TRACE

Food allergy is a common condition, affecting 5-8% of the childhood population and 2-4% of adults and the most frequent cause of fatal allergic reactions. Consequently, it is a problem of significant public health importance. We currently lack a disease-modifying therapy and hence, once diagnosed, food allergic individuals are advised to avoid the problem food(s) and in some instances are prescribed emergency medication for accidental ingestion.

To assist allergic consumers to practice food avoidance, EC legislation requires that certain important allergenic foods be labelled irrespective of the level they are added to a food. Such labelling laws govern deliberately added ingredients, and whilst Agency guidelines exist for managing the inadvertent presence of allergenic constituents, there is an increasing tendency for manufacturers to use precautionary labelling which consumers often find unhelpful and ignore, putting themselves at risk.

One of the difficulties in managing such contamination is that we do not know with certainty what level of contamination would be tolerated by the vast majority of allergic individuals and would therefore be ‘acceptable’.  Challenge thresholds vary between individuals by 4-5 orders of magnitude, and often do not correlate with clinical severity (e.g. low dose thresholds occur in the ‘milder’ oral allergy syndrome). No observed adverse effect levels (NOAEL) and lowest observed adverse effect levels (LOAEL) may be established for certain foods at a population level, but these parameters are not appropriate levels to inform risk management by industry as they impose unnecessarily low management thresholds, exacerbating the problem being addressed.

More preferable are probabilistic approaches which use the distribution of minimum eliciting doses (MEDs) together with consideration of the distribution of cross-contact allergen in food products to quantify the risk to public health. Investigation of changes in the distribution of MEDs may thus prove more suitable to compare the effect of extrinsic factors as they are more amenable to standard statistical methods. Peanut dose-distribution curves and MEDs are being established by EuroPrevall in other European countries, but this population is probably distinct from the UK population because of a high prevalence of oral allergy syndrome with cross-reacting antibodies to peanut.

Thresholds and severity may not be comparable to the UK peanut-allergic population. There are no published data to inform MEDs for peanut in the UK population. Regardless of the approach, translating MED data from challenges into safety thresholds for use by industry requires multiplication by an ‘uncertainty factor’ to account for intra-individual variation within the population. There are no published studies to help quantify this natural variability in threshold dose over time, within individuals. Further variability is thought to be caused by a number of extrinsic factors e.g. stress, exercise, alcohol and/or acute viral infections, identified at a recent Food Standards Agency stakeholder meeting. Clinical data supports this for a number of triggers, particularly asthma.

Translation of MED into acceptable levels of allergen contamination for the population requires consideration of a 'safety factor', to account for individual variability in dose threshold and severity. Data suggest such variability depends in part on extrinsic factors (including, exercise and sleep restriction). Each factor may have a different effect on scale and direction. We are a UK consortium, including clinical centres providing expertise in food challenges and EuroPrevall with expertise and materials to perform challenges and analysis. 100 adults will enter a cross-over trial, each undergoing a baseline peanut DBPCFC followed by three challenges with extrinsic factors in random order (repeat baseline, exercise, sleep restriction). These data will define MED for the UK population and a safety factor derived from a shift in the threshold, to inform industry and protect the allergic population.

The purpose of this study is to investigate whether common extrinsic factors reduce thresholds of reactivity and quantify the effect of those factors on the clinical severity in a group of representative food-allergic adults and children, as indicated by changes in threshold doses (individual NOAEL and LOAELs, distribution of MEDs) and severity of allergic reactions. Pilot work will determine the feasibility of studying extrinsic factors using volunteers to study (exercise and sleep restriction). The main study is a 52-week randomized cross-over peanut-challenge design. After an initial screening baseline double-blind, placebo-controlled food challenge (DBPCFC) without extrinsic factors, each subject will be randomized to one of six arms. In each arm, participants will then undergo a further three challenges (one baseline, and two with extrinsic factors- active arm only) at 12-week intervals. The order of challenges between the six groups will be balanced by employing a Latin square design for a six–by-three crossover trial of 72 subjects.