Wouldn't you figure that a randomized controlled trial published in the New England Journal of Medicine would yield information that is immediately useful in the clinical setting? Guess again. A combination of study design issues, and, likely, the wonderful but maddening inconsistencies in biology, frustrate us again.
Source: Perkin MR, Logan K, Tseng A, et al. Randomized trial of introduction of allergenic foods in breast-fed infants. N Engl J Med. 2016; 374:1733-43. doi:10.1056/NEJMoa1514210. See AAP Grand Rounds commentary by Drs. Neal LeLeiko and Michael Herzlinger (subscription required).
PICO Question: Among breastfed infants, does the introduction of allergenic foods at 3 months of age protect against the development of food allergy as compared those who exclusively breastfeed until approximately 6 months of age?
Question type: Intervention
Study design: Randomized controlled trial
As a sort of bookend to last week's post on peanut allergy, some of the same investigators present another facet to food allergy, namely whether early introduction of potential allergens can influence future development of food allergy. They enrolled a little over 1000 exclusively breast-fed infants who were then randomized to received early introduction of allergenic foods (cow's milk, peanut, egg, sesame, and whitefish; the early introduction group [EIG]), versus a standard introduction group (SIG) who received only breast milk until at least 5 months of age, then introduced foods as they wished.
A big problem was discovered at the end of the study: SIG infants adhered to the study protocol at a very high rate (93%), but EIG infants had only a 43% adherence rate. Of course this difference could seriously alter the strength of any conclusions. The authors tried to mitigate the effects of the situation, but it still leaves me with too many doubts to use their results in clinical practice. How did this happen?
If you look at how the study unfolded, it's not a surprise that large differences in adherence might result. First, this wasn't a blinded study. Although the infants were randomized to either SIG or EIG groups, both the families and the investigators knew which children were in which group. It would have required a great deal of effort to keep everyone blinded to group assignment; the investigators would have had to come up with identical food challenge capsules, or similar device, containing placebo for the SIG children and the specific allergen for the EIG enrollees.
They didn't do that. Furthermore, EIG enrollees underwent skin-prick testing for food allergy at the start of the study, whereas SIG enrollees did not. Ostensibly, the researchers didn't want to do this for the SIG patients because knowing the results could have influenced how parents introduced allergenic foods later on. This is certainly true, but this likely did influence both adherence and later food introduction in the EIG children and is yet another difference in how the 2 groups were managed.
The ideal randomized controlled trial compares 2 groups who are identical throughout the course of the study, except for the intervention being studied (in this case, early introduction of allergenic foods). The markedly different adherence rates between the 2 groups in this study make the results difficult to swallow (pun intended).
