Fluctuations in presleep thalamocortical connectivity predicted distinct features of subsequent sleep architecture, suggesting that presleep thalamocortical network dynamics may facilitate some aspects of later restorative sleep the same night.
Key Findings
Results
Increased thalamocortical connectivity with occipital, posterior middle temporal, and left frontal cortices during the early evening predicted greater time in N2 sleep that night.
Twenty adults with clinically significant insomnia symptoms participated (8 men, 12 women; age 19-39 years).
Two resting-state fMRI scans were conducted approximately 2 hours apart (~6:30 and 8:30 p.m.) before an overnight in-lab polysomnography study.
Whole-brain seed-to-voxel analyses used a bilateral thalamic region of interest to assess changes in functional connectivity between the two scans.
The connectivity changes were then correlated with polysomnographic sleep-stage metrics including time in wake, N1, N2, N3, and REM.
Results
Decreased thalamocortical connectivity with insula, putamen, and frontal regions before sleep predicted more N3 (slow-wave) sleep.
N3 sleep is considered the most restorative sleep stage.
The decrease in connectivity in these regions was observed between the ~6:30 p.m. and ~8:30 p.m. resting-state fMRI scans.
Putamen and insula are subcortical and cortical regions associated with arousal and sensory processing.
This finding suggests a disengagement of arousal-related circuitry may be predictive of deeper sleep.
Results
Reduced thalamic coupling with the left lateral occipital gyrus before sleep predicted greater REM sleep duration.
REM sleep is associated with memory consolidation and emotional processing.
The finding was specific to the left lateral occipital gyrus, indicating lateralized effects.
This presleep connectivity reduction was measured as the change between the two evening fMRI scans.
The association suggests that reduced visual cortex engagement before bed may be linked to later REM generation.
Results
Decreased thalamocerebellar connectivity before sleep was associated with increased wake time while in bed.
This finding links reduced thalamo-cerebellar functional coupling in the evening to worse sleep continuity.
Wake time while in bed is a key indicator of insomnia severity.
The sample specifically consisted of adults with clinically significant insomnia symptoms.
The direction of the effect (decreased connectivity predicting more wakefulness) is in contrast to the N3 finding, where decreased connectivity predicted better sleep.
Results
No presleep thalamocortical connectivity changes were associated with N1 sleep duration, total sleep time, or sleep efficiency.
Despite significant associations with N2, N3, REM, and wake time, N1 sleep showed no significant correlations with any connectivity changes.
Total sleep time and sleep efficiency were also unrelated to presleep thalamocortical coupling changes.
This pattern of selective associations suggests the presleep brain state specifically influences the architecture and composition of sleep rather than its overall quantity or efficiency.
The null findings for overall sleep metrics highlight the specificity of the connectivity-sleep stage relationships observed.
Methods
The study examined a sample of adults with clinically significant insomnia symptoms using a within-evening repeated fMRI design followed by polysomnography.
Twenty adults participated (8 men, 12 women; age 19-39 years).
Participants completed two resting-state fMRI scans approximately 2 hours apart in the early evening (~6:30 and ~8:30 p.m.).
An overnight in-lab sleep study with polysomnographic recording followed the evening fMRI sessions.
The study design allowed for assessment of short-term intra-evening changes in thalamocortical connectivity and their prospective relationship to sleep architecture.
What This Means
This research suggests that the way the brain's thalamus (a central relay station) communicates with the rest of the brain in the hours before bedtime can predict what kind of sleep a person will get that night. Researchers scanned the brains of 20 adults with insomnia symptoms twice during the evening—around 6:30 and 8:30 p.m.—and then monitored their sleep with standard sleep lab equipment. By looking at how the brain's connectivity patterns changed between those two evening scans, the researchers could identify which patterns were linked to more or less time in different sleep stages later that night.
The findings show that different patterns of 'winding down' in thalamocortical brain communication are linked to different aspects of sleep. For example, when brain regions involved in visual and cognitive processing became more connected to the thalamus during the evening, people later spent more time in light sleep (N2). Conversely, when regions associated with arousal and sensory awareness became less connected to the thalamus, people later got more deep, restorative sleep (N3). Reduced connectivity with a visual brain region predicted more REM sleep, while reduced connectivity with the cerebellum was linked to spending more time awake in bed. Notably, these evening brain patterns did not predict total sleep time or overall sleep efficiency—only the specific composition of sleep stages.
This research suggests that the brain actively 'prepares' for sleep during the evening hours through shifting connectivity patterns, and that these shifts have measurable consequences for sleep quality later that night. For people with insomnia, understanding how evening brain activity shapes subsequent sleep could eventually point toward new approaches for improving sleep—for instance, by identifying evening behaviors or interventions that promote the kinds of brain connectivity patterns associated with deeper, more restorative sleep.
Killgore W, Henderson-Arredondo K, Jankowski S, Patel S, Grandner M, Hildebrand L, et al.. (2026). Winding down for the night: changes in thalamocortical connectivity before bed are associated with subsequent sleep-stage duration.. Neuroreport. https://doi.org/10.1097/WNR.0000000000002245