Generalizability in Pragmatic Trials-Reply.

In Reply Our systematic review was 1 of 3 systematic reviews commissioned by the UK Food Standards Agency covering a range of early dietary exposures and risk of allergic or autoimmune disease.1 Many of the answers to the queries raised by Drs Nwaru and Virtanen can be found in the full Food Standards Agency report (https://www.food.gov.uk/science /research/allergy-research/fs305005b). For autoimmune conditions, others have reported on associations between cow’s milk exposure in infancy and risk of type 1 diabetes mellitus as well as timing of wheat or gluten exposure in infancy and risk of celiac disease or type 1 diabetes mellitus.2-4 There is less evidence for associations between other allergenic foods and autoimmune disease, but the environmental determinants of these diseases are largely unknown, so to inform national infant feeding guidance, we thought it was important to take a comprehensive and openminded approach. The definition of allergenic foods in our review was that used in the US Congress food allergen labeling law.5 We evaluated timing of introduction of other foods (including roots, fruits, and vegetables) in our third systematic review, currently under peer review. Nwaru and Virtanen highlight their DIPP cohort finding that early egg and fish introduction were associated with reduced risk of allergic sensitization to other allergens. We did include these data in our analyses, within the full Food Standards Agency report (“Allergic Sensitization Observational Data” section). The DIPP data for egg are not supported by some other observational studies such as the LISA cohort. The DIPP data for fish are supported by 2 other prospective cohort studies, and this very low-certainty evidence that early fish introduction is associated with reduced allergic sensitization was reported in the text and summary of findings (Table 4 in the article). In the intervention trials of allergenic food avoidance or early allergenic food introduction, we found no evidence that early introduction of one allergenic food can influence risk of either allergic sensitization or clinical allergy to a different allergenic food. We included wheeze data from the DIPP, SEATON, and Western Australian cohort studies, in addition to 27 other observational studies (eTable 4 in the Supplement to the article). The findings are in the Food Standards Agency report (“Wheeze Observational Data” section) and are summarized in eTable 11 in the Supplement to the article. Opportunities for pooling data were limited, partly because estimates of association reported as hazard ratios were meta-analyzed separately from odds ratios. We did not find consistent associations between timing of allergenic food introduction and wheeze across all the studies, and we found no consistent association in intervention trials. The LISA study did not report analyses of timing of introduction of the allergenic foods specified in our protocol and risk of wheezing. The investigators did report associations with timing of solid food introduction, which we have analyzed as part of our third systematic review. Finally, we did evaluate asthma as part of this systematic review but reported most asthma definitions within the category of recurrent wheeze, unless the definition of asthma was limited to a single episode of wheezing. Despo Ierodiakonou, MD, PhD Vanessa Garcia-Larsen, PhD Robert J. Boyle, MD, PhD