RESUMEN
BACKGROUND: Food specific immunoglobulin E (IgE) (fsIgE) cut points are used in the evaluation of food allergies. Concomitant measurement of total IgE (tIgE) is traditionally not obtained. We anecdotally observed elevations in fsIgE mirroring tIgE increases, which may confound accurate interpretation. OBJECTIVE: To determine whether changes in tIgE were associated with fsIgE and whether predictions of fsIgE could be formulated based on tIgE. METHODS: We studied children younger than 18 years who had both tIgE and fsIgE (egg, n = 136; milk, n = 123; peanut, n = 201; soy, n = 55) obtained simultaneously on 1 or more occasion between January 2008 and February 2011. After institutional review board approval, natural log-transformed (ln) tIgE and fsIgE levels were analyzed using univariate and multivariate regression models to assess associations and predict fsIgE using tIgE and other covariates. RESULTS: Soy IgE levels were strongly correlated (ρ = 0.85, P < .001), whereas egg, milk, and peanut IgE levels were substantially correlated (ρ = 0.69, 0.69, and 0.66, respectively, P < .001) with tIgE. A 1-unit increase in ln(tIgE) was significantly correlated with unit increases in ln(egg IgE) (0.77), ln(milk IgE) (0.84), ln(peanut IgE) (0.87), and ln(soy IgE) (0.89) (P < .001). The ln(tIgE)-based univariate models could predict fsIgE in the validation data with strong (soy) and substantial (egg, milk, and peanut) predictive ability (P < .001). CONCLUSION: Our study found significant and parallel relationships between tIgE and fsIgE levels to egg, milk, peanut, and soy. It underscores the importance of examining fsIgE levels in context of tIgE while making diagnostic and management decisions in children with food allergies.
