++
CLINICAL SCENARIO Does Soy Milk (or Soy Formula) Increase the Risk of Developing Peanut Allergy in Children?
You are a general practitioner examining a 29-year-old patient who is 8 months pregnant with her second child. Her first child, who is now 3 years old, had an intolerance to cow's milk as an infant. He was switched to soy formula and then soy milk, which he subsequently tolerated well. At 2 years of age, cow's milk was reintroduced without any problems, and he has been receiving cow's milk since. The mother was planning to start feeding her next child soy formula at birth but heard from a neighbor that it can increase the risk of peanut allergy in her child—a potentially serious and lifelong problem. She asks for your advice on the topic. Because you are not familiar with this issue, you inform the patient that you will examine the evidence and discuss your findings with her when she returns for her next prenatal visit in 1 week.
++
You formulate the relevant question: In infants, is there an association between exposure to soy milk and the subsequent development of peanut allergy? You search a point-of-care clinician evidence synthesis tool with the term “peanut allergy.” Under the subtopic “Causes and Risk Factors,” you see that “consumption of soy milk or soy formula” is identified as a possible risk factor and a reference is provided. You click on the hypertext link to view the relevant article.1
++
The article describes a case-control study that used a geographically defined cohort of 13971 preschool children. The investigators identified children with a convincing history of peanut allergy who reacted to a peanut challenge in which they were blind to whether they were being exposed to peanut protein or a “placebo.” They collected detailed information from the children's parents and from 2 groups of control parents (a random sample from the geographically defined cohort and from a subgroup of children from the cohort who had eczema in the first 6 months of life and whose mothers had a history of eczema).
++
Box 14-1 presents our usual 3-step approach to using an article about harm from the medical literature to guide your practice. You will find these criteria useful for a variety of issues that involve concerns of etiology or risk factors in which a potentially harmful exposure cannot be randomly assigned. These observational studies involve using cohort or case-control designs.
++