Idiopathic epilepsy burden in Asian children (1990-2023) and projections to 2045: A global burden of disease study analysis.

OBJECTIVE: To assess the burden of childhood idiopathic epilepsy across five Asian regions and 34 countries from 1990 to 2023, and to project future trends through 2045 using time-series modeling. METHODS: Data on disability-adjusted life years (DALYs), deaths, incidence, prevalence, years lived with disability (YLDs), and years of life lost (YLLs) were extracted from the Global Burden of Disease (GBD) 2023 database for the cause category "idiopathic epilepsy" in children aged younger than 1 year to 19 years. Age-specific rates per 100,000 population were analyzed across five age groups and stratified by sex. Regional and country-level heterogeneity was examined through descriptive analyses and visualizations. ARIMA-based forecasts to 2045 were evaluated using temporal holdout validation, rolling-origin validation, candidate-model comparison, Ljung-Box residual tests with Benjamini-Hochberg correction, and prespecified stability flags. RESULTS: In 2023, South Asia demonstrated the highest burden, with infant DALY rates reaching 789.8 per 100,000, while High-income Asia Pacific showed the lowest rates. Between 1990 and 2023, mortality rates declined by 20-90% in East Asia, High-income Asia Pacific, and South Asia. However, prevalence and incidence rates remained stable or increased modestly. Males consistently exhibited higher burden than females. Country-level analyses revealed substantial within-region heterogeneity. Forecast validation showed that only 5 of 150 regional age-specific series met all prespecified stability criteria. Therefore, long-term projections were interpreted cautiously as exploratory. CONCLUSIONS: Childhood idiopathic epilepsy burden in Asia demonstrates substantial regional heterogeneity. Despite significant mortality reductions, persistent prevalence and disability burden underscore the need for strengthened prevention strategies and enhanced epilepsy management programs. Long-term forecasts should be interpreted cautiously because most series showed limited out-of-sample robustness.