Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=igas20

Scandinavian Journal of Gastroenterology

ISSN: 0036-5521 (Print) 1502-7708 (Online) Journal homepage: http://www.tandfonline.com/loi/igas20

Identifying clinically relevant sliding hiatal hernias:a population-based endoscopy study

Bengt Wallner, Ove Björ, Anna Andreasson, Per M. Hellström, Anna M.Forsberg, Nicholas J. Talley & Lars Agreus

To cite this article: Bengt Wallner, Ove Björ, Anna Andreasson, Per M. Hellström, Anna M.Forsberg, Nicholas J. Talley & Lars Agreus (2018) Identifying clinically relevant sliding hiatalhernias: a population-based endoscopy study, Scandinavian Journal of Gastroenterology, 53:6,657-660, DOI: 10.1080/00365521.2018.1458896

To link to this article: https://doi.org/10.1080/00365521.2018.1458896

© 2018 The Author(s). Published by InformaUK Limited, trading as Taylor & FrancisGroup

Published online: 04 Apr 2018.

Submit your article to this journal

Article views: 132

View Crossmark data

ORIGINAL ARTICLE

Identifying clinically relevant sliding hiatal hernias: a population-basedendoscopy study

Bengt Wallnera , Ove Bj€orb , Anna Andreassonc,d,e , Per M. Hellstr€omf , Anna M. Forsbergg,Nicholas J. Talleyc,h and Lars Agreusc

aDepartment of Surgical and Perioperative Sciences, Surgery, Umeå University Hospital, Umeå, Sweden; bDepartment of Radiation Science,Division of Oncology, Umeå University, Umeå, Sweden; cDivision of Family Medicine and Primary Care, Karolinska Institutet, Huddinge,Sweden; dStress Research Institute, Stockholm University, Stockholm, Sweden; eDepartment of Psychology, Macquarie University, NorthRyde, NSW, Australia; fDepartment of Medical Science, Uppsala University, Uppsala, Sweden; gDepartment of Medicin Solna, KarolinskaInstitutet, Stockholm, Sweden; hFaculty of Health and Medicine, University of Newcastle, Newcastle, Australia

ABSTRACTObjectives: The clinical relevance of small to moderate sliding hiatal hernias is controversial. The aimsof the present study were to (1) investigate which symptoms are associated with sliding hiatal herniasand (2) define the length of a sliding hiatal hernia at which gastrointestinal symptoms occur.Methods: A study population representative of the general Swedish population answered a question-naire regarding gastrointestinal symptoms and was investigated with an upper endoscopy. The lengthof any sliding hiatal hernia was measured.Results: Only reflux-related symptoms were associated with length of the hiatal hernia (acid regurgita-tion OR 1.46, CI 1.19–1.79, heartburn OR 1.27, CI 1.05–1.54), and the association did not become signifi-cant until an axial hiatal hernia length of 2 cm.Conclusions: Only reflux symptoms could be attributed to sliding hiatal hernias. Hiatal hernias lessthan 2 cm should be considered clinically insignificant.

ARTICLE HISTORYReceived 13 February 2018Revised 20 March 2018Accepted 20 March 2018

KEYWORDSEndoscopy; sliding hiatalhernia; gastroesophagealreflux; gastroesopha-geal junction

Introduction

A hiatal hernia (HH) is diagnosed when elements of the abdom-inal cavity herniate through the esophageal hiatus into themediastinum. HH can be assessed with radiographic, manomet-ric and endoscopic investigations. Most commonly it is the stom-ach that herniates through the hiatus, and four different typesof HH have been described. The most common type of HH istype 1, sliding HH, which constitutes 85–95% of all cases [1–3].

Sliding HH is a common finding at endoscopy, with aprevalence of up to 25% in the general population [4]. A slid-ing HH is considered present when the gastroesophagealjunction (GEJ) is dislocated orally into the thoracic cavity,resulting in a discrepancy between the level of the GEJ andthe diaphragmatic hiatus. The GEJ and hiatus are, however,not static, but move several centimeters with breathing andswallowing, which leads to a physiological herniation [1,5].There is no consensus as to when a physiological herniationbecomes a pathological hiatal hernia, but a commonly usedcutoff point is 2 cm [1,6]. This is however less than thephysiological movement, why this diagnostic definition iscontroversial. Some authors continue to argue that a shortsegment hiatal hernia (SSHH) defined as less than 2 cm inlength is of clinical significance [7].

The clinical significance of HH has been controversial fordecades. It is accepted that HH probably leads to increasedgastroesophageal reflux (GER) [1,6,8,9]. There remains how-ever some controversy whether hiatal hernias may also leadto other symptoms, and some argue that HH can give rise todyspeptic symptoms or dysphagia [2,10].

It has been shown that the interobserver agreementregarding the endoscopic length measurement is poor evenunder ideal conditions [11]. As a result, the concept of meas-uring HH endoscopically suffers from this inherent weakness.Another way to assess the GEJ is by grading the gastroeso-phageal flap valve by using the Hill classification, anapproach that has been shown to be reproducible [12].

The aims of the present study were to (1) investigatewhich gastrointestinal symptoms are associated with slidinghiatal hernias and (2) define the length of a sliding HH atwhich gastrointestinal symptoms occur.

Materials and methods

A comprehensive description of the setting, population,esophagogastroduodenoscopy (EGD) procedure and symp-tom evaluation has been given elsewhere [13].

CONTACT Bengt Wallner bengt.wallner@umu.se Department of Surgical and Perioperative Sciences, Surgery Umeå University Hospital, Umeå SE-90185, Sweden� 2018 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License (http://creativecommons.org/licenses/by-nc-nd/4.0/),which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed, or built upon in any way.

SCANDINAVIAN JOURNAL OF GASTROENTEROLOGY2018, VOL. 53, NO. 6, 657–660https://doi.org/10.1080/00365521.2018.1458896

Study population

A population cohort in €Osthammar, Uppsala county, Sweden,has been studied for over two decades regarding gastrointes-tinal symptoms [14]. The study population is representativeof the general Swedish population in terms of gender, age,income and other potential selection factors. In 2012 thispopulation was invited to participate in the present study. Tobe invited to the EGD, the subjects had to have participatedin the 2011 study. Figure 1 illustrates the study populationand dropouts. Exclusion criteria for EGD were angina pectoris,myocardial infarction (last 6 months), congestive heart failure,severe lung disease, severe liver disease, esophageal varices,treatment with anticoagulants, need for anesthesia for theendoscopy, earlier surgery of the stomach, and those 80years old or above.

Of the 388 individuals who completed the EGD, 23 wereexcluded because of missing data, leaving 365 individuals inthe present study.

Symptoms were evaluated using the validated AbdominalSymptom Questionnaire (ASQ) including 13 upper gastro-intestinal symptoms [15]. The participants reported if theyhad been troubled by any of the symptoms during the lastthree months.

Endoscopy

The upper endoscopy followed a strict and structured endo-scopic protocol, investigating the esophagus, stomach andupper duodenum. The distance between the incisors andGEJ, and between the incisors and hiatus was measured. The

length of any sliding hiatal hernia was defined as the discrep-ancy between these two measurements. The GEJ was definedas the proximal margin of the gastric mucosal folds [16]. Thehiatus was defined as the level of the diaphragmatic pitch.The endoscopies were performed by five experienced endo-scopists (Among them the authors BW, PH, AF, and LA), whowere blinded to the presence of any symptoms. Two of theendoscopists were gastroenterologists, two were surgeons,and one was general practitioner.

Statistics

For identifying upper gastrointestinal symptoms independ-ently associated with the length of HH, multivariable logisticregression analyses were performed. Length of HH was inde-pendently associated with a symptom if a HH statistically sig-nificantly contributed to the prediction of the symptomcompared to a reduced model containing age, gender, bodymass index (BMI), Helicobacter pylori infection, and the use ofproton pump inhibitors (PPI). The comparison between themodel with HH included and the reduced model was basedon a likelihood-ratio test. In a first step, multivariable logisticregressions was conducted to identify symptoms independ-ently associated with herniation, the HH as a continuous vari-able. For symptoms associated with HH, and in order toassess at which length of a sliding HH gastrointestinal symp-toms occur, separate multivariable logistic models were per-formed with each unique combination of symptom as thedependent variable and binary HH cutoff of different lengthof a sliding HH as independent variable.

Data were analyzed with the statistical software packagesStata (version 13.1; StataCorp LP, College Station, TX). p val-ues (two-sided)<.05 were interpreted as statisticallysignificant.

Ethics

Approval was obtained from the Ethics Committee ofUppsala University (Dnr. 2010/443), and all participants gavetheir informed consent.

Results

A total of 365 patients were included in the final analysis,189 women and 176 men. The participants were between 20and 79 years of age (median 55 years). A total of 107 partici-pants reported acid regurgitations and/or heartburn. Twohundred and forty-nine participants had sliding HH, and fifty-eight participants had esophagitis. Of these 58 participants36 had grade-A esophagitis according to the LA classification,19 had grade B, and 3 had grade C (Tables 1 and 2).

The only symptoms that were associated with length ofthe HH were reflux-related (Table 3). Increasing length of slid-ing HH was significantly associated with a higher odds ofacid regurgitation (OR 1.46, CI 1.19–1.79), and heartburn (OR1.27 CI 1.05–1.54).

Study popula�on n = 1924, n=1757 < 80 years

Not available N=61 Moved abroad, unknown domes�c address,reported too ill, declined earlier, cannot read

Swedish.

Possible to mail the ASQ n = 1863

Non-responders n = 688≤80 years n=644,

Par�cipants non responsestudy n= 308

–

Responded to mailed ASQPar�cipants n = 1175

n=1073 < 80 years

Appropriate for EGD<80 years & live within 200 km

n = 1034

Par�cipants from 1988/89n = 901

New par�cipants 1995 n = 305

New par�cipants 2011n = 718

EGD par�cipants n = 388

Refused n = 545 Dropped ou� n theater n=14

Eligible for EGDn=947

Unreachable by phone n = 42Contraindica�ons n = 45

Excluded N=141:>80 years n=102, Lived to far n= 32

Replied to late n=5Not contacted by logis�c reasons n=2

Study par�cipants n = 365

Missing data n = 23

Figure 1. The study flow chart, illustrating the study population andthe dropouts.

658 B. WALLNER ET AL.

When evaluating the association between reflux symp-toms and length of HH in a binary fashion, the associationbecame significant at a length of 2 cm (Table 4).

Excluding those on current PPI-medication did not alterthe results.

Discussion

Sliding HH is a common finding at endoscopy, but there isno clear endoscopic definition of a clinically relevant HH.

There are furthermore controversies regarding which symp-toms that can be attributed to HH. In an endoscopic popula-tion-based study where validated symptom assessments anda standardized endoscopy were both carefully applied, wefound that hernias smaller than 2 cm endoscopically werenot related to symptoms, and sliding hernias 2 cm or largerwere only associated with reflux symptoms but not with anyother upper GI symptoms.

HH of different lengths are common findings at endos-copy and are considered present when there is an oral dis-placement of the GEJ and the proximal part of the stomachinto the thoracic cavity. However, the GEJ is a complex ana-tomical area that moves several centimeters with breathingand swallowing, giving rise to ‘physiological herniation’ ofthe GEJ. There has been no consensus as to when this‘physiological herniation’ becomes pathological but our pre-sent data suggest that small hernias are asymptomatic andof no clinical relevance.

In the present study, we have investigated a random sam-ple of the normal population, and not a group of patientsseeking medical care. In this way, we avoid any referral biasand minimizing selection bias based on high response rates.Since both HH and functional GI symptoms are common inthe population, results can be difficult to interpret when ana-lyzing the prevalence of HH in a symptomatic groupof patients.

Weaknesses of this study include the lack of an objectivemeasurement of GER, that is, 24-h pH monitoring. There arealso known weaknesses in the endoscopic length measure-ment in esophagus, where the interobserver agreement is apotential limitation for precision [11]. One way to avoid thislimitation when assessing the antireflux barrier of the GEJ, isto use the Hill classification, which has been used for gradingof the GEJ flap valve mechanism [12]. The Hill classificationhowever is less commonly used in clinical practice but it hasbeen proven to have at least as strong association withGERD as HH [17].

In summary, the present study demonstrates that the onlysymptoms attributable to HH are symptoms of heartburn andacid regurgitation. Furthermore, our study suggests that her-nias less than 2 cm should not be considered as clinically sig-nificant. Since there is a known limitation in endoscopicmeasurement of length, it may be of greater value to assessthe GEJ using the Hill classification than to measure thelength of small sliding HH, especially in the context of GERDwith or without esophagitis [17].

Acknowledgements

The present study was supported by Olympus Solna Sverige AB, whosupplied equipment for the realization of the study. The authors have noother affiliation with this company.

Disclosure statement

The authors declare that they have no conflicts of interest concerningthis article.

Table 1. Demographics of the study population.

Age (mean) 55Male:Female 48:52BMI (mean) 26.7H. pylori infection (%) 11.6PPI last week prior to endoscopy (%) 2.7HH of any length (%) 68.2HH 2 cm or larger (%) 43.1Esophagitis, any LA grade (%) 15.9

Table 2. The length of any sliding hiatal hernia (dis-tance between the gastroesophageal junction andthe diaphragmatic pinch) in the investi-gated population.

Length of hiatal hernia (cm) n %

0 116 31.81 92 25.22 84 23.03 42 11.54 21 5.85 5 1.46 2 0.67 2 0.68 1 0.3

Table 3. Associations between increasing length of hiatal hernia and differentupper GI-symptoms in ASQ.

OR CI p Value

Weight loss 0.87 0.41–1.81 .70Bad appetite 1.01 0.74–1.36 .97Uncomfortable feeling of fullness 1.15 0.94–1.41 .16Dysphagia 1.08 0.82–1.43 .58Eructation 1.01 0.82–1.23 .95Acid regurgitation 1.37 1.14–1.64 .001Early satiety 1.03 0.82–1.30 .79Nausea 1.07 0.85–1.34 .58Vomiting 1.10 0.71–1.72 .68Heartburn 1.21 1.02–1.44 .031Retrosternal pain 1.06 0.83–1.35 .62Burning sensation 1.01 0.81–1.26 .95Epigastric pain 0.72 0.49–1.04 .078

Logistic regression-analysis with symptoms as dependent variables and lengthof hiatal hernia as independent variable, adjusted for age, gender, body massindex, H. pylori infection, and use of PPI.

Table 4. In order to investigate at which length the HH becomes clinically sig-nificant, the association between gastroesophageal reflux and increasinglength of HH were evaluated.

Odds ratio 95% CI p Value

Acid regurgitation HH� 1 cm 1.61 0.87–3.01 .13HH� 2 cm 2.40 1.37–4.19 .002

Heartburn HH� 1 cm 1.64 0.90–2.97 .10HH� 2 cm 1.79 1.06–3.03 .030

When the length of the HH became�2 cm, the association with the symptomsbecame statistically significant. Logistic regression, adjusted for age, gender,body mass index, H. pylori infection, and use of PPI.

SCANDINAVIAN JOURNAL OF GASTROENTEROLOGY 659

ORCID

Bengt Wallner http://orcid.org/0000-0002-9234-7946Ove Bj€or http://orcid.org/0000-0002-6080-5973Anna Andreasson http://orcid.org/0000-0003-0203-7977Per M. Hellstr€om http://orcid.org/0000-0001-8428-0772Nicholas J. Talley http://orcid.org/0000-0003-2537-3092Lars Agreus http://orcid.org/0000-0002-6026-9883

References

[1] Kahrilas PJ, Kim HC, Pandolfino JE. Approaches to the diagnosisand grading of hiatal hernia. Best Pract Res Clin Gastroenterol.2008;22:601–616.

[2] Mittal RK. Hiatal hernia: myth or reality? Am J Med. 1997;103:33S–39S.

[3] Gordon C, Kang JY, Neild PJ, et al. The role of the hiatus hernia ingastro-oesophageal reflux disease. Aliment Pharmacol Ther.2004;20:719–732.

[4] Ronkainen J, Aro P, Storskrubb T, et al. High prevalence of gastro-esophageal reflux symptoms and esophagitis with or withoutsymptoms in the general adult Swedish population: a Kalixandastudy report. Scand J Gastroenterol. 2005;40:275–285.

[5] Lee YY, Whiting JGH, Robertson EV, et al. Measuring move-ment and location of the gastroesophageal junction: researchand clinical implications. Scand J Gastroenterol.2013;48:401–411.

[6] Hyun JJ, Bak Y-T. Clinical significance of hiatal hernia. Gut Liver.2011;5:267–277.

[7] Hyun JJ, Kim JH, Yeon JE, et al. Short segment hiatal hernia: is ita clinically significant entity? J Neurogastroenterol Motil. 2010;16:35–39.

[8] Petersen H, Johannessen T, Sandvik AK. Relationship betweenendoscopic hiatus hernia and gastroesophageal reflux symptoms.Scand J Gastroenterol. 1991;26:921–926.

[9] Sgouros SN, Mpakos D, Rodias M, et al. Prevalence and axiallength of hiatus hernia in patients, with nonerosive reflux disease:a prospective study. J Clin Gastroenterol. 2007;41:814–818.

[10] Tibbling L, Johansson M, Mj€ones AB, et al. Globus jugularis anddysphagia in patients with hiatus hernia. Eur ArchOtorhinolaryngol. 2009;267:251–254.

[11] Guda N, Partington S, Vakil NB. Inter- and intra-observer variabil-ity in the measurement of length at endoscopy: implications forthe measurement of Barrett’s esophagus. GastrointestinalEndoscopy. 2004;59:655–658.

[12] Hill LD, Kozarek RA, Kraemer SJM, et al. The gastroesophagealflap valve: in vitro and in vivo observations. GastrointestinalEndoscopy. 1996;44:541–547.

[13] Agreus L, Hellstr€om PM, Talley NJ, et al. Towards a healthy stom-ach? Helicobacter pylori prevalence has dramatically decreasedover 23 years in adults in a Swedish community. UEGJ. 3rd Ed.2016;4:686–696.

[14] Agr�eus L, Sv€ardsudd K, Nyren O. The epidemiology of abdominalsymptoms: prevalence and demographic characteristics in aSwedish adult population: a report from the Abdominal SymptomStudy. Scand J Gastroenterol. 1994;13:383–388.

[15] Agr�eus L, Sv€ardsudd K, Nyren O, et al. Reproducibility and validityof a postal questionnaire. The abdominal symptom study. Scand JPrim Health Care. 1993;11:252–262.

[16] Wallner B. Endoscopically defined gastroesophageal junction coin-cides with the anatomical gastroesophageal junction. SurgEndosc. 2009;23:2155–2158.

[17] Hansdotter I, Bj€or O, Andreasson A, et al. Hill classification issuperior to the axial length of a hiatal hernia for assessment ofthe mechanical anti-reflux barrier at the gastroesophageal junc-tion. Endosc Int Open. 2016;4:(3):E311–7.

660 B. WALLNER ET AL.