Background In food allergy, eliciting doses (EDs) of foods on a population level can improve risk management and labeling strategies for the food industry and regulatory authorities. Previously, data available for walnut were unsuitable to determine EDs. Objective The objective of this study was to determine EDs for walnut allergic adults and to compare with previously established threshold data for peanut and tree nuts. Methods Prospectively, adult subjects with a suspected walnut allergy underwent a low-dose double-blind, placebo-controlled food challenge. Individual no observed and lowest observed adverse effect levels were determined and log-normal, log-logistic, and Weibull models were fit to the data. Estimated ED values were calculated for the ED5, ED10, and ED50, the dose respectively predicted to provoke an allergic reaction in 5%, 10%, and 50% of the walnut allergic population. Results Fifty-seven subjects were challenged and 33 subjects were confirmed to be walnut allergic. Objective symptoms occurred in 20 of the positive challenges (61%). The cumulative EDs in the distribution models ranged from 3.1 to 4.1 mg for the ED05, from 10.6 to 14.6 mg walnut protein for the ED10, and from 590 to 625 mg of walnut protein for the ED50. Conclusions Our data indicate that population EDs for walnut are slightly higher compared with those for peanut and hazelnut allergy. Currently available data indicate that the ED values for hazelnut could be used as a conservative temporary placeholder when implementing risk management strategies for other tree nuts where little or no food challenge data are available.
