Breuer K, Heratizadeh A, Wulf A, et al. Clin Exp Allergy. 2004;34:817–824
Purpose of the Study.
To evaluate the frequency of late-phase atopic dermatitis (AD) reactions to foods during double-blind placebo-controlled food challenges (DBPCFCs) and correlate the results with food-specific IgE and patch tests.
Sixty-four children aged 1 to 10 years (median: 2 years) with mild to severe AD evaluated in an outpatient dermatology and allergy department in Germany.
The inclusion criterion was suspicion of food-related AD by parents and/or a referring physician. The children underwent testing for food-specific IgE (n = 64). Allergen patch testing (APT) to suspected foods was performed if they did not have a rash on their back (n = 41). The first day was an incremental food challenge up to a full serving as tolerated, and on the second day the children were given a full dose of the food/placebo. The children were observed for 48 hours after the challenges. Reactions occurring within 6 hours were considered immediate, and those occurring after >6 hours were considered late reactions.
A total of 106 food challenges were performed to milk, egg, wheat, or soy, with 64% of the children reacting to at least 1 food; of those who reacted, 83% reacted to only 1 food, 15% reacted to 2 foods, and 1 child reacted to 3 foods. The most common trigger was egg (62%), followed by milk (47%) and wheat and soy (35% each). Immediate reactions occurred in 88% of the challenges. Late AD reactions were seen in 28 of 49 (57%) of positive challenges. Sensitivity of history in predicting immediate reactions was only 34%, and for late reactions only 25%; reactions to milk had the highest sensitivity (50–67%), and soy had the lowest (0%). In general, sensitivity (77% vs 68%), specificity (60% vs 50%), and positive predictive values (57% vs 33%) were higher for immediate reactions versus late reactions. Diagnostic accuracy of food-specific serum IgE was greater for children <2 years old. There was no difference in sensitivity (67%) or specificity (38%) of APT for predicting immediate or late AD reactions. The positive predictive value of APT was greater for immediate reactions (38% vs 24%). It is notable that 19% of patients reacted on day 2 of the challenge, having previously tolerated the food on day 1.
Food allergy should be considered in any child with AD who is not responding to standard therapy, with DBPCFCs still considered the “gold standard” for this diagnosis.
This study adds support to the current literature demonstrating a link between food allergy and AD. The interesting finding of late reactions observed in this study should be considered, because most observation periods after food challenges are not generally that long.