NOTE: End date change to 3/31/2016 per NSBRI (Ed., 9/30/15)

NOTE: End date changed back to 9/30/2015 per NSBRI submission October 2013 (Ed., 10/16/13)

NOTE: Original end date was 9/30/2015; Changed to 9/30/13 per NSBRI 6/2/2012 (Ed., 6/4/12)

Specific Aim 1 identified core dimensions of neurobehavioral responses to chronic partial sleep loss. This was achieved by conducting dimensionality reduction using Principal Components Analysis (PCA) of a historical database of N=139 sleep-deprived subjects. The PCA included three cognitive performance outcomes: Psychomotor Vigilance Test (PVT) for behavioral alertness; Digit Symbol Substitution Test (DSST) for cognitive processing speed; and Digit-Span (DS) performance for working memory. PCA also included two primary subjective outcomes: Karolinska Sleepiness Scale (KSS), and Tiredness Visual Analog Scale (TVAS); and a measure of physiological sleep homeostatic drive (i.e., power spectral analyses [PSA] of delta frequency power in the NREM sleep EEG). The PCA included demographic covariates (i.e., age, gender, BMI, ethnicity), which contributed no variance to the neurobehavioral outcomes from sleep restriction identified by PCA. PVT performance and sleep homeostatic responses (i.e., slow wave activity in the non-REM EEG) were the two fully orthogonal components that emerged from the PCA dimensionality reduction of core neurobehavioral outcomes from sleep loss, indicating they were the more reliable and sensitive assays for detecting the phenotypic effects of sleep loss. Therefore, these two domains served as the sleep-deprivation outcomes for evaluation of biomarker prediction. Non-sleep-deprived PVT time on task performance and salivary amylase activity (sAA) were dropped as candidate predictors due to insufficient variance in baseline conditions.

Remaining predictor variables (i.e., potential biomarkers) to be evaluated in the study were derived from baseline (non-sleep-deprived) neuroimaging (i.e., resting arterial spin labeling [ASL fMRI] cerebral blood flow); resting heart rate variability (HRV); results of genome wide association analyses (GWAS); and the results of quantitative analyses of baseline sleep polysomnography (PSG).

As discovery capability for novel biomarkers of sleep-loss vulnerability has evolved over the last 4 years, the project has taken on new importance and opportunity (e.g., we had a recent preliminary finding of a metabolic marker of slept debt [Weljie et al. PNAS, 2015]). Instead of focusing on candidate genes as predictors of the vulnerability in a few subjects (which has not yielded a predictor relative to DQB1*0602, COMT Va1158Met, or PER3 VNTR genes), we shifted to conducting a genome-wide association study (GWAS) with our collaborator, Dr. Emmanuel Mignot, at Stanford University. Blood samples from N=358 unique human subjects (n=271 who underwent chronic partial sleep deprivation [PSD]; n=24 who underwent acute total sleep deprivation [TSD]; and n=63 who underwent both PSD and TSD) collected during our extensive laboratory studies supported by the National Institutes of Health (NIH), and the Office of Naval Research (ONR), were provided to Dr. Mignot for blinded GWAS analysis, to determine if a genetic variant is associated with the phenotypic trait of neurobehavioral vulnerability to sleep loss, as measured by PVT performance (the most sensitive neurobehavioral assay to sleep loss [Basner et al., 2013]). In addition, subjects' nocturnal polysomnographic (PSG) sleep records on pre-sleep-loss nights were subjected to power spectral analyses and machine learning algorithms to identify possible PSG biomarkers of PVT performance vulnerability to sleep loss.

The results are being correlated with the results of the genetic analyses of the blood samples. We are conducting whole genome single nucleotide polymorphism (SNP) typing to first study known polymorphisms that have recently been found to be associated with sleep disorders in multiple studies. Thus, our biomarker search in quantitative analyses of sleep physiology is predicated on the hypothesis that presumably healthy individuals who are more cognitively vulnerable to sleep loss may be so because of an (as yet) occult disturbance of sleep that reduces waking state stability (measured by PVT performance) not evident in human-scored sleep physiology.

Data analyses will be completed and papers prepared and submitted for publication in the coming months. Finding biomarkers of neurobehavioral vulnerability to sleep loss addresses multiple NASA BHP risk and gaps including indicators of vulnerabilities and resiliencies to sleep loss (Sleep Gap 4), methods to enhance behavioral health and prevent decrements during space flight (BMed1), characteristics of individuals resilient to neurobehavioral decrements from sleep loss (BMed5), and psychological measures that help select individuals for long-duration space flight (Team Gap 4). Finding valid biomarkers will help optimize crew resources and fatigue management during long-duration space flight, and it will have benefit for fatigue management in many Earth-based, safety-sensitive operations.

The project deliverables will be a biological or behavioral assay for discriminating those who are more resistant versus those more susceptible to the adverse effects of fatigue on neurobehavioral functions.

Specific Aim 1 has been accomplished by evaluating core dimensions of neurobehavioral responses to both chronic partial and acute total sleep loss. Through dimensionality reduction via Principal Components Analysis (PCA) of a historical database of cognitive, subjective, and physiological responses to acute and chronic sleep loss in N=139 healthy adults, it was determined that Psychomotor Vigilance Test (PVT) performance variables (especially response speed and total response errors), as well as EEG delta activity from non-REM sleep physiology (via polysomnography [PSG]) were the most reliable and sensitive assays for determining the effects of sleep loss. This was achieved by conducting a factor analysis of a historical database of cognitive, subjective, and physiological responses to sleep restriction in healthy adults. Analyses of the historical database again established that PVT performance outcomes were the most reliable and sensitive assays for determining the effects of sleep loss. These dimensions will serve as targets for prospectively assessing the predictive power (separately and in combination) of each of the objective markers being evaluated (i.e., resting brain activity, heart rate variability, GWAS results, and sleep physiology results).

Specific Aim 2 is being accomplished through the analysis of blood samples from N=358 unique human subjects (n=271 PSD; n=24 TSD; n=63 PSD/TSD) collected during our extensive laboratory studies of partial and chronic sleep restriction that were supported by NIH, NSBRI, and ONR. These samples are being evaluated in a genome-wide association study (GWAS), to determine if a genetic variant in different individuals is associated with the phenotypic trait of neurobehavioral vulnerability to sleep restriction. Phenotypic response will be evaluated using objective neurobehavioral responses to chronic partial sleep loss at the level it can be experienced in space flight. These trait response measures will include behavioral alertness assessed by Psychomotor Vigilance Test (PVT) performance and polysomnography during baseline. In the past year, the project (1) further confirmed the sensitivity of the PVT as a highly sensitive neurobehavioral measure of vulnerability to sleep loss; (2) identified a metabolic state marker of sleep debt (Weljie et al. PNAS, 2015); (3) completed (but not yet fully analyzed) GWAS on N=358 subjects who underwent sleep loss; and (4) initiated quantitative analyses of baseline sleep polysomnography.

NOTE: End date change to 3/31/2016 per NSBRI (Ed., 9/30/15)

NOTE: End date changed back to 9/30/2015 per NSBRI submission October 2013 (Ed., 10/16/13)

NOTE: Original end date was 9/30/2015; Changed to 9/30/13 per NSBRI 6/2/2012 (Ed., 6/4/12)

The project consists of two discrete specific Aims. Specific Aim 1 will identify core dimensions of neurobehavioral responses to both chronic partial and acute sleep loss. This will be achieved by conducting a factor analysis of a historical database of cognitive, subjective, and physiological responses to acute and chronic sleep loss in healthy adults (N=640). The historical data will be separated into two groups, one group of N=175 subjects studied during total sleep deprivation (TSD) and a second group of N=465 subjects studied during chronic partial sleep deprivation (PSD). The primary outcomes to determine the core dimensions of responses to sleep loss for both groups include performance on the Psychomotor Vigilance Test (PVT), Digit Symbol Substitution Task, Digit Span, Tower of London, and subjective ratings of mood, sleepiness, and tiredness through the Profile of Mood States, Visual Analog Scales of Fatigue and Tiredness, and Karolinska Sleepiness Scale. These dimensions will then serve as targets for prospectively assessing the predictive power (separately and in combination) of each of five objective markers that include physiological (brain activity, heart rate variability, salivary amylase), behavioral (time on task performance), and genetic (common polymorphisms) measures for susceptibility to neurobehavioral responses to sleep loss. Specifically, the five biomarkers are: (1) Resting arterial spin labeling (ASL fMRI) cerebral blood flow without sleep deprivation; (2) Resting psychomotor vigilance test (PVT) time on task (TOT) without sleep deprivation; (3) Heart rate variability (HRV) during PVT TOT; (4) Salivary a-amylase activity (sAA) during sleep deprivation; (5) Variations in genes regulating sleep, circadian, and cognitive functions.

The prospective validation is currently being accomplished by the addition of the predictor markers to three separate studies on sleep deprivation that are currently underway. Across the three studies being leveraged, there will be a total of N=120 healthy adults (diverse in age, gender, ethnicity) on which predictive validation will be performed.

Data collection on the five biomarkers continued during the second year of funding. During the second year of funding, data on predictor marker 1 (see above) have been collected on a total of N=18 unique subjects who completed a TSD protocol and N=25 unique subjects who completed a PSD protocol. Projected data collection on biomarker 1 during the second year funding period is expected to be a total of N=22 unique TSD subjects and N=29 PSD subjects. Data on predictor markers, 2, 3, 4, and 5 (see above) have been collected on a total of N=23 unique subjects who completed a TSD protocol and N=31 unique subjects who completed a PSD protocol. Projected data collection on biomarkers 2, 3, 4, and 5 during the second year funding period is expected to be a total of N=27 unique TSD subjects and N=35 PSD subjects. Total data collection on predictor marker 1, which includes years 1 and 2, have yielded collected data on a total of N=42 unique TSD subjects and N=49 unique PSD subjects. Total data collected to date on predictor markers 2, 3, and 5 have yielded collected data on a total of N=49 unique TSD subjects and N=57 unique PSD subjects. Total data collection on predictor marker 4 have yielded data on a total of N=45 unique TSD subjects and N=57 unique PSD subjects. During the third year of funding, the historical database of Specific Aim 1 will be complete and statistical analyses to determine the core dimensions of responses to acute and partial sleep deprivation will be executed. Data collection on the five biomarkers of Specific Aim 2 will continue to be collected resulting in a projected total of N=34 unique TSD subjects and N=34 unique PSD subjects during the third year of funding. Additionally, data from N=24 unique subjects were collected in both TSD and PSD conditions.

The identification of biomarkers addresses multiple NASA BHP risk and gaps by identifying indicators of vulnerabilities and resiliencies to sleep loss (Sleep Gap 4), defining characteristics of individuals resilient to the neurobehavioral decrements from sleep loss (BMed5), and by identifying psychological measures that can be used to select individuals for long duration space flight (Team Gap 4). Finding valid markers of susceptibility to neurobehavioral deficits from total and chronic partial sleep loss will make it possible to optimize crew resources and fatigue management during long-duration space flight, and it will have substantial benefits for fatigue management in many Earth-based, safety-sensitive operations. The project deliverable will be a technique (technology) for discriminating those who are more resistant versus those more susceptible to the adverse effects of fatigue on neurobehavioral functions. The functions to be predicted include separate and common variance among outcomes in the domains of behavioral and physiological alertness, cognitive performance, subjective fatigue/sleepiness, and homeostatic sleep response.

During the second year of funding, assessments of predictor markers were collected in three currently funded scientific experimental protocols, which expose participants to total and/or chronic partial sleep deprivation. The five predictor markers are: (1) Resting arterial spin labeling (ASL fMRI) cerebral blood flow without sleep deprivation; (2) Resting psychomotor vigilance test (PVT) time on task (TOT) without sleep deprivation; (3) Heart rate variability (HRV) during PVT TOT; (4) Salivary a-amylase activity (sAA) during sleep deprivation; (5) Variations in genes regulating sleep, circadian, and cognitive functions.

Collected data includes neuroimaging of non sleep deprived subjects at baseline, Psychomotor Vigilance Task (PVT) performance, electrocardiograms during the 20-minute PVT, saliva samples during PSD, and blood draws during both TSD and PSD. During year 2, data on predictor marker 1 have been collected on a total of N=18 unique TSD subjects and N=25 unique PSD subjects. Projected data collection on biomarker 1 for year 2 is expected to total of N=22 unique TSD and N=29 PSD subjects. Data on predictor markers, 2, 3, 4, and 5 have been collected on a total of N=23 unique TSD subjects and N=31 unique PSD subjects. Projected data collection on biomarkers 2, 3, 4, and 5 during year 2 is expected to total of N=27 unique TSD and N=35 PSD subjects. Data collection on the five biomarkers of Specific Aim 2 will continue in year 3 with a projected total of N=34 unique TSD subjects and N=34 unique PSD subjects. Additionally, data from N=24 unique subjects were collected in both TSD and PSD conditions.

NOTE: Original end date was 9/30/2015; Changed to 9/30/13 per NSBRI 6/2/2012 (Ed., 6/4/12)

The project consists of two discrete specific Aims. Specific Aim 1 will identify core dimensions of neurobehavioral responses to both chronic partial and acute sleep loss. This will be achieved by conducting a factor analysis of a historical database of cognitive, subjective, and physiological responses to acute and chronic sleep loss in healthy adults (N=640). The historical data will be separated into two groups, one group of N=175 subjects studied during total sleep deprivation (TSD) and a second group of N=465 subjects studied during chronic partial sleep deprivation (PSD). The primary outcomes to determine the core dimensions of responses to sleep loss for both groups include performance on the Psychomotor Vigilance Test (PVT), Digit Symbol Substitution Task, Digit Span, Tower of London and subjective ratings of mood, sleepiness and tiredness through the Profile of Mood States, Visual Analog Scales of Fatigue and Tiredness, and Karolinska Sleepiness Scale. These dimensions will then serve as targets for prospectively assessing the predictive power (separately and in combination) of each of five objective markers that include physiological (brain activity, heart rate variability, salivary amylase), behavioral (time on task performance), and genetic (common polymorphisms) measures for susceptibility to neurobehavioral responses to sleep loss. Specifically, the five biomarkers are: (1) Resting arterial spin labeling (ASL fMRI) cerebral blood flow without sleep deprivation; (2) Resting psychomotor vigilance test (PVT) time on task (TOT) without sleep deprivation; (3) Heart rate variability (HRV) during PVT TOT; (4) Salivary a-amylase activity (sAA) during sleep deprivation; (5) Variations in genes regulating sleep, circadian and cognitive functions. The prospective validation is currently being accomplished by the addition of the predictor markers to three separate studies on sleep deprivation that are currently underway. Across the three studies being leveraged, there will be a total of N=120 healthy adults (diverse in age, gender, ethnicity) on which predictive validation will be performed. Data collection on the five biomarkers has already begun during the first year of funding. Data on the predictor markers 1, 2, 3, and 5 (see above) have been collected on a total of N=22 unique subjects who completed a TSD protocol and N=22 unique subjects who completed a PSD protocol. Projected data collection on biomarkers 1, 2, 3, and 5 during the first year funding period is expected to be a total of N=28 unique TSD subjects and N=26 PSD subjects. Data on predictor marker 4 (sAA) has been collected on a total of N=18 unique TSD subjects and N=22 unique PSD subjects. Projected data collection on biomarker 4 during the first year of funding is expected to be N=24 unique TSD subjects and N=27 PSD subjects. During the second year of funding, the historical database of Specific Aim 1 will be complete and statistical analyses to determine the core dimensions of responses to acute and partial sleep deprivation will be executed.

Data collection on the five biomarkers of Specific Aim 2 will continue to be collected resulting in a projected total of N=22 unique TSD subjects and N=48 unique PSD subjects. Additionally, data from N=19 unique subjects were collected in both TSD and PSD conditions. The identification of biomarkers addresses multiple NASA BHP risk and gaps by identifying indicators of vulnerabilities and resiliencies to sleep loss (Sleep Gap 4), defining characteristics of individuals resilient to the neurobehavioral decrements from sleep loss (BMed5) and by identifying psychological measures that can be used to select individuals for long duration space flight (Team Gap 4). Finding valid markers of susceptibility to neurobehavioral deficits from total and chronic partial sleep loss will make it possible to optimize crew resources and fatigue management during long-duration space flight, and it will have substantial benefits for fatigue management in many Earth-based, safety-sensitive operations. The project deliverable will be a technique (technology) for discriminating those who are more resistant versus those more susceptible to the adverse effects of fatigue on neurobehavioral functions. The functions to be predicted include separate and common variance among outcomes in the domains of behavioral and physiological alertness, cognitive performance, subjective fatigue/sleepiness, and homeostatic sleep response.

In addition to the progress constructing the historical database, data collection on predictor biomarkers (Specific Aim 2) has begun. During the first year of funding, assessments of predictor markers were added (if not already present) to three currently funded scientific experimental protocols, which expose participants to total and/or chronic partial sleep deprivation. The five predictor markers are: (1) Resting arterial spin labeling (ASL fMRI) cerebral blood flow without sleep deprivation; (2) Resting psychomotor vigilance test (PVT) time on task (TOT) without sleep deprivation; (3) Heart rate variability (HRV) during PVT TOT; (4) Salivary a-amylase activity (sAA) during sleep deprivation; (5) Variations in genes regulating sleep, circadian and cognitive functions. Collected data includes neuroimaging of non sleep deprived subjects at baseline, Psychomotor Vigilance Task (PVT) performance, electrocardiograms during the 20-minute PVT, saliva samples during PSD, and blood draws during both TSD and PSD. Data on predictor markers 1, 2, 3, and 5 (see above) have been collected on a total of N=22 unique subjects who completed a TSD protocol and N=22 unique subjects who completed a PSD protocol. Projected data collection on biomarkers 1, 2, 3, and 5 during the first year funding period is expected to be N=28 unique TSD subjects and N=26 unique PSD subjects. Data on the predictor marker 4 (sAA) has been collected on N=18 unique TSD subjects and N=22 unique PSD subjects. Projected data collection on biomarker 4 during the first year of funding is expected to be N=24 unique TSD subjects and N=27 PSD subjects. Additionally, data from N=19 unique subjects were collected in both TSD and PSD conditions.

NOTE: Original end date was 9/30/2015; Changed to 9/30/13 per NSBRI 6/2/2012 (Ed., 6/4/12)