Effects of a dietary and environmental prevention programme on the incidence of allergic symptoms in high atopic risk infants: three years' follow-up.

A prospective case-control study is presented to assess an allergy prevention programme in children up to 36 months of age. Infants born at three maternity hospitals were followed from birth: 279 infants with high atopic risk (intervention group) were compared with 80 infants with similar atopic risk but no intervention (non-intervention group). The intervention programme included dietary measures (exclusive and prolonged milk feeding diet followed by a hypoantigenic weaning diet) and environmental measures (avoidance of parental smoking in the presence of the babies, day care > 2 years of life). Mothers in this group who had insufficient breast milk were randomly assigned to one of two coded formulas: either a hydrolysed milk formula (Nidina HA, Nestlé) or a conventional adapted formula (Nan, Nestlé). Other environmental measures remained the same as for the breastfeeding mothers. The non-intervention group were either breastfed or received the usual Italian milk feeding and weaning diet, without environmental advice. The main outcome measures were anthropometric measurements and allergic disease manifestations. Normal anthropometric data were observed both in the intervention group and in the non-intervention group. The incidence of allergic manifestations was much lower in the intervention group than in the non-intervention group at 1 year (11.5 versus 54.4%, respectively) and at 2 years (14.9 versus 65.6%) and 3 years (20.6 versus 74.1%). Atopic dermatitis and recurrent wheezing were found in both the intervention group and the non-intervention group from birth up to the second year of life, while urticaria and gastrointestinal disorders were only present in the non-intervention group in the first year of life. Conjunctivitis and rhinitis were present after the second year in both the intervention group and the non-intervention group. Relapse of the same allergic symptom was less in the intervention group (13.0%) than in the non-intervention group (36.9%). In comparison to the non-intervention group, there were fewer intervention group cases with two or more different allergic symptoms (8.7 versus 32.6%), and they were more likely to avoid steroid treatment (0 versus 10.8%) and hospital admission (0 versus 6.5%). Babies in the non-intervention group fed with adapted formula were more likely to develop allergies than breastfed babies in the same group. In the intervention group the breastfed infants had the lowest incidence of allergic symptoms, followed by the infants fed the hydrolysed formula (ns). Infants in the intervention group fed the adapted formula had significantly more allergies than the breastfed and hydrolysed milk fed infants, although less than their counterparts in the non-intervention group. Of the affected subjects in the intervention group, 80.4% were RAST and/or Prick positive to food or inhalant allergens. Total serum IgE values detected at birth in the intervention group were not predictive, but at 1 and 2 years of age, IgE values more than 2 SD above the mean in asymptomatic babies were found to predictive for later allergy. In breastfed babies the total IgE level at 1 and 2 years of age was lower than in the other two feeding groups. Of the various factors tested in the non-intervention group, the following were the most important in the pathogenesis of allergic symptoms: (i) formula implementation begun in the first week of life; (ii) early weaning (< 4 months); (iii) feeding beef (< 6 months); (iv) early introduction of cow's milk (< 6 months); and (v) parental smoking in the presence of the babies and early day care admission (< 2 years of life). All the preventive measures used in this study (exclusive breastfeeding and/or hydrolysed milk feeding, delayed and selective introduction of solid foods, and environmental advice) were effective at the third year of follow-up, greatly reducing allergic manifestations in high atopic risk babies in comparison with those not receiving these interventions