Associations between structural injury and task-based corticomuscular connectivity after stroke

Stroke-related damage to structural pathways and functional connections disrupts neural network communication, contributing to behavioral deficits. A critical next step is to determine whether observed relationships between connectivity and behavior align with established neurobiological frameworks. This involves investigating structural-functional relationships as structural connectivity provides the scaffold for functional communication. Prior work primarily explored structural-functional relationships at rest, particularly between structural measures and cortico-cortical functional connectivity. However, because stroke impacts both cortical and muscular systems, incorporating task-based functional connectivity measurements that reflect synchronous activity between cortex and muscle may offer additional insight. Therefore, in this study, we examined relationships between structural injury and integrity measures with task-based functional connectivity between electrodes overlying sensorimotor cortical regions and affected upper-extremity musculature (referred to as corticomuscular coherence; CMC).

Individuals with early subacute stroke admitted to an inpatient rehabilitation facility completed simultaneous electroencephalography (EEG) and electromyography (EMG) recordings during a grip task. Corticospinal tract (CST) injury and integrity were computed from structural and diffusion-weighted imaging. CMC measurements involving electrodes overlying ipsilesional motor areas and affected upper-extremity musculature were computed in frequency bands relevant to neural injury (delta, 1-3 Hz) and motor function (low beta, 13-19 Hz; high beta, 20-30 Hz). Correlational analyses were performed to ascertain relationships between structural and coherence measurements. To account for inter-individual heterogeneity, analyses were repeated for CST injury and integrity subgroups.

Of the 30 individuals enrolled, EEG data from 21 individuals who were able to complete the grip task were analyzed (10 females; 67.9 ± 9.8 years; 11.3 ± 4.1 days post-stroke). No significant structure-function associations were observed across the group. However, in the mild-moderate CST injury subgroup (n = 11), greater injury correlated with higher coherence between electrodes overlying the supplementary motor area and affected extensor digitorum (high beta: ρ = 0.83, p = 0.001). Similarly, in the subgroup depicting higher CST integrity (n = 9), CST integrity positively related to coherence between electrodes overlying the ipsilesional primary motor cortex and affected biceps (low beta: r = 0.94, p = 0.0001).

Findings exclusive to CST injury/integrity subgroups underscore the complexity of structure-function relationships in stroke. Associations between CMC measures in motor-relevant frequency bands with measures reflecting CST microstructure suggest that post-stroke structural injury modulates task-based corticomuscular connectivity. The identification of specific cortical regions and muscles depicts varying adaptive and/or compensatory neuroplastic-like mechanisms, providing mechanistic insights that could inform targeted rehabilitation strategies to optimize post-stroke recovery.