Cognitive control affects motor learning through local variations in GABA within the primary motor cortex

Neural substrates of choking under pressure: A 7T-fMRI study

Performance decrement under excessive psychological pressure is known as "choking," yet its mechanisms and neural foundations remain underexplored. Hypothesizing that changes in the internal model could induce choking, we conducted a 7 T functional MRI introducing excessive pressure through a rare Jackpot condition that offers high rewards for successful performance. Twenty-nine volunteers underwent a visual reaching task. We monitored practice and main sessions to map the task's internal model through learning. Participants were pre-informed of four potential reward conditions upon success at the beginning of the main session task. The success rates in the Jackpot condition were significantly lower than in other conditions, indicative of choking. During the preparation phase, activations in the cerebellum and the middle temporal visual area (hMT+) were associated with Jackpot-specific failures. The cluster in the cerebellar hemisphere overlapped with the internal model regions identified by a learning-related decrease in activation during the practice session. We observed task-specific functional connectivity between the cerebellum and hMT+. These findings suggest a lack of sensory attenuation when an internal model predicting the outcome of one's actions is preloaded during motor preparation. Within the active inference framework of motor control, choking stems from the cerebellum's internal model modulation by psychological pressure, manifested through improper sensory attenuation.

Acute stress differently modulates interneurons excitability and synaptic plasticity in the primary motor cortex of wild-type and SOD1G93A mouse model of ALS

Primary motor cortex (M1) network stability depends on activity of inhibitory interneurons, for which susceptibility to stress was previously demonstrated in limbic regions. Hyperexcitability in M1 following changes in the excitatory/inhibitory balance is a key pathological hallmark of amyotrophic lateral sclerosis (ALS). Using electrophysiological approaches, we assessed the impact of acute restraint stress on inhibitory interneurons excitability and global synaptic plasticity in M1 of the SOD1G93A ALS mouse model at a late pre-symptomatic stage (10-12.5 weeks). Based on their firing type (continuous, discontinuous, with accommodation or not) and electrophysiological characteristics (resting potential, rheobase, firing frequency), interneurons from M1 slices were separated into four clusters, labelled from 1 to 4. Among them, only interneurons from the first cluster, presenting continuous firing with few accommodations, tended to show increased excitability in wild-type (WT) and decreased excitability in SOD1G93A animals following stress. In vivo analyses of evoked field potentials showed that stress suppressed the theta burst-induced plasticity of an excitatory component (N1) recorded in the superficial layers of M1 in WT, with no impact on an inhibitory complex (N2-P1) from the deeper layers. In SOD1G93A mice, stress did not affect N1 but suppressed the N2-P1 plasticity. These data suggest that stress can alter M1 network functioning in a different manner in WT and SOD1G93A mice, possibly through changes of inhibitory interneurons excitability and synaptic plasticity. This suggests that stress-induced activity changes in M1 may therefore influence ALS outcomes. KEY POINTS: Disruption of the excitatory/inhibitory balance in the primary motor cortex (M1) has been linked to cortical hyperexcitability development, a key pathological hallmark of amyotrophic lateral sclerosis (ALS). Psychological stress was reported to influence excitatory/inhibitory balance in limbic regions, but very little is known about its influence on the M1 functioning under physiological or pathological conditions. Our study revealed that acute stress influences the excitatory/inhibitory balance within the M1, through changes in interneurons excitability along with network plasticity. Such changes were different in pathological (SOD1G93A ALS mouse model) vs. physiological (wild-type) conditions. The results of our study help us to better understand how stress modulates the M1 and highlight the need to further characterize stress-induced motor cortex changes because it may be of importance when evaluating ALS outcomes.

Inhibitory mechanisms in the prefrontal-cortex differentially mediate Putamen activity during valence-based learning

Learning from appetitive and aversive stimuli involves interactions between the prefrontal cortex and subcortical structures. Preclinical and theoretical studies indicate that inhibition is essential in regulating the relevant neural circuitry. Here, we demonstrate that GABA, the main inhibitory neurotransmitter in the central nervous system, differentially affects how the dACC interacts with subcortical structures during appetitive and aversive learning in humans. Participants engaged in tasks involving appetitive and aversive learning, while using functional magnetic resonance spectroscopy (MRS) at 7T to track GABA concentrations in the dACC, alongside whole-brain fMRI scans to assess BOLD activation. During appetitive learning, dACC GABA concentrations were negatively correlated with learning performance and BOLD activity measured from the dACC and the Putamen. These correlations were absent during aversive learning, where dACC GABA concentrations negatively correlated with the connectivity between the dACC and the Putamen. Our results show that inhibition in the dACC mediates appetitive and aversive learning in humans through distinct mechanisms.

MRS-assessed brain GABA modulation in response to task performance and learning

Gamma-aminobutyric acid (GABA), the most important inhibitory neurotransmitter in the human brain, has long been considered essential in human behavior in general and learning in particular. GABA concentration can be quantified using magnetic resonance spectroscopy (MRS). Using this technique, numerous studies have reported associations between baseline GABA levels and various human behaviors. However, regional GABA concentration is not fixed and may exhibit rapid modulation as a function of environmental factors. Hence, quantification of GABA levels at several time points during the performance of tasks can provide insights into the dynamics of GABA levels in distinct brain regions. This review reports on findings from studies using repeated measures (n = 41) examining the dynamic modulation of GABA levels in humans in response to various interventions in the perceptual, motor, and cognitive domains to explore associations between GABA modulation and human behavior. GABA levels in a specific brain area may increase or decrease during task performance or as a function of learning, depending on its precise involvement in the process under investigation. Here, we summarize the available evidence and derive two overarching hypotheses regarding the role of GABA modulation in performance and learning. Firstly, training-induced increases in GABA levels appear to be associated with an improved ability to differentiate minor perceptual differences during perceptual learning. This observation gives rise to the 'GABA increase for better neural distinctiveness hypothesis'. Secondly, converging evidence suggests that reducing GABA levels may play a beneficial role in effectively filtering perceptual noise, enhancing motor learning, and improving performance in visuomotor tasks. Additionally, some studies suggest that the reduction of GABA levels is related to better working memory and successful reinforcement learning. These observations inspire the 'GABA decrease to boost learning hypothesis', which states that decreasing neural inhibition through a reduction of GABA in dedicated brain areas facilitates human learning. Additionally, modulation of GABA levels is also observed after short-term physical exercise. Future work should elucidate which specific circumstances induce robust GABA modulation to enhance neuroplasticity and boost performance.

Enhanced Post-Movement Beta Rebound: Unraveling the Impact of Preplanned Sequential Actions

The Post-Movement Beta Rebound (PMBR) is the increase in beta-band power after voluntary movement ends, but its specific role in cognitive processing is unclear. Current theory links PMBR with updates to internal models, mental frameworks that help anticipate and react to sensory feedback. However, research has not explored how reactivating a preexisting action plan, another source for internal model updates, might affect PMBR intensity. To address this gap, we recruited 20 participants (mean age 18.55 ± 0.51; 12 females) for an experiment involving isolated (single-step) or sequential (two-step) motor tasks based on predetermined cues. We compared PMBR after single-step movements with PMBR after the first movement in two-step tasks to assess the influence of a subsequent action on the PMBR power associated with the first action. The results show a significant increase in PMBR magnitude after the first movement in sequential tasks compared to the second action and the isolated movements. Notably, this increase is more pronounced for right-hand movements, suggesting lateralized brain activity in the left hemisphere. These findings indicate that PMBR is influenced not only by external stimuli but also by internal cognitive processes such as working memory. This insight enhances our understanding of PMBR's role in motor control, emphasizing the integration of both external and internal information.

Examination of acute spin exercise on GABA levels in aging and stroke: The EASE study protocol

BACKGROUND

Changes in regional levels of the inhibitory neurotransmitter gamma-aminobutyric acid (GABA) may indicate the potential for favorable responses to the treatment of stroke affecting the upper extremity. By selectively altering GABA levels during training, we may induce long-term potentiation and adjust excitatory/inhibitory balance (E/I balance). However, the impact of this alteration may be limited by neural damage or aging. Aerobic exercise has been shown to increase GABA levels in the sensorimotor cortex and improve motor learning by widening the dynamic range of E/I balance. The cross-sectional project, Effects of Acute Exercise on Functional Magnetic Resonance Spectroscopy Measures of GABA in Aging and Chronic Stroke (EASE), is designed to assess the functional relevance of changes in GABA concentration within the sensorimotor cortex before and after an acute aerobic exercise session.

METHODS/DESIGN

EASE will enroll 30 participants comprised of healthy younger adults (18-35 years; n = 10), older adults (60+ years; n = 10), and persons with chronic stroke (n = 10) affecting distal upper extremity function. We will use resting magnetic resonance spectroscopy to measure all participants' GABA levels at rest before and after aerobic exercise. In addition, we will employ functional magnetic resonance spectroscopy using motor skill acquisition and recall tasks in healthy adults. We hypothesize that acute aerobic exercise will increase resting sensorimotor GABA concentration and that higher GABA resting levels will predict better motor learning performance on measures taken both inside and outside the magnet. We also hypothesize that a higher dynamic range of GABA during task-based spectroscopy in healthy adults will predict better motor skill acquisition and recall.

DISCUSSION

The EASE project will evaluate the effect of acute exercise on GABA levels as a biomarker of upper extremity motor skill learning with two populations (aging adults and those with chronic stroke). We predict that acute exercise, higher sensorimotor GABA levels, and broader dynamic range will be related to better motor skill acquisition.

Electroencephalogram Alpha Oscillations in Stroke Recovery: Insights into Neural Mechanisms from Combined Transcranial Direct Current Stimulation and Mirror Therapy in Relation to Activities of Daily Life

The goal of stroke rehabilitation is to establish a robust protocol for patients to live independently in community. Firstly, we examined the impact of 3 hybridized transcranial direct current stimulation (tDCS)-mirror therapy interventions on activities of daily life (ADL) in stroke patients. Secondly, we explored the underlying therapeutic mechanisms with theory-driven electroencephalography (EEG) indexes in the alpha band. This was achieved by identifying the unique contributions of alpha power in motor production to ADL in relation to the premotor cortex (PMC), primary cortex (M1), and Sham tDCS with mirror therapy. The results showed that, although post-intervention ADL improvement was comparable among the three tDCS groups, one of the EEG indexes differentiated the interventions. Neural-behavioral correlation analyses revealed that different types of ADL improvements consistently corresponded with alpha power in the temporal lobe exclusively in the PMC tDCS group (all rs > 0.39). By contrast, alterations in alpha power in the central-frontal region were found to vary, with ADL primarily in the M1 tDCS group (r = -0.6 or 0.7), with the benefit depending on the complexity of the ADL. In conclusion, this research suggested two potential therapeutic mechanisms and demonstrated the additive benefits of introducing theory-driven neural indexes in explaining ADL.

A preliminary study of dynamic neurochemical changes in the dorsolateral prefrontal cortex during working memory

Working memory (WM) is one of the fundamental cognitive functions associated with the dorsolateral prefrontal cortex (DLPFC). However, the neurochemical mechanisms of WM, including the dynamic changes in neurometabolites such as glutamate and GABA in the DLPFC, remain unclear. Here, we investigated WM-related glutamate and GABA changes, alongside hemodynamic responses in the DLPFC, using a combination of functional magnetic resonance spectroscopy (fMRS) and functional magnetic resonance imaging (fMRI). During a WM task, we measured Glx (glutamate + glutamine) and GABA levels using GABA editing MEscher-GArwood Point REsolved Spectroscopy (MEGA-PRESS) sequence and blood-oxygen-level-dependent (BOLD) signal changes. In the DLPFC, we observed elevated Glx levels and increased BOLD signal changes during a 2-back task. Specifically, the Glx levels in the DLPFC were significantly higher during the 2-back task compared with fixation, although this difference was not significant when compared with a 0-back task. However, Glx levels during the 0-back task were higher than during fixation. Furthermore, there was a positive correlation between Glx levels in the DLPFC during the 2-back task and the corresponding BOLD signal changes. Notably, higher Glx increases were associated with increased DLPFC activation and lower WM task performance in individuals. No notable changes in DLPFC GABA levels were observed during WM processing. These findings suggest that the modulation of glutamatergic activity in the DLPFC may play a crucial role in both working memory processing and its associated performance outcomes.

Neural correlates of transfer of learning in motor coordination tasks: role of inhibitory and excitatory neurometabolites

We aimed to investigate transfer of learning, whereby previously acquired skills impact new task learning. While it has been debated whether such transfer may yield positive, negative, or no effects on performance, very little is known about the underlying neural mechanisms, especially concerning the role of inhibitory (GABA) and excitatory (Glu) (measured as Glu + glutamine (Glx)) neurometabolites, as measured by magnetic resonance spectroscopy (MRS). Participants practiced a bimanual coordination task across four days. The Experimental group trained a task variant with the right hand moving faster than the left (Task A) for three days and then switched to the opposite variant (Task B) on Day4. The control group trained Task B across four days. MRS data were collected before, during, and after task performance on Day4 in the somatosensory (S1) and visual (MT/V5) cortex. Results showed that both groups improved performance consistently across three days. On Day4, the Experimental group experienced performance decline due to negative task transfer while the control group continuously improved. GABA and Glx concentrations obtained during task performance showed no significant group-level changes. However, individual Glx levels during task performance correlated with better (less negative) transfer performance. These findings provide a first window into the neurochemical mechanisms underlying task transfer.

Early excitatory-inhibitory cortical modifications following skill learning are associated with motor memory consolidation and plasticity overnight

Consolidation of motor memories is vital to offline enhancement of new motor skills and involves short and longer-term offline processes following learning. While emerging evidence link glutamate and GABA dynamics in the primary motor cortex (M1) to online motor skill practice, its relationship with offline consolidation processes in humans is unclear. Using two-day repeated measures of behavioral and multimodal neuroimaging data before and following motor sequence learning, we show that short-term glutamatergic and GABAergic responses in M1 within minutes after learning were associated with longer-term learning-induced functional, structural, and behavioral modifications overnight. Furthermore, Glutamatergic and GABAergic modifications were differentially associated with different facets of motor memory consolidation. Our results point to unique and distinct roles of Glutamate and GABA in motor memory consolidation processes in the human brain across timescales and mechanistic levels, tying short-term changes on the neurochemical level to overnight changes in macroscale structure, function, and behavior.

Plasticity-stability dynamics during post-training processing of learning

Learning continues beyond the end of training. Post-training learning is supported by changes in plasticity and stability in the brain during both wakefulness and sleep. However, the lack of a unified measure for assessing plasticity and stability dynamics during training and post-training periods has limited our understanding of how these dynamics shape learning. Focusing primarily on procedural learning, we integrate work using behavioral paradigms and a recently developed measure, the excitatory-to-inhibitory (E/I) ratio, to explore the delicate balance between plasticity and stability and its relationship to post-training learning. This reveals plasticity-stability cycles during both wakefulness and sleep that enhance learning and protect it from new learning during post-training processing.

Increased cortical inhibition following brief motor memory reactivation supports reconsolidation and overnight offline learning gains

Practicing motor skills stabilizes and strengthens motor memories by repeatedly reactivating and reconsolidating them. The conventional view, by which a repetitive practice is required for substantially improving skill performance, has been recently challenged by behavioral experiments, in which even brief reactivations of the motor memory have led to significant improvements in skill performance. However, the mechanisms which facilitate brief reactivation-induced skill improvements remain elusive. While initial memory consolidation has been repeatedly associated with increased neural excitation and disinhibition, reconsolidation has been shown to involve a poorly understood mixture of both excitatory and inhibitory alterations. Here, we followed a 3-d reactivation-reconsolidation framework to examine whether the excitatory/inhibitory mechanisms which underlie brief reactivation and repetitive practice differ. Healthy volunteers practiced a motor sequence learning task using either brief reactivation or repetitive practice and were assessed using ultrahigh field (7T) magnetic resonance spectroscopy at the primary motor cortex (M1). We found that increased inhibition (GABA concentrations) and decreased excitation/inhibition (glutamate/GABA ratios) immediately following the brief reactivation were associated with overnight offline performance gains. These gains were on par with those exhibited following repetitive practice, where no correlations with inhibitory or excitatory changes were observed. Our findings suggest that brief reactivation and repetitive practice depend on fundamentally different neural mechanisms and that early inhibition-and not excitation-is particularly important in supporting the learning gains exhibited by brief reactivation.

GABA, glutamate and excitatory-inhibitory ratios measured using short-TE STEAM MRS at 7-Tesla: Effects of macromolecule basis sets and baseline parameters

Rationale and objectives

Macromolecules (MMs) affect the precision and accuracy of neurochemical quantification in magnetic resonance spectroscopy. A measured MM basis is increasingly used in LCModel analysis combined with a spline baseline, whose stiffness is controlled by a parameter named DKNTMN. The effects of measured MM basis and DKNTMN were investigated.

Materials and methods

Twenty-six healthy subjects were prospectively enrolled and scanned twice using a short echo-time Stimulated Echo Acquisition Mode (STEAM) at 7-T. Using LCModel, analyses were conducted using the simulated MM basis (MMsim) with DKNTMN 0.15 and an MM basis measured inhouse (MMmeas) with DKNTMN of 0.15, 0.30, 0.60 and 1.00. Cramér-Rao lower bound (CRLB) and the concentrations of gamma-aminobutyric acid (GABA), glutamate and excitatory-inhibitory ratio (EIR), in addition to MMs were statistically analyzed. Measurement stability was evaluated using coefficient of variation (CV).

Results

CRLBs of GABA were significantly lower when using MMsim than MMmeas; those of glutamate were 2-3. GABA concentrations were significantly higher in the analysis using MMsim than MMmeas where concentrations were significantly higher with DKNTMN of 0.15 or 0.30 than 0.60 or 1.00. Difference in glutamate concentration was not significant. EIRs showed the same difference as in GABA depending on the DKNTMN values. CVs between test-retest scans were relatively stable for glutamate but became larger as DKNTMN increased for GABA and EIR.

Conclusion

Neurochemical quantification depends on the parameters of the basis sets used for fitting. Analysis using MMmeas with DKNTMN of 0.30 conformed best to previous studies and is recommended.

Are sleep disturbances a cause or consequence of autism spectrum disorder?

Autism spectrum disorder (ASD) is a neurodevelopmental disorder characterized by core symptoms such as atypical social communication, stereotyped behaviors, and restricted interests. One of the comorbid symptoms of individuals with ASD is sleep disturbance. There are two major hypotheses regarding the neural mechanism underlying ASD, i.e., the excitation/inhibition (E/I) imbalance and the altered neuroplasticity hypotheses. However, the pathology of ASD remains unclear due to inconsistent research results. This paper argues that sleep is a confounding factor, thus, must be considered when examining the pathology of ASD because sleep plays an important role in modulating the E/I balance and neuroplasticity in the human brain. Investigation of the E/I balance and neuroplasticity during sleep might enhance our understanding of the neural mechanisms of ASD. It may also lead to the development of neurobiologically informed interventions to supplement existing psychosocial therapies.

Contingent negative variation to tactile stimuli - differences in anticipatory and preparatory processes between participants with and without blindness

People who are blind demonstrate remarkable abilities within the spared senses and compensatory enhancement of cognitive skills, underscored by substantial plastic reorganization in relevant neural areas. However, little is known about whether people with blindness form top-down models of the world on short timescales more efficiently to guide goal-oriented behavior. This electroencephalography study investigates this hypothesis at the neurophysiological level, focusing on contingent negative variation (CNV) as a marker of anticipatory and preparatory processes prior to expected events. In sum, 20 participants with blindness and 27 sighted participants completed a classic CNV task and a memory CNV task, both containing tactile stimuli to exploit the expertise of the former group. Although the reaction times in the classic CNV task did not differ between groups, participants who are blind reached higher performance rates in the memory task. This superior performance co-occurred with a distinct neurophysiological profile, relative to controls: greater late CNV amplitudes over central areas, suggesting enhanced stimulus expectancy and motor preparation prior to key events. Controls, in contrast, recruited more frontal sites, consistent with inefficient sensory-aligned control. We conclude that in more demanding cognitive contexts exploiting the spared senses, people with blindness efficiently generate task-relevant internal models to facilitate behavior.

Decision heuristics in contexts integrating action selection and execution

Heuristics can inform human decision making in complex environments through a reduction of computational requirements (accuracy-resource trade-off) and a robustness to overparameterisation (less-is-more). However, tasks capturing the efficiency of heuristics typically ignore action proficiency in determining rewards. The requisite movement parameterisation in sensorimotor control questions whether heuristics preserve efficiency when actions are nontrivial. We developed a novel action selection-execution task requiring joint optimisation of action selection and spatio-temporal skillful execution. State-appropriate choices could be determined by a simple spatial heuristic, or by more complex planning. Computational models of action selection parsimoniously distinguished human participants who adopted the heuristic from those using a more complex planning strategy. Broader comparative analyses then revealed that participants using the heuristic showed combined decisional (selection) and skill (execution) advantages, consistent with a less-is-more framework. In addition, the skill advantage of the heuristic group was predominantly in the core spatial features that also shaped their decision policy, evidence that the dimensions of information guiding action selection might be yoked to salient features in skill learning.

Transcranial direct current stimulation enhances motor learning in Parkinson's disease: a randomized controlled trial

Writing training has shown clinical benefits in Parkinson's disease (PD), albeit with limited retention and insufficient transfer effects. It is still unknown whether anodal transcranial direct current stimulation (atDCS) can boost consolidation in PD and how this interacts with medication. To investigate the effects of training + atDCS versus training + sham stimulation on consolidation of writing skills when ON and OFF medication. Second, to examine the intervention effects on cortical excitability. In this randomized sham-controlled double-blind study, patients underwent writing training (one session) with atDCS (N = 20) or sham (N = 19) over the primary motor cortex. Training was aimed at optimizing amplitude and assessed during online practice, pre- and post-training, after 24-h retention and after continued learning (second session) when ON and OFF medication (interspersed by 2 months). The primary outcome was writing amplitude at retention. Cortical excitability and inhibition were assessed pre- and post-training. Training + atDCS but not training + sham improved writing amplitudes at retention in the ON state (p = 0.017, g = 0.75). Transfer to other writing tasks was enhanced by atDCS in both medication states (g between 0.72 and 0.87). Also, training + atDCS improved continued learning. However, no online effects were found during practice and when writing with a dual task. A post-training increase in cortical inhibition was found in the training + atDCS group (p = 0.039) but not in the sham group, irrespective of medication. We showed that applying atDCS during writing training boosted most but not all consolidation outcomes in PD. We speculate that atDCS together with medication modulates motor learning consolidation via inhibitory processes ( https://osf.io/gk5q8/ , 2018-07-17).

Vector-based analysis of cortical activity associated with dumbbell exercise using functional near-infrared spectroscopy

The mechanisms via which the brain and muscles work together remain poorly understood. The use of vector-based fNIRS, to propose a new metric and imaging method to understand neural activation during dumbbell-lifting exercises. This method can simultaneously measure oxyhemoglobin (oxyHb) and deoxyHb levels so that the angle k: Arctan (deoxyHb/oxyHb) represents the degree of oxygen exchange in the brain and can be used to quantify the distribution of oxygen consumption. The amplitude L of the vector reflects the intensity of the response caused by the amount of change in Hb. This study used vector-based fNIRS to simultaneously measure the left primary motor cortex (left M1), multiple peripheral regions, and the right biceps brachii muscle. The subjects were seven healthy adults. The task was a dumbbell-lifting exercise involving flexion and extension of the elbow joints of both arms. Dumbbell weights of 0 (no dumbbell), 4.5, and 9.5 kg were used. During dumbbell exercise, oxygen exchange increased in the left M1, indicating increased local oxygen consumption. Around the left M1, the cerebral oxygen exchange decreased, and oxygen supply increased without cerebral oxygen consumption. The spatial agreement between the maximum value of oxygen exchange k and L during the task was <20%. Therefore, the dumbbell-lifting exercise task study reported here supported the hypothesis that cerebral oxygen consumption associated with neural activation does not coincide with the distribution of cerebral oxygen supply. The relationship between the brain oxygen supply from the site of increased oxygen exchange in the brain and its surrounding areas can be quantified using the vector method fNIRS.

Theta-gamma coupling as a cortical biomarker of brain-computer interface-mediated motor recovery in chronic stroke

Chronic stroke patients with upper-limb motor disabilities are now beginning to see treatment options that were not previously available. To date, the two options recently approved by the United States Food and Drug Administration include vagus nerve stimulation and brain-computer interface therapy. While the mechanisms for vagus nerve stimulation have been well defined, the mechanisms underlying brain-computer interface-driven motor rehabilitation are largely unknown. Given that cross-frequency coupling has been associated with a wide variety of higher-order functions involved in learning and memory, we hypothesized this rhythm-specific mechanism would correlate with the functional improvements effected by a brain-computer interface. This study investigated whether the motor improvements in chronic stroke patients induced with a brain-computer interface therapy are associated with alterations in phase-amplitude coupling, a type of cross-frequency coupling. Seventeen chronic hemiparetic stroke patients used a robotic hand orthosis controlled with contralesional motor cortical signals measured with EEG. Patients regularly performed a therapeutic brain-computer interface task for 12 weeks. Resting-state EEG recordings and motor function data were acquired before initiating brain-computer interface therapy and once every 4 weeks after the therapy. Changes in phase-amplitude coupling values were assessed and correlated with motor function improvements. To establish whether coupling between two different frequency bands was more functionally important than either of those rhythms alone, we calculated power spectra as well. We found that theta-gamma coupling was enhanced bilaterally at the motor areas and showed significant correlations across brain-computer interface therapy sessions. Importantly, an increase in theta-gamma coupling positively correlated with motor recovery over the course of rehabilitation. The sources of theta-gamma coupling increase following brain-computer interface therapy were mostly located in the hand regions of the primary motor cortex on the left and right cerebral hemispheres. Beta-gamma coupling decreased bilaterally at the frontal areas following the therapy, but these effects did not correlate with motor recovery. Alpha-gamma coupling was not altered by brain-computer interface therapy. Power spectra did not change significantly over the course of the brain-computer interface therapy. The significant functional improvement in chronic stroke patients induced by brain-computer interface therapy was strongly correlated with increased theta-gamma coupling in bihemispheric motor regions. These findings support the notion that specific cross-frequency coupling dynamics in the brain likely play a mechanistic role in mediating motor recovery in the chronic phase of stroke recovery.