Evaluation of motor cortical excitability using evoked torque responses: A new tool with high reliability

Motor evoked potentials (MEPs) elicited by transcranial magnetic stimulation (TMS) are typically recorded via surface electromyography (EMG). However, another suitable alternative may be recording torque output associated with MEPs, especially when studying multiheaded muscles (e.g. quadriceps) for which EMG may not be ideal.

We recorded the motor evoked torque elicited by TMS along with conventional EMG-based MEPs (MEP_EMG) over a range of TMS intensities (100-140 % of active motor threshold [AMT]) from twenty healthy young adults on two different days. MEPs were normalized using different normalization procedures (raw, normalized to maximum voluntary isometric contraction [MVIC], and peak MEP). Additionally, motor evoked torque was normalized to TMS-evoked peripheral resting twitch torque. Intraclass correlation coefficients (ICCs) were determined for each of these variables to compute reliability.

Motor evoked torque showed good to excellent reliability (ICC: 0.65-0.90) at TMS intensities ≥ 110 % AMT, except when normalized by peak MEP. The reliability of raw MEP_EMG and MVIC normalized MEP_EMG was fair to excellent only at ≥ 130 % AMT (ICC: 0.42-0.82) and at ≥ 120 % AMT (ICC: 0.41-0.83), respectively. The reliability of both MEP_EMG and motor evoked torque generally increased with increasing TMS intensities, with motor evoked torque normalized to the resting twitch torque yielding the best ICC scores.

When compared with conventional MEP_EMG, motor evoked torque offers superior and reliable estimates of corticospinal excitability, particularly when normalized to resting twitch torque.

TMS-induced motor evoked torque can reliably be used to measure corticospinal excitability in the quadriceps muscles.