Early-life infections and childhood cancer risk: a nationwide cohort study of 2 million children.

BACKGROUND

The relationship between early-life infections and childhood cancer risk remains unclear despite its importance for risk stratification and clinical surveillance.

METHODS

We conducted a nationwide population-based cohort study using the South Korean National Health Information Database. Children born between January 1, 2002, and December 31, 2006, were followed through December 31, 2019. Two analytic cohorts were defined: Cohort 1 (n = 1,996,174) evaluated the number of infection episodes during ages 0–4 years, with a restricted analysis of infections occurring between ages 2–4 years; Cohort 2 (n = 1,999,065) evaluated infections during ages 0–1 year. Infection episodes were defined using healthcare visits with relevant ICD-10 diagnostic codes. Incident childhood cancer was ascertained from claims records. Multivariable Cox models estimated hazard ratios (HRs) adjusting for sex, residence, income quintile, and birth year.

RESULTS

Hazard ratios (HRs) were estimated relative to the lowest infection frequency category for each infectious disease. Upper respiratory tract infections, pneumonia, and influenza were not associated with increased cancer risk. At ages 0–1 year, risk increased with 2–4 lower respiratory infections (HR 1.06, 95% CI 1.00–1.12), ≥ 2 enteric infections (HR 1.11, 95% CI 1.01–1.22), ≥ 1 urinary infection (HR 1.21, 95% CI 1.09–1.33), and ≥ 2 mycoses (HR 1.19, 95% CI 1.03–1.38). At ages 2–4 years, risk remained elevated for ≥ 1 urinary infection (HR 1.13, 95% CI 1.02–1.25) and for skin infections (HR 1.09, 95% CI 1.02–1.17 for one episode; HR 1.13, 95% CI 1.03–1.24 for ≥ 2 episodes). Hospitalization showed a dose–response pattern: for ages 0–4 years, HRs were 1.07 (95% CI 1.00–1.14) for one admission, 1.17 (95% CI 1.04–1.32) for two, and 1.30 (95% CI 1.14–1.47) for ≥ 3; for ages 0–1 year, 1.07 (95% CI 1.00–1.15) for one and 1.22 (95% CI 1.05–1.41) for ≥ 2; for ages 2–4 years, 1.12 (95% CI 1.04–1.21) for one and 1.18 (95% CI 1.03–1.36) for ≥ 2. Limitations include potential outcome and exposure misclassification inherent to claims data, unmeasured confounding (e.g., genetic susceptibility, environmental factors), and use of hospitalization as a proxy for infection severity.

CONCLUSIONS

Infections managed in outpatient settings were not associated with increased cancer risk, whereas infections requiring hospitalization—used as a proxy for more severe infections—were associated with higher cancer incidence and showed a dose–response pattern. Infection severity and frequency may be relevant to childhood cancer development and risk stratification.