In a study published in Nature Medicine, researchers have unveiled LifeClock, a revolutionary biological clock model capable of predicting biological age from birth to old age using routine electronic health records (EHRs) and standard laboratory data.
Unlike traditional aging models that focus solely on adults, LifeClock fills a major scientific gap by mapping biological aging across all stages of life, from infancy through adulthood. Using EHRFormer, a deep learning framework, scientists analyzed an extraordinary dataset of over 24.6 million longitudinal clinical visits from nearly 9.7 million individuals, constructing a virtual representation of patients’ biological trajectories throughout their lifespan.
EHRFormer is an AI framework inspired by natural language processing (e.g., BERT, GPT).I t processes sequential, time-stamped clinical data like diagnoses, lab tests, and medical visits much like how LLM process words in a sentence. This allows the model to capture temporal patterns and interactions across millions of patient records.
The study revealed that the pediatric clock mirrors children growth and development, effectively predicting risks for conditions such as malnutrition, developmental delays, and growth abnormalities. In contrast, the adult clock tracks biological aging and predicts susceptibility to age-related diseases including type 2 diabetes, kidney failure, stroke, and cardiovascular disorders.
A particularly striking finding is the ability of LifeClock to detect when an individual’s biological age (BA) diverges from their chronological age (CA)—a signal that may indicate early or accelerated disease risk. By linking these deviations to clinical outcomes, researchers demonstrated that biological age can serve as a sensitive marker of overall health and future disease vulnerability. Beyond its predictive power, LifeClock provides a comprehensive framework for precision health, enabling continuous monitoring of biological aging and disease progression using everyday clinical data. This innovation paves the way for personalized prevention strategies, earlier interventions, and a deeper understanding of how humans age—from cradle to old age.
