J Am Heart Assoc. 2026 Jul 8:e047762. doi: 10.1161/JAHA.125.047762. Online ahead of print.
ABSTRACT
BACKGROUND: Breast cancer (BC) and atrial fibrillation (AF) represent increasing global health burdens with shared risk factors. However, their coincidence burden and global distribution among older women (≥55 years) remain unclear.
METHODS: This study integrated data from the Global Burden of Disease 2021 database spanning 204 countries and territories, covering incidence rates of BC and AF and exposure rates of 58 risk factors. Using a 4-quartile classification algorithm, we divided the globe into 3 regions (consistent, BC-dominant, and AF-dominant area). Risk factors were screened by using machine learning models (random forest, spatial statistics, and SHAP interpretability framework). Population-attributable fraction and a composite risk index model were generated to evaluate disease burdens and spatial risk localization of BC and AF.
RESULTS: The consistent area (80 countries, 39.60%) exhibited overlapping BC-AF incidence, while BC-dominant (65 countries, 32.18%) and AF-dominant (57 countries, 28.22%) regions highlighted geographic disparities. After screening of machine learning models, 2 shared factors (smoking and alcohol use) and 2 BC-specific factors (high body mass index and low physical activity) were included in the calculation of population-attributable fraction, potentially reducing global BC and AF incidence by 29.05% and 11.75% after shifting these factors to theoretical minimum exposure levels. For most factors, high-risk zones clustered in North and South America, Europe, and Oceania mainly correlate with dietary and lifestyle patterns.
CONCLUSIONS: The spatial coincidence of BC and AF present a significant global health challenge. Addressing both shared and region-specific risk factors can substantially reduce the dual burden of these diseases.
PMID:42417071 | DOI:10.1161/JAHA.125.047762
