|Home | About | Journals | Submit | Contact Us | Français|
Silvers et al say that our recent paper was written as if the study's main aim was to test the association between prolonged exclusive breast feeding and asthma and allergy.1 Every randomised controlled trial has a primary outcome that serves as the basis for estimating sample size requirements and for obtaining funding. But we also examined secondary outcomes in the first year of life and sought support for follow-up of the PROBIT cohort to assess important health outcomes at older ages that observational studies had associated with infant feeding. Cluster randomised allocation of the breastfeeding promotion intervention yielded two groups that differed greatly in the duration and exclusivity of breast feeding. This created a unique opportunity to study longer term child health outcomes, including growth, adiposity, neurocognitive development, behaviour, and dental caries, as well as asthma and allergy.2 3
In our recent BMJ paper reporting on the last two outcomes, my colleagues and I did not compare any breast feeding with no breast feeding.4 All PROBIT infants were breast fed at birth; the difference between the two randomised groups was limited to the duration and exclusivity of breast feeding. Our inference was thus not that “breast feeding has no effect,”1 but that prolonged and exclusive breast feeding did not protect against asthma and allergy. That inference is justified by the randomised design, intention to treat analysis, and observed results.
Finally, the wide confidence intervals noted by Silvers et al around the cluster adjusted odds ratios for the skin prick test results have nothing to do with “important confounding and predictor variables.”1 As shown in table 1, and as expected from the randomised allocation, the two groups had similar baseline characteristics.4 The wide confidence intervals are a function of the high degree of clustering for the skin prick test results (table 5). Clustering was far less evident for the allergic symptoms and diagnoses (table 4) and for the sensitivity analysis for the skin prick tests (table 6), with considerably narrower confidence intervals.
Competing interests: None declared.