Scripties UMCG - Rijksuniversiteit Groningen
 
English | Nederlands

The Eliciting Dose Curves of Placebo Reactions During Milk and Peanut Double-Blind Placebo-Controlled Food Challenges in Children

(2018) Bakhsh, F.R. (Faisal)

Introduction:
Food allergy is an important health issue because of its risks on health and its increasing
prevalence. There are different methods to diagnose food allergy; the double-blind placebocontrolled
food challenge (DBPCFC) is the gold standard to diagnose food allergy. Several
factors influence patient’s allergic reaction, one of which is eliciting dose, i.e. the dose of
allergen to which a patient reacts. Previous population threshold studies were based on verum
reactions, and little is known on placebo threshold distribution curves.
Objectives:
To produce threshold distribution curves based on placebo reactions in milk and peanut
challenges, to compare placebo threshold curves to previous verum threshold curves and
assess whether there is a need to modify the interpretation of verum threshold curves.
Methods:
Only double blinded challenges were included. Placebo reactions were classified into:
‘objective only’ and ‘any’ reactions (includes subjective symptoms). Interval-censoring
survival analysis was used for data analysis and to determine subjects’ no observed and low
observed adverse effect levels. The Statistical Analysis Software (SAS) was used to fit
parametric models to the interval-censored data based on results from previous studies. Pvalue
of 0.05 and 95% CI were used to determine significance
Results:
Placebo ‘objective only’ reactions occurred in 66 and 58 in milk and peanut challenges,
respectively. Placebo ‘any’ reactions occurred in 67 and 58 in milk and peanut challenges. In
milk and peanut challenges, there was no clumping of placebo reactions at the first or last
dose. Although a significant difference was predicted in only one distribution model for
objective reactions in milk (i.e. Weibull p=0.005), the difference was of no biological
relevance indicated by the 95% CI.
Conclusion:
No significant differences were predicted when comparing placebo to verum distribution
curves. Placebo reactions occur across the entire dose range in challenges. Therefore, there is
no need to correct or re-interpret verum population threshold distribution curves.






 
To top