Motor imagery (MI) has received a lot of interest during the last decades as its chronic or acute use has demonstrated several effects on improving sport performances or skills. The development of neuroimagery techniques also helped further our understanding of the neural correlates underlying MI. While some authors showed that MI, motor execution and action observation activated similar motor cortical regions, transcranial magnetic stimulation (TMS) studies brought great insights on the role of the primary motor cortex and on the activation of the cortico‐spinal pathway during MI. After defining MI and describing the TMS technique, a short report of MI activities only at cortical level is provided. Then, a main focus on the specificities of cortico‐spinal modulations during MI, investigated by TMS, is provided. Finally, a brief overview of sub‐cortical mechanisms gives importance to the activation of peripheral neural structures during MI.

Motor imagery (MI) is the mental simulation of movement, without the corresponding muscle contraction. Whereas the activation of cortical motor areas during MI is established, the involvement of spinal structures is still under debate. We used original and complementary techniques to probe the influence of MI on spinal structures. Amplitude of motor-evoked potentials (MEPs), cervico-medullary-evoked potentials (CMEPs), and Hoffmann (H)-reflexes of the flexor carpi radialis (FCR) muscle and of the triceps surae muscles was measured in young, healthy subjects at rest and during MI. Participants were asked to imagine maximal voluntary contraction of the wrist and ankle, while the targeted limb was fixed (static condition). We confirmed previous studies with an increase of FCR MEPs during MI compared with rest. Interestingly, CMEPs, but not H-reflexes, also increased during MI, revealing a possible activation of subcortical structures. Then, to investigate the effect of MI on the spinal network, we used two techniques: 1) passive lengthening of the targeted muscle via an isokinetic dynamometer and 2) conditioning of H-reflexes with stimulation of the antagonistic nerve. Both techniques activate spinal inhibitory presynaptic circuitry, reducing the H-reflex amplitude at rest. In contrast, no reduction of H-reflex amplitude was observed during MI. These findings suggest that MI has modulatory effects on the spinal neuronal network. Specifically, the activation of low-threshold spinal structures during specific conditions (lengthening and H-reflex conditioning) highlights the possible generation of subliminal cortical output during MI.

The Hoffmann reflex has been the tool most commonly used in exercise studies to investigate modulations in spinal excitability. However, the evolution of electromyographic responses with the increase in stimulation intensity has rarely been assessed when the muscle is active. The purpose of this study was thus to identify that part of the recruitment curve at which the investigation of the Hoffmann reflex is the most reliable in assessing spinal excitability during muscle contraction. Two recruitment curves were determined from the soleus and the medialis gastrocnemius, in passive and active (50% of maximal isometric voluntary contraction) conditions. No differences were found between the H reflexes in the two conditions in the ascending part of the recruitment curves, while the intensity necessary to elicit the same percentage of maximal H wave was different in the descending part of the curve, up to the maximal M wave. We concluded that during motor tasks, changes in spinal excitability should be assessed by recording H responses in the ascending part of the curve, where modulations do not depend either on the background electrical activity of the muscle tested or on methodological considerations.

The neural mechanisms explaining strength increase following mental training by motor imagery (MI) are not clearly understood. While gains are mostly attributed to cortical reorganization, the sub‐cortical adaptations have never been investigated. The present study investigated the effects of MI training on muscle force capacity and the related spinal and supraspinal mechanisms. Eighteen young healthy participants (mean age: 22.5 ± 2.6) took part in the experiment. They were distributed into two groups: a control group (n = 9) and an MI training group (n = 9). The MI group performed seven consecutive sessions (one per day) of imagined maximal isometric plantar flexion (4 blocks of 25 trials per session). The control group did not engage in any physical or mental training. Both groups were tested for the isometric maximal plantar flexion torque (MVC) and the rate of torque development (RTD) before and after the training session. In addition, soleus and medial gastrocnemius spinal and supraspinal adaptations were assessed through the recording of H‐reflexes and V‐waves, with electrical stimulations of the posterior tibial nerve evoked at rest and during MVC, respectively. After one week, only the MI training group increased both plantar flexion MVC and RTD. The enhancement of muscle torque capacity was accompanied by significant increase of electromyographic activity and V‐wave during MVC and of H‐reflex at rest. The increased cortical descending neural drive and the excitability of spinal networks at rest could explain the greater RTD and MVC after one week of MI training.

Muscle patterns reveal how humans and monkeys save muscle effort by taking advantage of gravity effects.