The effect of Helicobacter pylori eradication on gastro-oesophageal reflux


Tefera Dr Medical Department A, Haukeland Hospital, University of Bergen, N-5021 Bergen, Norway. E-mail:



: Increased prevalence of oesophagitis has been reported following eradication of Helicobacter pylori. We hypothesized that H. pylori eradication might increase gastro-oesophageal acid reflux in patients with reflux oesophagitis.


: Twenty-five consecutive patients (13 male, 12 female) with H. pylori infection and reflux oesophagitis grade I (22 patients) or II (three patients) were enrolled; mean age 49.9 (range 33–75) years. Twenty-four hour intra-oesophageal pH recording was performed before and 12 weeks after eradication of H. pylori, which was achieved using bismuth subnitrate suspension 150 mg q.d.s., oxytetracycline 500 mg q.d.s. and metronidazole 400 mg t.d.s. for 10 days. Eradication was confirmed by 14C-urea breath test 12 weeks after completion of treatment. The patients did not receive acid-suppressive medication.


: All patients had abnormal gastro-oesophageal reflux before anti-H. pylori treatment. After treatment, there was no significant change in the percentage of total time oesophageal pH < 4 (P=0.46) in the 23 patients in whom the infection had been cured. Nine of the cured patients had increased acid exposure, whereas 14 had decreased acid exposure. No significant change in reflux symptom scores was found. There was no relationship between change in acid exposure and symptom improvement.


: Twelve weeks after H. pylori eradication there was no consistent change in gastro-oesophageal acid reflux in patients with mild or moderate reflux oesophagitis.


In Western countries the prevalence of gastro-oesophageal reflux disease (GERD) is ≈ 8% when based on typical symptoms and ≈ 2% when based on endoscopic findings.1, 2 The prevalence increases with age and male gender is a risk factor. Pathogenetic factors in the oesophagus are: increased frequency of transient lower oesophageal sphincter relaxation, decreased lower oesophageal sphincter (LES) resting pressure, impaired oesophageal clearance, and presence of a hiatial hernia. Contributing gastric factors may be delayed gastric emptying and increased acid secretion.3, 4

Helicobacter pylori infection is important in the pathogenesis of peptic ulcer disease, gastric carcinoma, and low-grade mucosal-associated lymphoid type (MALT) lymphoma.5[6]–7 Its association with GERD is still unclear. Due to migration of the bacteria and development of atrophic gastritis in the proximal part of the stomach during treatment with proton pump inhibitors, it has been recommended that H. pylori infection should be eradicated before long-term therapy is initiated.8, 9 However, some recent observations suggest a protective effect of H. pylori infection in GERD: (i) a declining prevalence of H. pylori and associated diseases in the white population in Western countries,10 whereas the incidence of Barrett’s oesophagus and adenocarcinoma of the oesophagus, a complication of GERD, is increasing,11, 12 (ii) H. pylori infection seems to be associated with a lower incidence of cancer at the cardia (which may also be associated with Barrett’s oesophagus and longstanding GERD),13, 14 (iii) a lower prevalence of H. pylori infection among GERD patients than in controls,15, 16 and (iv) according to some reports, an increased prevalence of reflux oesophagitis after H. pylori eradication in duodenal ulcer patients.17, 18 Thus, Labenz et al. followed endoscopically 460 duodenal ulcer patients who had been treated for H. pylori infection.18 The cumulative incidence of reflux oesophagitis within the next 3 years was 25.8% in the 244 patients in whom the infection was cured, compared to 12.9% in the 216 patients with persistent infection, a statistically significant difference (P ≤ 0.001).

The most reliable method of quantifying acid gastro-oesophageal reflux is intra-oesophageal 24 h pH recording, in which oesophageal acid exposure can be evaluated at near physiological conditions.4 Using this method we wanted to test the hypothesis that H. pylori eradication might increase oesophageal acid exposure in patients with endoscopically verified reflux oesophagitis.



Eligible for the study were currently untreated patients with chronic recurrent heartburn or acid regurgitation, in whom a diagnosis of reflux oesophagitis and gastric H. pylori infection had been confirmed at endoscopy. Patients with current or previous grade 3 reflux oesophagitis, Barrett’s oesophagus, peptic ulcer disease, erosive duodenitis or gastritis, hepatobiliary disease, Crohn’s disease or scleroderma were excluded from the study. Patients who had used H2-receptor antagonists, proton pump inhibitors, prokinetics, sucralfate or anti-cholinergic drugs within the last 4 weeks were also ineligible.


A complete upper endoscopic examination was performed by one of the authors. Reflux oesophagitis was graded according to the Berstad classification (Table 1),19 and a rapid urease test for H. pylori colonization (one biopsy from the antral and one from the body mucosa) was read after 24 h.20 Hiatial hernia was defined as gastric folds seen at least 2 cm above the diaphragm.

Table 1.  . The Berstad grading of reflux oesophagitis19Thumbnail image of

Twenty-four hour pH-metry

Patients returned for manometry and their first 24 h pH-metry within 7 days of endoscopy. Manometry was performed with a 4.8 mm eight-channel PVC catheter (Arndorfer Medical Specialties, Greendale, WI) connected to a low-compliance hydrocapillary infusion pump. The proximal margin of the lower oesophageal sphincter was identified as the point at which the post-inspiratory pressure first reached the oesophageal baseline pressure. A monocrystalline antimony pH-sensitive electrode (Synthetics AB, Stockholm, Sweden) was placed 5.0 cm proximal to this point. Calibration was performed prior to each recording, as recommended by the manufacturer, in buffers of pH 1.07 and 7.01 at 22 °C and a temperature correction was introduced. An Ag/AgCl2 reference electrode with suitable gel (Hellige, Germany) was taped to the skin of the upper thorax. The electrode was passed transnasally and swallowed with a minimal volume of water. The catheter was initially placed in the stomach and then slowly retracted to the oesophageal body, while the pH, as recorded by the electrode, was closely followed to confirm the correct localization of the electrode. Sampling rate was 0.25 Hz, and data was stored in a solid-state datalogger (Digitrapper Mark III, Synthetics AB). At the end of the recording, digital values were transferred to an IBM-compatible computer for processing and editing using a dedicated software program (EsopHogram release 5.04; Gastrosoft Inc., Irving, TX).

Patients were untreated for the 4 weeks prior to the study, except for occasional use of antacids. During the study, patients were instructed to take meals in accordance with their daily routine and to return to work or, if this was thought impractical, to perform work in their home as similar as possible to their daily level of activity. They were not allowed to lie down, except during the night, and were instructed to avoid acid or alcoholic beverages and antacids. The recording lasted (23-) 24 h. A diary was kept covering the timing of meals, their composition, and periods in a recumbent position. The same meals were then repeated on the second pH-metry.

The following variables were calculated from the oesophageal recordings: percentage of the total recording period spent with oesophageal pH < 4.0 (reflux time) and also percentage of time in the upright and recumbent positions with pH < 4.0. The start of a reflux episode was defined as a drop in oesophageal pH below a threshold of 4.0, ending with an increase to above this level.

Symptom evaluation

Reflux symptoms during the last week before the two pH recordings were graded separately for heartburn, regurgitation and dysphagia on a scale ranging from 0 to 3, as shown in Table 2. The patients received neither acid-suppressive nor prokinetic medication but were allowed to use antacids when needed from 4 weeks before the first pH recording until the end of the study period.

Table 2.  . Grading of reflux symptoms used in the study Thumbnail image of


All patients who fulfilled the inclusion criteria received a modified classical triple anti-H. pylori therapy consisting of bismuth subnitrate suspension 150 mg q.d.s., oxytetracycline 500 mg q.d.s. and metronidazole 400 mg t.d.s. for 10 days.21 The eradication of H. pylori was assessed by 14C-urea breath test 12 weeks after cessation of treatment.22 In cases of severe symptom progression patients were withdrawn earlier than 12 weeks and patients who had failed to eradicate H. pylori were also excluded.


The statistical evaluation of differences in oesophageal acid exposure before and after H. pylori eradication was assessed by Wilcoxon signed rank sum test, a non-parametric test for paired data. The relationship between decreased reflux and symptom improvement was assessed by Fisher’s exact test.


Twenty-five patients were recruited to the study between March 1997 and December 1997.

Two patients were excluded from the analysis due to failure to eradicate H. pylori, but no patient was excluded because of symptom progression or complications during the follow-up period. The remaining 23 patients all had technically acceptable 24 h pH recordings before and after therapy. Demographic and baseline data, including endoscopic findings and pH-metry results, are presented in Tables 3 and 4.

Table 3.  . Demographic data for 23 patients with reflux oesophagitis who had successful H. pylori eradication therapy Thumbnail image of
Table 4.  . Median and interquartile range for 24 h pH-metry data in 23 patients with reflux oesophagitis before and after H. pylori eradication Thumbnail image of

All patients had abnormal 24 h intra-oesophageal pH-metry at baseline, with a median of 9.4% total time with pH < 4. Three months after H. pylori eradication, there was no statistically significant change (P=0.46) in acid exposure, with a median value of 9.6% ( Figure 1). Nine patients had increased total acid exposure, 14 patients had a decrease. There was no significant change in acid exposure either for the upright ( Figure 2) or for the supine recording period ( Figure 3). Median symptom score for heartburn was 2 before and 1 after treatment; for regurgitation and for dysphagia the median scores were 1 and 0, respectively, both before and after treatment. Heartburn improved in 12 patients, worsened in five, and remained unchanged in six patients (P=0.07) after eradication of H. pylori. There was no significant relationship between decreased reflux and improvement of heartburn (P=0.68).

Figure 1. Total gastro‐oesophageal acid reflux before and 3 months after H. pylori eradication in 23 patients with reflux oesophagitis. In nine patients the reflux increased, whereas in 1.

Figure 1. Total gastro-oesophageal acid reflux before and 3 months after H. pylori eradication in 23 patients with reflux oesophagitis. In nine patients the reflux increased, whereas in 1.

4 patients the reflux decreased. Individual results are combined with a line. Median values are indicated.

Figure 2. Upright gastro‐oesophageal acid reflux before and 3 months after H. pylori eradication in 2.

Figure 2. Upright gastro-oesophageal acid reflux before and 3 months after H. pylori eradication in 2.

3 patients with reflux oesophagitis. Individual results are combined with a line. Median values are indicated.

Figure 3. Supine gastro‐oesophageal acid reflux before and 3 months after H. pylori eradication in 23.

Figure 3. Supine gastro-oesophageal acid reflux before and 3 months after H. pylori eradication in 23.

patients with reflux oesophagitis. Individual results are combined with a line. Median values are indicated.


Our study is the first to investigate the effect of H. pylori eradication in patients with GERD. We found no significant change in the severity of oesophageal acid exposure or reflux symptoms 3 months after H. pylori eradication in patients with established reflux oesophagitis. We included only patients with grade 1 or 2 oesophagitis (according to the Berstad classification19) because patients with more severe disease could not be left untreated for an extended period of time. This group of patients comprises more than 90% of patients with reflux oesophagitis coming to endoscopy.23 The choice of bismuth subnitrate suspension, oxytetracycline and metronidazole for this study was based on our previous good experience; it is also inexpensive.20 Proton pump inhibitors as part of the eradication regimen were avoided because of the possibility of rebound acid hypersecretion.24

The findings of Carbone et al. were similar in terms of reflux pattern before and after H. pylori eradication in patients with dyspepsia and negative endoscopy.25 Contrary to our results, they found a statistically significant improvement in reflux symptoms. The improvement could not be explained on the basis of changes in reflux patterns. They studied a somewhat different group of patients, i.e. symptomatic patients without evidence of reflux oesophagitis.

Our results, and those of Carbone et al.,25 do not support the hypothesis that H. pylori eradication might induce or aggravate gastro-oesophageal acid reflux. It should be noted that although Labenz et al.18 observed an increased prevalence of reflux oesophagitis, this may not necessarily mean that these patients had more acid reflux after than before cure of the infection. Eradication of H. pylori may equally well remove a different and as yet undiscovered protective factor for GERD.

In H. pylori-positive patients, the secretory capacity and therefore the amount of acid that can reflux to the oesophagus, depends upon whether the antrum or the body of the stomach (or both) are involved. H. pylori-associated gastritis which is confined to the antrum stimulates gastrin release and acid secretion due to a deficiency of antral somatostatin.26 On the other hand, H. pylori gastritis involving the body of the stomach will lead to hyposecretion of gastric acid. Development of oesophagitis after eradication of H. pylori in patients with corpus gastritis might be due to a return to normal of the gastric acid secretory capacity. Disappearance of bacterial ammonia synthesis and gradual normalization of mucosal anatomy in the body would be expected to increase gastric acidity. A recent study by Murthy et al. suggests that in duodenal ulcer patients, increased incidence of reflux oesophagitis after H. pylori eradication might be due to the failure of the antral D-cell density to return to normal.27 Interruption of chronic therapy with antisecretory drugs for peptic disease might furthermore result in increased acid secretion and possibly lead to GERD. Changed drinking and eating habits after eradication of H. pylori might contribute to gastro-oesophageal reflux and weight gain. Increased acid secretory capacity would be expected to result in reflux oesophagitis, especially in patients who are predisposed to gastro-oesophageal reflux.

In our study, nine patients had increased acid exposure, whereas 14 patients had decreased acid exposure after cure of the infection. We did not take biopsies and do not know whether those who had increased reflux after eradication might have had H. pylori gastritis involving the gastric body, whereas those who had decreased reflux may have had an infection confined to the antrum prior to treatment. The changes we observed in these patients are just as likely to be due to day-to-day variation in oesophageal acid exposure.28 The trend towards improvement of reflux symptoms was not statistically significant, and individual improvement in heartburn could not be explained on the basis of changes in reflux patterns.

Acute inflammation, as indicated by infiltration of neutrophils, is a very sensitive indicator of the presence or absence of H. pylori and disappears within days of cure of infection, but chronic inflammatory cells have been shown to be slow to disappear and may take a year or more to fall to ‘normal’ levels.29[30]–31 Some researchers believe that, even several years after cure, the gastric mucosa, particularly in the antrum, remains infiltrated by a greater than normal number of chronic inflammatory cells. Reduced or even reversed mucosal atrophy due to recovery of gastric function 1 year after H. pylori eradication has recently been reported by Tucci et al.32 Our patients will therefore be followed up to look for further changes in reflux pattern with time. This will indicate whether H. pylori should be eradicated in patients with GERD. In the short term, however, it appears to be safe and unproblematic to eradicate H. pylori infection in GERD patients.


This study was supported by a grant from the University of Bergen, Norway.