Artificial Intelligence and Blood Tests: Unlocking the Secrets of Aging and Metabolic Health

Recent advancements in medical research have introduced innovative methods to assess aging and physical health through blood analysis combined with machine learning. A team led by researchers from the University of Vienna and Nankai University has employed metabolomics—a technique to analyze small molecules in blood—and cutting-edge AI models to uncover molecular markers linked to active aging.

Their study, published in npj Systems Biology and Applications, demonstrates that a specific metabolite, aspartate, serves as a key biomarker of physical fitness. By analyzing blood samples from older adults, the researchers identified a strong correlation between blood metabolite profiles and physical performance metrics such as endurance, strength, and balance. They developed a comprehensive 'Body Activity Index' based on physical tests and a 'Metabolomics Index' derived from blood metabolites, which showed a high correlation of 0.85, indicating that blood signatures reliably reflect physical fitness levels.

To further understand the underlying biological mechanisms, the researchers used machine learning models—ranging from simple statistical methods to advanced deep learning techniques—to classify active versus less-active individuals with over 91% accuracy. Consistently, eight metabolites, including aspartate, proline, and citrate, emerged as predictors of physical activity. Notably, aspartate was particularly prominent, confirming its importance in active aging.

Utilizing a novel network modeling approach called COVRECON, the team mapped metabolic interactions and identified enzymes like aspartate aminotransferase (AST) and alanine aminotransferase (ALT) as central hubs. These enzymes showed greater fluctuations in active individuals over time, signaling metabolic flexibility associated with physical activity.

Beyond physical health, aspartate plays a vital role in brain function as a neurotransmitter precursor involved in learning and memory. The study highlights potential links between physical activity, amino acid metabolism, and cognitive resilience. It suggests that engaging in regular exercise might not only maintain mobility but also support brain health through measurable metabolic changes.

Dr. Wolfram Weckwerth emphasizes that physical activity rewires metabolism at a molecular level. By decoding these changes, clinicians could better monitor and guide healthy aging processes, paving the way for personalized interventions to improve longevity and cognitive health.

This research underscores the power of combining metabolomics and AI to develop simple, blood-based biomarkers that reflect overall health and vitality in aging populations.