Infant feeding and allergy prevention: a review of current knowledge and recommendations. A EuroPrevall state of the art paper.

The relationship between infant feeding patterns and the later development of food allergies has been the focus of much debate and research over the last decade. National recommendations have been made by many countries on how to feed infants to reduce the risk of food allergy but due to the lack of firm evidence the recommendations differ widely. This review has been developed as part of EuroPrevall, a European multicentre research project funded by the European Union, to document the differing feeding recommendations made across Europe, to investigate the current evidence base for any allergy prevention feeding recommendations and to identify areas where further research is needed. This review will also provide information which, when combined with the infant feeding data collected as part of EuroPrevall, will give an indication of compliance to national feeding guidelines which can be utilised to assess the effectiveness of current dissemination and implementation strategies.

Approaches to risk assessment in food allergy: report from a workshop ''developing a framework for assessing the risk from allergenic foods".

A workshop was organised to investigate whether risk assessment strategies and methodologies used in classical/conventional toxicology may be used for risk assessment of allergenic foods, to discuss the advantages and limitations of different approaches and to determine the research needed to move the area forward. Three possible approaches to safety assessment and risk assessment for allergenic foods were presented and discussed: safety assessment using NOAEL/LOAEL and uncertainty factors, safety assessment using Benchmark Dose and Margin of Exposure (MoE), and risk assessment using probabilistic models. The workshop concluded that all the three approaches to safety and risk assessment of allergenic foods should continue to be considered. A particular strength of the MoE and probabilistic approaches is that they do not rely on low-dose extrapolations with its inherent issues. Probabilistic modelling is considered to be the most promising approach for use in population risk assessment (which is a particular focus for risk managers). For all approaches, further improvement of input data is desirable, particularly data on consumption patterns/food choices in food allergic consumers, data on minimum eliciting doses and data that can be used to evaluate whether the whole population at risk has been modelled accurately. Specific research topics were identified.