Replying to Kewin Tien Ho Siah Nature Communicationshttps://doi.org/10.1038/s41467-020-17831-z (2020).

We cordially thank Dr. Siah for their comments on our publication1, giving us the opportunity to clarify some points and to discuss limitations inherent in observational studies in general and analysis of claims data in particular.

Our publication “Country-wide medical records infer increased allergy risk of gastric acid inhibition”2 confirmed smaller previous studies in adults and children3,4,5. The relationship between anti-acid drug prescriptions followed by anti-allergy medication prescriptions was specific and not observed after prescriptions of other commonly prescribed drugs. With this in mind, we agree that the sequence of prescriptions is important and was among the first things we checked in our data base. However, it has to be considered that prescription rates of anti-acid drugs are around three times higher than anti-allergy medication prescriptions. It follows, that without any influence of an anti-allergy medication prescription on subsequent anti-acid drug prescriptions, this sequence must occur more frequently than the opposite one. However, our analyses of the country-wide and regional data sets showed that prescription of anti-allergy substances did not result in any elevated risk for subsequent anti-acid drug prescriptions. Contrarily, the hazard ratio for prescription of gastric acid-inhibiting drugs after anti-allergy drugs was below 1 (0.82 [95% confidence interval 0.80–0.84]; p < 0.001) when compared with other commonly prescribed drug classes used as controls, clearly implying an asymmetric relationship.

Considering the unidirectional phenomenon of an increased risk of anti-allergy medications after an anti-acid drug prescription, we agree that future research might reveal relationships linking dyspepsia and allergic diseases. Dr. Siah stressed correctly that observational studies and especially analyses of claims data, provide on their own no firm basis for cause-and-effect statements. Therefore, we referred the reader to the consistency of our findings with experimental data. Although we also agree that randomized controlled studies (RCTs) could, in principle, clarify whether or not the observed association is causal, we doubt that such RCTs will ever be conducted owing to a lack of commercial interest and because of the ethical issues.

As medical doctors, we regard gastric acid-inhibiting drugs as invaluable therapeutic options in a variety of dyspeptic diseases, if indication and reasonable prescription duration are considered. Also as MDs, we are trained to handle patients’ many anxieties, and must take the responsibility to explain, prescribe, and discontinue the anti-acid treatments in a controlled manner. We further stress that guidelines, pointing to a discontinuation of acid-inhibiting drug prescriptions as soon as there is no further indication, should be adhered to.