Functional imaging of working memory following normal sleep and after 24 and 35 h of sleep deprivation: Correlations of fronto-parietal activation with performance
Introduction
Sleep deprivation (SD) is associated with performance decline in a wide range of cognitive tasks (Pilcher and Huffcutt, 1996, Durmer and Dinges, 2005). Recent evidence indicates substantial inter-individual differences in this performance decline, and that this differential vulnerability to cognitive impairment following SD may be trait-like (Van Dongen et al., 2004). The cognitive vulnerability to sleep loss clusters on three distinct neurobehavioral dimensions: (1) self-evaluation of sleepiness, fatigue and mood; (2) cognitive processing capability (e.g., working memory) and (3) behavioral alertness as measured by sustained attention performance (Van Dongen et al., 2004).
The first goal of the present study was to examine how inter-individual differences in brain activation correspond to working memory task performance in the context of sleep deprivation. We were specifically interested in whether functional imaging can serve as a marker for predicting performance decline following SD. We evaluated working memory as this cognitive domain engages fronto-parietal networks (Curtis and D'Esposito, 2003) whose activation is influenced by sleep deprivation (Bell-McGinty et al., 2004, Chee and Choo, 2004, Chee et al., 2004, Habeck et al., 2004, Mu et al., 2005a, Mu et al., 2005b). Of particular interest, Mu et al. (2005a), using a Sternberg-type working memory task, reported that individuals who were vulnerable to working memory impairment following SD were those with higher global brain activation prior to SD (also see Caldwell et al., 2005) and less reduction in global activation following SD, suggesting the viability of using imaging to predict resistance to cognitive decline following SD.
In the present study, we evaluated working memory decline in SD, using a different set of working memory tasks that were presented in a counterbalanced manner. We reasoned that if brain imaging is to be useful in predicting differential vulnerability to SD, results should replicate across different tests evaluating the same cognitive domain. Such reproducibility has been evaluated in behavioral testing (Frey et al., 2004, Van Dongen et al., 2004, Frey et al., 2005), but not with functional imaging.
Concurrent with the goal of identifying the neural correlates of resistance or vulnerability to working memory performance decline following SD, we also investigated the effect of scanning at different time points following SD on brain activation. Previous imaging studies have involved 24 (Thomas et al., 2000, Chee and Choo, 2004, Choo et al., 2005), 30 (Mu et al., 2005b), 35 (Drummond et al., 2000, Doran et al., 2001, Drummond and Brown, 2001, Drummond et al., 2004) and 48 h (Bell-McGinty et al., 2004, Habeck et al., 2004) of total SD but, to date, imaging has been performed only at a single time point following SD. The non-uniformity of imaging results obtained from different studies (Table 1) has been attributed to task differences (Drummond and Brown, 2001). However, it is also possible that scanning at different time points (SD24 and SD35) could have contributed to the difference in imaging results.
Cognitive performance in the context of SD is known to be modulated by the interaction of two effects: an endogenous, cyclically varying circadian effect as well as an homeostatic effect related to the increasing maintenance of wakefulness (Van Dongen and Dinges, 2005). Depending on how the two effects interact, performance after 35 h of wakefulness, while impaired relative to baseline, can stay the same (Van Dongen and Dinges, 2005) or improve (Doran et al., 2001) relative to 24 h of SD. Replicating SD effects at two different circadian phases increases confidence that the brain changes are due to the elevation of homeostatic sleep drive.
In the present study, we evaluated each volunteer after rested wakefulness as well as after 24 and 35 h of SD, the two most commonly used test times in imaging studies, to determine the comparability of studies performed at SD24 and SD35. A null result in a study recruiting a large number of volunteers would facilitate the interpretation of data emerging from different laboratories that are constrained to perform imaging at particular times for operational reasons. Otherwise, investigators in this field might have to agree on a standardized time to perform such studies.
Section snippets
Volunteer characteristics
Of an original cohort of 41 recruits, 28 healthy, right-handed adults aged between 19–25 years (14 men) completed this study, and of these, 26 volunteers' data were fully analyzed. They were recruited through advertisements placed in university halls of residence. Ethical approval for this study was obtained from the Singapore General Hospital IRB and volunteers were reimbursed for the completion of the experimental protocol. To qualify for recruitment, volunteers had to declare that they:
- 1.
Sleep
In-scanner working memory tasks
Performance accuracy declined significantly during the performance of the LTR, PLUS and PLUS-L working memory tasks following SD. The difference in accuracy between SD24 and SD35 was not significant. RT was longer in the various tasks following SD, the largest change being between RW and SD24. The difference in RT between SD24 and SD35 was not significant (Table 2).
The difference in RT change between RW and SD24 for SD-R and SD-V was also significant (t(14) = 2.80, P = 0.01), reflecting a
Imaging findings at SD24 and SD35 differ from RW but do not differ significantly from one another
The in-scanner effects on sleepiness were pronounced at both SD24 and SD35 relative to RW. As such, lying in the scanner provided constant conditions that unmasked (accentuated) the effects of the homeostatic sleep drive at SD24 and SD35. This is ideal for comparing the effects of elevated sleep drive at the two circadian phases (Van Dongen and Dinges, 2005). Replicating SD effects at two different circadian phases increases confidence that the brain changes are due to the elevation of
Acknowledgments
This work was supported by NMRC Grants 2000/0477, BMRC Grant 014, DMERI, the Shaw Foundation and AFOSR FA9550-05-1-0293 (DFD). Joanna Sze and Shen Li-Juan collected and preprocessed the imaging data. We thank the editors and three anonymous reviewers for their helpful comments.
References (48)
- et al.
Body posture affects electroencephalographic activity and psychomotor vigilance task performance in sleep-deprived subjects
Clin. Neurophysiol.
(2003) - et al.
Dissociation of cortical regions modulated by both working memory load and sleep deprivation and by sleep deprivation alone
NeuroImage
(2005) - et al.
Persistent activity in the prefrontal cortex during working memory
Trends Cogn. Sci.
(2003) - et al.
The effects of total sleep deprivation on cerebral responses to cognitive performance
Neuropsychopharmacology
(2001) - et al.
The activation of attentional networks
NeuroImage
(2005) - et al.
Circadian and wake-dependent modulation of fastest and slowest reaction times during the psychomotor vigilance task
Physiol. Behav.
(2004) - et al.
An event-related fMRI study of the neurobehavioral impact of sleep deprivation on performance of a delayed-match-to-sample task
Brain Res. Cogn. Brain Res.
(2004) - et al.
Valid conjunction inference with the minimum statistic
NeuroImage
(2005) - et al.
The neural basis of executive function in working memory: an fMRI study based on individual differences
NeuroImage
(2004) - et al.
Neural correlates of visual working memory: fMRI amplitude predicts task performance
Neuron
(2002)
Circadian rhythm in sleepiness, alertness and performance
Subjective and objective sleepiness in the active individual
Int. J. Neurosci.
Identification and differential vulnerability of a neural network in sleep deprivation
Cereb. Cortex
Are individual differences in fatigue vulnerability related to baseline differences in cortical activation?
Behav. Neurosci.
Complexity of prefrontal cortical dysfunction in schizophrenia: more than up or down
Am. J. Psychiatry
Functional imaging of working memory after 24 hr of total sleep deprivation
J. Neurosci.
Left insula activation: a marker for language attainment in bilinguals
Proc. Natl. Acad. Sci. U. S. A.
Local infusion of scopolamine into intraparietal cortex slows covert orienting in rhesus monkeys
J. Neurophysiol.
Cumulative sleepiness, mood disturbance, and psychomotor vigilance performance decrements during a week of sleep restricted to 4–5 hours per night
Sleep
Sustained attention performance during sleep deprivation: evidence of state instability
Arch. Ital. Biol.
Psychomotor vigilance performance: a neurocognitive assay sensitive to sleep loss
Sleep deprivation-induced reduction in cortical functional response to serial subtraction
Neuroreport
Altered brain response to verbal learning following sleep deprivation
Nature
Increasing task difficulty facilitates the cerebral compensatory response to total sleep deprivation
Sleep
Cited by (225)
Acute sleep deprivation in humans
2023, Encyclopedia of Sleep and Circadian Rhythms: Volume 1-6, Second EditionImpact of spaceflight stressors on behavior and cognition: A molecular, neurochemical, and neurobiological perspective
2022, Neuroscience and Biobehavioral ReviewsCitation Excerpt :Similarly, a cortical “sustained-attention network”, involving prefrontal, motor and parietal cortical regions, and subcortical structures such as the basal ganglia, and “DMN”, involving a network of medial prefrontal and medial cortical regions, have both shown alterations in fMRI activity following sleep deprivation (Chee et al., 2006; Chen and Zheng, 2018; Chen et al., 2018; Kaufmann et al., 2016; Strangman et al., 2005; Wang et al., 2015). Alterations in task-based functional MRI activity have been also observed in the anterior cingulate, middle occipital gyrus, inferior frontal gyrus, medial frontal cortex, parietal cortex, and thalamus (Chee et al., 2006; Choo et al., 2005; Kilgore, 2010; Saletin et al., 2019). Anterior cingulate activation following sleep deprivation is consistent with findings studying divided attention and working memory, implicated in detecting situations where errors are likely.