Lowest Observed Adverse Effect Level

“Highly allergic” or “severely allergic” can mean low threshold for reacting, or severe previous reaction, or both!

Lowest observed adverse effect level is the formal term for the minimum dose at which you react! An eliciting dose (ED) can then be calculated through multiple food challenges for a stated proportion of the allergic population. For example, 1 peanut =150mg protein. ED 1 (where 1% of allergic population will react) is 1.3mg. But range of reported values – type of peanut allergy? Entry criteria? Boiled or roasted? Criteria for stopping challenge?

Higher eliciting dose predicts future tolerance, at least for milk and peanut, maybe not for egg.

Children probably more sensitive although ED 10 similar for children and adults.

Conflicting evidence on severity of reaction vs threshold.  Certainly trace amounts can cause anaphylaxis in some cases.  Higher fat content delays absorption, so the reaction (when it happens) can still be severe…

ED05 of 1.5mg for peanut seems safe [Hourihane].

Being cross allergic to other things might help predict – in French study, 3 phenotypes emerged from cluster analysis:

  • Cluster 1 have high level of rAra h 2 (mean 81), low threshold reactive doses for peanut and high proportion of asthma;
    • Cluster 2 (mostly boys), have high threshold, milder symptoms, and the lowest proportion of asthma/AR and cross-allergy to TN and/or legumes;
  • Cluster 3 have low Ara h 2, high risk of cross-allergy to TN and/or legumes, and most patients suffer from eczema.  [Matthias Cousin, Lille, PAI]

Complicated though – thresholds vary up to 10x for individuals in sequential challenges (n=14)!

On the day factors – alcohol/meds, infection.  Exercise is a factor in 15-20% of anaphylaxis episodes according to German/Austrian registries.

TRACE peanut study – 45% reduction in threshold for exercise and sleep deprivation (independent) esp at lower eliciting doses. ED1 and ED5 remain above 0.2 though.

Sleep deprivation (sleep overs!) worsened severity of reactions too, by 48% – exercise 28%, not significant. Repeated challenges also seemed to increase severity. [score for severity using Practall criteria. JACI 2022 Dua]

Lower thresholds had higher BAT and SPT – 8mm cut off had 100% sensitivity and NPV for severe reactions, and 92% specificity. Has nomogram combining all tests! SPT 6mm had 92% sensitivity, 95% specificity and 100% NPV for low threshold.  [JACI 2020, Santos study – LEAP cohort]

Particulate contamination risk seems to be a particular problem for milk in chocolate – big range of values found cf peanut.

UK and Europe look at processes as the way to assess risk of contamination with allergens. But not consistent between countries of the EU, let alone globally.

Quantitative methods would  appear to make sense – how much allergen is actually present in a given sample? But of course one sample may differ from another (“particulate contamination”).  Also difficult to establish minimum eliciting or threshold dose (consider denaturation of the allergen during processing, derived ingredients eg glucose syrup from wheat, soya lecithin, effect of food matrix, individual factors). 

Australia and NZ use lowest observable adverse event level (LOAEL) already – warnings required within 10x concentration of LOAEL (“VITAL” threshold, as in toxicology).  Studies from Europe and US have found most advisory warnings used for ingredients below the VITAL threshold.

But then it depends on how much you eat, as well as presence of other co-factors that are known to contribute to risk of anaphylaxis.

[BMJ 2011;343:830][Allergy 2021, Paul Turner]