Modulations of interlimb and intralimb cutaneous reflexes during simultaneous arm and leg cycling in humans

Clinical Neurophysiology 117 (2006) 1301–1311 www.elsevier.com/locate/clinph

Modulations of interlimb and intralimb cutaneous reflexes during simultaneous arm and leg cycling in humans Masanori Sakamoto a, Takashi Endoh a, Tsuyoshi Nakajima a, Toshiki Tazoe a, Shinichiro Shiozawa b, Tomoyoshi Komiyama a,b,* a

Division of Health and Sport Education, United of Graduate School of Education, Tokyo Gakugei University, 1-33 Yayoi-cho, Inage-Ku, Chiba City 263-8522, Japan b Department of Health and Sports Sciences, Faculty of Education, Chiba University, 1-33 Yayoi-cho, Inage-Ku, Chiba City 263-8522, Japan Accepted 5 March 2006 Available online 2 May 2006

Abstract Objective: We investigated to what extent intralimb and interlimb cutaneous reflexes are altered while simultaneously performing arm and leg cycling (AL cycling) under different kinematic and postural conditions. Methods: Eleven subjects performed AL cycling under conditions in which the arm and leg crank ipsilateral to the stimulation side were moved synchronously (in-phase cycling) or asynchronously (anti-phase cycling) while sitting or standing. Cutaneous reflexes following superficial radial or superficial peroneal nerve stimulation (2.0–2.5 times radiating threshold, 5 pulses at 333 Hz) were recorded at 4 different pedal positions from 12 muscles in the upper and lower limbs. Cutaneous reflexes with a peak latency of 80–120 ms were then analyzed. Results: The magnitude of interlimb and intralimb cutaneous reflexes in the arm and leg muscles was significantly modulated depending on the crank position for the relevant limb (phase-dependent modulation). A significant correlation between the magnitude of the cutaneous reflex and background EMG was observed in the majority of muscles during static contraction, but not during AL cycling (task-dependent modulation). No significant difference was found in comparisons of the magnitude of intralimb and interlimb cutaneous reflexes obtained during in- and anti-phase AL cycling. Qualitatively, the same results were obtained during AL cycling while sitting or standing. In addition, the modulation of cutaneous reflexes in arm muscles was identical among in-phase, anti-phase and isolated arm cycling. Results were the same for leg muscles. Conclusions: Cutaneous reflexes in arm muscles are little influenced by rhythmic movement of the legs and vice versa during AL cycling. It is likely that neural components that control interlimb reflexes are loosely coupled during AL cycling while sitting or standing. Significance: Our results provide a better understanding of the coordination between the upper and lower limbs during rhythmic movement. q 2006 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. Keywords: Interlimb coordination; Cutaneous reflex; Cycling; Human

1. Introduction Around a century ago, original physiological evidence indicating the possible contribution of neural connections subserving interlimb and intralimb reflexes to locomotion was provided by Sir Charles Sherrington (Sherrington, 1910; Sherrington and Laslett, 1903). Later, it was shown in

* Corresponding author. Tel./fax: C81 43 290 2621. E-mail address: [email protected] (T. Komiyama).

the cat that stimulation of the cutaneous nerve produces general flexor excitation with the exception of muscles directly beneath the stimulated region of the skin (local sign, Hagbarth, 1952), and that contact between the dorsum of the hindlimb paw and an obstacle during the swing phase of locomotion evokes coordinated ankle flexion and knee extension in an attempt to avoid tripping (Forssberg et al., 1975). The latter reflex pattern has been termed the ‘stumbling corrective reaction’ and also occurs during human walking (Van Wezel, 1997; Zehr et al., 1998). To date, intralimb cuntaneous reflexes were demonstrated to

1388-2457/$30.00 q 2006 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.clinph.2006.03.005

1302

M. Sakamoto et al. / Clinical Neurophysiology 117 (2006) 1301–1311

have task dependency (phasic locomotor vs. tonic maintained activity), intensity dependency (noxious vs. nonnoxious stimulation), load dependency, phase dependency (swing vs. stance), and laterality dependency (ipsilateral vs. contralateral effects) (review see, Dietz, 2002; Dietz and Duysens, 2000; Zehr and Duysens, 2004). Stimulation of the skin was shown to produce a widespread reflex effect in various motoneuron pools (for review, see Dietz, 2002; Zehr and Duysens, 2004). For example, coordinated muscle activation in intra- and intersegments following stimulation of the skin during on-going locomotor activities was shown in the cat (Miller et al., 1975; 1977; Schomburg et al., 1977). Stimulation of cutaneous nerves in the human median and sural nerves also evokes reflex changes in the upper and lower limbs (Delwaide and Crenna, 1984; Kearney and Chan, 1979; 1981). Recent advances in the study of interlimb reflexes in humans have provided ample evidence that they are crucial for generating rapid compensatory movements in remote limbs in response to a sudden disturbance of limb movement during various locomotor activities (Dietz, 2002; Dietz et al., 2001; Duysens and Tax, 1994; Haridas and Zehr, 2003; Wainner et al., 2001; Zehr et al., 2003; 2001a; Zehr and Duysens, 2004). The magnitude of interlimb reflexes during rhythmic movement shows strong phase modulation and task-dependent modulation compared to static movement, suggesting the possible contribution of a central pattern generator (CPG, for review, see Dietz, 2002; Zehr and Duysens, 2004). Rhythmic bipedal cycling is a powerful tool for investigating human locomotor systems (Brooke et al., 1997; Zehr et al., 2001b; Zehr and Duysens, 2004). In leg muscles, cutaneous reflexes following sural nerve stimulation are modulated in a phase-dependent manner depending on the biomechanical function of the limb muscles during cycling (Mileva et al., 2004). Also, the Hoffmann (H-) reflex and cutaneous reflexes evoked in forearm muscles are dependent on the movement phase during rhythmic arm movement (Zehr and Kido, 2001; Zehr et al., 2003). Furthermore, the size of the reflexes is proportional to the background EMG (B.EMG) during the stationary condition but not during cycling (Zehr and Kido, 2001; Zehr et al., 2001a, b). These results suggest a contribution of CPGs to the regulation of locomotor activity not only in the lower limbs but also in the upper limbs. Thus, rhythmic bipedal cycling can be used as a model for investigating the function of a CPG in humans. The presence of interlimb reflexes was shown in spinal cord injured subjects (Calancie et al., 1996; 2002), indicating that propriospinal pathways coupling the cervical and lumbosacral enlargements of the spinal cord contribute to interlimb coordination (Dietz et al., 2001; Frigon et al., 2004; Haridas and Zehr, 2003; Wannier et al., 2001; Zehr et al., 2001a). We hypothesized that if there exists a strong neural link between the upper and lower limbs, interlimb and intralimb cutaneous reflexes would show phase- and

task-dependent modulations when human subjects simultaneously performed arm and leg cycling (AL cycling). In addition, cutaneous reflexes are differently modified during AL cycling from that during arm or leg cycling alone and during two types of AL cycling in which the upper and lower limbs are rotated synchronously (in-phase) or asynchronously (anti-phase). Furthermore, it is possible that cutaneous reflexes are modulated if AL cycling is performed under the different postural conditions of sitting and standing because the magnitude of cutaneous reflexes is known to strongly depend on the motor task (Burke et al., 1991; Do et al., 1990; Duysens et al., 1990; Komiyama et al., 2000; Thoumie and Do, 1996). Therefore, in the present study, interlimb and intralimb cutaneous reflexes in the upper and lower limb muscles following nonnoxious electrical stimulation of superficial radial (SR) and superficial peroneal (SP) nerves were examined during inphase and anti-phase AL cycling in sitting and standing positions.

2. Methods 2.1. Subjects Eleven male volunteers aged 22–29 years participated in this study. All subjects gave their informed consent according to the Declaration of Helsinki before participating in the experimental procedures. No subject had a neurological deficit nor had been involved in any resistance training programs. This study was approved by the local ethics committee. 2.2. Nerve stimulation Electrical pulses were delivered from a constant current stimulator (SS-100, Nihon Kohden, Tokyo, Japan) controlled by a pulse generating system (SEN7201, Nihon Kohden). Cutaneous reflexes were evoked by applying electrical stimulation via rectangular pulses 1 ms in duration to the right SR and SP nerves, respectively. Both nerves were stimulated with trains of 5 pulses at 333 Hz. Ag/Ag–Cl disk electrodes (f 1 cm, NE-101, Nihon Kohden) for SR nerve stimulation were placed on the dorsal surface of the right forearm just proximal to the radial head. The same electrodes were used to stimulate the SP nerve and were placed on the anterior surface of the right leg just near the crease of the ankle joint. The threshold of stimulation (radiating threshold; RT) was defined as the stimulus voltage at which the subject reported just being able to distinctly feel radiating paresthesia occurring along the dorsal surface of the hand towards the index finger and thumb (SR nerve) and dorsal surface of the foot towards the first toe (SP nerve). The RT was determined when the crank position for both the arm and leg was located at 12 o’clock. The stimulation intensities were set at 2.0w2.5!RT and

M. Sakamoto et al. / Clinical Neurophysiology 117 (2006) 1301–1311

did not produce pain in any subject. The RT was rechecked between trials throughout the experiment to confirm consistency of the electrical simulation. 2.3. Electromyography (EMG) EMG signals were recorded from the medial gastrocnemius (MG), tibialis anterior (TA), anterior deltoid (AD) and posterior deltoid (PD) bilaterally and vastus lateralis (VL), biceps femoris (BF), biceps brachii (BB) and triceps brachii (TB) ipsilateral to the site of nerve stimulation with Ag/Ag–Cl electrodes (f 1 cm, NE-101, Nihon Kohden). EMG electrodes were placed longitudinally 3 cm apart over each muscle and fixed by surgical tape. EMG signals were amplified (!2000), band-pass filtered at 50–300 Hz (Model 1206, NEC Sanei, Tokyo, Japan) and then digitally fullwave rectified. All signals were converted into digital data via an A/D converter system at a sampling rate of 1 kHz for later analysis (CED 1401 interface with Spike2 software, CED, Cambridge, UK) 2.4. Task and stimulus procedures 2.4.1. Ergometer settings The subjects performed AL cycling while sitting or standing. There was no coupling between the ergometers for the upper (Matsushita, EU6210) and lower limbs (COMBI, Power Max V); therefore the subjects could independently move both ergometers. The length of the ergometer pedal crank for the upper and lower limbs was 15 cm. The arm ergometer was positioned in front of the subject and the height of the axis of rotation of the arm ergometer was set at shoulder level to cause the elbow to become semiflexed (w308) at 3 o’clock (see Fig. 1) The axis of rotation of the ergometer for the legs was located 30 cm back from that of the ergometer for the arms. The location of the leg ergometer was adjusted so that the knee would be semiflexed (w308) at 6 o’clock. Eight of 11 subjects performed more than one session on different days of AL cycling while standing. For AL cycling while standing, subjects were made to lean forward slightly, and the height of the axis of rotation of the arm ergometer was set at the chest. The axis of rotation of the leg ergometer was located about 60 cm back from that of the arm ergometer. All reference positions during the movement cycle were defined with respect to clock positions. Cutaneous reflexes were evoked at 4 different positions: 12, 3, 6 and 9 o’clock. For example, ‘top dead center’ was defined as 12 o’clock and 3 and 9 o’clock were the locations at which the subject’s right elbow was at extended and flexed positions, respectively. Electrical stimulation was delivered at 1 of 4 positions once every 3 or 4 cycles. Stimulation was triggered by a TTL pulse generating from a photocell (PS-102, Cocoreserach, Co., LTD, Tokyo) that detected the position of the reflecting plate attached to 1 of 4

1303

predetermined positions on the wheel gear for the lower limb ergometer. 2.4.2. Motor tasks The subjects were asked to rhythmically rotate their upper and lower limbs with the isolated ergometers at 45 rpm. AL cycling was performed as follows: Experiment 1. Upper and lower limbs were rotated synchronously (in-phase). Therefore, arm and leg cranks ipsilateral to the stimulation sites were moved in a similar manner. Experiment 2. Upper and lower limbs were rotated asynchronously (anti-phase). Therefore, the arm and leg cranks ipsilateral to the stimulation sites were moved in an opposite manner (1808 out of phase). The subjects underwent a training session in performing in- and anti-phase AL cycling in a sitting or standing posture before the test trials. The training session consisted of 8 sets for 2 min with a 5 min rest period between each of the 4 different types of AL cycling. In the test session, the subjects performed 16 trials of 4 different stimulus positions!2 nerves!2 motor tasks (in- and anti-phase)!1 postural condition (sitting or standing). On another day at least 7 days later, subjects underwent the same test session but with a different postural condition. Electrical stimulations were delivered 20 times at 1 of 4 points in each trial. The position of the stimulation was determined randomly beforehand. The subjects performed approximately 80 revolutions per trial to record a stimulated EMG and nonstimulated control EMG. Each trial lasted around w100 s. The order of the trial was randomized across the subjects. 2.4.3. Crank position-dependent changes in cutaneous reflexes of iAD and iTA during static contraction For 5 subjects, cutaneous reflexes were elicited from iAD or iTA by stimulation of the SR or SP nerve, respectively, to determine whether the magnitude of the intralimb cutaneous reflex depended on crank position. The subjects first performed maximum voluntary contraction (MVC) of iAD and iTA to record the maximal EMG. To perform the MVC of iAD (shoulder flexion) and iTA (ankle dorsiflexion), the subjects used an immobile iron frame that was constructed to fix the arm ergometer. For iTA, the subjects performed the MVC while standing. Then the subjects performed static contraction of iAD and iTA at 4 different crank positions (12, 3, 6 and 9 o’clock) with the arm and lower leg ergometers. The cranks and pedals of the ergometers were immobilized by the experimenters to induce isometric contraction. The level of EMG during static contraction was set at 20% of the maximal EMG. The rectified and averaged EMG signal (low pass filtered at 100 Hz) from the target muscle was displayed on an analog voltmeter (AX-313TR, Sanwa, Tokyo, Japan) to provide visual feedback.

1304

M. Sakamoto et al. / Clinical Neurophysiology 117 (2006) 1301–1311

Fig. 1. Typical recordings obtained from a single subject of changes in EMG activity and elbow and knee angles during in- (thick lines) and anti-phase (thin lines) cycling while sitting (left) and standing (right). EMG data are shown as full-wave rectified and averaged signals. EMG traces of the upper and lower limb muscles were aligned based on the right arm. EMG traces of the lower limb muscles lagged by 1808 between in- and anti-phase cycling. Vertical calibrations of the EMG indicate 5% of EMG max in each muscle obtained during static contraction. Displacement of the elbow and knee joint angles is shown as degrees (full extension position, 08).

2.4.4. Activity-dependent changes in gain of cutaneous reflexes during static contraction and AL cycling To determine whether the magnitude of cutaneous reflexes was proportional to that of B.EMG, intralimb and interlimb cutaneous reflexes were evoked while the subjects performed graded isometric contractions of the upper and lower limb muscles. These recordings were conducted after subjects completed the AL cycling test session. The subjects performed MVC to record the maximal EMG in all muscles. To achieve the MVC in arm muscles, the subjects used an immobile iron frame that was constructed to fix the arm

ergometer. For the lower leg muscles, the subjects performed the MVC while standing. In a preliminary study, we confirmed that the magnitude of EMG values in all muscles during AL cycling while sitting was below 50% of the maximal EMG. Therefore, subjects were asked to perform from 4 to 5 grades of static contraction (2w50% of the maximal EMG) with visual feedback as mentioned above. 2.4.5. Isolated arm or leg cycling To compare reflex modulations between in- and antiphase cycling in greater detail, 8 subjects performed isolated

M. Sakamoto et al. / Clinical Neurophysiology 117 (2006) 1301–1311

arm or leg cycling while standing. Cutaneous reflexes were recorded from arm muscles by stimulating the SR nerve during isolated arm cycling while standing. Similarly, in the standing position, cutaneous reflexes were recorded from leg muscles by stimulating the SP nerve during isolated leg cycling. 2.5. Data analysis The EMG was full-wave rectified and smoothed by a moving average technique with 10 ms (10 points) bins, in which the averaged EMG data at time t with time resolution of 1 ms was the average value of the original EMG data points for tG5 ms. We set the time window at 230 ms with 30 ms pre-stimulus and 200 ms post-stimulus time intervals. The electrical stimulation was applied 20 times, and thus EMG signals were averaged 20 times with respect to the time at which the electrical stimuli were given. The control EMG obtained during pedaling without electrical stimulation was then subtracted from the stimulated EMG to obtain ’pure’ reflex responses. The EMG pattern used in the subtraction procedure was obtained from averaged EMG data for 50 cycles without electrical stimulation. The size of the cutaneous reflexes was determined by measuring the base-to-peak amplitude within the pre-set time window from 80 to 120 ms after electrical stimulation. A reflex component was considered present if its peak amplitudes rose above or fell below a 3-SD of the prestimulus EMG variation, calculated from the subtracted EMG, for at least 5 ms in at least 1 of 4 crank positions. The means and SDs. of the B.EMG for all conditions were calculated from data obtained K30 to 0 ms before electrical stimulation. Then, to investigate the modulation of the cutaneous reflex in more detail, each response was plotted against the B.EMG. The amplitude of the cutaneous reflexes was expressed as a percentage of the maximal EMG that was obtained during static contraction (see Section 2.4.4 Static muscle contraction). Changes in elbow and knee angles were measured by hand-made goniometers. To estimate kinematic correlations of elbow and knee positions during in- and anti-phase AL cycling, the waveform correlation was applied. Spike2 software (CED, UK) was used to calculate these statistics. In-phase (or anti-phase) AL cycling was correctly performed if the correlation co-efficient was significant and the time-lag was around 08(or 1808). 2.6. Statistics For the modulation of cutaneous reflexes across the pedal cycle, one-way repeated measures analysis of variance (ANOVA) was performed. To investigate whether the reflex responses were significantly facilitated or inhibited from zero, multiple comparisons were conducted using Dunnett’s test. For the comparison of differences in cutaneous reflex amplitudes between in- and anti-phase cycling, two-way

1305

ANOVA with repeated measures was performed for the stimulation phases (12, 3, 6, and 9 o’clock) and tasks (inand anti-phase cycling). For intralimb cutaneous reflexes, two-way ANOVA with repeated measures was performed for the stimulation phases and tasks of in-phase, anti-phase and only arm or leg cycling. The significance of the F values was obtained after Greenhouse–Geisser correction when appropriate, and then a correction coefficient epsilon was determined. Linear regression analysis was used to calculate the correlation coefficient between the magnitude of the cutaneous reflexes and B.EMG during static contraction and AL cycling. Significance was set at P!0.05. Data were expressed as the meanGstandard deviation (SD).

3. Results 3.1. General EMG activity and changes in joint angles during in- and anti-phase AL cycling while sitting and standing Fig. 1 shows schematic illustrations of the motor tasks and typical EMG and joint angle recordings during in-phase (thick lines) and anti-phase (thin lines) cycling while sitting (left panels) and standing (right panels) for a single subject. Note that the traces were arranged with respect to the arm crank position so that the EMG and joint angle patterns in the leg muscles showed a 1808 phase lag between in- and anti-phase cycling. Changes in the in-phase and anti-phase EMG patterns were the same for sitting and standing except for the iAD and iBB. In the iAD, muscle activation was seen at 12 o’clock when sitting but not while standing. In the iBB, muscle activation occurred between 6 and 9 o’clock while sitting, but was also seen except at 6 o’clock while standing. For lower limb muscles, although there were no obvious differences in the EMG pattern between sitting and standing, the activation levels of lower limb muscles were much higher while standing than while sitting as reflected in the differences in the calibration scales. 3.2. Kinematic correlations of the elbow and knee joints during in- and anti-phase AL cycling Waveform correlation analyses between the angles of the knee and elbow joints showed highly significant correlation coefficients for all 4 motor tasks (0.93w0.97, all P!0.001). The time lag of the peak of the correlation coefficient during in- and anti-phase cycling while sitting appeared at 0.07G 0.08 s (18.9G21.68 out of phase) and 0.52G0.09 s (140.4G 24.38), respectively. Furthermore, those values during inand anti-phase while standing were 0.03G0.02 s (8.1G 5.48) and 0.60G0.07 s (162.0G18.98), respectively. Cadence variability was found to be quite small with a range of 45.6G1.5 rpm to 46.1G0.2 rpm (mean; 45.7G

1306

M. Sakamoto et al. / Clinical Neurophysiology 117 (2006) 1301–1311

1.3 rpm), and there was no significant difference among the 4 motor tasks (F(3,34)Z0.78, PO0.05). 3.3. Phase-dependent modulation of interlimb and intralimb reflexes during AL cycling Fig. 2 depicts typical recordings obtained from a single subject of EMG responses in the iTA (left) and iPD (right) following SP nerve stimulation during in-phase cycling while standing. In the iTA, a large inhibitory intralimb reflex was observed at 12 o’clock. For iPD, in general, the amplitude of the interlimb reflex was smaller than that of the intralimb reflex, but a prominent excitatory response was seen at 3 o’clock. Fig. 3A illustrates group means (GSD) of the amplitude of cutaneous reflexes following SR nerve stimulation during the in- (upper panels) and anti-phase (lower panels) in arm and leg muscles during AL cycling while standing. Note that the abscissa indicates the movement phase for the right arm so that leg movement during anti-phase cycling was about 1608 out of phase compared to that during in-phase. Thus, for the iMG and iBF during anti-phase cycling (lower panels), the bars at 12 o’clock indicate that the right arm and leg were located at around 12 and 6 o’clock, respectively. In this figure, selected muscles that showed a prominent phasedependent modulation are illustrated. During in-phase cycling (upper panels), a significant main effect for the crank position on the magnitude of the cutaneous reflex was seen in the iMG (F(1.50,10.45)Z6.65, P!0.05), iTA

(F(1.33,9.32)Z3.17, P!0.05) and iAD (F(1.42,9.92)Z11.76, P!0.01). The same was true during anti-phase cycling (lower panels) for the iMG (F(1.20,7.19)Z9.18, P!0.05), iBF (F(3,18)Z4.27, P!0.01) and iAD (F(3,18)Z11.78, P! 0.001). Significant phase-dependent modulations following SR nerve stimulation were found in all muscles except for the iBF during in-phase cycling and in all muscles except for the iTA, iVL, iPD and cMG during anti-phase cycling while standing. For iMG, iTA and iVL, interlimb and intralimb cutaneous reflexes following SP nerve stimulation also showed prominent phase-dependent modulation during inand anti-phase AL cycling (Fig. 3B) (all P!0.05). However, for the iPD during in- and anti-phase cycling, there was no significant phase-dependent modulation. The reflex amplitudes at 3 and 12 o’clock for in- and anti-phase cycling, respectively, were significantly increased compared to the unstimulated EMG value. A significant phase-dependent modulation of the cutaneous reflexes following SP nerve stimulation was found in 12 muscles, with the exceptions of the iAD, iBB, iTB and cPD during inphase cycling and the iMG, iTA, iVL, iPD and cMG during anti-phase cycling. Qualitatively identical results were obtained during in- and anti-phase cycling while sitting (not illustrated). 3.4. Effects of crank positions on the magnitude of intralimb cutaneous reflex In 5 subjects, the effects of crank positions at 12, 3, 6 and 9 o’clock on the magnitude of the intralimb cutaneous reflex were examined in the iTA following SP nerve stimulation and the iAD following SR nerve stimulation while performing static contraction at each point with an identical level of B.EMG (not illustrated). There were no significant main effects from the crank position on the reflex amplitudes in the iAD (F(3,12)Z1.46, PO0.05) and iTA (F(3,12)Z1.59, PO0.05). 3.5. Relationship between reflex amplitude and B.EMG

Fig. 2. Typical recordings obtained from a single subject of cutaneous reflexes for 4 different cycling phases in the iTA (left panels) and iPD (right panels) evoked by stimulating the SP nerve during in-phase cycling while standing. At 12 o’clock, a large intralimb cutaneous reflex was seen in the iTA. In the iPD, an excitatory reflex was seen at 3 o’clock. The shaded areas indicate from 80 to 120 ms after nerve stimulation.

Table 1 summarizes results of linear regression analysis between the magnitude of cutaneous reflexes and that of B.EMG during isometric contraction and AL cycling. For SR nerve stimulation, significant correlation coefficients were found in 10 of 12 muscles tested during static contraction. In contrast, significant correlations were found only in 5 and 3 of 12 muscles tested during antiphase cycling while sitting and standing, respectively. For SP nerve stimulation, significant correlation coefficients were found in 8 of 12 muscles tested during static contraction. However, conversely, significant correlations were found in 3 and 6 of 12 muscles tested during anti-phase cycling while sitting and standing, respectively. For inphase AL cycling, qualitatively the same results were obtained after SR and SP nerve stimulation.

M. Sakamoto et al. / Clinical Neurophysiology 117 (2006) 1301–1311

1307

Fig. 3. Grand means (and SD) of cutaneous reflexes elicited by stimulating the SR (A) and SP (B) nerves during in-phase (upper panels) and anti-phase (lower panels) cycling while standing. Selected muscles that showed a prominent phase-dependent modulation in response to SR or SP nerve stimulation are illustrated. Abscissa shows 4 different stimulation phases for right arm movement (A) and right leg movement (B). Asterisks indicate a significant difference from zero. *P!0.05, **P!0.01, ***P!0.001.

3.6. Comparison of in- and anti-phase cycling Fig. 4 illustrates comparisons of interlimb reflexes in leg muscles exhibiting prominent reflex modulation following SR nerve stimulation during in- (black bars) and anti-phase (white bars) cycling in sitting (A) and standing (B) positions. While sitting, there was a significant interaction between the crank position and tasks (in- vs. anti-phase cycling) only in the iMG (F(3,30)Z6.11, P!0.01). In addition, no significant interaction between these factors was found in the other muscles tested. While standing, although there were significant main effects from crank position on reflex amplitudes (iMG: F(3,21)Z9.25, P! 0.001; iTA: F(3,21)Z5.85, P!0.01; iVL: F(3,21)Z6.87, P!

0.01; cTA: F(3,21)Z25.91, P!0.001), there were no significant interactions between crank position and tasks during in- and anti-phase cycling in any of the muscles tested (iMG: F(3,21)Z0.98, PO0.05; iTA: F(3,21)Z0.35, PO0.05; iVL: F(3,21)Z0.75, PO0.05; cTA: F(3,21)Z0.59, PO0.05). It is, however, noteworthy that some inhibitory responses while sitting changed into facilitatory ones while standing, such as for the iMG and iVL at 12 and 3 o’clock and the iTA at 9 o’clock. Fig. 5 illustrates comparisons of intralimb reflexes in leg muscles obtained by stimulating the SP nerve. Similar to interlimb reflexes (Fig. 4), no significant interactions between crank positions and tasks during in- and antiphase cycling were noted. However, in the iMG the

1308

M. Sakamoto et al. / Clinical Neurophysiology 117 (2006) 1301–1311

Table 1 Correlation coefficients between the reflex amplitude and background EMG during static contraction, in- and anti-phase cycling while sitting and standing after SR and SP nerve stimulation Muscle

SR nerve

SP nerve In-phase

iMG iTA iVL iBF iAD iPD iBB iTB cMG cTA cAD cPD

Anti-phase

In-phase

Anti-phase

Static

Sitting

Standing

Sitting

Standing

Static

Sitting

Standing

Sitting

Standing

0.35*** 0.14* NS 0.65*** 0.19** 0.32*** 0.79*** 0.15** 0.63*** 0.73*** 0.20** NS

NS 0.53*** NS NS 0.47*** NS NS NS NS 0.43*** 0.43*** NS

NS NS 0.26** NS 0.40*** NS 0.18* 0.19* NS 0.32*** 0.52*** 0.31**

NS 0.21*** 0.40*** NS 0.50*** 0.48*** 0.24*** NS NS NS 0.45*** NS

NS NS NS 0.18* NS NS NS NS NS NS 0.13* 0.46***

0.21** 0.19** 0.52*** NS NS NS 0.46*** 0.17** 0.53*** 0.12* NS 0.19**

NS 0.33*** 0.19** NS NS 0.18** NS 0.25*** NS 0.21** NS NS

0.49*** 0.82*** NS 0.30** 0.21** 0.28** NS NS NS NS 0.53*** NS

NS 0.32*** 0.25*** NS NS NS NS NS 0.20** NS NS NS

0.52*** 0.64*** NS 0.20* NS NS NS 0.23** 0.33*** NS NS 0.23**

Significant correlation coefficients (r2) are indicated by asterisks (*P!0.05, **P!0.01, ***P!0.001; NS, no significant correlation).

Fig. 4. Grand means (and SD) of the magnitude of cutaneous reflexes following SR nerve stimulation for leg muscles averaged across all subjects during in- (black bars) and anti-phase (white bars) cycling while sitting (A) and standing (B). Abscissa shows crank position of the right leg. Arm movement was 1808 out of phase for anti-phase cycling with respect to inphase cycling.

Fig. 5. Grand means (and SD) of the magnitude of cutaneous reflexes following SP nerve stimulation for leg muscles averaged across all subjects during in- (black bars) and anti-phase (white bars) cycling while sitting (A) and standing (B). Abscissa shows crank position of the right leg. Arm movement was 1808 out of phase for anti-phase cycling with respect to inphase cycling.

M. Sakamoto et al. / Clinical Neurophysiology 117 (2006) 1301–1311

intralimb reflex at 6 o’clock showed facilitation while sitting, but changed to inhibition while standing even when the same crank position was maintained. Similar reflex reversal was observed at 12 o’clock in the iVL. For intralimb and interlimb reflexes in arm muscles, no significant interaction for crank position vs. task was seen during in- and anti-phase cycling while sitting and standing (not illustrated). 3.7. Cutaneous reflexes during isolated arm cycling while standing To compare the reflex modulations between in- and antiphase cycling in more detail, cutaneous reflexes were recorded in arm muscles by stimulating the SR nerve during isolated arm cycling while standing. In addition, cutaneous reflexes were recorded in leg muscles by stimulating the SP nerve during isolated leg cycling while standing. Fig. 6

Fig. 6. Typical recordings obtained from a single subject of cutaneous reflexes in the iMG evoked by stimulation of the SP nerve and changes in the EMG in the iMG and knee angle during in-phase (thin lines), anti-phase (gray lines) and isolated leg cycling (thick lines) while standing. Shaded area indicates from 80 to 120 ms after nerve stimulation. EMG data are shown as full-wave rectified and averaged signals. Flexion and extension are shown as downward and upward deflections, respectively, of the angular displacement in the trace.

1309

depicts typical cutaneous reflexes recorded from the iMG following SP nerve stimulation obtained from a single subject during in-phase (thin lines), anti-phase (gray lines) and isolated leg cycling (thick lines). Inhibitory responses were evident at 6 o’clock (see Fig. 5B). The waveforms and amplitudes of cutaneous reflexes in the iMG and kinematics in the ankle joint were almost the same for all tasks across the pedal cycle. There were no significant interactions between crank position and task in any arm muscles with SR nerve stimulation nor in the 6 leg muscles except for the iTA s for SP nerve stimulation.

4. Discussion 4.1. Methodological considerations The validity of our results substantially depends on the extent of how precisely the subjects performed the motor tasks. In the present study, subjects began the in- and antiphase AL cycling only after sufficient familiarization with motor tasks. This procedure would lead to significant correlation coefficients between changes in elbow and knee joint movement for all motor tasks. During in-phase AL cycling while sitting and standing, the time lags of the peak of the waveform correlation were 0.07 s (18.98 out of phase) and 0.03 s (8.18), respectively. These values indicate that the elbow and knee joints rotated in almost perfect synchrony. For anti-phase AL cycling while sitting and standing, the calculated time lags were 0.52 s (140.28) and 0.60 s (162.08), respectively. These values did not perfectly match the expected values (0.667 s and 1808), suggesting that anti-phase cycling is somewhat difficult because of its biomechanical demands. However, when comparing time lags between in- and anti-AL cycling, the difference was large enough to say that the elbow and knee joints were almost asynchronously rotated during anti-phase cycling. The subjects performed AL cycling while assuming two different postures: sitting and standing. These postural differences led to differences in the activation level of proximal and distal muscles (see Fig. 1). For example, the magnitude of iPD, iBF and iMG activity was substantially greater during AL cycling while standing than while sitting. These differences may be due to differences in biomechanical demands, especially differences in load, between the postures. The differences in EMG activities between the two postures may lead to a differential modulation pattern of cutaneous reflexes for both tasks as demonstrated in previous studies (Bastiaanse et al., 2000; Fouad et al., 2001; Li and Caldwell, 1998; Stephens and Yang, 1999). However, despite this, no significant difference was seen in the modulation pattern of intra-and interlimb cutaneous reflexes during in- and anti-phase AL cycling (see Figs. 4 and 5). These results suggest that the neural mechanisms responsible for patterning the intra- and interlimb cutaneous reflexes were less affected by phase differences of rhythmic

1310

M. Sakamoto et al. / Clinical Neurophysiology 117 (2006) 1301–1311

movement in the upper and lower limbs. Thus, we favor the explanation that the magnitudes of the intra- and interlimb reflexes in a muscle strongly depend on the timing at which the electrical stimulation is applied irrespective of the position or phase of the other limb. 4.2. Phase-dependent modulation of cutaneous reflexes during AL cycling We found that the magnitude of the cutaneous reflexes following stimulation of the same (intralimb reflexes) and different (interlimb reflexes) segmental nerve strongly depended on the crank position of the arm and leg (phasedependent modulation) during AL cycling. These results correspond well with the modulation of intralimb cutaneous reflexes during isolated arm cycling (Zehr and Kido, 2001), isolated leg cycling (Mileva et al., 2004) and walking (Duysens et al., 1990; Yang and Stein, 1990). Interlimb reflexes are modulated according to the step cycle during fictive locomotion in the cat (Schomburg et al., 1977) and during stepping in the decerebrate cat (Miller et al., 1977). In humans, interlimb cutaneous reflexes with an 80–120 ms latency in arm and leg muscles evoked by stimulation of the hand (SR) and foot (SP) are modulated in a phase-dependent manner while walking (Haridas and Zehr, 2003). In the present study, we found that reflex modulation patterns were slightly different during AL cycling while sitting and standing. For example, an excitatory response in the iMG at 6 o’clock following SP nerve stimulation (see Fig. 5) during AL cycling while sitting was reversed to an inhibitory response while standing. Differences in the excitability of the homonymous and heteronymous muscles could at least in part account for the reflex reversal in the iMG, because a substantial difference in the amplitude of the iMG and activation pattern of the upper thigh muscles was notable during the two tasks (see Fig. 1). 4.3. Mechanisms underlying phase- and task-dependent modulation In the current study, we found that in the majority of muscles a significant linear correlation was observed during isometric contraction but not during AL cycling while sitting or standing (Table 1). These results suggest that the phase-dependent modulation of cutaneous reflexes observed during AL cycling was not simply due to a change in motoneuron excitability. This indicates that a contribution of CPGs to the phase-dependent modulation of interlimb and intralimb cutaneous reflexes during AL cycling may exist as seen during walking and cycling in humans (Yang and Stein, 1990; Zehr and Kido, 2001; Zehr et al., 2001b). The activity of the CPGs is thought to play an important role in the phase- and task-dependent modulation of cutaneous reflexes via premotoneuronal gating of afferent

feedback while walking (Duysens and Tax, 1994; Duysens and Van de Crommert, 1998; Haridas and Zehr, 2003; Van Wezel et al., 1997). During isolated arm or leg cycling, cutaneous reflexes show phase- and task-dependent modulation, suggesting that the modulation of cutaneous reflexes also at least in part results from activity of the CPGs during arm or leg cycling as well as during walking (Mileva et al., 2004; Zehr and Kido, 2001; Zehr et al., 1998). Although the possibility that a transcortical pathway contributes to cutaneous reflexes cannot be fully denied (Christensen et al., 1999; 2000), it seems likely that the phase-dependent reflex modulations observed during AL cycling are a manifestation of CPG activity. 4.4. Coupling of CPGs between upper and lower limbs during AL cycling Active and passive cycling of the contralateral leg dramatically inhibits the soleus H-reflex in the stationary leg (Cheng et al., 1998; Collins et al., 1993; McIlroy et al., 1992), suggesting that there exists a neural link between the right and left legs during cycling and that the reflexes evoked in one limb are influenced by the movement of the other. We hypothesized that if there is strong coupling in CPGs between the upper and lower limbs, then cutaneous reflexes in the arm muscles can be influenced by rhythmic movement of the legs and vice versa. However, we found that modulation patterns of interlimb and intra limb cutaneous reflexes were similar during in- or anti-phase cycling while sitting or standing. In addition, there were no significant differences in the magnitude of cutaneous reflexes among in-phase, anti-phase cycling and isolated arm or leg cycling while standing (Fig. 6). Thus, the magnitude of cutaneous reflexes during AL cycling was largely dependent on the combination of the position and activity of the muscle in which the reflex is evoked. Carroll et al. (2005) recently reported that the modulation of cutaneous reflexes evoked in a particular muscle during arm cycling is determined by the functional state of the limb in which the muscle resides irrespective of whether the stimulation is given ipsilateral or contralateral to the muscle. This implies a loose connection between the CPGs for both sides of the arm that regulates muscle activity and the reflex amplitude during rhythmic movement. Taking this into consideration, the modulation of cutaneous reflexes during AL cycling is most probably determined by CPGs that regulate the activity of a particular limb irrespective of whether the movement of other limbs is in-phase or 1808 out of phase. It is likely that the mutual interaction of CPGs between the upper and lower limbs is weak during AL cycling and that just rhythmically or simultaneously moving the arms and legs is insufficient for setting up a tight coupling of CPGs between them (Dietz, 2002).

M. Sakamoto et al. / Clinical Neurophysiology 117 (2006) 1301–1311

References Bastiaanse CM, Duysens J, Dietz V. Modulation of cutaneous reflexes by load receptor input during human walking. Exp Brain Res 2000;135: 189–98. Brooke JD, Cheng J, Collins DF, McIlroy WE, Misiaszek JE, Staines WR. Sensori-sensory afferent conditioning with leg movement: gain control in spinal reflex and ascending paths. Prog Neurobiol 1997;51:393–421. Burke D, Dickson HG, Skuse NF. Task-dependent changes in the response to low-threshold cautaneous afferent volleys in the human lower limb. J Physiol 1991;432:445–58. Calancie B, Lutton S, Broton JG. Central nervous system plasticity after spinal cord injury in man: interlimb reflexes and the influence of cutaneous stimulation. Electroencephalogr Clin Neurophysiol 1996; 101:304–15. Calancie B, Lavoie BA, Broton JG. Interlimb reflexes and synaptic plasticity become evident month after human spinal cord injury. Brain 2002;125:1150–61. 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 2005;161:133–44. Cheng J, Brooke JD, Misiazek JE, Staines WR. Crossed inhibition of the soleus H reflex during passive pedaling movement. Brain Res 1998; 779:280–4. Christensen LOD, Morita H, Petersen N, Nielsen J. Evidence suggesting that a transcortical reflex pathway contributes to cutaneous refleses in the tibialis anterior muscle during walking in man. Exp Brain Res 1999; 124:59–68. Christensen LOD, Petersen N, Andersen JB, Sinkjær T, Nielsen JB. Evidence for transcortical reflex pathways in the lower limb of man. Prog Neurobiol 2000;62:251–72. Collins DF, McIlroy WE, Brooke JE. Contralateral inhibition of soleus H reflexes with different velocities of passive movement of the opposite leg. Brain Res 1993;603:96–101. Dietz V. Do human bipeds use quadrupedal coordination? Trends Neurosci 2002;25:462–7. Dietz V, Duysens J. Significance of load receptor input during locomotion: a review. Gait Posture 2000;11:102–10. Dietz V, Fouad K, Bastiaanse CM. Neuronal coordination of arm and leg movements during human locomotion. Eur J Neurosci 2001;14: 1906–14. Do MC, Bussel B, Breniere Y. Influence of plantar cutaneous afferents on early compensatory reactions to forward fall. Exp Brain Res 1990;79: 319–24. Delwaide PJ, Crenna P. Cutaneous nerve stimulation and motoneuronal excitability: II. Evidence for non-segmental influences. J Neurol Neurosurg Psychiatry 1984;47:190–6. Duysens J, Tax T. Interlimb reflexes during gait in cat and human. In: Swinnen SP, Massion J, Heuer H, Caesar P, editors. Interlimb coordination: neural, dynamical, and cognitive constraints. San Diego, CA: Academic Press; 1994. p. 97–126. Duysens J, Van de Crommert HW. Neural control of locomotion; part 1: the central pattern generator from cats to humans. Gait Posture 1998;7: 131–41. Duysens J, Trippel M, Horstmann GA, Dietz V. Gating and reversal of reflexes in ankle muscles during human walking. Exp Brain Res 1990; 82:351–8. Forssberg HS, Grillner S, Rossignol S. Phase dependent reflex reversal during walking in chronic spinal cats. Brain Res 1975;85:103–7. Fouad K, Bastiaanse CM, Dietz V. Reflex adaptations during treadmill walking with increased body load. Exp Brain Res 2001;137:133–40. 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 2004;91:1516–23.

1311

Hagbarth KE. Excitatory and inhibitory skin areas for flexor and extensor motoneurones. Acta Physiol Scand Suppl 1952;94:1–58. Haridas C, Zehr EP. Coordination interlimb compensatory responses to electrical stimulation of cutaneous nerves in the hand and foot during walking. J Neurophysiol 2003;90:2850–61. Kearney RE, Chan CW. Reflex response of human arm muscles to cutaneous stimulation of the foot. Brain Res 1979;170:214–7. Kearney RE, Chan CW. Interlimb reflexes evoked in human arm muscles by ankle displacement. Electroencephalogr Clin Neurophysiol 1981;52: 65–71. Komiyama T, Zehr EP, Stein RB. Absence of nerve specificity in human cutaneous reflexes during standing. Exp Brain Res 2000;133:267–72. Li L, Caldwell GE. Muscle coordination in cycling: effect of surface incline and posture. J Appl Physiol 1998;85:927–34. McIlroy WE, Collins DF, Brooke JD. Movement features and H-reflex modulation. II. Passive rotation, movement velocity and single leg movement. Brain Res 1992;582:85–93. Mileva K, Green DA, Turner DL. Neuromuscular and biomechanical coupling in human cycling: modulation of cutaneous reflex responses to sural nerve stimulation. Exp Brain Res 2004;158:450–64. 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 1975;91:217–37. Miller S, Ruit JB, Van der Meche´ FGA. Reversal of sign long spinal reflexes dependent on the phase of the step cycle in the high decerebrate cat. Brain Res 1977;128:447–59. Schomburg ED, Roesler J, Meinck HM. Phase-dependent transmission in the excitatory propriospinal reflex pathway from forelimb afferents to lumbar motoneurones during fictive locomotion. Neurosci Lett 1977;4: 249–52. Sherrington SC. Flexion-reflex of the limb, crossed extension reflex, and reflex stepping and standing. J Physiol 1910;40:28–121. Sherrington SC, Laslett EE. Observations on some spinal reflexes and the interconnection of spinal segments. J Physiol 1903;29:58–96. Stephens MJ, Yang JF. Loading during stance phase of walking in humans increases the extensor EMG amplitude but does not change the duration of the step cycle. Exp Brain Res 1999;124:363–70. Thoumie P, Do MC. Changes in motor activity and biomechanics during balance recovery following cutaneous and muscular deafferentation. Exp Brain Res 1996;110:289–97. Van Wezel BM, Ottenhoff FA, Duysens J. Dynamic control of locationspecific information in tactile cutaneous reflexes from the foot during human walking. J Neurosci 1997;15:3804–14. Wannier T, Bastiaanse C, Colombo G, Dietz V. Arm to leg coordination in humans during walking, creeping and swimming activities. Exp Brain Res 2001;141:375–9. Yang JF, Stein RB. Phase-dependent reflex reversal in human leg muscles during walking. J Neurophysiol 1990;63:1109–17. Zehr EP, Duysens J. Regulation of arm and leg movement during human locomotion. Neuroscientist 2004;10:347–61. Zehr EP, Kido A. Neural control of rhythmic, cyclical human arm movement: task dependency, nerve specificity and phase modulation of cutaneous reflexes. J Physiol 2001;537:1033–45. Zehr EP, Stein RB, Komiyama T. Function of sural nerve reflexes during human walking. J Physiol 1998;507:305–14. Zehr EP, Collins DF, Chua R. Human interlimb reflexes evoked by electrical stimulation of cutaneous nerves innervating the hand and foot. Exp Brain Res 2001a;140:495–504. Zehr EP, Hesketh KL, Chua R. Differential regulation of cutaneous and H-reflexes during leg cycling in human. J Neurophysiol 2001b;85: 1178–84. Zehr EP, Collins DF, Frigon A, Hoogenboom N. Neural control of rhythmic human arm movement: phase dependence and task modulation of hoffmann reflexes in forearm muscles. J Neurophysiol 2003;89:12–21.