AUG 06, 2022
New machine learning approach can identify your circadian rhythm from a blood sample
SEP 11, 2021
Doctors do not currently monitor a person’s circadian rhythms because there is not an efficient way to measure them
Many of the body’s physiological activities, including hunger, wakefulness, and metabolism, run on 24 hour cycles called circadian rhythms. These cycles are primarily controlled by the release of chemical messengers into the bloodstream from the brain and have been linked to cardiovascular disease, neurodegenerative disorders, and weight-gain.
Measuring circadian rhythms could significantly improve medical care. Doctors could better prevent and treat illness by more accurately assessing individual risk of disease and recommending times to eat, take medication, and rest. Circadian rhythms are not used as a clinical indicator at present because there is not an efficient way to measure them. Based on the results of a new study, however, that could change soon.
The results were recently reported in the Journal of Biological Rhythms by a team of researchers at the University of Colorado at Boulder. The researchers sought to develop a circadian rhythm test that is more efficient than the standard dim-light melatonin assessment, which requires hourly collection of blood over the course of an entire day to measure the amount of a sleep-inducing molecule in the body called melatonin. The test is accurate, but it is impractical for clinical use.
To develop an efficient circadian rhythm test, the researchers invited 16 participants into their sleep lab. Over the next two weeks, the researchers took regular blood samples and analyzed them to measure the levels of melatonin and approximately 4,000 metabolites, molecules that are produced by biochemical reactions in the body. Then, they used machine learning to determine which metabolites are associated with different phases of the circadian rhythm. When the analysis was complete, the researchers could measure circadian rhythms by testing a single blood sample for 65 metabolites with similar results to the dim-light melatonin assessment.
The new test does have some limitations; it worked best for people who had adequate sleep and whose food intake was controlled, which limits its practical application, and it would be more efficient if fewer metabolites had to be analyzed. Nonetheless, the study results are exciting and suggest an efficient circadian rhythm test may be available soon.
Natural Language Processing