Scientists from the University of Surrey in England trained a neural network to gauge the body’s internal clock, aiming to create personalized sleep and meal plans. The study’s findings were published in the journal PNAS. The approach analyzes blood samples to map blood metabolites and determine the timing of the human circadian system. Today, melatonin production timing is commonly used to gauge circadian rhythm, but the researchers say their method is on par with current techniques and could become even more precise as the science advances.
The team collected blood samples from 24 healthy, non-smoking adults who maintained a seven-day, consistent sleep schedule before the study. More than 130 metabolic rhythms were tracked and fed into artificial intelligence to predict each participant’s circadian timing.
The researchers note that this work could enable genuinely personalized sleep and meal plans. If people follow these plans, there could be a reduced risk of serious illnesses linked to poor sleep and irregular eating patterns, the authors explain.
Looking ahead, the team plans to refine the neural network further and expand the data set to enhance predictive accuracy and individual customization.