Statistics from Altmetric.com
Nwokolo et al have demonstrated that following eradication of Helicobacter pylori from asymptomatic patients, plasma ghrelin “increases profoundly” (Gut 2003;52:637–40). Although we find these results interesting, we cannot agree with the conclusion that this may be causally linked to epidemiological observations of the rising incidence of obesity and oesophageal adenocarcinoma in Western populations. In particular, the present study in fact demonstrates that after H pylori eradication, ghrelin merely returns to levels detected in non-obese control patients using the same hormone assay.1 It would seem likely that H pylori infection, leading to oxyntic gland atrophy,2 is associated with at most a mild suppression of plasma ghrelin, which recovers after treatment. This seems unlikely to have a profound effect on calorie intake, particularly as obese patients have a lower mean plasma ghrelin concentration than matched non-obese controls.1 While it is possible, although unproven, that the virtual abolition of plasma ghrelin seen after roux-en-Y gastric bypass surgery may contribute to the paradoxical reduction in hunger observed in these patients, it is simplistic to suggest that the moderate reduction in ghrelin, as seen in the H pylori infected group in this study, is protective against the development of obesity and its associated conditions.
Macadam et al have rightly questioned whether our novel observation is merely epiphenomenal or of pathophysiological significance. Their “gut feeling” is that it is the former as the changes are “mild” and ghrelin concentrations after Helicobacter pylori cure are no different from those seen in a non-obese Western population. They also suggest that the “moderate” reduction of ghrelin in H pylori positive subjects (which they attribute to oxyntic gland atrophy) is unlikely to protect these individuals from obesity.
There is no doubt that cure of H pylori increases plasma ghrelin in “healthy subjects”. The real questions are: whether plasma ghrelin concentrations are higher in H pylori negative individuals and, if so, whether the higher ghrelin concentrations cause weight gain, and whether any weight gained exacerbates gastro-oesophageal acid reflux enough to induce Barrett’s oesophagus and cancer. There are no satisfactory answers to these questions based on first class evidence but in our discussion, we considered some indirect evidence. Firstly, populations with a high prevalence of H pylori have a relatively high proportion of thin children and adults, and those with a low prevalence have a higher proportion of obese individuals; we acknowledge the numerous other confounding factors in our paper.
Secondly, Furuta et al showed that patients cured of H pylori gained weight.1 Lane et al, continuing their reporting of the large Bristol Helicobacter project, showed that at the end of six months, individuals who received treatment for H pylori increased their weight by 0.6 kg over and above a matched group that received placebo.2 Finally, in the only published study of its kind, infusion of ghrelin into healthy subjects was associated with increased appetite and food intake.3
In the presence of H pylori, abnormalities in the function of the gastric neuroendocrine cell population can be detected long before gastric atrophy occurs. “Inappropriate” hypergastrinaemia and disturbances in D cell function have been described; these are fully reversible, returning to normal soon after H pylori cure.45 Similarly, gastric atrophy which is irreversible in the short term is unlikely to be the mechanism that mediates hypo-ghrelinaemia in H pylori positive subjects, given that reversion to normal non-obese concentrations occurred 6–12 weeks after cure, which was the time course of our study. Also, the median age of our subjects was 36 years; the fact that they had normal gastrin concentrations and 24 hour intragastric acidity makes it unlikely that they had significant gastric atrophy.
In general, single factors rarely account for large epidemiological trends. We do not believe that everything can be explained by ghrelin; that would really be simplistic. However, we believe that H pylori positive subjects with low ghrelin may have decreased appetite and food intake and are thinner than their H pylori negative counterparts in the Western world. Their poor nutritional status would be exaggerated by coexisting dyspepsia due to peptic ulceration, concurrent infection, and poor diet. They would have relatively lower intragastric acidity. Taken together, these factors could protect these individuals from gastro-oesophageal reflux disease (GORD). Conversely, having normal ghrelin, a good appetite, and normal intragastric acidity could make GORD more likely, possibly leading to Barrett’s oesophagus and oesophageal adenocarcinoma.