Plasma proteomic signature of frailty in 50,506 adults

作者：Xueqing Jia, Weijing Gao, Hampus Hagelin, Yanjie Zhao, Jingyun Zhang, Xingqi Cao, Liming Zhang, Youheng Wu, Lina Ma, Liangkai Chen, Liang Sun, Huan Guo, Jianhua Zhang, Juulia Jylhävä, Zhi Hu, Emiel O. Hoogendijk, Sara Hägg, Z D Liu · 发表于：Cell Metabolism · 年份：2026 · DOI：10.1016/j.cmet.2026.02.013 · 被引用次数：8 · 研究领域：Frailty in Older Adults、Genetics, Aging, and Longevity in Model Organisms、Nutrition and Health in Aging

Proteomics enables the systematic elucidation of biological mechanisms underlying frailty. Through a large proteome-wide association study of 2,911 plasma proteins from 50,506 UK Biobank participants, we identified 1,339 proteins significantly associated with frailty, highlighting collagen-containing extracellular matrix and vesicle lumen pathways. Replication in the TwinGene study confirmed partial but consistent associations. Mendelian randomization analyses identified five potentially causal proteins for frailty. Moreover, we developed a proteomic frailty score (PFS) that showed strong predictive performance for 199 incident diseases across 13 categories and broad responsiveness to 84 modifiable risk factors. Longitudinal analyses revealed accelerated PFS progression with advancing age and increasing baseline frailty severity. An online tool ( https://zipoa.shinyapps.io/frailty/ ) was created for public PFS calculation. Finally, we observed a biphasic pattern of frailty-associated proteomic dysregulation across the lifespan, with peaks at ages ∼50 and ∼63. Together, we establish PFS as a biomarker of biological aging while identifying critical windows and molecular targets for frailty interventions. • Delineate the most comprehensive plasma proteomic landscape of frailty • Develop different versions of proteomic frailty scores via the LASSO algorithm • PFS predicts multiple incident diseases and responds to modifiable factors • Uncover dynamic changes in frailty-associated...