Clinical records of 286 consecutive patients reacting positively with objective symptoms to double-blind, placebo-controlled oral peanut challenges at University Hospital, Nancy, France were examined for individual No Observed Adverse Effect Levels (NOAELs) and Lowest Observed Adverse Effect Levels (LOAELs). After fitting to a log-normal probability distribution model, the ED(10) and ED(05) were 14.4 and 7.3mg (expressed as whole peanut), respectively, with 95% lower confidence intervals of 10.7 and 5.2mg, respectively. Compared to results from a previous study where the ED(10) was based upon individual peanut thresholds gleaned from 12 publications, a statistically significant difference was observed between the ED(50)'s, but not the ED(10)'s of the two probability distribution curves. The Nancy patient group contains more sensitive subjects than the group from the published literature thus contributing to the observed differences. Minimum eliciting dose-distributions for patients with histories of more severe reactions (grade 4 or 5; 40 subjects) did not differ significantly from those of patients with histories of less severe reactions (grades 1-3; 123 subjects). These data and this modeling approach could be used to establish population thresholds for peanut-allergic consumers and thereby provide a sound basis for allergen control measures in the food industry.
Copyright (c) 2009 Elsevier Ltd. All rights reserved.