The peripheral origin of tap-induced muscle contraction revealed by multi-electrode surface electromyography in human vastus medialis

Abstract

It is well established that muscle percussion may lead to the excitation of muscle fibres. It is still debated, however, whether the excitation arises directly at the percussion site or reflexively, at the end plates. Here we sampled surface electromyograms (EMGs) from multiple locations along human vastus medialis fibres to address this issue. In five healthy subjects, contractions were elicited by percussing the distal fibre endings at different intensities (5-50 N), and the patellar tendon. EMGs were detected with two 32-electrode arrays, positioned longitudinally and transversally to the percussed fibres, to detect the origin and the propagation of action potentials and their spatial distribution across vastus medialis. During muscle percussion, compound action potentials were first observed at the electrode closest to the tapping site with latency smaller than 5 ms, and spatial extension confined to the percussed strip. Conversely, during tendon tap (and voluntary contractions), action potentials were first detected by electrodes closest to end plates and at a greater latency (mean ± s.d., 28.2 ± 1.7 ms, p < 0.001). No evidence of reflex responses to muscle tap was observed. Multi-electrode surface EMGs allowed for the first time to unequivocally and quantitatively describe the non-reflex nature of the response evoked by a muscle tap.

Figure 1

Experimental setup and possible EMG propagation patterns. (a) Schematic representation of the experimental setup. Medial view of the human left thigh. Two linear arrays of 32 electrodes (5 mm inter-electrode distance) were positioned longitudinally and transversally with respect to the distal fibres of the vastus medialis (VM) muscle. Muscle taps were delivered at the distal end of the longitudinal electrode array. (b) Generation and propagation patterns of EMG potentials associated with the direct idiomuscular (dashed arrow) and reflex (solid arrows) activation of VM fibres, as detected by the longitudinal array of electrodes. The latencies L (L_idio and L_reflex) between the tap onset and the first detected EMG potentials are shown on the time axis. W1 and W2 are the time windows, within which the amplitudes of the idiomuscular and reflex responses were evaluated, respectively (see methods).

Figure 2

Patterns of propagation of the EMG signal. (a) Propagation patterns of experimental, single-differential (SD) EMG responses detected with the longitudinal electrode array in the three experimental conditions: muscle tap (MT), tendon tap (TT), and voluntary contraction. The tap onset is indicated by a red arrow in the two panels showing the responses to percussions (MT and TT). IZ indicates the innervation zone position as identified in voluntary and tendon tap contractions. (b) EMG potentials induced by muscle tap (solid line) and by tendon tap (dashed line). The two grey-shaded boxes around EMG responses picked up by electrode 5 indicate the two time windows (W1 and W2), in which the amplitudes of the idiomuscular (W1) and reflex (W2) responses were evaluated.

Figure 3

Delays between consecutive single differential EMG signals computed along the fibre direction. Boxplots represent the delay distribution obtained from voluntary contractions (grey boxes) and from muscle taps (MT, white boxes) for all subjects. The sign of the delay depends on the propagation direction. Delays with constant sign along the entire array indicate unidirectional propagation, while a sign change denotes bi-directional propagation. Dashed lines indicate the range of physiological delays associated with a conduction velocity between 3 and 6 m/s. An attenuation of action potentials along the fibres could be observed for all subjects (e.g. Fig. 2a) possibly because of misalignment between fibres and electrodes and difference in propagation velocity between excited fibres. Therefore the delay values are represented for the maximal number of consecutive channels (n = 16) for which we were able to successfully track the propagation of action potentials.

Figure 4

Latencies between the tap and EMG onsets for muscle tap (MT) and tendon tap (TT). Data is shown for each individual subject. Note the broken ordinate. (**p < 0.001, Mann-Withney test).

Figure 5

Effect of the muscle-tap force on the induced EMG amplitude for each participant. RMS values were computed for two windows (W1 and W2, see Fig. 2) defined as the time intervals where the induced potential was expected to appear if induced at the muscle tap location (W1) or through a reflex (at the IZ channel – W2). For each graph, the dashed horizontal line indicates the background noise level, while the solid line is the regression line fitting RMS estimates in W1. Pearson correlation coefficients (r) are reported within each scatter plot.

Figure 6

Transversal distribution of the EMG signal in response to muscle tap. (a) Amplitude distributions of the EMG responses induced by muscle tap and detected by the transversal electrode array for one representative subject. Different traces represent the amplitude distributions for different muscle tap forces. The abscissa refers to the numbers of electrodes and therefore spans 16 cm. (b) Average values of means and standard deviations of the Gaussian distributions fitting the transversal amplitude profiles for all subjects and tap forces. The ordinate refers to the medio-lateral position along the transversal electrode array (med = medial; lat = lateral).