Background Infant feeding guidelines in Australia changed in 2016 to recommend introduction of common allergy causing foods by age 1 year to prevent food allergy. Although the majority of Australian infants now eat peanut and egg by age 6-months, some still develop food allergy despite introducing allergens early. Objective We aimed to describe the prevalence of food allergy in a cohort recruited after introducing the nation-wide allergy prevention recommendations; identify characteristics of infants who developed allergy despite early introduction of allergens; and estimate the causal effect of modifiable exposures on food allergy prevalence and whether this differed between infants who were introduced to allergen before or after age 6 months. Methods We recruited a population-based sample of 12-month-old infants in Melbourne, Australia. Infants had skin prick tests to 4 foods and parents completed questionnaires. Infants with evidence of sensitisation were offered oral food challenges. Prevalence estimates were adjusted using inverse probability weighting. Results In a cohort of infants (n=1420) where nearly all infants had been introduced to common allergens such as egg, milk and peanut by one-year-of age, the prevalence of food allergy remained high at 11.3% (95% CI 9.6-13.4%). Infants who developed food allergy despite introducing the allergen by age 6-months were more likely to have Asian-born parents. Early-onset moderate/severe eczema was associated with an increased odds of food allergy, irrespective of whether allergens were introduced before or after age 6 months. Among infants who were introduced to peanut ≤6m, antibiotic use by age 6 months was associated with an increased odds of peanut allergy (aOR 6.03 (95%CI 1.15-31.60). Conclusion In a cohort where early allergen introduction was common, the prevalence of food allergy remained high. Infants who developed food allergy despite introducing the respective allergen by 6 months were more likely to have Asian parents and early-onset eczema. New interventions are needed for infants with a phenotype of food allergy that is not amenable to early allergen introduction.

Prevalence and determinants of food allergy in the era of early allergen introduction: the EarlyNuts population-based study