Published recipes for double-blind placebo-controlled food challenges
Version of Record online: 2 AUG 2005
Volume 60, Issue 9, page 1212, September 2005
How to Cite
Vlieg-Boerstra, B. and Dubois, A. (2005), Published recipes for double-blind placebo-controlled food challenges. Allergy, 60: 1212. doi: 10.1111/j.1398-9995.2005.00935.x
- Issue online: 2 AUG 2005
- Version of Record online: 2 AUG 2005
With great interest we read the publication by van Odijk et al. (1), entitled: ‘double-blind placebo-controlled challenges for peanut allergy the efficiency of blinding procedures and the allergenic activity of peanut availability in the recipes.’ We have a few comments on this well-conducted study.
First, the authors state that, although difficulty in masking suspected ingredients has been elucidated by Vlieg-Boerstra et al. (2), the recipes have not been published. This statement is incorrect: recipes for double-blind placebo controlled food challenges (DBPCFCs) containing milk, egg, peanut, nuts, soy, and wheat belonging to this paper can be viewed at the Online Repository of the Journal of Allergy and Clinical Immunology, as is mentioned in the text of the article. Furthermore, we can provide the recipes by mail on request.
Secondly, van Odijk et al. describe how they validated two of the recipes they developed by sensory testing. We think they succeeded in developing good tasting recipes in a small volume. However, sensory testing was conducted in a non-professional setting by volunteers (students). We have shown that it is important to perform sensory testing in a professional food laboratory, using professional panelists for optimal test results (2). We compared the results of validation of recipes in a non-professional setting using hospital volunteers (as van Odijk did) to validation in a food laboratory by professional panelists. In our study, only 17 of the 27 recipes validated by the volunteer panelists could subsequently be validated in the professional food laboratory. Thus, ten of the 27 recipes ‘validated’ by the hospital volunteers were decoded by the professional panelists. Since the achievable power of validation studies is relatively limited, using a food laboratory designed for sensory testing aimed at minimizing panelist's biases and to maximize panelist's sensory abilities yields the best possible results. A food laboratory is free of crowding, odors, noise, and is quiet, temperature controlled and comfortable. There is no contact between the panelists during the testing. Furthermore, panelists are prepared for their job. The panelists are experienced in participating in sensory testing and are paid for it. They are not allowed to wear perfumes or cosmetics on the day of the testing; they should be in a good physical condition and should refrain from smoking and consuming alcohol, coffee, and spicy food before they participate in sensory testing.
Thus, validating recipes in a non-professional environment is likely to overestimate the validity of recipes. We would stress the use of food laboratories and professional panelists in validating recipes for DBPCFC.