Introduction:

Poor cardiovascular health has been linked to structural brain changes, including brain atrophy, cerebral microbleeds, total brain volume, and white matter hyperintensities, which have also been shown to play a role in dementia. Govindarajan et al. (2025) created machine learning derived MRI Spatial Patterns of Abnormalities for Recognition (SPARE) scores for the cardiometabolic and vascular risk factors (CVMs) of diabetes, hyperlipidemia, hypertension, obesity, and smoking. Scores reflect how strongly one’s brain structure expresses the risk-factor–related patterns of brain atrophy. The proteome’s role in mediating the relationship between CVMs and brain atrophy patterns requires further investigation. In this study, we identified protein mediators of the CVM-specific brain atrophy, leveraging proteomic and MRI data from multiple cohorts including the UK Biobank (UKB), Global Neurodegeneration Proteomics Consortium (GNPC), and the Baltimore Longitudinal Study of Aging (BLSA).