Aging is commonly associated with a decline in motor control and neural plasticity. Tuning cortico–cortical interactions between premotor and motor areas is essential for controlling fine manual movements. However, whether plasticity in premotor–motor circuits predicts hand motor abilities in young and elderly humans remains unclear. Here, we administered transcranial magnetic stimulation (TMS) over the ventral premotor cortex (PMv) and primary motor cortex (M1) using the cortico–cortical paired-associative stimulation (ccPAS) protocol to manipulate the strength of PMv-to-M1 connectivity in 14 young and 14 elderly healthy adults. We assessed changes in motor-evoked potentials (MEPs) during ccPAS as an index of PMv-M1 network plasticity. We tested whether the magnitude of MEP changes might predict interindividual differences in performance in two motor tasks that rely on premotor-motor circuits, i.e., the nine-hole pegboard test and a choice reaction task. Results show lower motor performance and decreased PMv-M1 network plasticity in elderly adults. Critically, the slope of MEP changes during ccPAS accurately predicted performance at the two tasks across age groups, with larger slopes (i.e., MEP increase) predicting better motor performance at baseline in both young and elderly participants. These findings suggest that physiological indices of PMv-M1 plasticity could provide a neurophysiological marker of fine motor control across age-groups.

Humans tend to spontaneously imitate others’ behavior, even when detrimental to the task at hand. The action observation network (AON) is consistently recruited during imitative tasks. However, whether automatic imitation is mediated by cortico-cortical projections from AON regions to the primary motor cortex (M1) remains speculative. Similarly, the potentially dissociable role of AON-to-M1 pathways involving the ventral premotor cortex (PMv) or supplementary motor area (SMA) in automatic imitation is unclear. Here, we used cortico-cortical paired associative stimulation (ccPAS) to enhance or hinder effective connectivity in PMv-to-M1 and SMA-to-M1 pathways via Hebbian spike-timing-dependent plasticity (STDP) to test their functional relevance to automatic and voluntary motor imitation. ccPAS affected behavior under competition between task rules and prepotent visuomotor associations underpinning automatic imitation. Critically, we found dissociable effects of manipulating the strength of the two pathways. While strengthening PMv-to-M1 projections enhanced automatic imitation, weakening them hindered it. On the other hand, strengthening SMA-to-M1 projections reduced automatic imitation but also reduced interference from task-irrelevant cues during voluntary imitation. Our study demonstrates that driving Hebbian STDP in AON-to-M1 projections induces opposite effects on automatic imitation that depend on the targeted pathway. Our results provide direct causal evidence of the functional role of PMv-to-M1 projections for automatic imitation, seemingly involved in spontaneously mirroring observed actions and facilitating the tendency to imitate them. Moreover, our findings support the notion that SMA exerts an opposite gating function, controlling M1 to prevent overt motor behavior when inadequate to the context.

Motor inhibition is the ability to suppress inappropriate motor responses. Dual-site transcranial magnetic stimulation studies suggest that the right inferior frontal gyrus (rIFG) can inhibit the primary motor cortex (M1), but it remains unclear whether inducing plasticity along this pathway affects motor inhibition. To address this issue, we used cortico-cortical paired associative stimulation (ccPAS) in healthy participants and assessed behavioral and electrophysiological markers of inhibition. In two sessions, we administered a ccPAS protocol to boost rIFG-M1 connections via Hebbian-like plasticity (ccPAS-rIFG-M1) and a reverse-order protocol (ccPAS-M1-rIFG) as a control. Consistent with well-documented inter-individual variability in response to brain stimulation, we divided participants into higher- and lower-sensitivity groups based on the progressive increase in motor excitability during ccPAS-rIFG-M1. Behavioral performance in the Go-NoGo task did not change as a function of ccPAS. However, independently of the ccPAS protocol, higher-sensitivity participants maintained an efficient proactive inhibitory strategy over time, whereas lower-sensitivity participants displayed increased response readiness at the expense of inhibitory control, suggesting stable individual differences rather than direct effects of stimulation. Furthermore, electrophysiological findings revealed that, following ccPAS-rIFG-M1, higher-sensitivity individuals showed a state-dependent increase of rIFG inhibitory influence over M1 only when viewing NoGo-cues. No comparable modulation was found for Go-cues, in lower-sensitivity individuals, or after ccPAS-M1-rIFG. These findings indicate that ccPAS-rIFG-M1 selectively modulates rIFG-M1 inhibitory interactions in a state- and sensitivity-specific manner, supporting a specific involvement of the targeted circuit in motor inhibition and linking individual plasticity capacity to stable differences in proactive inhibitory control.