"Leveraging the Metabolic Fingerprint of Sleep Deprivation and Sleep Restriction for Forensic Applications: A Machine Learning Study in Oral Fluid Metabolomics" Journal of Proteome Research
Sleep loss dulls alertness and coordination, and it can produce effects similar to severe intoxication, making actions like driving incredibly risky. But there's no clinical test for determining when someone is dangerously sleep deprived. Now, researchers report a step toward a non-invasive test for sleep deprivation in ACS' Journal of Proteome Research. In a study of 20 men, they identified molecular differences in saliva after a full night's rest and 24 hours without sleep.
Until now, sleep deprivation has been impossible to measure biochemically - and yet it is one of the greatest burdens of our time." - Thomas Kraemer
Drowsy driving contributes to tens of thousands of crashes each year in the U.S. As a result, some states have instituted laws to deter tired drivers. So, Thomas Kraemer and colleagues wanted to see if saliva contains metabolites that change after sleep loss, with the hopes of one day developing a test for sleepiness that could be conducted anywhere from the roadside to a clinical setting.
"Until now, sleep deprivation has been impossible to measure biochemically - and yet it is one of the greatest burdens of our time," says Kraemer, the corresponding author of the study. "This study introduces the first direct biomarkers of sleep loss in saliva under real-world conditions, marking a milestone in forensic investigations."
The team recruited 20 healthy young adult males who usually sleep seven to nine hours every night. Participants completed three sleep scenarios in a random order, each separated by a week: deprivation (one night without sleep), restriction (four nights with two hours' less sleep than usual) and well rested (around eight hours' sleep).
Researchers collected saliva before and after each scenario and analyzed the samples' metabolite compositions. Using statistical analyses, they determined 10 molecular differences between sleep-deprived and well-rested samples. In contrast, the sleep-restricted state showed no significant metabolic difference from the rested state. Then the team developed and trained a predictive model based on the varying saliva metabolites, and it correctly identified samples from sleep-deprived individuals 94% of the time.
The mistakes that the model made were likely attributable to individual metabolic processes. For example, after being awake for a day, some participants didn't return to a fully rested metabolic profile even after eight hours of sleep, suggesting that may not be enough time for everyone to fully recover.
These findings demonstrate that a collection of salivary metabolites forms a "sleepiness fingerprint," which holds potential for situations such as roadside checks, where accurately detecting sleep deprivation is needed. Therefore, the team is next undertaking a large-scale international assessment of the predictive model, expanding tests to over 1,000 samples collected from shift workers, women and frequent drivers, says Kraemer.
The authors acknowledge funding from the Fund for Road Safety (Fonds für Verkehrssicherheit FVS).
Some of the authors have filed a patent related to the method for determining the metabolic sleepiness markers.
