outpatients referred for endoscopy, without histologically confirmed BE	None	47	224	SSBE
Jonaitis et al.³⁸ (2011)	2011	Cross-sectional	Lithuania	Consecutive patients aged 18 years and over referred for upper endoscopy from primary and secondary settings due to upper GI and/or "alarm" symptoms, with histologically confirmed BE	Consecutive patients aged 18 years and over referred for upper endoscopy from primary and secondary settings due to upper GI and/or "alarm" symptoms, without histologically confirmed BE	Ulcer and/or stricture of esophagus, age, smoking (> 10 cigarettes per day), H. pylori status, BMI, male gender	33	4 032	Any Length BE
Kuo et al.³⁹ (2009)	2009	Cross-sectional	Taiwan	Consecutive patients who underwent upper endoscopy for a variety of GI, with histologically confirmed BE	Consecutive patients who underwent upper endoscopy for a variety of GI, without histologically confirmed BE	Gender, age, GERD duration, smoking history, alcohol use, BMI, the presence of erosive esophagitis	13	736	Any Length BE
Lee et al.⁴⁰ (2010)	2010	Cross-sectional	Korea	BE patients who had visited outpatient clinics to receive an upper GI endoscopy and who had clinical symptoms	Patients without BE who had visited outpatient clinics to receive an upper GI endoscopy and who had clinical symptoms	None	21	2 048	Any Length BE
Lord et al.⁴¹ (2009)	2009	Cross-sectional	USA	Patients with symptoms suggestive of reflux, with histologically confirmed BE	Patients with symptoms suggestive of reflux, without histologically confirmed BE	None	44	160	Any Length BE
Mathew et al.⁴² (2011)	2011	Cross-sectional	India	Consecutive GERD patients with histologically confirmed BE	Consecutive GERD patients without histologically confirmed BE	Age, duration of symptoms, presence of dysphagia, presence of eructation	25	278	Any Length BE
Moons et al.⁴³ (2008)	2008	Case control	The Netherlands	Caucasian patients referred for the evaluation of reflux-related symptoms, odynophagia or dysphagia, suspected extra-esophageal manifestations of GERD, with histologically confirmed BE	Caucasian patients referred for the evaluation of reflux-related symptoms, odynophagia or dysphagia, suspected extra-esophageal manifestations of GERD, without histologically confirmed BE	Age, gender, tested cytokine polymorphisms	255	502	Any Length BE
Nasseri-Moghaddam et al.⁴⁴ (2003) 1a	2003	Cross-sectional	Iran	All patients 18 years of age and older scheduled for upper GI endoscopy with histologically confirmed BE	All patients 18 years of age and older scheduled for upper GI endoscopy without histologically confirmed BE	None	45	269	SSBE
Nasseri-Moghaddam et al.⁴⁴ (2003) 1b	2003	Cross-sectional	Iran	All patients 18 years of age and older scheduled for upper GI endoscopy with histologically confirmed BE	All patients 18 years of age and older scheduled for upper GI endoscopy without histologically confirmed BE	None	8	269	LSBE
Nasseri-Moghaddam et al.⁴⁴ (2003) 1c	2003	Cross-sectional	Iran	All patients 18 years of age and older scheduled for upper GI endoscopy with histologically confirmed BE	All patients 18 years of age and older scheduled for upper GI endoscopy without histologically confirmed BE	None	68	269	Any Length BE
Oberg et al.⁴⁵ (1998) 1a	1998	Cross-sectional	USA	Patients with symptoms of foregut disease and no previous history of gastric or esophageal surgery, with histologically confirmed BE	Patients with symptoms of foregut disease and no previous history of gastric or esophageal surgery, without histologically confirmed BE	None	30	262	SSBE
Oberg et al.⁴⁵ (1998) 1b	1998	Cross-sectional	USA	Patients with symptoms of foregut disease and no previous history of gastric or esophageal surgery, with histologically confirmed BE	Patients with symptoms of foregut disease and no previous history of gastric or esophageal surgery, without histologically confirmed BE	None	32	262	LSBE
Oberg et al.⁴⁵ (1998) 1c	1998	Cross-sectional	USA	Patients with symptoms of foregut disease and no previous history of gastric or esophageal surgery, with histologically confirmed BE	Patients with symptoms of foregut disease and no previous history of gastric or esophageal surgery, without histologically confirmed BE	None	62	262	LSBE
Odemis et al.⁴⁶ (2009)	2009	Cross-sectional	Turkey	Consecutive patients referred for endoscopy for any clinical indication, with histologically confirmed BE	Consecutive patients referred for endoscopy for any clinical indication, without histologically confirmed BE	None	12	1 000	Any Length BE
Peng et al.⁴⁷ (2009)	2009	Cross-sectional	China	Consecutive individuals aged 18–75 who underwent routine upper endoscopy as part of their regular medical examination, with histologically confirmed BE	Consecutive individuals aged 18–75 who underwent routine upper endoscopy as part of their regular medical examination, without histologically confirmed BE	Reflux symptoms, alcohol consumption	27	2 580	Any Length BE
Rajendra et al.⁴⁸ (2004)	2004	Cross-sectional	Malaysia	Consecutive patients undergoing elective gastroscopy for predominantly upper abdominal or reflux complaints, with histologically confirmed BE	Consecutive patients undergoing elective gastroscopy for predominantly upper abdominal or reflux complaints, without histologically confirmed BE	None	123	1 985	Any Length BE
Rajendra et al.⁴⁹ (2007) 1a	2007	Cross-sectional	Malaysia	Patients referred to endoscopy unit, found to have histologically confirmed BE	GERD and non-GERD controls combined	None	25	188	LSBE
Rajendra et al.⁴⁹ (2007) 1b	2007	Cross-sectional	Malaysia	Patients referred to endoscopy unit, found to have histologically confirmed BE	GERD and non-GERD controls combined	None	25	188	SSBE
Rajendra et al.⁴⁹ (2007) 1c	2007	Cross-sectional	Malaysia	Patients referred to endoscopy unit, found to have histologically confirmed BE	GERD and non-GERD controls combined	None	30	188	Any Length BE
Rajendra et al.⁴⁹ (2007) 1d	2007	Cross-sectional	Malaysia	Patients referred to endoscopy unit, found to have histologically confirmed BE	GERD controls	None	25	188	LSBE
Rajendra et al.⁴⁹ (2007) 1e	2007	Cross-sectional	Malaysia	Patients referred to endoscopy unit, found to have histologically confirmed BE	GERD controls	None	25	188	SSBE
Rajendra et al.⁴⁹ (2007) 1f	2007	Cross-sectional	Malaysia	Patients referred to endoscopy unit, found to have histologically confirmed BE	GERD controls	None	30	188	Any Length BE
Rajendra et al.⁴⁹ (2007) 1g	2007	Cross-sectional	Malaysia	Patients referred to endoscopy unit, found to have histologically confirmed BE	Non-GERD controls—patients without histologically confirmed BE undergoing upper GI endoscopy for reasons other than reflux symptoms, BE surveillance, or any form of dyspepsia	None	25	188	LSBE
Rajendra et al.⁴⁹ (2007) 1h	2007	Cross-sectional	Malaysia	Patients referred to endoscopy unit, found to have histologically confirmed BE	Non-GERD controls—patients without histologically confirmed BE undergoing upper GI endoscopy for reasons other than reflux symptoms, BE surveillance, or any form of dyspepsia	None	25	188	SSBE
Rajendra et al.⁴⁹ (2007) 1i	2007	Cross-sectional	Malaysia	Patients referred to endoscopy unit, found to have histologically confirmed BE	Non-GERD controls—patients without histologically confirmed BE undergoing upper GI endoscopy for reasons other than reflux symptoms, BE surveillance, or any form of dyspepsia	None	30	188	Any Length BE
Ringhofer et al.⁵⁰ (2008)	2008	Cross-sectional	Austria	Patients investigated for symptoms of GERD with histologically confirmed BE	Patients investigated for symptoms of GERD without histologically confirmed BE	None	19	102	Any Length BE
Sarr et al.⁵¹ (1985)	1985	Cross-sectional	USA	Patients investigated for symptoms compatible with GERD, including regurgitation, heartburn, epigastric distress, and dysphagia, with histologically confirmed BE	Patients investigated for symptoms compatible with GERD, including regurgitation, heartburn, epigastric distress, and dysphagia, without histologically confirmed BE	None	44	362	Any Length BE
Sgouros et al.⁵² (2007) 1a	2007	Case control	Greece	Patients aged 18–79 who presented to the clinic for investigation of heartburn and/or acid regurgitation, with histologically confirmed BE	Patients with an endoscopic diagnosis of peptic ulcer disease who presented during the same period as cases, without symptoms typical of reflux and without BE	None	17	863	Any Length BE
Sgouros et al.⁵² (2007) 1b	2007	Case control	Greece	Patients aged 18–79 who presented to the clinic for investigation of heartburn and/or acid regurgitation, with histologically confirmed BE	GERD patients without BE	None	17	863	Any Length BE
Sgouros et al.⁵² (2007) 1c	2007	Case control	Greece	Patients aged 18–79 who presented to the clinic for investigation of heartburn and/or acid regurgitation, with histologically confirmed BE	GERD and non-GERD controls combined	None	17	863	Any Length BE
Toruner et al.⁵³ (2004)	2004	Cross-sectional	Turkey	Consecutive dyspeptic patients, never previously investigated, who were referred for upper endoscopy in whom histologically confirmed BE was found	Consecutive dyspeptic patients, never previously investigated, who were referred for upper endoscopy, without histologically confirmed BE	None	29	395	Any Length BE
Trujillo-Benavides et al.⁵⁴ (2005)	2005	Cross-sectional	Mexico	Consecutive patients undergoing endoscopy in whom histologically confirmed BE was found	Consecutive patients undergoing endoscopy without histologically confirmed BE	None	10	109	Any Length BE
Veldhuyzen Van Zanten et al.⁵⁵ (2006)	2006	Cross-sectional	Canada	Patients > 18 years of with a primary complaint of at least 3 months of either continuous or intermittent dyspepsia of any severity, with histologically confirmed BE	Patients > 18 years of with a primary complaint of at least 3 months of either continuous or intermittent dyspepsia of any severity, without histologically confirmed BE	None	25	1 040	Any Length BE
Wang et al.⁵⁶ (2008)	2008	Cross-sectional	USA	Patients undergoing an upper endoscopy for any indication, with histologically confirmed BE	Patients undergoing an upper endoscopy for any indication, without histologically confirmed BE	Gender, age, race, length of BE	1215	2 511	Any Length BE
Xiong et al.⁵⁷ (2010)	2010	Cross-sectional	China	Consecutive patients aged 18–88 receiving an endoscopy for upper GI symptoms, with histologically confirmed BE	Consecutive patients aged 18–88 receiving an endoscopy for upper GI symptoms, without histologically confirmed BE	Age, gender, reflux esophagitis, BMI, heartburn	21	2 022	Any Length BE
Yilmaz et al.⁵⁸ (2006)	2006	Cross-sectional	Turkey	Patients who underwent upper endoscopy, with histologically confirmed BE	Patients who underwent upper endoscopy, without histologically confirmed BE	None	84	18 766	Any Length BE

BE, Barrett's esophagus; BMI, body mass index; EGD, esophagogastroduodenoscopy; GERD, gastroesophageal reflux disease; GI, gastrointestinal; LES, lower esophageal sphincter; LS, long segment; NSAIDs, non-steroidal anti-inflammatory drugs; SS, short segment.

Table 2. Meta-analysis by different study characteristics
Factor	SSBE OR (95% CI)	Any Length BE OR (95% CI)	LSBE OR (95% CI)
Control group
GERD controls	1.86 (1.16–2.97)	3.65 (2.74–4.85)	11.17 (6.38–19.55)
Non-GERD controls	7.93 (2.05–30.59)	13.72 (3.54–53.22)	13.44 (8.26–21.87)
Adjusted ORs
Adjusted for any variables	–	2.99 (2.24–3.96)	13.70 (9.61–19.54)
Adjusted for BMI	–	2.63 (1.85–3.76)	–
Adjusted for reflux	–	3.35 (2.25–4.39)	13.84 (5.19–36.89)
Not Adjusted	2.85 (1.43–5.67)	4.26 (2.82–6.43)	11.97 (5.53–25.88)
Continent
Europe	3.40 (0.53–21.89)	7.93 (3.39–18.58)	–
North America	2.82 (2.35–3.37)	3.38 (1.95–5.86)	13.22 (9.63–18.14)
South America	–	12.14 (3.27–45.13)	–
Asia	–	3.84 (2.48–5.94)	–
Study type
Case control	5.35 (1.74–16.47)	4.38 (2.91–6.61)	13.35 (9.62–18.53)
Cross-sectional	1.96 (1.29–2.99)	3.66 (2.72–4.91)	11.30 (5.06–25.21)

BE, Barrett's esophagus; BMI, body mass index; CI, confidence interval; GERD, gastroesophageal reflux disease; OR, odds ratio; LS, long segment; SS, short segment.

processing....

Hiatal Hernia and the Risk of Barrett's Esophagus

Abstract and Introduction

Abstract

Introduction

Tables

Tables

References

Authors and Disclosures

Authors and Disclosures

Email This

Feedback