Real-world data from the Japanese National Health Insurance System enable fine phenotyping in a 14K-scale population-based study

The use of medical databases, known as real-world data (RWD), enables accurate and efficient estimation of disease prevalence in population-based studies, making it a potential game-changer in epidemiology. However, the lack of standardized data formats across hospitals complicates integration across institutions. In Japan, health insurance claims data are standardized under a nationally unified format, providing a reliable source of structured RWD. We evaluated the utility of insurance claims data in epidemiological research. Incorporating both diagnosis and prescription information into case definitions resulted in four- and six-fold increases in the estimated prevalence of Alzheimer’s disease (AD) and Parkinson’s disease (PD), respectively, compared with conventional self-reported definitions. Subsequent genome-wide association studies (GWAS) for AD showed increased model log-likelihood and identified a characteristic APOE signal, findings observed only with extended case definitions. The APOE effect size was consistent with large case–control studies, while standard errors remained comparable to smaller studies. These results indicate that claims-based phenotyping improves case identification without loss of accuracy and supports scalable approaches for genomic epidemiology and public health surveillance.