Prevalence and risk factors for milk allergy overdiagnosis in the BEEP trial cohort

Abstract

AbstractBackgroundCow's milk allergy (CMA) overdiagnosis in young children appears to be increasing and has not been well characterised. We used a clinical trial population to characterise CMA overdiagnosis and identify individual‐level and primary care practice‐level risk factors.MethodsWe analysed data from 1394 children born in England in 2014–2016 (BEEP trial, ISRCTN21528841). Participants underwent formal CMA diagnosis at ≤2 years. CMA overdiagnosis was defined in three separate ways: parent‐reported milk reaction; primary care record of milk hypersensitivity symptoms; and primary care record of low‐allergy formula prescription.ResultsCMA was formally diagnosed in 19 (1.4%) participants. CMA overdiagnosis was common: 16.1% had parent‐reported cow's milk hypersensitivity, 11.3% primary care recorded milk hypersensitivity and 8.7% had low‐allergy formula prescription. Symptoms attributed to cow's milk hypersensitivity in participants without CMA were commonly gastrointestinal and reported from a median age of 49 days. Low‐allergy formula prescriptions in participants without CMA lasted a median of 10 months (interquartile range 1, 16); the estimated volume consumed was a median of 272 litres (26, 448). Risk factors for CMA overdiagnosis were high practice‐based low‐allergy formula prescribing in the previous year and maternal report of antibiotic prescription during pregnancy. Exclusive formula feeding from birth was associated with increased low‐allergy formula prescription. There was no evidence that practice prescribing of paediatric adrenaline auto‐injectors or anti‐reflux medications, or maternal features such as anxiety, age, parity and socioeconomic status were associated with CMA overdiagnosis.ConclusionCMA overdiagnosis is common in early infancy. Risk factors include high primary care practice‐based low‐allergy formula prescribing and maternal report of antibiotic prescription during pregnancy.