Disturbed laterality of non-rapid eye movement sleep oscillations in post-stroke human sleep: a pilot study

Ethics, consent and permissions

This research was conducted in accordance with and approval of the Cedars-Sinai Medical Center Institutional Review Board (IRB). All research participants and/or their surrogates provided informed consent to participate in the study.

Inclusion/exclusion criteria

Retrospective chart review of the Cedars Sinai EEG database was done to identify patients with acute middle cerebral artery strokes (MCA strokes; with high probability of stroke lesion affecting sensorimotor regions in the brain) who also received EEG monitoring as part of their hospital stay. We selected patients who received EEG in the acute period (2-3 days) post-stroke. Other inclusion criteria were that this should be the first stroke for the patient, they should be within 50-80 years of age, and the patients should not have any sleep disorders or circadian /diurnal rhythm disruption. Subjects were excluded if they met the following criteria: currently pregnant or diagnosed with uncontrolled medical conditions. Five patients were retrospectively identified for this study, with notable limited availability of EEG studies done within 2-3 days after an MCA distribution stroke. Of the 5 patients, 3 were female and 2 male, all within the age range of 50-80 years old (see Table 1 for other details regarding demographic and clinical information). Indications for EEG were universal for altered mental status after acute stroke. Unlike all other patients, P4 had subcortical involvement in stroke. It is important to note that spindle oscillations are postulated to have have a subcortical (thalamocortical) origin³². Also of note is that P5 had a hemorrhagic stroke (ruptured right MCA aneurysmal stroke) and was on norepinephrine due to shock which improved within 24 hours. P2 had partial status epilepticus involving the right temporal lobe. We excluded seizure related epochs based on manual inspection of recordings. This inspection was done by epileptologist (CMR) and seizures were excluded based on no evolving seizure pattern across electrodes (10-20 EEG system). Hence, all of our presented data was from sleep periods in all the five patients (even in the patient with status epilepticus). An average of ∼5.9801 ± 1.2563 hours (or 358.8036 ± 75.4 mins, mean ± s.e.m.) of NREM sleep was identified and analyzed in each of the five patients. Additionally, we analyzed a dataset of 3 healthy subjects from Cox et. al, Sleep medicine reviews, 2020^33,34 where average NREM sleep analyzed was 3.0652 ± 0.1396 hours (or 183.9100 ± 8.3773 mins) for 3 subjects. We were not able to analyze REM/ wake periods in these recordings due to inability to differentiate these periods due to the lack of EMGs/ video recordings.

EEG analysis and identification of NREM oscillations

Patients with overnight EEG studies 2 to 3 days post-stroke, with appropriate clinic follow up, were included. The data, obtained by a Natus Xltek EEG and Sleep System, was de-identified and made compatible for analysis with MATLAB. Each 30-second epoch was individually and manually marked for NREM sleep by an expert scorer (C.M.R. and B.K.S.). EEG epochs were analyzed for NREM sleep in a bipolar montage. The following analysis was done with EEG data in a referential montage, referenced to the auricle electrodes. Spindles, SOs, and δ waves were extracted from these NREM epochs using custom code in MATLAB (details below). This allowed for the identification of specific sleep waveforms and how they nested temporally and topographically during NREM sleep. An emphasis was placed on assessing spindles and their nesting to SOs and δ waves. Topographical maps of the average density of these sleep waveforms allowed us to visualize the average densities with respect to electrode location, especially their lateral symmetry between hemispheres.

From the healthy subjects dataset, we used the common linked mastoids referenced data³³ and analyzed NREM sleep. We also identified and only selected 20 electrode channels in similar locations as stroke patient data for further analysis (because the healthy subject data had more electrodes than our stroke patient dataset). Similar to stroke EEG data, spindles, SOs, and δ waves were extracted from these NREM epochs using custom code in MATLAB and analyzed.

EEG Data processing

For stroke patients, NREM-marked EEG data from all channels was referenced with respect to the average of the auricular electrodes (A1 & A2, Fig. 1A) while the heathy control dataset had common linked mastoids referenced EEG data. Any high amplitude artifact in the differential EEG signal was removed. This data was filtered into the frequency ranges of 0.1-4 Hz for δ/SOs identification and 10-16 Hz for spindle detection. We used previously used methods for automatic detection of these NREM oscillations^6,14,35. For δ/SOs detection, signal was first passed through a 0.1 Hz high-pass filter and then a 4 Hz low-pass Butterworth filter. All positive-to-negative zero crossings, previous peaks, following troughs, and negative-to-positive zero crossings were identified. A wave was considered a δ wave if its trough was lower than the negative threshold and preceded by a peak that was lower than the positive threshold, within 500 ms (Fig. 1B, E, H). SOs were classified as waves with troughs lower than a negative threshold (the bottom 40 percentile of the troughs) and preceding peaks higher than a positive threshold (the top 15 percentile of the peaks; Fig. 1C, F, I). Duration between peaks and troughs was between 150 ms and 500 ms. For spindle detection, EEG data was filtered using a 10 Hz high-pass Butterworth filter and a 16 Hz low-pass Butterworth filter. A smoothed envelope of this signal was calculated using the magnitude of the Hilbert transforms with convolving by a Gaussian window (200 ms). Epochs with signal amplitude higher than the upper threshold (mean, µ + 2.5*s.d., σ) for at least one sample and amplitude higher than the lower threshold (µ + 1.5*σ) for at least 500 ms were considered spindles (Fig. 1 D, G, J). The lower threshold was used to define the duration of the spindle. Nested SO-spindles (similar to k-complexes studied in humans) were identified as spindle peaks following SO peaks within 1.5 s duration (Fig. 1K). The same criterion was used to identify δ wave-nested spindles (Fig. 1L).

• Download figure
• Open in new tab

Figure 1. Stroke versus contralateral mirror/non-mirror electrode assignment and NREM sleep oscillations.

A, 10–20 system for EEG (used in stroke patients) showing locations of all electrode locations recorded with an illustration of stroke. Grey shaded area shows a representative stroke perilesional region. Blue shaded circles represent auricular electrodes (A1, A2) that were used for referencing in the stroke patients. Red circles indicate identified stroke electrodes based on proximity to the perilesional area. Green circles indicate identified contralateral mirror (CM) electrodes which are contralateral and mirrored to identified stroke electrodes. Yellow circles indicate identified contralateral non-mirror (CNM) electrodes which are electrodes other than contralateral mirror (CM) electrodes in non-stroke hemisphere. B, Mean δ– wave along with s.e.m. (standard error of mean) bands (blue) for all identified δ–waves from an example stroke electrode channel from EEG data recording for one stroke patient. C, Same as B for SO waveforms. D, Same as B for spindle waveforms. E, F, G, Same as B, C, D for one example contralateral mirror electrode channel for a stroke patient. H, I, J Same as B, C, D for one example channel for a healthy subject. All waveforms are centered around the detected states. K, Illustration of SO-spindle nesting. Nesting window was –0.5 to +1.0 s from SO’s UP state as shown. L, Illustration of δ–wave-spindle nesting. Nesting window was –0.5 to +1.0 s from δ UP state as depicted.

Data Analysis

We generated topographical maps of these different waveforms using plot_topography function in MATLAB³⁶ as shown in Fig. 2. The patients were separated into 3 different groups based on concurrent medications, as detailed in Table 1. Patient 1 was assigned to Group 1, who was on continuous propofol and dexamethasone injections every four hours. Group 2, patients 2 and 5, was administered levetiracetam (Keppra) twice daily; and Group 3, comprised of patients 3 and 4, was not on medications known to significantly modulate excitatory/inhibitory neural transmission.

• Download figure
• Open in new tab

Figure 2. Imaging data and topographical density plots for different NREM oscillations.

Top to bottom: Imaging data: CT (computed tomography) image for patient P1, T2 sequences of MRI (magnetic resonance imaging) images for patients P2 to P5; no imaging data available for healthy subjects (P6 to P8). Radiologic imaging has been flipped horizontally to align with topographic density maps; i.e., image left, and right are ipsilateral to patient left and right. Left and right are marked in imaging figures (P1-P5) and apply to density topographical maps below them; Topographical maps for detected spindle density (count/min) during NREM sleep for all subjects; Topographical maps for detected SO density (count/min) during NREM sleep for all subjects; Topographical maps for detected δ waves’ density (count/min) during NREM sleep for all subjects; Topographical maps for detected nested SO-spindle density (count/min) during NREM sleep for all subjects; Topographical maps for detected δ wave-nested-spindle density (count/min) during NREM sleep for all subjects. Color map shown at right for all the panels in a row.

Perilesional electrodes were identified by analyzing post-stroke MRI and CT brain imaging. We marked ‘Stroke electrodes’ as the electrodes covering the perilesional region of the brain as shown in Fig. 1A. The mirror opposite electrodes on the contralateral side were marked as “Contralateral mirror (CM) electrodes’ for further analysis (Fig. 1A). The non-mirror opposite electrodes on the contralateral side were marked as “Contralateral non-mirror (CNM) electrodes’.

We compared the symmetry in NREM oscillations’ density across hemispheres for stroke patients and healthy control using a laterality index (Fig. 3A-F). Laterality index of 1 meant the average density being analyzed for electrode locations selected across hemisphere is exactly same. For stroke patients, laterality index is defined as the ratio of mean of stroke electrodes’ NREM densities to all contralateral electrodes’ NREM densities. For healthy subjects, laterality index is defined as the ratio of mean of left hemisphere electrodes’ NREM densities to right hemisphere electrodes’ NREM densities. We also compared the ratio of SO-nested spindles laterality index to δ wave-nested spindles laterality index for stroke vs healthy subjects.

• Download figure
• Open in new tab

Figure 3. NREM oscillations’ laterality in stroke patient’s vs healthy controls; and NREM oscillations’ densities for different patient groups on stroke verses contralateral mirror (CM) electrodes.

For stroke patients’ laterality index (LI) is defined as ratio of mean of stroke electrode NREM densities to all contralateral electrode NREM densities. For healthy subjects’ laterality index is defined as ratio of mean of left hemisphere electrode NREM densities to right hemisphere electrode NREM densities. A, LI for SO density for stroke patients and healthy controls. Black line connects the mean of stroke and control group. Dots represent different patients/subjects; blue dots: Patients in propofol medication group; orange dots: Patients in levetiracetam medication group; green dots: Stroke patients in other medication group; yellow dots: Healthy subjects. B, Same as A for δ wave density LI. C, Same as A for spindle density LI. D, Same as A for nested SO-spindle density LI. E, Same as A for Nested δ wave-spindle density LI. F, Ratio of LI for nested SO-spindle density and nested δ wave-spindle density. G, Table showing selected stroke and contralateral mirror electrodes (CM) for all patients. H, Comparison of δ wave density (count/min) on stroke versus CM electrodes for patients on different medications. Thick black line shows the mean values within the group. Thinner black lines join pair of stroke and CM electrode. Dots represent the NREM oscillations’ density for single electrode. I, Same as H for SO density. J, Same as H for spindle density. K, Same as H for nested δ wave-nested spindle density. L, Same as H for SO-nested spindle density. *: statistically significant p values for two-tailed t-test.

Statistical Analysis

We performed a linear mixed effect analysis for all patients comparing the Stroke electrodes density vs Contralateral (CM/CNM) electrodes density for different waveforms using the fitlmematrix function in MATLAB. The linear mixed effect model was fitted by maximum likelihood using the formula below (1) for all the different waveforms identified during EEG data processing. Medication groups were defined as the 3 groups as mentioned earlier. This model considers fixed effects of stroke vs contralateral (CM/CNM) electrodes, and the random effect of electrodes and medication groups depending on the patient and can be represented as following equation: The above formula/equation is written in a format similar to documentation for fitlmematrix Matlab function. We compared the Stroke electrodes density vs contralateral (CM/CNM) electrodes density within each medication group using a two-tailed t-test. Contralateral electrodes chosen were the exactly mirrored electrodes (Fig. 3G–L) or non-mirrored (Supp. Fig. 2A-F). One-way ANOVA was used to compare the stroke electrodes’ NREM oscillations’ density of the 3 different medication groups.

We calculated r-squared (R²) and the Cohen’s d values for the overall linear mixed effect model generated. However, the p-values were specifically assessed for fixed effect of electrodes (stroke vs CM/CNM). Cohen’s d was used to evaluate if the nested data (all data combined) for NREM oscillations had a small, medium or large experimental effect (Cohen’s d = 0.20, 0.50 or 0.80, respectively)³⁷. Effect size indicates if such research findings have practical significance. Metrics such as Cohen’s d are better at the planning stage for pilot studies such as the one here to determine optimal sample sizes for sufficient power in bigger clinical trials³⁸. We have summarized the linear mixed effects models results in a table in the Supplementary Information.

NREM oscillation densities symmetry is disturbed acutely in stroke

We found that stroke patients had laterality differences (densities were higher or lower in stroke hemisphere) for all NREM oscillations, while the healthy subject NREM oscillation density looked more symmetrical across hemispheres (Fig. 2). Comparing the laterality index (LI) (as defined in methods), we found that the LI was closer to 1 on average for healthy subjects and had low variance but there were differences in stroke patients. SO density LI, stroke: 1.7761 ± 0.3387 and healthy: 1.0520 ± 0.0587 (Fig. 3A). For δ wave density LI, stroke: 1.9251 ± 0.4368 and healthy: 1.0521 ± 0.0592 (Fig. 3B). For spindle density LI, stroke: 1.6478 ± 0.2736 and healthy: 1.0498 ± 0. 0.0669 (Fig. 3C). For SO-nested with spindles LI, stroke: 1.6325 ± 0.2876 and healthy: 1.0884 ± 0.0882 (Fig. 3D). For δ wave-nested with spindles LI, stroke: 1.6299 ± 0.3387 and healthy: 1.0525 ± 0.0608 (Fig. 3E). The ratio of nested SO-spindles LI and δ wave-nested spindle LI, stroke: 0.8948 ± 0.1201 and healthy: 1.0280 ± 0.0344 (Fig. 3F).

SO and δ wave density increased in perilesional electrodes

Next, we wanted to closely look at stroke-affected electrodes in stroke patients vis-à-vis the contralateral hemisphere electrodes in same stroke subjects. In the contralateral hemisphere, we looked at exactly mirrored electrodes (CM, as defined in the methods; Fig. 3G), or non-mirrored electrodes (CNM, as previously in methods; Supp. Fig. 2A). Consistent with previous reports, we found that stroke electrodes had higher < 4 Hz oscillations (Fig. 3H,I; and Supp. Fig. 2B,C)³⁹. Our mixed-effects model showed a significant fixed effect of stroke vs CM and CNM electrodes for a subset of NREM oscillations and overall medium to large effect sizes which included random effects of concurrent pharmaceuticals. Overall, we observed higher δ wave density in the perilesional electrodes (Fig. 3H; Supp. Fig. 2B; Supp. Table 1 and 2 provide statistical details for stroke versus CM or CNM: p-value is provided for the fixed effect (’electrode’), R² and Cohen’s d are for the overall model with fixed and random effects, coventions same henceforth). Our comparison of laterality index of SOs and δ wave showed that LI was higher in stroke patients compared to healthy subjects: Mean LI SOs, stroke: 1.7761 ± 0.3387 and healthy: 1.0520 ± 0.0587; Mean LI δ wave, stroke: 1.9251 ± 0.4368 and healthy: 1.0521 ± 0.0592. We also observed that the stroke patients on propofol and dexamethasone (Group-1) and Group-3 both had higher δ wave density on stroke electrodes than the levetiracetam group (Fig. 3H and Supp. Fig. 2B; stroke electrodes’ δ wave density-Group 1: 11.2315 ± 2.5270 counts min⁻¹ (mean ± s.e.m.); Group 2: 9.0653 ± 1.3206 counts min⁻¹; Group 3: 12.2454 ± 1.5866 counts min⁻¹, see Supp. Table 3 for details). Levetiracetam (Group-2) and Group-3 showed high density of δ waves in the stroke electrodes vs CM/ CNM electrodes (Fig. 3H and Supp. Fig. 2B). For SOs as well, there was a significant fixed effect of stroke vs contralateral electrodes (Fig. 3I; Supp. Fig. 2C; Supp. Table 1 and 2 provide p-values and Cohen’s d). When we looked at stroke patients with different drugs, we observed that the patients in Group-1 did not show a significant difference between stroke or contralateral electrode SO density; patients in Group-2 showed elevation in SO on stroke electrodes when compared to CM electrodes (Fig. 3I). The patients in Group-3 showed more SOs on stroke electrodes when compared for CM/CNM electrodes (Fig. 3I; Supp. Fig. 2C; stroke electrodes’ SO density: Group 1: 2.9134 ± 0.7147 counts min⁻¹; Group 2: 2.4204 ± 0.3701 counts min⁻¹; Group 3: 3.2858 ± 0.4464 counts min⁻¹; see Supp. Table 3 for details).

For spindle oscillations, LI was higher in stroke patients (Mean LI spindles, stroke: 1.6478 ± 0.2736 and healthy: 1.0498 ± 0. 0.0669). Interestingly, in one patient with subcortical involvement with stroke (P4), spindles were higher in contralesional hemisphere (Fig. 3J). Linear mixed-effects model did not show a significant fixed effect of spindle density on stroke versus contralateral electrodes, however, overall it was a medium effect based on the Cohen’s d (Fig. 3J and Supp. Fig. 2D; see Supp. Table 1 and 2 for p-value and Cohen’s d). Upon looking at different stroke patients, spindle density was the highest on the stroke electrodes in the patient in Group-1 (8.0037 ± 0.8763 counts min⁻¹), followed by the patients in Group-2 (6.8267± 0.7872 counts min⁻¹), and then patients in Group-3 (5.6120 ± 0.4366 counts min⁻¹) (Fig. 3J and Supp Fig. 2D; see Supp. Table 3 for details).

δ wave-nested spindles and SO-nested spindles

Next we wanted to look at nested oscillations, namely, δ wave-nested spindles and SO-nested spindles oscillations that were shown to have a competing role recently and inverse trend during stroke recovery^6,14. LI for both nested oscillations was higher in stroke subjects: Mean LI SO-nested spindle, stroke: 1.6325 ± 0.2876 and healthy: 1.0884 ± 0.0882; Mean LI δ wave-nested spindle, stroke: 1.6299 ± 0.3387 and healthy: 1.0525 ± 0.0608. While the mixed effects models on δ wave-nested spindles and SO-nested spindles did not show a significant difference between stroke and contralateral electrodes, overall these models still had large and medium effect sizes, respectively (Supp. Table 1 and 2, Fig. 3K and Supp. Fig. 2E, δ wave-nested spindle density on stroke electrodes: Group-1: 3.4911 ± 0.3032 counts min⁻¹; Group-2: 3.2476 ± 0.4770 counts min^{− 1}; Group-3: 2.7031 ± 0.2032 counts min⁻¹, also see Supp. Table 3; SO-nested spindle density on stroke electrodes: Group 1: 0.9150 ± 0.1111 counts min⁻¹; Group 2: 0.8583 ± 0.1751 counts min⁻¹; Group 3: 0.6828 ± 0.0576 counts min⁻¹; see Fig. 3L; Supp. Fig. 2F; and Supp. Table 3). Notably, the ratio of SO-nested spindle LI to δ wave-nested spindle LI was lower in stroke subjects compared to heathy subjects (Mean LI ratio, stroke: 0.8948 ± 0.1201 and healthy: 1.0280 ± 0.0344), which indicates relatively increased δ wave-nested spindles when compared to SO-nested spindles (the oscillations that have a competing role in forgetting vs strengthening, respectively) in the perilesional areas for stroke brain compared to healthy brain.

Together, these results show that lateral symmetry of NREM oscillations is disturbed in stroke (Fig. 3A-F), when compared to healthy subjects. These results also indicate that there is an elevation of SO, δ wave, spindles, and spindle nesting to SOs or δ waves in the perilesional areas post-stroke.

Sleep and plasticity post-stroke

Preclinical and clinical studies that have evaluated local-field potentials (LFPs) in animals^40,41 and EEG in human patients^22,42,43 have found increased low-frequency power during awake, spontaneous periods after a stroke. These studies postulate that this increased low-frequency activity could be a marker of cortical injury and loss of subcortical inputs⁴⁴. Our findings on increased SOs and δ waves on stroke electrodes are indicative of similar phenomena. We also found an increase in SO-nested spindles and δ wave-nested spindles on stroke electrodes along with a lower ratio of SO-nested spindle LI/ δ wave-nested spindle LI (Fig. 3F). There is growing evidence that temporal coupling of spindles to SOs is a primary driver of sleep-related plasticity and memory consolidation^{6,30,31,45–48}. SO-nested spindles are linked to spike-time dependent plasticity⁴⁹. These events are also related to reactivation of awake experiences^30,47,50. Importantly, disruption of this coupling can impair sleep-related memory consolidation of awake experiences⁶. This same work showed that SO-nested spindles and δ wave-nested spindles compete with each other to either strengthen or forget a memory. Our results indicate that balance of SO-nested spindle density and δ wave-nested spindle density is disturbed hemispherically in stroke patients compared to healthy subjects. These might be related to impaired sleep-processing that may impact recovery. Interestingly, we observed large-to medium effect sizes in our mixed-effects models for δ wave-nested spindle and SO-nested spindle where we considered fixed effects of electrodes and random effects of drugs and patients. It is worth noting that drugs like propofol can impact such nested sleep oscillations^51,52. It may be important to consider effects of drugs on sleep oscillations when modulating sleep for stroke recovery.

Propofol and Levetiracetam: effect on sleep

We made some observations on different medications that stroke patients received during sleep EEG recordings. Group-1 patient received propofol, which is one of the most commonly used anesthetics in neurologic intensive care units after stroke or traumatic brain injury⁵³. It exerts its action by potentiating the activity of chloride currents through GABA receptors while blocking voltage-gated sodium channels^54–56. The patient on propofol received less than 10 mcg dose of propofol at which does it is not known to impact sleep^57,58. Group-2 patients received levetiracetam (keppra), which is a relatively newer anti-seizure drug. The exact mechanism for its anti-seizure mechanism is unclear, but it is believed to exert its effect through synaptic vesicle glycoprotein 2A⁵⁹. Through this mechanism, levetiracetam is capable of modulating neurotransmission by inhibiting calcium currents⁶⁰. A study has shown that levetiracetam has minimal effect on sleep parameters like total sleep duration, sleep latency, and sleep efficiency in both healthy humans and partial epilepsy patients⁶¹. However, observations have been made that levetiracetam can reduce motor activity and cause daytime drowsiness in patients^61,62. Propofol, by its GABAergic action, causes greater loss of faster frequencies during induction with a shift in alpha frequencies to the frontal regions that reverses post-awakening^63–65. Since our mixed-effects model found large to medium effects when considering random effects of drugs on all NREM oscillation, it may be useful to explore effects of drugs on NREM sleep densities with larger patient cohorts in the future. Since our data shows disturbed laterality of NREM oscillations, different drugs that stroke patients are on might need to be factored in, when restoring physiological sleep oscillations post-stroke.

Sleep processing and stroke rehabilitation

Recent rodent work profiled SO-nested and δ wave-nested spindles during the course of stroke recovery and found links between these nested structures and motor performance gains during recovery⁶. This work specifically looked into gains on reach-training, but clinical rehabilitation approaches can be varied^66–68. It is likely that the sleep features of nested oscillations and their putative pathological or physiological roles need to be factored in when considering timing for rehabilitation, irrespective of training type. Previous human and rodent studies have also suggested critical periods in training can offer long-term benefits^69–71. Past studies that have found awake low-frequency power in stroke patients might be related to our findings of increased SO and δ waves densities. Future studies where EEG data is captured over the long term may delineate a transition of δ wave LI, SOs LI, δ wave-nested spindles LI (pathological sleep) and SO-nested spindle LI (physiological sleep), and its relation to critical periods post-stroke for optimal timing of rehabilitation. For example, SO-nested spindles LI and δ wave-nested spindles LI proportions between hemispheres could be targeted to be brought closer to unity as in healthy subjects, to accelerate recovery.

Modulation of sleep as a therapeutic intervention

The results we have presented can form the basis of translational studies in the future that target modulation of sleep post-stroke. Animal studies have suggested that modulation of GABAergic transmission (specifically GABA_A-receptor mediated tonic inhibition) in the perilesional cortex can serve as a therapeutic target to promote recovery, and that blocking of GABA_A-mediated tonic inhibition promoted motor recover maximally in the first 1 to 2 weeks post-stroke^72,73. Both short-term (acute) and long-term chronic infusion of GABA_A inhibiting compounds have been tested, and long-term infusion was shown to be better⁷². Long-term pharmacologic modulation as shown by Clarkson and colleagues may be essential to achieve observable motor benefits in human patients. Benefits of long-term infusion include effect of the drug not only with rehabilitation-specific online (awake) training, but also during offline memory consolidation during sleep.

Studies such as ours can also help guide electric stimulation-based neuromodulation for augmenting recovery. SOs and δ waves can be easily monitored using EEG in stroke patients. Non-invasive brain stimulation during sleep^30,47,74,75 can be used to modulate specific NREM oscillations. Invasive stimulation approaches such as epidural stimulation⁷⁶ can also focus on sleep state to optimize sleep neural processing. Similar approaches have shown that direct epidural motor cortical electric stimulation can enhance awake performance and neural activity^77,78 and epidural stimulation of subcortical regions can also modulate low-frequency oscillations in the motor cortex⁷⁹, however such approaches have not been applied during sleep. A recent study suggested modulating UP states during sleep can enhance recovery¹⁸. It is plausible that future approaches targeting sleep, when delivered in a closed-loop fashion, optimize both awake task performance and its consequent sleep processing, and may lead to greater long-term benefits during rehabilitation. Indices such as laterality index that we pursued here may serve a utilitarian purpose in long-term sleep evaluation post-stroke with different treatments. Our pilot observations here also suggest that concurrent pharmacologic drugs may affect NREM oscillations and hence they should also be considered when personalizing sleep stimulation.

Limitations

One of the limitations of our study is the lack of a link between sleep architecture and motor status. Future work that studies sleep over several days post-stroke and assesses motor functionality longitudinally may find more robust links between sleep processing and related gains in motor performance. It is also possible that, with more effective task performance and associated awake neural dynamics^77,78,80, efficacy of sleep may change. Precise disruption of sleep processing, specifically SO-spindle coupling in healthy animals, was sufficient to prevent offline performance gains, even when awake task learning was robust⁶. This work also showed that precise modulation of the extent of sleep spindle-SO coupling in healthy animals could either enhance or impede sleep processing. While extension of this work in stroke animals has shown SO-spindle nesting resurges with recovery¹⁴, future animal studies that modulate sleep microarchitecture can study if artificial manipulation of SO-nested spindles or δ wave-nested spindles after stroke are sufficient to enhance or impair motor recovery. Our work here showed that both SO-nested spindles and δ wave-nested spindles increased in stroke affected hemisphere acutely post-stroke. Future work that monitors these oscillations for longer periods will be needed to assess if SO-nested spindles should increase relatively to δ wave-nested spindles for better recovery in human stroke patients.

One more limitation of this study is the limited sample size with varying lesion location and size. But ours is a pilot study. While we focused on getting patients with cortical lesions, sleep may have been impacted differently for one patient with a primarily subcortical stroke. For example, a stroke in the white matter that impacts thalamocortical networks may impact spindles (as we saw in our P4, with higher spindles in contralesional hemisphere), believed to have a thalamocortical origin. Future work with larger sample sizes and incorporation of motor task rehabilitation training and drug manipulation, may provide stronger links to engineer sleep to benefit motor recovery post-stroke.