HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: 97/2/1809    most recent 01038.2006v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (17) Citing Articles via Google Scholar Google Scholar Articles by Balter, J. E. Articles by Zehr, E. P. Search for Related Content PubMed PubMed Citation Articles by Balter, J. E. Articles by Zehr, E. P.

Neural Coupling Between the Arms and Legs During Rhythmic Locomotor-Like Cycling Movement

¹Rehabilitation Neuroscience Laboratory, University of Victoria, Victoria; and ²Human Discovery Science, International Collaboration on Repair Discoveries (ICORD), Vancouver, British Columbia, Canada

Submitted 28 September 2006; accepted in final form 23 October 2006

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Neuronal coupling between the arms and legs allowing coordinated rhythmic movement during locomotion is poorly understood. We used the modulation of cutaneous reflexes to probe this neuronal coupling between the arms and legs using a cycling paradigm. Participants performed rhythmic cycling with arms, legs, or arms and legs together. We hypothesized that any contributions from the arms would be functionally linked to locomotion and would thus be phase-dependent. Reflexes were evoked by electrical stimulation of the superficial peroneal nerve at the ankle, and electromyography (EMG) was recorded from muscles in the arms and legs. The main finding was that the relative contribution from the arms and legs was linked to the functional state of the legs. For example, in tibialis anterior, the largest contribution from arm movement [57% variance accounted for (VAF), P < 0.05] was during the leg power phase, whereas the largest from leg movement (71% VAF, P < 0.05) was during leg cycling recovery. Thus the contribution from the arms was functionally gated throughout the locomotor cycle in a manner that appears to support the action of the legs. Additionally, the effect of arm cycling on reflexes in leg muscles when the legs were not moving was relatively minor; full expression of the effect of rhythmic arm movement was only observed when both the arms and legs were moving. Our findings provide experimental support for the interaction of rhythmic arm and leg movement during human locomotion.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Locomotion (e.g., walking, running, or swimming) requires coordinated movement of the limbs. Coordination between the fore- and hindlimbs (i.e., interlimb coordination) has been well documented during quadrupedal locomotion (Miller and Van Der Meche 1975; Miller et al. 1975) and has been attributed to neural linkages connecting cervical and lumbosacral networks in the spinal cord (Gernandt and Megirian 1961; Gernandt and Shimuamura 1961). Although there is little direct evidence for interlimb locomotor linkages in humans, it has been suggested that elements of quadrupedal interlimb coordination are conserved during bipedal locomotion in humans (Dietz 2002). Indirect evidence from reflex studies can be used as a tool to probe such neural activity (Burke 1999). Recent research has investigated interlimb reflexes during locomotion, which provides some insight on the neural connections between the rhythmic centers controlling arm and leg movement. During walking, interlimb reflexes in both the arms and legs after stimulation of cutaneous nerves at the hand and foot were phase-modulated during the walking cycle (Haridas and Zehr 2003; Zehr and Haridas 2003). Reflexes observed in arm and leg muscles on both sides of the body at the same portion of the step cycle suggest that a coupling between segmental CPGs regulating activity in the arms and legs may be the mechanisms giving rise to the observed patterns of reflex modulation (Haridas and Zehr 2003). However, although evidence suggests the existence of flexible, task-dependent neuronal coupling between the arms and legs during gait (Dietz et al. 2001), the specific interaction between the arms and legs remains unknown.

It was suggested that the pattern of arm to leg coordination during walking, creeping, and swimming tasks in humans are due to the activity of coupled neural pattern generators (i.e., CPGs; 2 controlling arm movements and 2 controlling leg movements) (Wannier et al. 2001). Dietz and colleagues have shown task-modulated interlimb effects and have also ascribed these to the output of CPG elements (Dietz et al. 2001). We recently provided additional evidence for interlimb coupling when we showed cutaneous reflexes in both the arms and legs evoked by stimulation of nerves in the hand and foot to be phase-modulated during the walking cycle and to be also task dependent (Haridas and Zehr 2003). There was a reciprocally organized pattern and coordination of reflex responses from hand to foot and from foot to hand. These data provide evidence of strong reflex effects of sensory feedback evoked by activation of distant skin fields during movement and also of the influence of CPG activity regulating the arms and legs during walking (Zehr and Haridas 2003). However, examination of interactions between control of the arms and legs is difficult during locomotion because of the "interference" between rhythmic arm and leg activity. That is, it is not simple to determine if the modulation of reflexes in leg muscles evoked by stimulation in the hand during walking (e.g., Haridas and Zehr 2003) is due to control of rhythmic arm or leg activity. Indeed it may not even be possible to remove any effect of rhythmic neural activity related to the arms on leg muscles during walking since, even with the arms restrained, rhythmic EMG patterns exist in arm muscles (Fernandez-Ballesteros et al. 1965).

A way to probe for interaction between the neuronal oscillators presumed to contribute to movement of the arms and legs is to examine their coupling effects during rhythmic movement. In this context, coupling effects are operationally defined as a measurable effect of movement of one limb or limb pair on background or reflex muscle activity in another limb or limb pair (e.g., the effect of arm swing on activity in the legs). Within this framework and using a recumbent stepping paradigm (Huang and Ferris 2004), active arm movement led to an increase in neuromuscular activation in passively moving legs. Further, muscle activity amplitudes measured in the leg increased as arm exertion increased by either increasing the resistance (Huang and Ferris 2004) or the movement frequency (Kao and Ferris 2005). A different perspective on neural coupling related to rhythmic arm activity was shown using a reflex paradigm. In this regard, rhythmic arm cycling was shown to significantly suppress soleus H-reflex amplitude by 20% when the legs were stationary (Frigon et al. 2004). Interestingly, arm cycling also cancelled the presynaptic facilitation of H-reflex amplitude induced by stimulation of the cutaneous sural nerve and enhanced presynaptic inhibition of H-reflex amplitude induced by common peroneal nerve stimulation (Frigon et al. 2004). In addition, we were unable to detect any large changes in sural nerve cutaneous reflexes observed in the rectified EMG. In sum these observations led to the speculation that the effect of the arm cycling was to modulate Ia presynaptic inhibition in the soleus H-reflex pathway. However, the failure to modulate the cutaneous sural reflex amplitudes has remained puzzling because the coupling effects on the H-reflex pathway are robust. We speculated that detecting effects on the cutaneous pathways may require more subtle analysis techniques, including those that could allow partitioning of the effects of multiple movement sources (e.g., arm and leg movement) such as multiple regression.

Here we examined the effects of rhythmic arm activity on background and cutaneous reflex EMG modulation patterns during an active rhythmic task using both the arms and legs (i.e., arm&leg cycling). This allowed for the individual separation of arm and leg movement to determine the individual contributions of the arms and legs. We hypothesized that significant neural coupling would be observed during the arm&leg task. Specifically, this would be manifested as phase-related contributions from the arms and the legs to the coupling effects observed during the combined arm and leg task.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
The experimental methodology and protocol are similar to that described in previous experiments involving reflex modulation during arm (Zehr and Chua 2000; Zehr et al., 2001a,b) and leg cycling (Zehr et al. 2001b).

Participants

Fourteen participants (ages: 24–44; 10 females and 4 males), free of any known history of neurological or metabolic disorders, participated in this study. The participants provided informed written consent in a protocol approved by the Human Research Ethics Committee at the University of Victoria and performed in accordance with the Declaration of Helsinki.

Protocol

Participants completed three movement tasks: arm cycling with legs stationary and the feet placed on the base of the cycle ergometer while the knees were bent at a 90° angle (ARM); leg cycling with stationary arms at the side (LEG); and combined arm and leg cycling (ARM&LEG). Participants also performed static contractions with the limbs held stationary at four positions from the movement tasks (i.e., 12, 3, 6, and 9 o'clock for arm, leg, and arm&leg tasks). During all cycling and static tasks, a consistent EMG level was held in the tibialis anterior muscle, ipsilateral to the site of stimulation (iTA). To aid in maintaining this contraction, participants wore an ankle-foot orthosis (AFO) on their right side and were given visual feedback of this rectified EMG on an oscilloscope.

The experiment was conducted using an arm and leg cycle ergometer (PRO II, SCIFIT Systems, Tulsa OK) in which the arm and leg cranks were mechanically coupled to maintain a constant rigid relation between arm and leg movement (see Fig. 1). This mechanical coupling maintained the arms and legs 90° out-of-phase with one another during the combined arm and leg cycling condition (see Fig. 1). Rhythmic cycling was performed in a clockwise direction at a comfortable pace (60 rpm or 1 Hz cycling frequency) for 8 min for each cycling task. This rate is similar to those used during leg cycling (Brown and Kukulka 1993) and arm cycling (Zehr and Chua 2000) and is considered to be equivalent to a typical walking cadence.

View larger version (34K): [in this window] [in a new window]   FIG. 1. Schematic representing the arm and leg positions and corresponding functional phases of movement across the movement cycle. All positions and phases throughout the text are in reference to the position of the right leg. Leg positions are labeled on the outside of the circle with 12 o'clock at the top center position. Note that the arm crank is 90° ahead of the leg crank, meaning that when the leg is at 12 o'clock, the arm is at 3 o'clock (see arm positions labeled inside the circle). Also, the functional phases of leg cycling (i.e., power and recovery) are labeled and indicated by the dark and light arrows, respectively. For reference, an image of the equipment used (SCIFIT Pro II) is pictured in the lower right corner of the figure.

Cutaneous reflexes were evoked throughout the movement cycle with trains (5 x 1.0-ms pulses at 300 Hz) of isolated constant current stimulation applied pseudorandomly using a Grass S88 stimulator (Grass Instruments, AstroMed) connected in series with a SIU5 isolator and a CCU1 constant current unit. This electrical stimulation was applied to the superficial peroneal (SP) nerve on the participant's right side using disposable UNI-GEL electrodes (Thought Technologies) placed on the anterior surface of the leg near the crease of the ankle joint. Stimulus intensity was set at approximately two times [2.1 ± 0.18 (mean ± SD)] the threshold at which a clear radiating paresthesia [radiating threshold (RT)] was detected in the innervation area of the dorsum of the foot.

Electromyography

EMG was collected from six muscles ipsilateral (i) to the site of stimulation: posterior deltoid (PD), flexor carpi radialis (FCR), vastus medialis (VM), biceps femoris (BF), medial gastrocnemius (MG), and TA. Muscle activity and reflexes were collected in 14 participants for iTA, 13 participants for iPD, 12 participants for iFCR; and 11 participants for iVM, iBF, and iMG.

Sites over the selected muscles were cleaned with rubbing alcohol, and disposable 1-cm surface EMG electrodes (Thought Technologies) were applied in a bipolar configuration with a 3-cm inter-electrode distance. Ground electrodes for EMG recordings were placed on bony landmarks near the selected muscles. EMG recordings were preamplified and band-pass filtered at 100–300 Hz (P511 Grass Instruments, AstroMed)

Cycle timing and kinematics

Positions of the cranks throughout the movement cycle were obtained from two (i.e., 1 each for the arms and legs) custom-made optical encoders for counting the sprocket teeth and that were mounted to the interior frame of the ergometer. The movement cycle was divided into 12 phases, equivalent to a clock-face with 12 o'clock at the top center position (see Fig. 1). Phases of movement are referenced to the position of the legs. In addition, angular positions of the elbow (n = 12) and knee (n = 13) were measured using lightweight electro-goniometers (Twin Axis SG150 and SG110, Biometrics, Gwent, UK).

Data acquisition and analysis

Data were acquired at a sampling rate of 1,000 Hz with a 12-bit A/D converter connected to a computer running custom-written (Dr. Timothy Carroll, University of New South Wales, Australia) Lab View software (National Instruments, Austin, TX). Off-line, using custom-written software programs (Matlab, The Mathworks, Natick, MA) data were separated into 12 phases of the movement cycle, beginning with the top center position (12 o'clock) and were averaged throughout the movement cycle.

At each phase of movement, individual subtracted EMG traces from each muscle were analyzed in terms of reflex amplitudes and latencies. These reflex traces were obtained by subtracting the average trace of nonstimulated cycles from the corresponding average trace of stimulated cycles. The stimulation artifact was removed and the EMG trace was low-pass filtered using a third-order Butterworth dual pass filter set at 40 Hz. Reflex amplitudes were analyzed at middle latencies (80–120 ms to peak) only, although the exact latency of this epoch was based on the overall pattern of responses from all participants separately and was not determined a priori. Peak reflex amplitudes were obtained by averaging a 10-ms window centered on the peak of each reflex response. Reflex amplitudes were considered significant if at least one of the epochs exceeded a 2 SD band calculated from recorded prestimulus activity subtraction error.

Cutaneous reflex amplitudes and background EMG for each subject were normalized to the peak value of the control (unstimulated) EMG for that muscle across the movement cycle. Activity in arm muscles were normalized to the peak value during the ARM task, while the peak value from the LEG task was used to normalize reflex amplitudes and background EMG in the leg muscles.

Statistical analysis

Based on our previous experience (Frigon et al. 2004; Zehr et al. 2004) and our a priori expectation that any coupling effects between the arms and legs would be weak, we exploited both standard (e.g., ANOVA) and more sophisticated analysis and statistical procedures (e.g., multiple linear regression and absolute value analysis).

Approach common to all data

In all cases, analysis was conducted on the averaged normalized values for each subject from each phase of the movement cycle. STASTICA software (version 6.1, StatSoft, Tulsa, OK) was used to perform repeated measures analyses of variance (RM ANOVA) tests to identify significant main effects for task and movement phase on background EMG (bEMG) levels and middle latency reflex amplitudes during static contraction and cycling trials. Subsequently, Tukey's HSD tests were performed post hoc on significant differences found.

Linear regression analysis using Pearson's correlation coefficients (r) was used to determine relationships between reflex amplitudes and background EMG levels for each muscle during all tasks.

To gauge the relative contributions of ARM and LEG to the combined ARM&LEG task, a series of forward stepwise multiple regressions were performed for ipsilateral TA, VM, BF, and MG muscles. Across the entire movement cycle and within each specific phase, middle latency reflex amplitudes evoked during the arm&leg task (i.e., criterion variable) were compared with those obtained using two predictor variables: the ARM task and the LEG task. Additionally, for results from iTA, a Hotellings t-test was used to assess the differences between the two dependent correlation coefficients (Brace et al. 2003).

Descriptive statistics included means ± SE (except in Table 1 where SDs are used). For all tests, statistical significance was set at P < 0.05.

Because ongoing muscle activity (i.e., bEMG) was experimentally controlled in iTA during all three tasks, the statistical analysis for this muscle was different from the analysis performed for the other muscles. That is, one-way RM ANOVAs were conducted for each task to determine whether middle latency reflexes were significantly modulated dependent on the phase of movement within that task. Additionally, planned comparisons were conducted at all phases within the cycle where bEMG was matched (that is, not significantly different) across tasks to determine if task-dependent differences in reflex amplitudes exist. Despite it being fairly well established that muscle activity per se does not influence cutaneous reflex amplitudes during rhythmic movement of either the arms or legs, we thought it important to include this step to be more convincing and to ensure that the targeted EMG activity in iTA would not contaminate our results. When attempts have been made to find association between background EMG and cutaneous reflex amplitude during rhythmic movements by our laboratory (Haridas and Zehr 2003; Hundza and Zehr 2006; Komiyama et al. 2000; Zehr and Haridas 2003; Zehr and Hundza 2005; Zehr and Kido 2001) and others (notably see the recent paper by Sakamoto et al. 2006 who used a very similar procedure to our own) little relation has been found.

As mentioned in the INTRODUCTION, the effect of ARM on cutaneous reflex amplitudes in TA was predicted to be small, subtle to detect, and not likely to be phase-dependent (Frigon et al. 2004). This is in contrast with the effect of LEG or ARM&LEG. Therefore to clearly detect any conditioning effect of arm movement on cutaneous reflexes in the legs, we also applied an absolute value analysis to test for significant independent effects for ARM, LEG, and ARM&LEG. This approach is similar to one we have used recently to determine more subtle context-dependent effects on cutaneous reflexes in leg muscles during walking (see Haridas et al. 2005). This procedure involved taking the absolute value of the difference of the ARM, LEG, and ARM&LEG from corresponding reflexes during static contraction at 3, 6, 9, and 12 o'clock and calculating a two-tailed t-ratio with a hypothesized mean difference of zero (see also Zehr et al. 1997). This is similar in principle to the procedure of evaluating changes in H-reflex amplitude as a percentage change from control.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Reflex latencies

Reflex latencies as determined by the time to peak response are shown in Table 1 for all muscles in each task (mean ± SD). Reflexes were examined at similar time windows among tasks within a muscle for a given individual because no significant differences were found among the reflex latencies for different tasks.

Control data from static trials

Analysis of static trials was only conducted for iTA (n = 13), and a prominent middle latency inhibition was consistently observed across the 12 static trials (i.e., 12, 3, 6, and 9 o'clock for arm, leg, and arm&leg tasks). On average this static value was –20 ± 7.5% bEMG (mean ± SE, normalized to the peak EMG value from the unstimulated leg task). This inhibition during static contraction was not modulated by position or across tasks. That is, there was no significant effect for static task or position on this middle latency inhibitory reflex.

Phase- and task-dependent reflex modulation in the target muscle (iTA)

Reflex traces in iTA muscle for a representative individual subject during each cycling task are shown in Fig. 2. The phase of movement is noted by the numbers found to the left of each panel. Phase-modulation of the middle latency reflex (85–110 ms after stimulation) can be seen within each task. The most prominent phase-dependent modulation (refer to Fig. 2) occurred during the leg (middle) and arm&leg (right) tasks when compared with that seen during the arm task (left). Identical scales have been used for the subtracted reflex traces to allow for comparison among the tasks. In addition to depicting phase-dependent reflex modulation, Fig. 2 also shows task-dependent reflex modulation in iTA for a single subject (the 6 shaded boxes highlight the 3 phases where significant task differences were observed in the group data; see Fig. 3). Examination of middle latency reflex amplitudes (highlighted by the shaded boxes) across the three tasks for this subject reveals a decrease in inhibitory amplitude during the leg and arm&leg tasks (middle and right, respectively) when compared with the arm task (left).

View larger version (15K): [in this window] [in a new window]   FIG. 2. Subtracted reflex electromyographic (EMG) traces ipsilateral tibialis anterior (iTA) muscle from 1 subject evoked during the arm, leg, and arm&leg tasks. In these traces the stimulus artifact has been removed and the phases (in a clockwise direction) are plotted vertically starting with the arm at 3 o'clock and/or the leg at 12 o'clock as the top plot. In addition to the visible middle latency reflex phase modulation (85–110 ms), also note the decrease in inhibition observed at 12, 1, and 11 o'clock (gray boxes) when comparing the middle latency reflex amplitudes during the leg and arm&leg tasks to that of the arm task.

View larger version (34K): [in this window] [in a new window]   FIG. 3. Background EMG (bEMG) and middle latency reflexes in iTA during the 3 tasks averaged across all participants (n = 14). bEMG (shaded plots) and reflex amplitudes (line plots) have been normalized to the peak value of the background (unstimulated) EMG obtained in the movement cycle for the leg task in each subject and are expressed as percentages. The white box covering the data at 2 and 3 o'clock indicate where the bEMG was found to be significantly different across the tasks. The asterisk denotes the tasks in which significant phase modulation of middle latency reflex amplitudes was observed. Middle latency reflex amplitudes in iTA were phase modulated during the leg and arm&leg tasks (P < 0.05), whereas no significant phase modulation was observed during the arm task.

View larger version (22K): [in this window] [in a new window]   FIG. 4. Results of the absolute value analysis performed on middle latency reflexes in iTA during the 3 tasks. The bars represent the average absolute difference (cycling –static) in normalized amplitude of the middle latency reflexes at the 12, 3, 6, and 9 o'clock positions. *, indicates whether the t-ratio was significant, meaning that the difference in amplitude between cycling and static tasks was significantly different from 0 (df = 12, 2-tail, P < 0.05). Note that within the arm task, cycling had a significant effect on middle latency reflex amplitudes in iTA when the arm was at 3 and 6 o'clock. Three phases were also significant during leg, however during the combined arm&leg task the conditioning effect of movement was significant at all phases.

Figure 5 shows bEMG (left) and middle latency reflex amplitudes (right) across all participants plotted according to the phase of movement in which they were evoked. Significant main effects for task and phase, and significant task x phase interactions (P < 0.05) are discussed in the following text and labeled on Fig. 5. Phase-dependent modulation was found for middle latency reflex amplitudes in iBF and iMG (P < 0.05; indicated by the "phase" written in the panels for each muscle on Fig. 5). Table 3 summarizes the phases in which bEMG were matched (i.e., not statistically different) between tasks in the first panel, and asterisks denote significant task-dependent differences (P < 0.05) in middle latency reflex amplitudes in the second panel. Although subsequent discussion of our findings focuses on results in iTA, these are supported by similar findings in the other leg muscles (i.e., iVM, iBF, and iMG).

View larger version (43K): [in this window] [in a new window]   FIG. 5. Phase- and task-dependent modulation of bEMG (line plots) and middle latency reflexes (bar plots) in all muscles ipsilateral to the site of stimulation. Amplitudes for middle latency reflexes and bEMG have been normalized to the peak value of the background (unstimulated) EMG obtained in the movement cycle for the leg task for leg muscles and arm muscles are normalized to the peak bEMG from the arm task in each subject and are expressed as percentages. Additionally, kinematics for the elbow and knee are represented in degrees. Significant main effects for task and phase and significant task x phase interactions are indicated for each muscle. Refer to Table 3 for the phases with matched bEMG among tasks and task-dependent differences in middle latency reflex amplitudes. Abbreviations as in Table 1.

For the remaining muscles, there were fewer significant task-dependent differences in middle latency reflex amplitudes when compared with task-dependent differences in bEMG. A main effect for task (P < 0.05) on middle latency reflex amplitudes was observed in iFCR, iVM, and iMG (indicated by the "task" in Fig. 5, right). More specifically, reflex amplitudes during the ARM task were significantly different from the LEG task for iFCR, iVM, and iMG, and the arm task was significantly different from the arm&leg task for iMG. For leg muscles, middle latency reflex amplitudes (much like bEMG) between the leg and arm&leg tasks were similarly modulated throughout the movement cycle. This is depicted in Fig. 5, right, where for the leg muscles the LEG (light gray bars) and ARM&LEG (black bars) follow the same pattern of modulation. The only significant task by phase modulation was observed at 1–4 o'clock in iMG (see the asterisks in Table 3 and on Fig. 5).

Relation between level of background EMG and amplitude of middle latency reflexes

For all muscles, results from a linear regression analysis for bEMG and middle latency reflex amplitudes demonstrated that of the 18 correlations examined, only 3 were statistically significant (Table 2, asterisks). These significant correlations were found during the arm task in iBF and iMG (compare the dark gray triangle symbols and bars in both panels of Fig. 5 for these muscles and notice their almost identical patterns) and during the leg task in iPD (compare the light gray circles in both panels of Fig. 5 for this muscle). During the arm&leg task, no muscles demonstrated a significant correlation between bEMG and middle latency reflex amplitudes.

Relative contributions from the arms and legs

REFLEX EXPRESSION IN ITA.  The individual contribution from the arms or legs (statistically, the weighting of each factor) in the combined arm and leg cycling condition was determined using forward stepwise multiple regression analysis across the movement cycle (i.e., 12 phases of movement). This analysis indicated that the leg task was the main contributor to middle latency reflex amplitudes expressed during the arm&leg task throughout the entire movement cycle. Specifically, the leg predictor variable accounted for 33% (R² change = 0.33; df = 2, 156; F-change = 80.58, P < 0.05) of the total variance within the arm&leg condition, whereas the ARM accounted for an additional 5% of the variance (R² change = 0.05; df = 2, 156; F-change = 50.87, P < 0.05). Both LEG and ARM made statistically significant contributions to the arm&leg task [beta coefficients for each predictor variable were leg = 0.497; arm = 0.245 (P < 0.05)]).

To determine any phase-dependent contributions from the arms and legs, additional forward stepwise multiple regressions were performed on middle latency reflex amplitudes in iTA for all participants for each phase of movement. Figure 6C depicts the relative contributions from the arm and leg to reflex expression during the arm&leg task. Significant adjusted R² values for each predictor variable, according to the movement phase, are indicated by the asterisks in Fig. 6C, and the "L" and "A" denote significant contributions from the leg and arm, respectively. For example, the greatest contribution from the arms was 57% at 11 o'clock (R² change = 0.57; df = 2, 11; beta = 0.63; F-change = 16.17, P < 0.05; t = 0.85, P < 0.05 indicated by the asterisk and "A"), and the greatest contribution from the legs was 71% at 9 o'clock (R² change = 0.71; df = 2, 11; beta = 0.84, P < 0.05; t = –1.41, P < 0.05 indicated by the asterisk and "L"). Overall, the LEG (light gray bars) accounts for the greatest amount of variance during the arm&leg task (see the black asterisks and "L" on Fig. 6 at 4, 6, 7, 8, 9, and 11 o'clock), whereas ARM (black bars) contributes significantly to the overall variance in ARM&LEG only at certain phases of the movement cycle (see the white asterisks and "A" on Fig. 6 at 2, 10, and 11 o'clock). Additionally, results from a Hotellings t-test revealed that the strength of the correlation between each predictor variable with the criterion variable was significantly greater for the LEG and ARM&LEG comparison at 4, 6–9 o'clock whereas the correlation between the ARM and ARM&LEG variable was significantly stronger at 11 o'clock. Interestingly, these phase-dependent contributions from the arm and leg were found to correspond to the functional phases of cycling (see Figs. 1 and 6B). That is, the largest contribution from the arm to the expression of reflexes in the legs during the arm&leg task was during the power phase of leg movement, whereas the largest contribution from leg was during the recovery phase of leg cycling.

View larger version (29K): [in this window] [in a new window]   FIG. 6. Phase-dependent arm and leg contributions to reflex expression in iTA and the functional phases of the legs during cycling and walking. Using the previously defined terms of limb "flexion" and "extension" (Ting et al. 1998, 1999) (A), the recumbent cycling paradigm in this present study can be interpreted functionally (B) and compared with similar phases within the gait cycle (D). Additionally, the relative contributions from the arm and leg to reflex expression in iTA during the ARMandLEG task [% variance accounted for (VAF)] can be addressed according to the functional state of the legs (C). Significant contributions from the arm and leg variables (P < 0.05) are indicated (asterisk) and "A" and "L" refer to significant contributions from the arm and leg variables, respectively. The gray arrow on C highlights the largest contribution from the arm at heel strike which is a crucial point in the gait cycle.

Additional forward stepwise multiple regressions were performed at the leg position that corresponded to the position at which the greatest contribution from the ARM to reflex expression in iTA (i.e., 11 o'clock). Results in iVM and iMG correspond with those reported for iTA. That is, at 11 o'clock the arm was found to make a significant contribution to the reflex expression during the arm&leg task (iVM: R² change = 0.42; df = 1, 9; beta = 0.65; F-change = 6.47, P < 0.05; iMG: R² change = 0.47; df = 1, 9; beta = –0.68; F-change = 7.95, P < 0.05). However, for iBF no significant contribution from the ARM was found at this phase (R² change = 0.38; df = 1, 8; beta = 0.61; F-change = 4.2, P > 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
The locomotor interaction between the control of rhythmic arm and leg movement was probed by using the modulation of middle latency cutaneous reflexes during arm, leg, and arm&leg cycling. The main finding was that rhythmic arm movement has a significant phase-dependent contribution to reflex expression in the legs during a combined arm and leg task. This effect is superimposed on the dominant effect of leg movement. We interpret this coupling between the arms and legs in terms of functional relationships during similar rhythmic activities, such as walking.

Interaction between the control of rhythmic arm and leg movement

The arm&leg task was used as the task of reference to determine the interaction between the individual control of the arms and that of the legs and the relative contribution of each during a combined arm and leg task. The neuronal coupling between the arms and legs has been previously examined in terms of neuromuscular activation during recumbent stepping at various levels of resistance (Huang and Ferris 2004) and speeds (Kao and Ferris 2005). In these studies, the active arm and leg stepping condition served as the task of reference just as the arm&leg task was used in this present study. In contrast to Huang & Ferris (2004), who reported that active upper limb movement increases neuromuscular activation in passively moving lower limbs, no significant increases in muscle activity in the lower limbs were observed during the combined cycling task when compared with the leg cycling task (see the phases with matched bEMG for the leg muscles in the 1st panel of Table 3 and the comparison plots in Fig. 5). This is likely a result of the differing activity states of the lower limbs (i.e., passive vs. active movement). Additionally, examination of reflex amplitudes during the leg and arm&leg tasks in iTA and other muscles where bEMG was matched across tasks, revealed no significant differences between the two tasks (refer to Figs. 2 and 5).

Interestingly, in a recent study similar to our own, Sakamoto et al. (2006) examined the effects of rhythmic arm and leg cycling on cutaneous reflexes in arm and leg muscles. This paper also addressed the issue of voluntary postural changes on reflex modulation. They found little effect on reflex amplitudes of these voluntary manipulations. An important point is that, in contrast to the current study, there was no mechanical link between arm and leg motion during the arm&leg task in Sakamoto et al. (2006). Additionally, many other parameters including inphase (i.e., arm and leg positions identical during movement) and antiphase (i.e., as performed here) cycling were performed. However of major interest, the conclusion reached by Sakamoto et al. was that cutaneous reflexes in leg muscles are not influenced strongly by arm movement. This contrasts somewhat with our observations showing a subtle but real interaction of arm and leg movement. We believe this discrepancy between our conclusions and those of Sakamoto et al. to be largely due to the more detailed analytical and statistical procedures employed here as well as the difference in mechanical linkage between the arms and legs.

Previous investigations into the influences of rhythmic arm movement on motoneuronal excitability in the legs (reviewed in Zehr and Duysens 2004) suggested that presynaptic inhibition may be a likely mechanism responsible for the subtle interlimb reflex modulation observed. Consequently, it might be expected that any influence of the remote rhythmic arm activity on reflex modulation in the legs observed in this current study would also require more subtle analysis techniques (e.g., absolute value T-ratios, planned comparisons and multiple regression analysis). Indeed, a significant conditioning effect from movement of the arms was detected in the arm task at two movement phases. Most interestingly, full expression of the arm conditioning effect was only observed when both the arms and legs were moving (i.e., ARM&LEG) and the movement conditioning was then significant at all four phases examined (see Fig. 4). In addition, multiple regression analysis across the entire movement cycle in iTA during ARM&LEG revealed that the leg variable accounted for the largest amount of variance, whereas the arm variable had a minor yet significant effect. Therefore reflex modulation during combined rhythmic activity indicates that rhythm generation in the legs (i.e., the lumbar region) remains dominant over rhythm generators for the arms (i.e., the cervical region) while still being sensitive to some contribution from the arms. This finding represents a complimentary observation to a recent study in the neonatal rat preparation (Juvin et al. 2005). Juvin et al. (2005) examined the nature of propriospinal interactions between cervical and lumbar locomotor CPGs in an isolated spinal cord preparation. In addition to revealing independent rhythmogenic capabilities in the cervical and lumbar locomotor regions, these authors also reported dominance in locomotor drive from the lumbar generators over the cervical counterparts. Furthermore, this finding was suggestive of an ascending "caudorostral excitability gradient" mediating interlimb coordination because the ascending influence of lumbar CPGs on the cervical counterparts increased as a function of the number of thoracic segments exposed (Juvin et al. 2005). The ARM&LEG human cycling paradigm we used is clearly not identical to the neonatal rat preparation. However, we suggest that the general interaction between rhythmic arm and leg movement parallels the observation from the rat preparation. We interpret this as evidence for commonalities in neural regulation among mammalian tetrapods and is a predicted outcome for the interactions between CPGs regulating arm and leg movement during human walking (Zehr 2005; Zehr and Duysens 2004).

Contributions from the arms depend on the phase of leg movement

An interesting observation in iTA is that the relative contributions from the arms and legs were phase-dependent (refer to Fig. 6C, which represents these contributions in relation to the functional phase of leg movement). That is, the relative contribution from the arms was the largest during the power phase of leg cycling (i.e., limb extension), whereas the leg contribution was dominant during the recovery phase (i.e., limb flexion). Interestingly, this outcome parallels similar observations in the cat in which descending propriospinal paths were found to facilitate extension in the hindlimbs (Lloyd and McIntyre 1948; Miller et al. 1975). Additionally, Ting et al. (1998) suggested that "the default strategy during locomotor tasks, such as pedalling and walking, may be to modulate the gain of afferent pathways such that they are strongly effective during limb extension, or the power phase, and ineffective during flexion, or the recovery phase. " Our current findings corroborate these suggestions because the arm contribution was significant only during limb extension. Further examination of the phase-dependent contribution from the arm showed that the greatest contribution was at a leg position that is comparable to the transition from swing to stance phase during walking (indicated by the large arrow on Fig. 6C). It has been suggested that an increase in inhibition within the pathway to tibialis anterior is essential in balancing the increased excitation experienced during heel strike at this specific phase of the gait cycle (Duysens et al. 2004; Zehr and Duysens 2004). This significant contribution from the arms to reflex expression in the legs at this crucial point in the gait cycle may be explained by a possible increased reliance on multisensory integration to help guide the swing limb and ensure proper foot placement.

Furthermore, the question arises as to whether this significant contribution from the legs is due to the fact that the legs are moving, or because the sensory input (i.e., SP nerve stimulation at the ankle) arises from the legs. It appears that the activity state of the limb receiving the input (i.e., movement or stimulation) is given priority over inputs evoked from stationary limbs (Carroll et al. 2005; Sakamoto et al. 2006). We suggest that commands from the control centers regulating arm movement are given access to the legs at certain phases of movement (i.e., during limb extension or the power phase of cycling), although the overall contribution from the arms is lower in comparison to the contribution from rhythmic leg movement.

Functional implications

In the present study, significant phase-dependent modulation of middle latency reflexes in iTA was only observed during the leg and arm&leg tasks (see Fig. 5), and appeared unaffected by the focused contraction in TA. It is of interest to note that recent experiments in the spinal cat have demonstrated that cutaneous reflexes arising from SP nerve (which in the cat and man is the main substrate mediating the "stumble corrective response") are "hard wired" into the lumbar spinal locomotor CPG (Quevedo et al. 2005). Therefore the significant phase-dependent reflex modulation present only during the leg and arm&leg tasks makes functional sense and it is directly related to the cat work because both of these tasks are presumed to have considerable contributions from the lumbar CPG. Although no significant phase-dependent modulation was detected during the arm task, it should be restated that a significant conditioning effect from arm movement was detected during the arm task. A similar conditioning effect from leg movement was observed during the leg task, although this conditioning effect was enhanced and fully expressed only when all limbs were active (i.e., during the arm&leg task).

Cutaneous reflexes were used as a tool to probe the interaction between rhythmic arm and leg movement. The paradigm allowed for the partitioning out of the relative contributions from arm and leg movement to the reflexes expressed during a combined task. The main finding is that during a combined rhythmic arm and leg task, the arms have a significant phase-dependent effect on reflexes expressed in the legs. Specifically, the contribution from the arms to reflex expression in the legs appears to be gated according to the functional state of the legs. That is, during the power phase of leg cycling, there may be an increased reliance on feedback from the arms as evidenced by a significant contribution observed from the arms onto the legs. Therefore the activity state of the legs appears to be the most important consideration for coupling between the arm and legs. Furthermore, the observation that the expression of reflexes in iTA during the arm&leg task throughout the recovery phase of movement is dominated by the legs can possibly be explained by the fact that input from the arms (via the CPGs for the arms) may come via common interneuron(s) and have limited access to the motorneurons responsible for the observed CPG output during leg cycling. This speculation would explain why no significant differences in reflex amplitudes were observed when comparing the leg and arm&leg tasks. However, it must also be mentioned that coupling between the arms and legs is likely to be differentially affected based on the mode of locomotion (e.g., swimming, walking, arm/leg cycling, etc.) (e.g., see (Dietz 2002).

	GRANTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
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, the Michael Smith Foundation for Health Research, and the Rick Hansen Man in Motion Foundation.

	FOOTNOTES

 
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.

Address for reprint requests and other correspondence: E. P. Zehr, Rehabilitation Neuroscience Laboratory, PO Box 3010 STN CSC, University of Victoria, Victoria, BC, Canada, V8W 3P1 (E-mail: pzehr{at}uvic.ca, www.zehr.ca)

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Brace N, Kemp R, Sneglar R. Multiple Regression in SPSS for Psychologists: A Guide to Data Analysis Using SPSS for Windows. Hampshire, UK: Palgrave MacMillan, 2003.

Brown DA, Kukulka CG. Human flexor reflex modulation during cycling. J Neurophysiol 69: 1212–1224, 1993.[Abstract/Free Full Text]

Burke RE. The use of state-dependent modulation of spinal reflexes as a tool to investigate the organization of spinal interneurons. Exp Brain Res 128: 263–277, 1999.[CrossRef][Web of Science][Medline]

Carroll TJ, Zehr EP, Collins DF. Modulation of cutaneous reflexes in human upper limb muscles during arm cycling is independent of activity in the contralateral arm. Exp Brain Res 161: 133–144, 2005.[CrossRef][Web of Science][Medline]

Dietz V. Do human bipeds use quadrupedal coordination? Trends Neurosci 25: 462–467, 2002.[CrossRef][Web of Science][Medline]

Dietz V, Fouad K, Bastiaanse CM. Neuronal coordination of arm and leg movements during human locomotion. Eur J Neurosci 14: 1906–1914, 2001.[CrossRef][Web of Science][Medline]

Duysens J, Bastiaanse CM, Smits-Engelsman BC, Dietz V. Gait acts as a gate for reflexes from the foot. Can J Physiol Pharmacol 82: 715–722, 2004.[CrossRef][Web of Science][Medline]

Fernandez-Ballesteros ML, Buchtal F, Rosenfalck P. The pattern of muscular activity during the arm swing of natural walking. Acta Physiol Scand 63: 296–310, 1965.[Web of Science][Medline]

Frigon A, Collins DF, Zehr EP. Effect of rhythmic arm movement on reflexes in the legs: modulation of soleus H-reflexes and somatosensory conditioning. J Neurophysiol 91: 1516–1523, 2004.[Abstract/Free Full Text]

Gernandt BE, Megirian D. Ascending propriospinal mechanisms. J Neurophysiol 24: 364–376, 1961.[Free Full Text]

Gernandt BE, Shimuamura M. Mechanisms of interlimb reflexes in cat. J Neurophysiol. 24: 665–676, 1961.[Free Full Text]

Haridas C, Zehr EP, Misiaszek JE. Postural uncertainty leads to dynamic control of cutaneous reflexes from the foot during human walking. Brain Res 1062: 48–62, 2005.[CrossRef][Web of Science][Medline]

Haridas C, Zehr EP. Coordinated interlimb compensatory responses to electrical stimulation of cutaneous nerves in the hand and foot during walking. J Neurophysiol 90: 2850–2861, 2003.[Abstract/Free Full Text]

Huang HJ, Ferris DP. Neural coupling between upper and lower limbs during recumbent stepping. J Appl Physiol 97: 1299–1308, 2004.[Abstract/Free Full Text]

Hundza S, Zehr E. Cutaneous reflexes during rhythmic arm cycling are insensitive to asymmetrical changes in crank length. Exp Brain Res 168: 165–177, 2006.[CrossRef][Web of Science][Medline]

Juvin L, Simmers J, Morin D. Propriospinal circuitry underlying interlimb coordination in mammalian quadrupedal locomotion. J Neurosci 25: 6025–6035, 2005.[Abstract/Free Full Text]

Kao PC, Ferris DP. The effect of movement frequency on interlimb coupling during recumbent stepping. Mot Control 9: 144–163, 2005.[Web of Science][Medline]

Komiyama T, Zehr EP, Stein RB. Absence of nerve-specificity in human cutaneous reflexes during standing. Exp Brain Res 133: 267–272, 2000.[CrossRef][Web of Science][Medline]

Lloyd DPC, McIntyre AK. Analysis of forelimb-hindlimb reflex activity in acutely decapitate cats. J Neurophysiol 11: 455–470, 1948.[Free Full Text]

Miller S, Van Der Burg J, Van Der Meche FGA. Coordination of movements of the hindlimbs and forelimbs in different forms of locomotion in normal and decerebrate cats. Brain Res 91: 217–237, 1975.[CrossRef][Web of Science][Medline]

Miller S, Van Der Meche FGA. Movements of the forelimbs of the cat during stepping on a treadmill. Brain Res 91: 255–269, 1975.[CrossRef][Web of Science][Medline]

Quevedo J, Stecina K, Gosgnach S, McCrea DA. Stumbling corrective reaction during fictive locomotion in the cat. J Neurophysiol 94: 2045–2052, 2005.[Abstract/Free Full Text]

Sakamoto M, Endoh T, Nakajima T, Tazoe T, Shiozawa S, Komiyama T. Modulations of interlimb and intralimb cutaneous reflexes during simultaneous arm and leg cycling in humans. Clin Neurophysiol 117: 1301–1311, 2006.[CrossRef][Web of Science][Medline]

Ting LH, Kautz SA, Brown DA, Zajac FE. Phase reversal of biomechanical functions and muscle activity in backward pedaling. J Neurophysiol 81: 544–551, 1999.[Abstract/Free Full Text]

Ting LH, Raasch CC, Brown DA, Kautz SA, Zajac FE. Sensorimotor state of the contralateral leg affects ipsilateral muscle coordination of pedaling. J Neurophysiol 80: 1341–1351, 1998.[Abstract/Free Full Text]

Wannier TMJ, Bastiaanse CM, Columbo G, Dietz V. Arm to leg coordination in humans during walking, creeping and swimming activities. Exp Brain Res 141: 375–379, 2001.[CrossRef][Web of Science][Medline]

Zehr EP. Neural control of rhythmic human movement: the common core hypothesis. Exerc Sport Sci Rev 33: 54–60, 2005.[Web of Science][Medline]

Zehr EP, Chua R. Modulation of human cutaneous reflexes during rhythmic cyclical arm movement. Exp Brain Res 135: 241–250, 2000.[CrossRef][Web of Science][Medline]

Zehr EP, Collins DF, Chua R. Human interlimb reflexes evoked by electrical stimulation of cutaneous nerves innervating the hand and foot. Exp Brain Res 140: 495–504, 2001a.[CrossRef][Web of Science][Medline]

Zehr EP, Duysens J. Regulation of arm and leg movement during human locomotion. Neuroscientist 10: 347–361, 2004.[Abstract/Free Full Text]

Zehr EP, Haridas C. Modulation of cutaneous reflexes in arm muscles during walking: further evidence of similar control mechanisms for rhythmic human arm and leg movements. Exp Brain Res 149: 260–266, 2003.[CrossRef][Web of Science][Medline]

Zehr EP, Hesketh KL, Chua R. Differential regulation of cutaneous and H-reflexes during leg cycling in humans. J Neurophysiol 85: 1178–1185, 2001b.[Abstract/Free Full Text]

Zehr EP, Hoogenboom N, Frigon A, Collins DF. Facilitation of soleus H-reflex amplitude evoked by cutaneous nerve stimulation at the wrist is not suppressed by rhythmic arm movement. Exp Brain Res 159: 382–388, 2004.[CrossRef][Web of Science][Medline]

Zehr EP, Hundza SR. Forward and backward arm cycling are regulated by equivalent neural mechanisms. J Neurophysiol 93: 633–640, 2005.[Abstract/Free Full Text]

Zehr EP, Kido A. Neural control of rhythmic, cyclical human arm movement: task dependency, nerve specificity and phase modulation of cutaneous reflexes. J Physiol 537: 1033–1045, 2001.[Abstract/Free Full Text]

Zehr EP, Komiyama T, Stein RB. Cutaneous reflexes during human gait: electromyographic and kinematic responses to electrical stimulation. J Neurophysiol 77: 3311–3325, 1997.[Abstract/Free Full Text]

This article has been cited by other articles:

				A. Frigon, G. Barriere, H. Leblond, and S. Rossignol Asymmetric Changes in Cutaneous Reflexes After a Partial Spinal Lesion and Retention Following Spinalization During Locomotion in the Cat J Neurophysiol, November 1, 2009; 102(5): 2667 - 2680. [Abstract] [Full Text] [PDF]

				N. Kawashima, D. Nozaki, M. O. Abe, and K. Nakazawa Shaping Appropriate Locomotive Motor Output Through Interlimb Neural Pathway Within Spinal Cord in Humans J Neurophysiol, June 1, 2008; 99(6): 2946 - 2955. [Abstract] [Full Text] [PDF]

				E. P. Zehr, M. Klimstra, K. Dragert, Y. Barzi, M. G. Bowden, B. Javan, and C. Phadke Enhancement of Arm and Leg Locomotor Coupling With Augmented Cutaneous Feedback From the Hand J Neurophysiol, September 1, 2007; 98(3): 1810 - 1814. [Abstract] [Full Text] [PDF]

				E. P. Zehr, J. E. Balter, D. P. Ferris, S. R. Hundza, P. M. Loadman, and R. H. Stoloff Neural regulation of rhythmic arm and leg movement is conserved across human locomotor tasks J. Physiol., July 1, 2007; 582(1): 209 - 227. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 97/2/1809    most recent 01038.2006v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (17) Citing Articles via Google Scholar Google Scholar Articles by Balter, J. E. Articles by Zehr, E. P. Search for Related Content PubMed PubMed Citation Articles by Balter, J. E. Articles by Zehr, E. P.