0
0
King’s College London Researchers Use AI to Assess Aging and Health Risks
Researchers at the Institute of Psychiatry, Psychology & Neuroscience (IoPPN) at King’s College London have made a breakthrough in the field of aging and health prediction. Their recent study explores the potential of artificial intelligence-based aging clocks, which estimate a person’s biological age by analyzing blood metabolites—small molecules generated during metabolism. This cutting-edge research offers new insights into the aging process and how it may be modulated.
The study, published in Science Advances on December 18, 2024, reveals that AI models can predict not only a person’s biological age but also offer a clearer picture of their overall health and potential lifespan. By using data from the UK Biobank, which includes blood samples and health information from over 225,000 participants aged 40 to 69, the researchers trained and tested 17 different machine learning algorithms to create aging clocks.
The central finding of the research is that “metabolomic age”—dubbed “MileAge”—is an effective measure of biological age. Unlike chronological age, which is fixed, MileAge reflects the internal aging process based on blood metabolites. The difference between a person’s predicted MileAge and their actual age, known as the MileAge delta, shows whether someone is aging faster or slower than average.
Individuals with accelerated aging—those whose MileAge is greater than their chronological age—tended to be more frail, had poorer health ratings, and were at higher risk for chronic illnesses. This group also exhibited shorter telomeres, which are indicators of cellular aging and are linked to age-related diseases like atherosclerosis. On the other hand, those with a decelerated biological age—having a MileAge younger than their chronological age—showed only weak links to better health outcomes.
The study also revealed that non-linear machine learning algorithms, particularly Cubist rule-based regression, were the most effective in predicting aging and health outcomes. These algorithms excelled at identifying complex patterns in the data, which linear models struggled to capture. According to the researchers, this suggests that biological aging is not a simple, linear process, but one influenced by various interacting factors.
A Tool for Prevention and Proactive Health
The findings of this research could have significant implications for early detection of health risks. By providing a more accurate picture of biological age, AI-powered aging clocks may allow individuals and healthcare providers to identify potential health issues before they manifest. This could lead to personalized, preventative strategies, including lifestyle adjustments and early interventions aimed at promoting healthier aging.
Dr. Julian Mutz, King’s Prize Research Fellow at IoPPN and lead author of the study, highlighted the potential of these aging clocks: “Metabolomic aging clocks have the potential to provide insights into who might be at greater risk of developing health problems later in life. Unlike chronological age, which cannot be changed, our biological age is potentially modifiable.” He emphasized the role of these clocks in shaping better health choices and guiding public health strategies.
Professor Cathryn Lewis, co-senior author of the study, noted that the research is a significant step forward in understanding biological aging. “This study is an important milestone in establishing the potential of biological aging clocks and their ability to inform health choices,” she said. The use of big data analytics is vital in refining these tools and making them more accurate and accessible for widespread use.
Looking Ahead
While the ability to reverse biological age remains a distant goal, AI aging clocks offer a promising tool for better understanding aging and improving health outcomes. As the technology matures, it could become a powerful component in managing aging and extending healthy lifespans.
For now, the research offers a new way to assess health risks and could empower individuals to take more control over their aging process—helping people live healthier, longer lives.
Reference: Mutz, J., Iniesta, R., & Lewis, C. M. (2024). Metabolomic age (MileAge) predicts health and life span: A comparison of multiple machine learning algorithms. Science Advances. DOI: 10.1126/sciadv.adp3743.
