AI Creates Multi-Organ Aging Clocks Using MRI Scans to Predict Disease Risk
Researchers developed seven organ-specific aging clocks from MRI data that predict disease risk and mortality better than chronological age alone.
Summary
Scientists from the MULTI Consortium developed seven MRI-based biological aging clocks for different organs (brain, heart, liver, kidney, spleen, pancreas, and adipose tissue) using data from over 313,000 individuals. These AI-powered clocks measure how fast each organ is aging compared to chronological age, creating "biological age gaps" that predict disease risk and mortality. The study linked these aging patterns to thousands of proteins, metabolites, and genetic variants, identifying potential drug targets for anti-aging treatments.
Detailed Summary
This groundbreaking study represents the largest systematic effort to create organ-specific aging clocks using medical imaging. While brain aging clocks have been used for years to assess neurological health, this research extends the concept to six additional organs, providing a comprehensive view of biological aging across body systems.
The researchers analyzed MRI scans from 313,645 participants, using artificial intelligence to identify imaging patterns that correlate with aging in each organ. They developed biological age gaps (MRIBAGs) that measure the difference between an organ's apparent biological age and the person's chronological age. A positive gap indicates accelerated aging, while a negative gap suggests slower aging.
The study's scope extended far beyond imaging, incorporating analysis of 2,923 plasma proteins, 327 metabolites, and over 6 million genetic variants. This multi-omics approach revealed 53 genetic loci associated with organ aging and identified nine potentially druggable genes for future anti-aging therapies. The research also demonstrated significant genetic correlations between different organ aging patterns and connections to 525 disease endpoints.
Clinically, the MRI aging clocks showed strong predictive power for future disease risk and mortality. Participants with more youthful organ profiles had better health outcomes over time. Notably, in Alzheimer's disease patients, those with younger-appearing brain scans showed different cognitive decline trajectories during treatment, suggesting these tools could help personalize medical interventions.
The findings support the concept that aging occurs at different rates across organ systems within the same individual, opening new possibilities for targeted interventions and personalized medicine approaches to healthy aging.
Key Findings
- Seven organ-specific MRI aging clocks predict disease risk and mortality beyond chronological age
- 53 genetic loci identified as associated with accelerated organ aging patterns
- Nine potentially druggable genes discovered as targets for anti-aging treatments
- Brain aging patterns predicted different cognitive decline trajectories in Alzheimer's patients
- Multi-organ aging approach reveals organ-specific and cross-organ aging connections
Methodology
The study used nested cross-validation with AI/ML models (LASSO regression and support vector regression) trained on 313,645 participants from the MULTI Consortium. Age prediction models achieved moderate correlation coefficients (0.23-0.77) with approximately 5-year mean absolute errors across organs.
Study Limitations
Some abdominal organ clocks showed limited performance due to high feature collinearity. The study was primarily conducted in UK Biobank participants, potentially limiting generalizability to other populations. Causal relationships between aging patterns and disease outcomes require further validation.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.