Specialists from Peking University have published the results of their research on LifeClock, or “the Clock of Being.” This is a clock model based on artificial intelligence that not only determines biological age but also predicts diseases and life expectancy. Scientists have developed a clock that predicts illnesses years before they occur © Liudmila Chernetska/iStock.com To train the system, data from the medical records of 9.6 million individuals was used. 184 variables at various stages of life, from birth to the end of life, were taken into account, including anthropometric data, medical history, and results from laboratory and instrumental studies. Analysis of the data demonstrated that biological age is divided into two phases: establishment (approximately up to 18 years) and formation. The biomarkers that govern these processes are different. If the establishment clock is most strongly influenced by indicators of growth, creatinine, and total protein, then for the aging clock, the concentration of urea, albumin, and the red blood cell distribution width (RDW—an indicator of the proportion of abnormally sized cells in a blood sample) are more significant. In trials, the Clock of Being correctly predicted growth delay in children, developmental deviations, obesity, pituitary hypofunction, and other conditions. In adults, the system accurately forecasted type 2 diabetes mellitus, myocardial infarction, renal failure, and heart disease. In theory, such a calculation of biological age would allow doctors to identify patients in the risk zone and begin health improvement measures before the first signs appear. This is especially timely for middle-aged people. Age is considered a risk factor for many diseases, but chronological age from the date of birth does not reflect all the accumulated damage and functional characteristics of the body. Therefore, biological age is more applicable for assessing health status. Methods that allow for the determination of biological age have existed for a long time, but they are expensive and complex. They require genetic analyses and other molecular data. Furthermore, they assess a person’s condition at a specific moment in time without accounting for past developmental stages and the sequence of impairments, and thus cannot predict diseases, unlike LifeClock.
About the Author
