Objective: To review and discuss the relationship between cow's milk allergy (CMA) and some gastrointestinal manifestations, such as gastroesophageal reflux, constipation, food protein-induced enterocolitis, and food-induced eosinophilic proctocolitis, with respect to diagnostic strategies that might eliminate the need for a double-blind, placebo-controlled oral food challenge (DBPCFC).
Data sources: A review of pertinent PubMed articles, published during the past 10 years, was performed.
Study selection: To obtain positive and negative predictive values known as posterior probabilities and to calculate the likelihood ratio, only those studies including both patients and control subjects were selected for analysis.
Results: With respect to gastroesophageal reflux, a typical 24-hour esophageal pH monitoring pattern might obviate the performance of a DBPCFC in patients with symptoms of reflux suspected of having CMA, provided this pH pattern is confirmed in other studies. A relationship between CMA and constipation has been reported in only one prospective controlled study; the clinical and laboratory variables of perianal lesions, histologic abnormalities, and signs of hypersensitivity had likelihood ratios of 2.2, 2.4, and 3.7, respectively, and posttest probabilities of 83, 84, and 88%, respectively. Therefore, a DBPCFC is warranted. In reference to food protein-induced enterocolitis, clinical and laboratory criteria suggested in the literature for defining a food challenge as positive have not been prospectively evaluated in the untreated state. Some simple stool tests, such as fecal tumor necrosis factor-alpha and alpha1-antitrypsin determination, might be candidates for diagnostic studies in patients with food protein-induced enterocolitis, if prospectively evaluated. In infants with food-induced eosinophilic proctocolitis, rectal biopsy invariably shows eosinophilic infiltration and thus makes performance of a DBPCFC unnecessary.
Conclusion: Although the current diagnosis of gastrointestinal manifestations of CMA usually depends on a DBPCFC, investigators continue to study other options for confirming the diagnosis.