Statistics from Altmetric.com
Editor,—The joint statement of the European Society for Paediatric Allergology and Clinical Immunology (ESPACI) and the European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN),1 deserves some comment.
Firstly, on the use of soy based formulas for the treatment, as well as for the prevention of food allergy: I was disappointed that no word about this subject appeared in the conclusions of the statement. Many have claimed that the use of soy bean formulas in infancy is an efficient way of treating or preventing food allergic disorders, but more recent prospective and randomised clinical studies have shown that soy protein is as allergenic as cow's milk protein.2 As the matter remains controversial,3 I believe that the conclusions should have been that soy based formulas are not recommended for the treatment or prevention of food allergy until more data are available.
The second issue concerns the use of partially hydrolysed formulas for preventing food allergy. A recent five year follow up prospective, randomised, and controlled study by Chandra,4 which showed a beneficial preventive effect of a partial whey hydrolysed formula in high risk infants, was ignored. The only study where the preventive effect of an extensively hydrolysed formula was compared with the effect of a partially hydrolysed one, showed that the former was superior to the second.5 This paper, however, has a possible methodological shortcoming: the manufacturer (Mead Johnson, Evansville, Indiana, USA) provided both a commercially available extensively hydrolysed formula (Nutramigen) and a non-commercially available (at least in Sweden where the study was undertaken) partially hydrolysed formula, prepared by mild (how mild?) enzymatic hydrolysis. In future, such studies should only use commercially available formulas of either the same or different brands. I consider that current data are insufficient to allow a firm view. Therefore, I believe the conclusions should have stated that no clear recommendation can be made for the use of a partially hydrolysed formula to prevent food allergy.
Conclusions of consensus statements are generally considered as guidelines for the practitioner. Omissions, as in the case of soy based formulas, or ambiguities, as in the case of partially hydrolysed formulas, do not clarify the issues so should be avoided. I believe that modified conclusions, as referred to above, would have been more in agreement with the literature and more helpful to the reader.
Drs Koletzko and Host comment:
We thank Professor Salazar-de-Sousa for his insightful comments on the joint comment of ESPACI and ESPGHAN.
We kept our conclusions brief and did not repeat all the considerations discussed earlier in the text but, rather, focused on the practically most relevant advisable measures to treat and prevent food allergy. In the text of the comment it is stated that, based on information currently available, we do not recommend the use of soy protein based formulas as a first line choice to prevent food allergy in infants. However, we also note that different views exist on this issue and that further studies may be useful to extend the rather limited database available, in order to clarify the allergenicity of soy formulas in infants with allergy risks.
The data presented in one of the studies by Chandra1-1referred to by Professor Salazar-de-Sousa were not ignored. However the committee felt that neither this study, nor many similar studies allowed definitive conclusions on all the issues. Since our comment was not intended to be an extensive review of all available publications, we did not cite this particular paper or the many other original papers on this topic, but referred to a recent editorial considering these and other data.1-2
We agree with Professor Salazar-de-Sousa that currently available data are insufficient to allow a firm conclusion on the relative effects of partially versus extensively hydrolysed formulas for the prevention of food allergy, an important issue for clinical practice. Hence, we concluded that more studies are needed.