As people drift into sleep, their brains do not shut down all at once. Instead, different brain regions change their activity and energy use at different times and in distinct ways. A new study published in Nature Communications provides detailed evidence that sensory and cognitive regions of the brain follow separate paths into sleep, with important implications for how the brain remains responsive to the environment even while unconscious.
The findings provide new insight into the complex processes that govern sleep. They help explain how the brain can maintain a degree of responsiveness to sounds or other sensory inputs during light sleep, potentially allowing a person to wake up if something important happens nearby.
Although scientists have long known that sleep involves changes in brain activity and energy use, it has been less clear how these changes unfold across different parts of the brain. Earlier studies have suggested that sleep supports functions such as memory consolidation and waste removal. However, the coordination of various physiological processes like blood flow, metabolism, and electrical activity during sleep has not been well understood.
One challenge has been the limitations of previous brain imaging methods. While tools like functional magnetic resonance imaging (fMRI) can track blood flow changes, they do not directly measure energy use. Positron emission tomography (PET), which can assess glucose metabolism, typically lacks the time resolution to capture rapid changes. As a result, researchers have had a hard time observing how the brain’s energy use and activity fluctuate together during the transition from wakefulness to sleep.
To address this gap, the team behind the current study combined several imaging techniques to monitor the brain’s activity in real time as participants moved from being awake to asleep. Their goal was to better understand how different physiological systems—neural activity, blood flow, and glucose metabolism—interact during the sleep process.
“We were motivated by the expectation that different physiological processes in the brain—such as metabolism, blood flow, and neural activity—must be tightly coordinated during sleep. Yet most human studies have examined these processes separately, leaving the nature of their coordination largely unknown,” said study author Jingyuan Chen, an assistant investigator at Massachusetts General Brigham and an assistant professor of radiology at Harvard Medical School.
“Until recently, capturing multiple physiological processes simultaneously was technically difficult. Now, with advances in neuroimaging, we can use dynamic PET to track metabolic changes over time and combine it with EEG-fMRI to simultaneously measure neural activity and blood flow. These tools allow us to directly observe how these different systems interact during the transition from wakefulness to NREM sleep.”
Twenty-six healthy adults took part in the study, which took place during the afternoon after participants had slept only four hours the night before. Participants lay in a brain scanner with their eyes closed and were encouraged to fall asleep.
The researchers monitored their brain activity throughout the session. For those undergoing EEG, an expert later identified when participants were awake or in various stages of non-REM sleep. Others were evaluated based on behavioral signs, such as a keypress task that tracked responsiveness.
The data showed that as participants transitioned into NREM sleep, the brain exhibited a drop in glucose metabolism—its primary source of energy. At the same time, large, slow fluctuations in blood flow became more pronounced. These changes were not random. They followed a clear pattern that corresponded to the participant’s level of wakefulness, as measured by EEG signals or behavior.
“We were surprised to see much stronger oscillatory blood flow activity in sensory regions during NREM sleep compared to quiet wakefulness,” Chen told PsyPost. “One possible explanation is that during sleep, neural activity becomes highly synchronized, producing large-scale, rhythmic fluctuations, and these oscillations—around 0.02 Hz—are particularly effective at driving vascular responses, which may explain why we see such pronounced oscillations in blood flow during sleep.”
One of the key findings was that not all brain regions responded to sleep in the same way. The default mode network, which includes areas involved in introspective thought and memory processing, showed a marked drop in glucose metabolism during sleep. In contrast, sensory and motor areas remained more metabolically active and showed stronger blood flow fluctuations.
This means that while the brain as a whole becomes less active and consumes less energy during sleep, some parts stay relatively alert. These preserved dynamics in sensory regions may serve a protective function, allowing people to wake up in response to sounds or other stimuli even while most of the brain is at rest.
The researchers also observed that slow brain waves and other physiological patterns associated with sleep were tightly linked to the observed changes in metabolism and blood flow. For example, large fluctuations in blood flow in sensory regions tended to occur alongside infraslow oscillations in electrical activity. These patterns have been associated with varying degrees of arousal and may help explain how the brain remains periodically receptive to external inputs during sleep.
“Our main message is that the brain’s physiological processes—energy use, blood flow, and neural activity—interact in complex and dynamic ways as we fall asleep,” Chen explained. “We found that different brain regions ‘fall asleep’ at different times. Sensory regions remain active and metabolically dynamic during light and intermediate sleep, while higher-order cognitive regions are more strongly suppressed. This pattern may help the brain remain responsive to important sensory cues, facilitating awakenings during lighter and intermediate sleep stages.”
While the study provides evidence for region-specific changes in the brain during sleep, it also comes with some limitations. The sleep sessions took place in the afternoon and involved only partial sleep deprivation. This setup may have limited the amount of deep sleep achieved by participants, which means the results are mostly applicable to light and intermediate stages of NREM sleep. The loud environment of the MRI scanner may have also influenced the quality of sleep.
The imaging approach used in this study represents a significant advance in measuring real-time changes in the brain, but the tools are complex and not yet widely available. “Because imaging measures are inherently noisy, we need to average results across many participants to obtain statistically robust effects,” Chen noted. “However, since sleep has such a strong influence on brain activity and metabolism, clear effects can even be observed in some individual participants.”
Looking ahead, the researchers plan to explore how these processes are altered in people with sleep disorders or neurological diseases such as Alzheimer’s. Sleep is thought to play a role in clearing waste products from the brain, and disruptions in this process could contribute to disease progression.
“Our current findings in healthy young adults provide a kind of ‘blueprint’ for how physiological processes in the brain are coordinated during normal sleep,” Chen said. “Moving forward, we want to explore how these delicate interactions between metabolism, blood flow, and neural activity are disrupted in sleep and neurological disorders—such as Alzheimer’s disease. Understanding these disruptions could eventually help identify early markers of disease or guide interventions that restore healthy brain physiology during sleep.”
The study, “Simultaneous EEG-PET-MRI identifies temporally coupled and spatially structured brain dynamics across wakefulness and NREM sleep,” was authored by Jingyuan E. Chen, Laura D. Lewis, Sean E. Coursey, Ciprian Catana, Jonathan R. Polimeni, Jiawen Fan, Kyle S. Droppa, Rudra Patel, Hsiao-Ying Wey, Catie Chang, Dara S. Manoach, Julie C. Price, Christin Y. Sander, and Bruce R. Rosen.
