Association between the use of infant formula and risks for infants' health is seldom studied in western countries. We set up a historical cohort based on record linkage analysis, combining the data from administrative databases providing individual data. Infants receiving the second dose of pediatric immunization between 2015 and 2017 were included. The main outcome measure was antibiotic prescriptions from enrolment up to 24 months of age, by infant feeding category at enrolment. The extended Cox regression technique was used to account for recurrent events. The infants' cohort included 40,258 5-month-old infants; during the study period, 60,932 antibiotic prescriptions were filled. Compared with infants fully breastfed, children fed with both maternal milk and formula received 106 more antibiotic prescriptions per 1000 children/year, whereas infants receiving formula only had 138 excess prescriptions per 1000 children/year. The association with infant feeding was statistically significant and remained unchanged after adjustment for common confounders (adjusted hazard ratio, HR, for complementary feeding vs full breastfeeding 1.09; 95%CI 1.05 to 1.12; formula only versus full breastfeeding adj. HR 1.12; 95%CI 1.08 to 1.16).Conclusion: In our cohort, we observed a positive association between infant formula use considered a proxy of infections antibiotic prescription rate, considered a proxy of infections. The association followed a gradient.What is Known:• Formula feeding is associated with increased morbidity and mortality even in western countries, but still, it is common.• Information on formula are seldom unbiased; thus, public perception of risks is distorted.What is New:• In a large Italian cohort of infants, formula feeding at 5 months of age results to be associated with an increased rate of antibiotic prescription (considered to be a proxy of infection) up to 24 months of age: the association follows a dose-response relationship.• Record linkage analysis using administrative databases provides useful information at a limited cost.
Keywords: Breastfeeding; Infections; Italy; Record linkage analysis.