Due to the age when it becomes apparent and the treatment needed, cow's milk proteins (CMP) allergy requires an accurate diagnosis to avoid labelling infants falsely as allergic and subjecting them to unnecessary diets. The objective of this multi-centre study carried out at the Allergy Units of 14 Children's Hospitals was to discover the epidemiological, clinical and evolutionary characteristics of cow's milk protein allergy (CMPA).
Infants suspected of CMPA who attended allergy clinics at the hospitals taking part during the study period were studied and a detailed clinical history was collected on all of them. Prick tests were done with cow's milk and its proteins and specific IgE anti-bodies were determined by means of CAP with the same allergens as the Prick test. The challenge test with cow's milk was carried out unless contraindicated by the diagnostic protocol. Two different challenge regimens were used: one of them carried out in 3 days and the other in one day. 409 infants with suspected CMPA were included and the diagnostic challenge test was performed on 286 patients (70 %) and not carried out on 123, as it was not indicated according to the protocol. IgE-mediated allergy was confirmed in 234 infants (58 %) and in 15 (4 %) non-IgE-mediated hypersensitivity was diagnosed. The two challenge regimens were equally secure. The average age when the reaction to cow's milk formula took place was 3.5 months (10 days-10 months). The symptoms appeared in the first week of introduction in 95 % of cases and appeared in 60 % with the first feeding with the formula. The most frequent clinical signs were cutaneous in 94 % of cases and the majority of cases appeared within 30 minutes of the feed. 99 % had been breast fed and 44 % had received some cow's milk supplement during the lactation period. Sensitization to egg not given in the feed was noted in 30 % and to beef in 29 %, being well tolerated in all of these.
Carrying out an appropriate diagnostic protocol in infants attending for suspected CMPA allows allergy to be ruled out in a high percentage of cases.