Salivary biomarkers of physical fatigue as markers of sleep deprivation

Abstract

Study objective: Determine whether a salivary biomarker of physical fatigue, referred to as the fatigue biomarker index (FBI), can discriminate a control group from a sleep deprived group when saliva is collected under controlled conditions. The study expands on previous work examining changes in the composition of saliva during periods of prolonged exercise.

Methods: Thirty (30) young adults (14 Control [CON]; 16 Sleep Deprived [SDEP]) were monitored for mood state (Profile of Mood States [POMS]), cognitive performance (Stroop Color-Conflict Tests), and salivary biomarkers of physical fatigue over a 48-h period with sampling at 3-h intervals. Trials lasted from 06:00 on day 1 (time = -3 h) to 09:00 on day 3 (time = 48 h). Levels of salivary biomarkers were calculated from liquid chromatography-mass spectrometry (LC-MS) data. Statistical comparisons were made using Wilcoxon rank sum tests with a Bonferroni correction to limit type 1 error. Receiver-operator characteristic (ROC) analysis was used to evaluate the ability of the various parameters to distinguish the SDEP population from the CON population.

Results: Longitudinal analysis demonstrated significant between-group differences in all three parameters. ROC analysis demonstrated that cognitive performance tests and salivary biomarkers of physical fatigue distinguish the SDEP population from the CON population.

Conclusions: A previously identified salivary biomarker of physical fatigue may provide an alternative method for discriminating sleep deprived from rested individuals. The salivary biomarker of physical fatigue holds promise as an objective measure of sleep deprivation, perhaps eventually removing the reliance on self-reported sleep diaries and/or repeated polysomnographs for longitudinal tracking of sleep quality and/or diagnosis of sleep disorders.

Keywords: Sleep deprivation; analytical chemistry methods; biomarkers; chromatography; fatigue; liquid; mass spectrometry.

Figure 1. Sleep deprivation led to significant changes in self-reported fatigue level, cognitive performance, and salivary biomarkers of physical fatigue

Subjects were evaluated every 3 h during the course of the study. (A) Levels of the salivary biomarker of physical fatigue remained similar across groups through the first hours of the study, and then changed significantly during the hours after the first night of sleep. Both groups showed a significant increase around midnight on the first day of the study, while only the CON group showed a similar, though muted, increase on the second day of the study. (B) Cumulative scores on the Stroop tests, shown as percentages relative to the subject's initial scores, remained similar across groups until the early morning hours of the second day. Scores for individuals in the SDEP group dropped significantly after the first evening, whereas scores for individuals in the CON group remained relatively constant. The initial upward slope seen in both groups is likely to do a learning effect. The absence of significant differences during these early time points suggests that the size of the learning effect was similar for both groups. (C) Self-reported fatigue levels drifted higher with time in both groups. Previous analysis using linear mixed-effects modeling showed a significant positive slope for both groups. With the statistical approach used here, groups did not differ significantly until late on the second day. *indicates significant difference between groups.

Figure 2. Salivary biomarkers of physical fatigue efficiently detected the effects of sleep deprivation

Values for self-reported fatigue level, cognitive performance and salivary biomarkers of physical fatigue are shown by group for the data collected between 09:00 and 15:00 on the second day of the study. Each point represents average data for one individual. Receiver-operator characteristic (ROC) analysis suggested that salivary biomarkers of physical fatigue can identify sleep-deprived patients with reasonable specificity and selectivity (area under curve, 92%). Self-reported fatigue levels and cognitive performance tests also performed reasonably (area under curve, 77% and 88%, respectively).