Blood Test Screening for Food Allergy Has Limited Use in Infant Eczema

When an infant experiences atopic dermatitis, also known as eczema, the child may not be the only one who gets itchy. Parents worry that the skin condition is triggering food allergies in their children, and they want quick answers from pediatricians. Too often, new research shows, they may jump to conclusions based on a screening blood test that has limited usefulness.

Eczema occurs in about 10 to 20 percent of all infants, and its symptoms include itchy, red, dry, flaky skin caused by inflammation. Children with eczema are at higher risk for asthma and other atopic diseases.

Researchers at The Children’s Hospital of Philadelphia wanted to determine how common it is for clinicians to actually see food allergies occur in patients with eczema. And they wanted to find out if performing a blood test early could help to identify which of these patients was most likely to develop a food allergy.

Jonathan Spergel, MD, PhD, chief of the Allergy section of The Children’s Hospital of Philadelphia, led a large, multisite study of 1,000 infants ages 3 months to 18 months with mild-to-severe eczema but no history of food allergy and followed them prospectively over three years for allergy development. The study results, published in Pediatrics, revealed two key insights.

First, the researchers demonstrated how prevalent food allergies are in children with eczema from 36 clinics in the United States. Overall, about 16 percent of the study participants developed a food allergy. Allergy to peanut was most common (6.6 percent), followed by cow’s milk (4.3 percent) and egg white (3.9 percent). Allergies to wheat, seafood, and soybean were rare.

For infants with mild eczema, the percentage of study participants who developed a food allergy was 10 percent, which is close to the occurrence in the average population. On the other end of the severity range, for infants with the worst cases of eczema, 25 percent developed a food allergy.

Second, the investigators wanted to evaluate the usefulness of a radioallergosorbent test, a blood test commonly called RAST that measures IgE antibody levels to specific foods (sIgE). It is widely used to screen for potential food allergy in young children with eczema. Yet, the study results showed that RAST is not sensitive or specific enough to be used alone for the diagnosis of food allergy in infants without a history of reaction to a food.

“Generally, the testing is very limited,” Dr. Spergel said. “Typically, if you’re negative, you’re not allergic. But we showed that even at the highest (sIgE) numbers, for example when we tested for milk allergies, 50 percent of the infants were allergic, and 50 percent were not. So a lot of young children (in particular to infants) who are positive are not really allergic to the food. This supports what allergists have recommended to pediatricians for years: Don’t do it as a screening test because you’ll get too many false positives, and you’ll remove important parts of children’s diets for no reason.”

Obtaining a through patient history of allergic reactions to a specific food and closely tracking clinical symptoms is more reliable than a blood sample to pinpoint the factors associated with a child’s eczema, Dr. Spergel said. These fundamental components of good pediatric care can help families to detect a food allergy and intervene early.

“You shouldn’t do a broad screen that gives too much information,” said Dr. Spergel, who also is an associate professor of pediatrics at the Perelman School of Medicine at the University of Pennsylvania. “Instead, a good patient history will help screen for what the family reports is causing the problem.”

As this study builds on evidence that there is a weak relationship between measured sIgE and clinical food allergy, Dr. Spergel suggested future research is needed to develop a better diagnostic test for food allergies. He added that researchers also could delve into why the 25 percent of infants with the worst cases of eczema developed a food allergy. For example, is there a certain clinical feature that tends be more prominent in those children? Finding out these answers could eventually improve clinician’s track record for prediction of food allergy development.