Rhythmic arm movement reduces Hoffmann (H)-reflex amplitudes in leg muscles by modulation of presynaptic inhibition in group Ia transmission. To date only the acute effect occurring during arm movement has been studied. We hypothesized that the excitability of soleus H-reflexes would remain suppressed beyond a period of arm cycling conditioning. Subjects used a customized arm ergometer to perform rhythmic 1-Hz arm cycling for 30 min. H-reflexes were evoked before, during, and after arm cycling via stimulation of the tibial nerve in the popliteal fossa. The most important finding was that the H-reflex amplitudes were significantly suppressed during and ≤20 min after arm cycling had been terminated. Thus remote arm cycling can induce adaptive plasticity in the soleus H-reflex pathway that persists beyond the period of conditioning. In an additional experiment, the prolonged effect of arm cycling combined with cutaneous superficial radial (SR) nerve stimulation was investigated. Cutaneous stimulation cancelled the prolonged suppression of H-reflex amplitude induced by arm cycling. Because SR nerve stimulation facilitates soleus H-reflex via reducing the level of Ia presynaptic inhibition, persistence in presynaptic inhibitory pathways is suggested as the underlying neural mechanism. The simplest explanation of this observation is plateau potential-like behavior of interneurons mediating presynaptic inhibition of Ia afferent transmission.
During movement Hoffmann (H)-reflexes are dynamically and phasically modulated by central rhythmic patterning and by afferent feedback arising from the movement itself (Brooke et al. 1997; Zehr and Duysens 2004; Zehr and Stein 1999). Leg movements occurring during rhythmic locomotor activities of stepping, cycling, or walking can induce strong suppression of H-reflex amplitude (Brooke et al. 1995; Capaday and Stein 1987). Also movement that is remote from the site where the reflex is evoked (such as arm cycling when reflexes in the legs are examined) can affect the excitability of spinal circuitry. Interestingly, as with leg movement, arm cycling suppresses soleus H-reflex amplitude in stationary legs (Frigon et al. 2004; Loadman and Zehr 2007). Remote leg movement can also acutely suppress H-reflex amplitude in forearm muscles (Zehr et al. 2007). The main mechanism leading to modulation of soleus H-reflex amplitude during both leg or arm movement is presynaptic inhibition (PSI) of Ia afferent transmission (Brooke et al. 1997; Frigon et al. 2004; Zehr et al. 2004).
Some limited evidence suggests that the acute and immediate effect of rhythmic movement on reflex amplitude can persist beyond the period of movement conditioning itself. After passive leg cycling, H-reflex amplitude remained significantly suppressed for ≤4 s after cycling terminated (Misiaszek et al. 1995). Another study found that 20 min of active and passive leg cycling could lead to suppression of soleus H-reflex amplitude lasting for 30 min after cycling stopped (Motl et al. 2003). Additionally, FES stimulation to the common peroneal nerve applied at the swing phase during 30 min of walking led to significant suppression of H-reflex amplitude for ≤30 min after the conditioning period (Thompson et al. 2006). These findings suggest that leg movement can induce short-term plasticity of H-reflexes in the leg muscles that persists beyond the period of movement conditioning. The primary objective of this study was to test the hypothesis that arm cycling, a remote conditioning input that acutely suppresses H-reflex amplitude during movement (Frigon et al. 2004), could produce prolonged suppression of the soleus H-reflex.
Stimulation applied at the wrist to the cutaneous superficial radial (SR) nerve innervating the dorsum of the hand facilitated soleus H-reflex amplitudes in static positions and could partially counteract the suppressive effect of arm cycling (Zehr et al. 2004). This effect was ascribed to reduced presynaptic inhibition in the Ia afferent-alpha motoneuronal pathway (Zehr et al. 2004). Thus the second objective of the present study was to test the hypothesis that remote cutaneous input evoked by SR nerve stimulation would interfere with any prolonged suppressive effect of arm cycling on soleus H-reflex amplitude.
Thirteen subjects (aged 22–43 yr; 8 females and 5 males) with no history of neurological disorders were recruited in this study. Participants provided informed written consent in a protocol approved by the Human Research Ethics Committee at the University of Victoria and in accordance with the declaration of Helsinki.
The experimental methodology and protocol are similar to that described in previous experiments involving arm cycling (Frigon et al. 2004; Loadman and Zehr 2007; Zehr et al. 2004). Unwanted movements in the trunk and lower limb were minimized by having participants sit in a custom-adapted chair with support provided for their back, trunk, and legs and with the feet fixed to foot plates positioned at a neutral ankle angle (∼90°). A custom-made hydraulic arm ergometer (described in Zehr et al. 2003) was positioned directly in front of the subjects. Subjects were asked to hold the handgrips firmly but comfortably with the forearms pronated and performed 30 min of rhythmic arm cycling while maintaining a relaxed posture in the rest of the body. The handles of the ergometer were constrained to move together but were 180° out of phase. Arm cycling was performed in a clockwise direction (viewed from the right side of the body) in which the 3 o'clock position corresponded to the largest extended elbow angle and maximal shoulder flexion (∼70° in front of the mid-axillary line). Last, participants cycled at a comfortable pace (∼60 rpm) and were provided with visual feedback on an oscilloscope (Hameg 20 MHz, HM205-3, Frankfurt/Main, Germany) displaying cycling frequency.
During each experimental session, two sets of experiments were conducted both involving arm cycling and static (pre- and postcycling) trials. In experiment 1 (subject: n = 12), the prolonged effect of arm cycling on soleus H-reflex amplitude was investigated. In experiment 2 (n = 8), cutaneous (SR nerve) conditioned H-reflexes were evoked after each unconditioned H-reflex test in static trials to evaluate the interaction of arm cycling with cutaneous input after termination of movement. Experiment 2 was conducted subsequent to experiment 1 after H-reflex amplitudes returned to the control values. For three subjects, experiment 2 was performed on a different day from experiment 1. An additional control experiment was conducted with identical protocol to the first experiment except that subjects (n = 4) did no arm cycling.
To evoke H-reflexes in the soleus muscle, the tibial nerve was stimulated with single 1-ms square wave pulses at the left popliteal fossa using bipolar surface electrodes (Thought Technologies) and a Grass S88 stimulator (Grass Instruments, AstroMed) connected in series with a Grass SIU5 isolation unit and a CCU1 constant current unit. Nerve stimulation was delivered pseudo-randomly between 3 and 5 s apart when the hand ipsilateral to the stimulation was at 3 o'clock (as described in the preceding text). Thus H-reflex tests consisted of 10 stimulations applied to the tibial nerve approximately every three or four cycles of arm movement. In both experiments, H-reflexes were evoked at 5-min intervals, starting prior to cycling, continuing throughout the 30-min cycling period and ≤30 min after cycling termination. In experiment 2, each H-reflex test was followed by an SR conditioned H-reflex for all the static trials. Participants were asked to keep the soleus muscle at rest during all stimulations.
At the beginning of both experiments, H-reflex and M-wave recruitment curves of 40 stimulation sweeps were constructed to determine the maximum M-wave (Mmax; mean of 3 largest M-wave values) and maximum H-reflex sizes (Hmax) under control, nonmoving conditions. Subsequently, stimulation intensity was set to evoke an H-reflex size of 70–80% of Hmax on the ascending limb of the recruitment curve. This position evoked a corresponding M-wave of ∼3% of Mmax. This M-wave amplitude was monitored and adjusted on-line to assure consistent stimulation intensity. Current was measured by a mA-2000 Noncontact Milliammeter (Bell Technologies, Orlando, FL).
In addition to two recruitment curves at the beginning and end of both experiments, two more (n = 30 stimulations) were recorded at 1 min after start and end of cycling in experiment 1. These curves were used to determine Mmax amplitudes to normalize data (see data analysis) and control and allowed us to check for variation in Mmax over time to properly assess stimulus and normalized reflex amplitudes over the course of the experiment.
SR nerve stimulation
In experiment 2, a cutaneous conditioning of H-reflexes was added to the protocol. As such, static trials were collected as in experiment 1 but were alternated with an H-reflex conditioned with SR nerve stimulation. A condition-test (C-T) interval of 100 ms was used for the SR conditioned reflexes (Zehr et al. 2004). Cutaneous nerve stimulation made use of the same stimulator instrumentation as for the H-reflexes except that trains of 5 × 1.0-ms pulses at 300 Hz were delivered to the SR nerve. Flexible electrodes (Thought Technologies) for SR nerve stimulation were placed on the dorsal surface of the forearm just proximal to the radial head and the crease of the wrist joint. Stimulus intensity was set at twice the threshold at which a clear radiating paresthesia (radiating threshold, RT) into the innervation area of the nerve (dorsal surface of the hand toward the index finger and thumb) was reported (Zehr et al. 2001).
Electromyographic (EMG) signals were recorded with surface electrodes (Thought Technologies) placed in bipolar configuration with ground electrodes placed on the patella. The activity of soleus, tibialis anterior (TA), vastus lateralis (VL), biceps femoris (BF), and anterior deltoid (AD) muscles were recorded ipsilateral to H-reflex stimulation. The prestimulus (20 ms) EMG from each muscle was used to determine the level of background muscle activation at the time of reflex sampling. For soleus, EMG signals were preamplified at 500 times and band-pass filtered 100-1,000 Hz (P511 Grass Instruments, AstroMed). For the other muscles, EMG was amplified by 5,000 times, band-pass filtered 100–300 Hz, and full-wave rectified.
Data acquisition and analysis
Data were sampled at 5,000 Hz with a 12- bit A/D converter controlled by a custom-written Labview (National Instruments, Austin, TX) computer program. For all the trials, 10 sweeps (70-ms duration) were collected. Peak-to-peak amplitudes of M-waves and H-reflexes were determined off-line (custom-written software, Matlab) from the single sweeps of soleus EMG and were averaged and normalized to the Mmax values (see following text) to reduce intersubject variability (Frigon et al. 2007). In experiment 1, data were normalized to the corresponding Mmax obtained from the recruitment curve which was within 20 min of sampling. In experiment 2, data were typically normalized to the average Mmax from the beginning and end of the experiment.
In both experiments, STATISTICA (StatSoft, Tulsa, OK) was used to perform repeated-measures ANOVA to identify significant main effects for time intervals on the amplitudes of the soleus H-reflexes, M waves, and prestimulus EMG levels. The mean amplitudes of the H-reflexes prior to cycling were used for the control values (H-control). In experiment 1, a two-tailed paired t-test was used to compare the H-reflex amplitude averaged across cycling trials with the average from the postcycling trials. In both experiments 1 and 2, Dunnett's post hoc test was used to compare mean values of H-reflex amplitudes from each cycling and postcycling trial with H-control. For experiment 2, control values of the H-reflex and conditioning test were compared using two-tailed paired t-test. To confirm that the conditioning of reflex amplitude induced by arm cycling was similar in the two experiments, two-tailed paired t-test were used to compare H-reflex amplitudes in experiment 1 at different time intervals (precycling, cycling, and the first postcycling trials) with those in experiment 2 across subjects who participated in both experiments (n = 7). Finally, for the background EMG, M-wave, and Mmax values, Tukey's HSD post hoc tests were applied. Statistical significance was set at P < 0.05.
Experiment 1: prolonged effect of arm cycling on soleus H-reflex amplitude
The effect of prolonged arm cycling on H-reflex amplitude is shown in Fig. 1 A for a single subject at precycling, 5 min of cycling, and 10 min postcycling. H-reflex amplitudes were significantly decreased during cycling (P < 0.001) and remained suppressed ≤20 min after cycling termination (P < 0.04). In Fig. 1B, group data are shown for H-reflex amplitudes evoked during (solid bars) and after (dashed bars) cycling. The H-control amplitude is shown as the solid horizontal line. Despite similar M-waves and background EMG levels (not shown), H-reflex amplitudes were significantly reduced during cycling and remained suppressed for ≤20 min after cycling ended. Arm cycling caused a sudden attenuation of H-reflex amplitudes by ∼18% (P < 0.001; not plotted). The average suppression of H-reflex amplitudes during cycling and 20 min post cycling were not significantly different.
In experiment 1, there were no significant differences in background EMG levels for soleus, TA, VL, and BF muscles across the different time intervals. AD background EMG was not significantly different among various time intervals in static trials (except at time 0 after cycling, P < 0.001).
Experiment 2: effect of SR stimulation on prolonged suppression of soleus H-reflex amplitude
This second experiment was intended as a test to determine whether any long-lasting suppression of the H-reflex was mediated by presynaptic inhibition. Accordingly, in this part of the experiment, SR conditioning of H-reflexes was applied immediately after each H-reflex test during the postcycling period. Thus the first H-reflex trial immediately after cycling (i.e., 0 at postcycling) was not preceded by any SR conditioning. As is shown in Fig. 2 A, H-reflex amplitudes remained suppressed only for the first H-reflex trial after cycling termination (see Fig. 2A, *). H-reflex amplitudes subsequent to the initial application of SR stimulation had returned to control amplitude.
Comparison of experiment 1 to 2 at different time intervals showed that averaged H-reflex amplitudes were not significantly different at control, cycling, and the first postcycling trials. Therefore the earlier recovery of the H-reflexes to control amplitude seen in experiment 2 is attributed to the cutaneous input as applied with SR stimulation.
The size of the SR nerve conditioned control H-reflexes (i.e., amplitude of conditioned H-reflex by SR before cycling) was significantly larger than H-reflex control (P < 0.001), indicating the presence of anticipated significant facilitation of the H-reflex pathway due to the cutaneous conditioning (Zehr et al. 2004). SR conditioning tests did not show any significant change from pre- to postcycling period (see Fig. 2B).
Control study: effect of number of stimuli and time course on the H-reflex size
Across the full time course of the control experiment, H-reflex amplitudes, levels of background EMG, M-wave, and Mmax amplitudes did not change significantly (P = 0.18 to P > 0.99) when participants did no cycling.
The main finding of this study is that the suppressive effect of arm cycling on soleus H-reflex amplitude can persist beyond the period of cycling. These findings support the working hypothesis that interlimb coupling between arms and legs affects neuronal circuitry resulting in short term changes in the excitability of the H-reflex pathway. Furthermore, it was found that the short term adaptive plasticity of the H-reflex pathway was strongly influenced by cutaneous volleys from remote sources.
The amplitude of the direct motor response (M-wave) was kept constant to gauge similar stimulus input to the tibial nerve and therefore consistent activation of the Ia afferents (Brooke et al. 1997; Zehr 2002). H-reflex and M-wave amplitudes were normalized to the Mmax to reduce intersubject variability (Zehr 2002). However, the amplitude of Mmax may vary in the relaxed muscles during the course of an experiment (Crone et al. 1999). To minimize the effect of minor Mmax variations, H-reflex and M-wave values were normalized to data from recruitment curves recorded within 20 min of H-reflex sampling. However, small changes in Mmax did not significantly affect the size of M-waves or stimulus input over the course of the experiment. Therefore it is improbable that prolonged H-reflex suppression was due to a change in the afferent volley used to evoke the reflex.
To control the possible effects of heteronymous muscle activity on the soleus H-reflex pathway (Iles and Roberts 1987; Meunier et al. 1993) participants were seated in a chair equipped so that the trunk and lower limb joints were supported to prevent any perceptible movements. There were no differences in the background EMG of the leg muscles across experimental trials. Previous activation of Ia afferents may cause a depression in the H-reflex amplitude which may last from a few milliseconds to several seconds (Rossi-Durand et al. 1999; Sabbahi and Sedgwick 1982; Stein and Thompson 2006). We used pseudo-random stimulus intervals of 3–5 s to avoid any such prolonged effects from previous activation of the test pathway. Importantly, the lack of H-reflex attenuation in the control study with no arm cycling demonstrates that any effect of repetitive activity in the H-reflex pathway was insignificant. Thus we are confident that arm cycling and not methodological issues was responsible for the prolonged suppression of H-reflex amplitude shown here.
Remote movement induces prolonged suppression of H-reflex amplitude
The main finding of this study was that the suppressive effect of arm cycling on the soleus H-reflex pathway persists after termination of cycling. Thus remote movement can induce prolonged changes in reflex excitability. It has previously been shown that during rhythmic movement of one limb pair (arms or legs), the opposite stationary limb pair (legs or arms) shows a general suppression of H-reflex excitability (Frigon et al. 2004; Loadman and Zehr 2007; Zehr et al. 2004, 2007). Suppression of soleus H-reflex amplitude during arm cycling has been attributed to increased segmental Ia presynaptic inhibition. In the present study, soleus H-reflex amplitudes were suppressed during arm cycling consistent with the previous findings. The prolonged effect of remote movement on H-reflex excitability has not been studied before. However, there is some evidence that supports persistent activity-induced plasticity of soleus H-reflex amplitudes following leg cycling. After one revolution of slow (10 rpm) passive leg cycling, soleus H-reflexes remained suppressed ≤1–4 s after cessation of movement while tonically maintained background EMG did not significantly change (Misiaszek et al. 1995). This suggests that the H-reflex was suppressed via changes in presynaptic inhibition during cycling and also after termination of movement. Arm cycling also induces a generalized suppression of H-reflex amplitude via the modulation of Ia presynaptic inhibition (Frigon et al. 2004). We suggest the simplest explanation for the source of our results to be that increased Ia presynaptic inhibition persists beyond the period of arm cycling and well into the postcycling period. It should be noted that this is not likely to be some general systemic effect of arm cycling exercise. Motl and Dishman (2003) showed that prolonged leg cycling suppressed only H-reflexes in the leg muscle soleus that was activated and not the flexor carpi radialis in the arm that was at rest. This strongly argues against a systemic or hormonal effect.
Although the mechanism that leads to the persistence of Ia presynaptic inhibition remains unclear, the present findings include some features that suggest persistent inward currents of spinal interneurons mediating presynaptic inhibition. These could lead to plateau-like behavior that could generate self-sustained firing leading to persistent activity in this pathway (Collins et al. 2001, 2002; Gorassini et al. 1998; Hounsgaard et al. 1988; Kiehn and Eken 1997). Thus sustained suppression of H-reflex amplitude could occur with sustained persistent inward currents in Ia presynaptic inhibitory interneurons.
Cutaneous stimulation cancels out the movement-induced plasticity of H-reflex amplitude
After termination of cycling, cutaneous nerve stimulation immediately returned H-reflex amplitudes to control levels (see Fig. 2A). This is in line with our previous study in which SR nerve stimulation significantly facilitated H-reflex amplitudes during static contractions and counteracted the suppression of reflex amplitude induced by arm cycling (Zehr et al. 2004). In the present study, after application of SR stimulation, unconditioned H-reflex size was not significantly different from control value and it returned completely to the control size after 15 min. As plateau-like behavior is suggested to be terminated by inhibitory postsynaptic inputs (Lee and Heckman 1998a,b), we speculate that SR stimulation was able to cancel plateau-like behavior of the Ia PSI interneuron. Interestingly, there were no significant changes in the amplitudes of H-reflexes conditioned by SR stimulation after termination of cycling (Fig. 2B). Therefore the modulation of transmission of volleys from afferents in the cutaneous SR nerve to the Ia afferent soleus motoneuronal terminals was not influenced by the prolonged effect of rhythmic movement. We suggest the simplest explanation is that this remote cutaneous input hyperpolarizes interneuronal membrane excitability to reduce the plateau-like behavior. Plateau potentials have been described in many central neurons (Fraser and MacVicar 1996; Morisset and Nagy 1999) and have been well verified in mammalian motoneurons (Bennett et al. 1998; Collins et al. 2001, 2002; Kiehn and Eken 1997, 1998). While indirect, our results suggest the existence of such behavior at interneuronal levels in the human. This is in agreement with previous findings in some classes of spinal interneuron in the ventral horn of the turtle (Hounsgaard and Kjaerulff 1992) and the dorsal horn of rats, and turtles (Morisset and Nagy 1999; Russo and Hounsgaard 1996).
The results of the present study have translational implications for incorporating rhythmic arm movements in postneurotrauma rehabilitation strategies. Most importantly, the result of the present study might be useful in mitigation of spasticity. Muscle spasticity is related to exaggerated reflex excitability after CNS disorders such as stroke (Pierrot-Deseilligny 1990) and is correlated with H-reflex size in the neurologically impaired population (Levin and Hui-Chan 1993). Because rhythmic arm movement can produce sustained suppression in the soleus H-reflex pathway, it could be proposed as a helpful therapeutic tool for altering spasticity. An underlying assumption in any neurologic rehabilitation is the possibility of inducing persistent change beyond any rehabilitative intervention. The current results demonstrate the capacity for the spinal H-reflex pathway to express short-term neural plasticity. This is an essential feature which would permit a prolonged effect of any related rehabilitative interventions. Additional work remains to determine if rhythmic arm movement can still effectively access neural coupling between arms and legs and induce a prolonged suppressive effect on the H-reflex pathway in the neurologically impaired population.
This work was supported by grants to E. P. Zehr from the Natural Sciences and Engineering Council of Canada, the Heart and Stroke Foundation of Canada (BC and Yukon), and the Michael Smith Foundation for Health Research.
We thank H. Murray for assistance during data collection.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 by the American Physiological Society