Participants

Fourteen recreational active individuals: 5 males (age 23.6 ± 1.5 years, body mass 75.2 ± 7.3 kg, stature 180 ± 7.9 cm) and 9 females (age 24.3 ± 2.0 years, body mass 58.0 ± 3.0 kg, stature 166 ± 4.8 cm) were recruited for this study. Volunteers abstained from any physical activity within 24 h before the testing. The study was approved by the Ethics Committee at the University of Graz (GZ. 39/77/63 ex 2013/14), all experiments were performed in accordance with relevant guidelines and regulations and written informed consent was obtained from all volunteers before the onset of the experimental procedures.

Study design

Participants attended the laboratory on two separate occasions: the first, lasting about 90 minutes, for familiarising with testing procedures and equipment and the second for the actual testing session. The testing sessions lasted about two and half hours and were run in the morning (13 out of 14 participants started before 1pm and only one started in the afternoon at 4pm).

The experiment consisted in the measurement of MEPs after TMS during: during passive dynamic ankle DF, at different time points during SS, during a passive dynamic ankle PF and following stretching.

Ten TMS stimulations were performed at each of the following time points into stretching: 3, 6, 9, 18, 21 and 25 seconds in random order. Moreover 16 MEPs were elicited during dynamic DF and 14 during dynamic PF. Further stimulations were performed between 10 and 20 seconds at the end of the stretching procedure.

Experimental procedures

The experimental procedures have already been described in our previous works [28,38]. Subjects were sitting on an isokinetic dynamometer (CON-TREX MJ, CMV AG, Duebendorf, Switzerland) with the standard setup for ankle joint rotation individually adjusted. Participants had their right knee fully extended and the foot resting on the dynamometer footplate, the ankle joint aligned with the dynamometer rotation shaft and the ankle angle set at 100° (10° plantar flexion deviating from a neutral position at 90°). Volunteers sat with the trunk at 110° and the head supported by a cushion (dentafix^®, pro medico HandelsGmbH, Graz, Austria) that once positioned could be deflated allowing the formation of a stable form molded on the volunteers´ head and neck shapes. By using a remote control, the volunteers were instructed to adjust the dorsiflexion isokinetic rotation operated by the dynamometer around the foot plate until the point of perceived maximal DF (therefore the maximal dorsiflexion was not fixed at a specific value, but was different for each volunteer). Participants were asked to keep their knees extended and to relax during the procedures.

Once the maximal individual dorsiflexion was defined, participants were prepared for surface electromyographic recording (EMG) from tibialis anterior muscle (TA) of both right and left leg as well as from the right soleus muscles (SOL). The volunteers then returned to sit on the dynamometer chair in the position described above and were instructed to relax meanwhile coil position and stimulation intensity for TMS was determined. Subsequently a trigger-driven sequence of stimulations started and continued for 150 seconds until 30 MEPs were collected for baseline reference values.

In order to perform all measurements at 100° during dynamic PF, following baseline recordings, the foot was passively rotated to a starting position of 110° plantar flexion. In this way the stimulations during dorsiflexion movements (20°/s) could be delivered when the ankle reached 100° PF (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Stretching protocol and stimulation points.

https://doi.org/10.1371/journal.pone.0284289.g001

The sequence of stimulations during the intervention is represented in Fig 1. A first stimulation was delivered during dynamic DF, then, once individual maximal dorsiflexion was reached, the position was kept for 30 s and during this period a second stimulation was delivered at one of the investigated time points (3, 6, 9, 18, 21 and 25 s into stretching). After this first 30 s SS bout, the foot was passively rotated (20°/s) back to 110° PF and a third stimulation was delivered during the PF movement as the ankle was at 100° PF. This procedure was repeated without pauses a second time with only one more stimulation delivered at maximal dorsiflexion and, after the second 30 s stretch, further stimulations were performed with the ankle joint back at 100°.

To collect a sufficient number of stimulations at each given time point, this 2 x 30 s stretching block was repeated 30 times; 50 second rest with the volunteer sitting were allowed between every block and two minutes rest with the volunteer standing up every 12 stretching applications.

Surface electromyography

For preparation of EMG recording, skin was shaved, abraded, cleaned with alcohol, and left to dry. Surface EMG was collected from the SOL and TA of the right leg. Additionally, the EMG signal was recorded from the tibialis anterior of the left leg to ensure the correct directional positioning of the coil. Electrodes (Blue Sensor N, Ambu A/S, Ballerup, Denmark) were placed in monopolar configuration for recording from the SOL muscle (as suggested by Hadoush et al. [39]); the electrodes for TA were placed in bipolar configuration with an interelectrode distance of 20 mm. One ground electrode per leg was placed over the tibial bone medial surface.

Transcranial magnetic stimulation

All stimulations at baseline, during dynamic DF and PF and after stretching were performed with the ankle joint at 100° PF, stimulations during stretching were performed at individual maximal DF position (Fig 1).

Motor evoked potentials in response to single pulse TMS were induced by a Magstim 200, (Magstim Company Ltd., UK) using a double cone coil (110mm coil diameter). The coil was placed over the M1 of the leg area, 1–2 cm posterior from the vertex and slightly shifted to the left side in order to obtain the largest response from the contralateral right SOL. Resting motor threshold was determined as the minimum stimulator intensity able to evoke MEPs of at least 50μV amplitude in more than five out of ten consecutive trials [13]. To ensure a constant coil positioning throughout the experiments, subjects were wearing an electroencephalography cap on which the optimal coil position was marked with a soft pen. MEPs were elicited at 5-s intervals with stimulation intensity equal to 120% of the resting motor threshold.

Data monitoring

EMG was monitored online to control that the subjects were not actively contracting their plantar flexors against the stretch. However, a more careful control of this parameter was then repeated off line during data analysis (see Data analysis).

During the recordings, the trigger signal responsible for the activation of the TMS was simultaneously sent to the A/D converter. This served off line for data analysis and online for signal monitoring, as the A/D software (DeweSoft) was programmed similarly to an oscilloscope to freeze the screen within a time window of 80ms before and 80ms after the trigger event to give the experimenter the possibility to control every single MEP. Additionally, 2 cursors were placed on the screen approximately in correspondence with the MEPs average value; this allowed to detect any extra-ordinary MEP, and, in case, lead to the decision of collecting extra data for that specific series. The variability of the measures was additionally checked offline (see Data analysis), therefore, the extra recording, when done, only increased the chance to retain more MEPs within the acceptable variability range for the subsequent analysis.

Data analysis

Electromyography, foot displacement and trigger signals were synchronized (Dewetron 7.0 recording system), digitized with a sampling frequency of 10 KHz, stored on a PC and analysed using custom algorithms developed in Matlab (R2014b).

Baseline MEPs with peak to peak amplitude exceeding by ±2 standard deviations the average baseline value were discarded [40]. Similarly, consistency was checked within the 60 MEPs (10 stimulations at each of the 6 time points) collected during stretching and the MEPs collected during DF and PF (16 and 14 MEPs respectively). Because voluntary activation increases MEP amplitude [41], additional discards were applied in all those cases where MEPs were elicited in a pre-activation state of the target muscle (defined as EMG activity larger than 50 μV in the 50 ms preceding the TMS pulse) [42]. All the remaining waves were retained and when a minimum number of 5 MEPs per data set was still available, the average value was used for statistical analysis.

Statistical analysis

Measurements were checked for normal distribution by the Shapiro-Wilk test. Since data were not normally distributed, Friedman test followed by Wilcoxon signed-rank tests with Bonferroni Holm’s adjustment was used to compare values at baseline, during dynamic DF, during SS (all values merged), during dynamic PF and post stretching. The effect sizes (ES) were calculated for every comparison done with the Wilcoxon signed-rank tests that was still significant following Bonferroni Holm’s adjustment. These were calculated dividing the z value by the root square of the product: conditions x observations [43].

The time course of MEPs amplitude throughout the stretching was tested by a Friedman test to compare values at different time points into stretching and further assessed with Pearson ’s correlation for linear regression analysis.

All statistical analysis was completed using PASW Statistic 18.0.0.

Response modulation during SS

Cortico-spinal excitability is affected by muscle length and limb posture [44–49], and such alterations are commonly attributed to afferent projections to cortical and/or spinal paths [17,50]. Potential sites of modulation of the MEP in response to TMS include intrinsic cortical inputs to the pyramidal tract neurons in M1, the activity of the α-motoneurons and interneuron network at spinal level and the afferent signals arising from activated sensory receptors. At least in the upper limb, MEPs amplitude tends to increase when the muscle is at a shorter length. For example, greater MEPs in the biceps brachii are observed when this is in a shorter position with the forearm supinated, compared to when the biceps brachii is in a lengthened position, with the forearm pronated [45–47], and this effect seems related to the amount of stretch [49]. Although results on the lower limb don´t always corroborate this observation [44], our data confirm that TA cortico-spinal excitability increases when the muscle is shorter. Additionally, the present study tried to extend the current knowledge by investigating whether this expected facilitation in the shortened muscle would have remained constant throughout the entire stretching period or not. MEPs were therefore collected at six time points: 3, 6, 9, 18, 21 and 25 seconds into stretching. Since we observed a clear facilitation during the DF movement required to bring the foot to maximal individual static stretched position, we could have expected to capture a transient effect within the first measurement at 3 seconds. Looking at Fig 2 this seems to be the case only in 6 of the 13 subjects. Although, the group average seems to show a pattern consisting in an initial sharp facilitation at 3 seconds into stretching, followed by a reduction of this from 3 to 6 seconds, and finishing with a more or less linear increase, this pattern is mostly affected by 3 subjects who had remarkable changes from baseline to 3 seconds into stretching. The absence of a time-dependent trend in the increased excitability is confirmed by the absence of statistical differences between time points into stretching.

Response modulation during dynamic DF

The observed corticomotor facilitation during DF was registered for both the elongated (SOL) and the shortened (TA) muscles. While the increased corticomotor excitability of TA during DF is in line with studies showing an increased MEP amplitude during passive muscle shortening [15–19,23], the simultaneous facilitation in the elongated muscle SOL is not supported by previous findings, which commonly report reduced MEP amplitude during lengthening of the wrist and elbow flexors and extensors [15–19,51]. However, differently to the upper limb, responses in the lower limb are sometimes contradictory being susceptible to more variables [20–24]. For example, Škarabot and colleagues reported a facilitation in the shortened TA in young [23] but not in old participants [24] and no effect on the lengthened SOL [23,24], whilst Hultborn and colleagues [8] did not report any facilitation in neither TA nor SOL, and in the present study we witnessed a facilitation in both TA and SOL.

Considering that TMS activates predominantly monosynaptic pathways to the α-motoneurons and these connections are not exposed to presynaptic inhibition [17], it seems that the observed excitability alterations during DF can be attributed either to cortical inputs to the corticospinal pathways or to postsynaptic inputs directly modulating the α-motoneurons. However, a recent study showed that during passive ankle movements, the excitability at the lumbar spinal segmental level was not modulated in neither TA nor SOL [23], suggesting supraspinal rather than postsynaptic contributions to the observed MEP facilitation.

Supraspinal structures can indeed be affected by proprioceptive afferents [17,50]. During passive lengthening movements, Ia fibres from muscle spindles respond to changes in muscle length by increasing their firing rate [52,53]. The respective inflows projecting primarily to the area 3a in the primary somatosensory cortex (S1) [54], can activate indirectly the cortical inputs to the pyramidal tract neurons in M1. Such co-activation within the S1 area and the motor areas (M1 and SMA) in response to passive proprioceptive stimulation (passive fingers’ flexion and extension) was demonstrated in a fMRI study [55]. It is known that the finely scaled topographic maps of S1 and M1 enable both areas to have highly specialized responses to changes in the periphery [56,57]. As shown in a sensorimotor slice [58], the anatomical and functional sensorimotor connections are reciprocal and their localization in cortical layers V and VI allow descending outputs also to the spinal cord [59]. In our study we can expect that the increase in Ia afferents activity from SOL during DF reached the somatosensory area (3a), which transmits these inputs further to the M1 output neurons, increasing the excitability of the descending corticomotor pathway.

Such increased excitability originating from the cortex with related facilitation on the descending drive, could be a compensation mechanism for the spinal inhibition (typically observed as decrease in H-reflex during lengthening movements for review see [34]), as shown in a study with active muscle lengthening [19]. However, if an increased Ia activity provoked an increase in corticospinal excitability or a reduction of cerebello-cortical inhibition on projections on the homonymous muscle, we should have observed not only greater MEPs in the SOL during DF, but also in the TA during PF movement. Alternatively, in case of projections on the antagonist muscles, we should have observed greater MEPs in the TA during DF together with greater MEPs in the SOL during PF movement, but neither of these two scenarios was the case (Fig 4).

A possible explanation of this not muscle-specific response could be attributed to the awareness of the subjects about the passive displacements of the foot, that, for methodological requirements, was repeated 60 times. Indeed, in healthy individuals, passive movements are shown to activate not only the primary somatosensory cortex but also the primary motor cortex, supplementary motor area, and posterior parietal cortex as well as the secondary somatosensory cortex (S2) (for review see [60]). Moreover, passive movements can selectively increase cortical excitability in relation to the duration and velocity of the movements, the presence of rest, and whether attention is directed to the movement. Attention to the stimulated side during an intervention (peripheral afferent stimulation, passive movements) decreases the activity of the inhibitory cortical circuits, and thus increases corticospinal excitability. In our study the subjects were instructed not to observe the stretching. Nevertheless, the effect of attention to the movement throughout the experimental procedure cannot be completely excluded. In addition, passive movements repeated with the same amplitude and velocity for a certain time, might induce attempted movement, thereby activating neurons in M1 [61] and resulting in increased MEPs during DF for both TA and SOL muscles.

Despite the precise mechanisms responsible for the increased excitability cannot be clarified with this study, the findings may have clinical potential in neurorehabilitation, where passive movements are commonly applied in physiotherapy. For example in patients with paresis of the lower extremity passive dorsi- and plantar flexions can be repeatedly performed using robotic devices. It is expected that the augmented proprioceptive input facilitating motor cortical excitability, may promote motor activation and thus improve motor recovery.