No evidence that introduction of allergenic foods prior to 6 months of age increases reports of wheeze or eczema in young childhood
- Correspondence to S Hasan Arshad
Department of Clinical and Experimental Sciences, Southampton General Hospital, Tremona Road, Southampton, Hants SO16 6YD, UK;
Implications for practice and research
A delay in the introduction of solid foods beyond 6 months does not protect from allergic disease and should not be recommended.
A randomised controlled trial (RCT) investigating the specific effect of early versus delayed introduction of solid foods is needed.
A delay in introduction of solid foods has been suggested for decades as a means of preventing atopic diseases due to infants' immature immune system and hence inability to handle a wide variety of food antigens.1 The normal response to food antigens is that of immune tolerance. However, early exposure to significant amounts of food antigens might result in abnormal or allergic immune responses. This is important as restricting appropriate solid foods can lead to inadequate nutrition, especially if it is for longer than 6 months. Early observational studies were supportive of this concept, more so in reducing eczema and food allergy than asthma or rhinitis.2 On this basis, a number of allergy prevention guidelines recommend delaying the introduction of solid foods for 4–6 months after birth.3 However, RCTs are lacking and, hence, the quality of the evidence for this recommendation remains poor. A systematic search of the literature in 2010 found only two poor-quality controlled trials.3 Primary prevention trials have generally combined the delay in introduction of solid foods with other preventive measures such as breastfeeding, restriction of maternal diet, or avoidance of other allergens and, therefore a positive outcome cannot necessarily be regarded as due to avoidance of solid foods.4
Tromp et al conducted a large birth cohort, named ‘The GenerationR study’ with a sample size of nearly 7000 preschool children. Although the attrition rate was relatively high with a loss to follow-up of 30–40%, they managed to enlarge the sample size back to the original 6905 children by using statistical method of imputation of data. When the children were 6 and 12 months old, parents were asked about introduction of common foods including cow's milk, egg, peanut, tree nuts soy and gluten. When children were aged 2, 3 and 4 years, information was collected on the presence of asthma or eczema using standardised questionnaires.
The authors looked at the role of introduction of solid foods, in the context of duration of breastfeeding, and on the development of allergic disease. They adjusted for a number of environmental factors that may influence the outcome. More importantly, the sample was stratified for atopic heredity and infants' (reported) cows' milk allergy. These factors may truly modify the effect of solid food introduction, as the child's immune system may respond differently in those with personal or family history of atopy.
The authors concluded that the introduction of common allergenic foods such as cows' milk, egg, nuts, soy and gluten before the age of 6 months does not increase the risk of eczema or wheeze up to the age of 4 years. This effect was independent of parental atopy, the presence of cow's milk allergy in the child and a number of other confounders.
This study supports many previous observational studies and current guidelines that a delay in the introduction of solids foods beyond 6 months does not protect from eczema or wheeze, irrespective of the family history of atopy and should not be recommended. This is significant as it has implications for children's nutrition. As the sample is large and data were prospectively collected, the conclusions are sound. However, the possibility of a reporting bias causing reverse causation cannot be totally excluded. This is because parents of children, or the child, with an atopic disease in the family, may recognise and report symptoms of eczema and wheeze more often than parents of children without such history. Being an observational study, using questionnaire data, it also suffers from the possibility of unknown confounders, misclassification of disease and reporting bias. Thus, the level of evidence remains moderate at best.