Chronological age is the most important factor determining risk of disease and death in adults. However, life expectancy can vary considerably among individuals with similar chronological age.
In this study, the researchers used data from 51,408 participants across three large population studies to develop the most powerful clock to date that captures biological age and predicts the risk of premature death and numerous diseases. They analysed nearly 3,000 proteins in blood samples from participants in the UK Biobank study to develop a machine learning model that uses 204 proteins to estimate a person’s biological age.
This protein-based biological age model was also shown to be able to accurately estimate the biological age of participants in the other two studies, the China Kadoorie Biobank, and FinnGen (based in Finland), who have a very different genetic makeup and lifestyles compared with people living in the UK.
The researchers compared the participants’ chronological age with their biological age based on blood proteins to calculate the ‘protein age gap’ as a biological indicator of how fast a person is ageing. Within UK Biobank, they could link the protein age gap to a wide range of health outcomes to see if it could reliably predict age-related physical and mental wellbeing, risk of disease and death.