Abstract
Objective: To investigate the association between recurrent childhood and adolescent antibiotic exposure and the risk of adult-onset pneumonia in individuals aged 40 and over, using data from the United Kingdom Biobank (UKB). Methods: A total of 170 496 UKB participants without a history of pneumonia at baseline were included. Based on a baseline questionnaire, participants were categorized into an exposed group (recurrent early-life antibiotic use), a non-exposed group, or an unknown-exposure group. Cox proportional hazards regression models were applied to calculate HRs and their 95%CIs for pneumonia incidence, with the non-exposed group as the reference. Subgroup analyses were conducted to test for effect modification and heterogeneity, and the robustness of the findings was assessed through multiple sensitivity analyses. Results: During a median follow-up of 13.8 years, 6 041 incident pneumonia cases were recorded. After adjusting for major sociodemographic and lifestyle factors, a significantly increased risk of pneumonia was observed in both the exposed group (HR=1.42, 95%CI: 1.32-1.52) and the unknown-exposure group (HR=1.21, 95%CI: 1.11-1.31) compared with the non-exposed group. Subgroup analysis revealed that this association was stronger in individuals with baseline cancer but weaker in those with chronic obstructive pulmonary disease (COPD) (both P for interaction <0.05). The results remained robust across multiple sensitivity analyses, and the association varied with increasing age at pneumonia onset. Conclusions: Recurrent childhood and adolescent antibiotic exposure is associated with a significantly increased risk of adult-onset pneumonia in individuals aged 40 and older. Baseline health status, such as the presence of cancer or COPD, may act as an effect modifier for this association, and its strength appears to be influenced by age of onset.</p>