Whole population or high-risk group?

Allergen avoidance seems to be a logical way to protect high-risk children from the development of allergic diseases. If we assume that allergic diseases are predictable, we firstly need to define which allergic disease and what age we are discussing. Then the predictive capacity of risk factors, their clinical value and their prevalence in the population has to be considered. In newborns, the positive predictive value of family history and an elevated cord blood immunoglobulin E for atopic disorders in the first 2 yrs of life was well under 50% in high-risk cohorts. In the multicentre allergy study (MAS)-90 in Germany, 40% of children 3-5 yrs of age whose parents both had asthma, also had asthma themselves. However, when extrapolated to the population as a whole, most children with asthma would have no parent with asthma. Risk factors suitable for secondary prevention, such as presence and persistence of food antibodies or early atopic dermatitis together with an atopic family history may predict up to 80% of aeroallergen sensitization by 5 yrs, However, the sensitivity of these factors is also low. We therefore need better predictors for asthma and a policy for intervention strategies according to age and risk.