Cutaneomuscular reflexes recorded from the lower limb

Transcription

1 Journal of Physiology (1995), 487.1, pp Cutaneomuscular reflexes recorded from the lower limb in man during different tasks J. Gibbs, Linda M. Harrison * and J. A. Stephens Department of Physiology, University College London, Gower Street, London WC1E 6BT, UK 1. Cutaneomuscular reflexes have been recorded in ten adult subjects from extensor digitorum brevis (EDB), tibialis anterior (TA), soleus (Sol), quadriceps femoris (Quad) and erector spinae (ES) following electrical stimulation of the digital nerves of the second toe. 2. Recordings were made while subjects were instructed to activate voluntarily the relevant muscles and also when these muscles were active posturally. 3. Reflex responses could comprise three components: an initial increase in EMG (El), followed by a decrease (I1), followed by a second increase (). El and I1 were confined to muscles acting at the ankle and in the foot. was most pronounced in EDB but also found in TA, Sol, Quad (1 subject) and ES. No responses were recorded contralateral to the stimulus. 4. was significantly larger when the reflex was recorded during voluntary contraction of the muscle, rather than when the muscle was active posturally. 5. El and I1 components are mediated via spinal pathways. requires the integrity of the dorsal columns, sensorimotor cortex and corticospinal tract (Jenner & Stephens, 1982; Rowlandson & Stephens, 1985 b). The present study suggests that one or more of these supraspinal pathways is more active during voluntary contraction of lower limb muscles than when these muscles are active posturally. Modest electrical stimulation of the digital nerves of the upper or lower limb produces reflex modulation of on-going EMG recorded from nearby muscles. In the upper limb, stimulation of the digital nerves of the index finger produces reflex responses in the intrinsic hand muscles and in the muscles of the forearm (Caccia, McComas, Upton & Blogg, 1973; Jenner & Stephens, 1982; Issler & Stephens, 1983; Evans, Harrison & Stephens, 1989). In the lower limb, stimulation of the digital nerves of the second toe produces reflex responses in the intrinsic foot muscles (Jenner & Stephens, 1982) and in muscles of the leg (Choa & Stephens, 1982; Rowlandson & Stephens, 1985a, b). In the adult, this cutaneomuscular reflex (CMR) response typically has three identifiable early components comprising an initial short latency increase in EMG, the El component, followed by a decrease, the II component, followed by a second increase, the component. The first two components are believed to be mediated by activity in oligosynaptic spinal pathways whereas the component requires the integrity of the dorsal columns, sensorimotor cortex and corticospinal tract (Jenner & Stephens, 1982; Choa & Stephens, 1982; Rowlandson & Stephens, 1985a, b). In the upper limb, the amplitudes of the El and components change as muscles participate in different tasks (Evans, Harrison & Stephens, 1989). In tasks involving relatively independent finger movements, the component recorded from finger muscles was prominent and larger than the El component. When recorded during a power grip involving the co-operative action of all the fingers, the component was smaller. In the present study we have carried out a comparable study in the lower limb and have found that similar taskdependent effects can be demonstrated for foot, leg and trunk muscles. In both the upper and lower limb, the component of the CMR is more pronounced during those tasks which are believed to involve more activity in the motor cortex. * To whom correspondence should be addressed.

2 238 J Gibbs, L. M. Harrison and J A. Stephens J Physiol A preliminary account of these experiments has been presented to The Physiological Society (Gibbs, Shiers, Mallik, Harrison & Stephens, 1993). METHODS Recordings were made from ten healthy adults (seven female) aged from 21 to 43 years with written informed consent and local ethical committee approval. EMG was recorded using bipolar surface electrodes constructed from neonatal ECG electrodes (ARBO Pink H82 V) cut down to provide a recording area of 25 cm2. Electrodes were placed, with poles positioned 2-5 mm apart, on the right leg over the extensor digitorum brevis (EDB), tibialis anterior (TA), soleus (Sol), and quadriceps (Quad) muscles and on the left leg over the TA and Sol muscles. Further electrodes were placed over the lower aspect of the right erector spinae muscle just above the posterior superior iliac spine. With the subject relaxed, the digital nerves of the right second toe were stimulated using ring electrodes at 3 Hz using 1 ms duration pulses from a constant current stimulator (Medelec, Sapphire) to determine the threshold current required for perception. Reflex responses were elicited throughout using a stimulus strength twice threshold for perception. Such a stimulus is not painful and produces a readily identifiable reflex response after a few minutes of averaging. The threshold for perception was rechecked prior to each recording. The first series of recordings was made with subjects lying supine or prone and voluntarily contracting lower limb and trunk muscles. Subjects were instructed to contract voluntarily the muscles under investigation to produce a level of EMG 1-2% of that achieved during a maximal voluntary contraction (MVC). They were aided in this by visual feedback of the root mean squares of the EMG signals. EMG was amplified and filtered (-3 db at 15 Hz and 5 khz) using a Medelec Sapphire and stored on magnetic tape (Racal 4 Store) for off-line analysis. EMG recordings could only be made from two or three muscles simultaneously, thus several sets of recordings were required in order to include all the muscles listed above. Each set of recordings lasted about 3 min. A period of rest was allowed between recordings to avoid fatigue. The threshold for stimulus perception was rechecked prior to each recording with the subject relaxed. Dorsal lower limb muscles (ES, Sol) were voluntarily contracted while lying prone, whereas the ventral muscles (EDB, TA and Quad) were contracted with subjects lying supine. Erector spinae muscles were activated as subjects attempted to raise the trunk while prone; 1-2% activity could be maintained in these muscles without actually needing to raise the trunk from the supporting surface. A second series of recordings was made from the same muscles, using the same stimulus strength, while these were active A Extensor digitorum brevis D Quadriceps 2% E3 B 2% E Erector spinae C Soleus Figure 1. Cutaneomuscular reflexes recorded from lower limb muscles Averaged rectified EMG from extensor digitorum brevis (A), tibialis anterior (B), soleus (C), quadriceps (D) and erector spinae (E) muscles following stimulation of the digital nerves of the second toe at 3 s-1 (pulse width 1 ms) at a strength twice threshold for perception during voluntary activation by subject. Same subject throughout. 5 sweeps. Stimulation at time zero. Vertical calibration bar indicates 2% modulation of background EMG. El, It, and E3 indicate components of the CMR.

3 J Phy8iol Lower limb cutaneous reflexes in man 239 posturally. EMG was recorded from Sol and ES as subjects stood, leaning slightly forwards and from EDB, TA and Quad as subjects stood and leant slightly backwards. Subjects were encouraged to maintain background EMG at 1-2 % MVC, i.e. background EMG was similar to that recorded during a voluntary contraction, by adjusting the degree of lean. Data analysis EMG was rectified and averaged time-locked to the stimulus for 5 sweeps using averaging software (SigAvg Program, Cambridge Electronic Design, Cambridge, UK). The latency of each component was measured as it intersects the mean background EMG level. Mean background EMG (usually between 2 to 4 1sV) was taken as the mean of the on-going EMG during the 3 ms preceding presentation of the stimulus. In this study, as in previous studies from our laboratory, the size of each reflex component was quantified by measuring the amplitude. We had found previously that a similar trend was seen whether the amplitude or the area of each component was measured; hence either measurement leads to the same conclusions (Evans, 1992). Since the size of a reflex component depends upon the level of the on-going background EMG (Jenner, 1981; Evans, Harrison & Stephens, 199), the peak amplitude of each component was measured and normalized with respect to the mean background EMG. Amplitudes and latencies of the El, It and components of the CMR recorded from muscles voluntarily contracted during lying and from the same muscles posturally active were compared using Student's paired t tests. RESULTS Figure 1 shows CMR responses recorded from muscles ipsilateral to the stimulus in one of the subjects while these muscles were contracted voluntarily whilst prone or supine. The components of the CMR can be seen as modulations of the background EMG level. The individual components in the distal muscles were identified, taking account of the subject's height, from the typical latencies reported in adults for these components when recorded from the EDB muscle (Jenner & Stephens, 1982) and the TA muscle (Rowlandson & Stephens, 1985a, b). In more proximal muscles, components were identified by adjusting the latencies of these components in the distal muscles for the relatively shorter efferent conduction pathway of the respective proximal muscles. For the subject whose CMR responses are shown in Fig. 1, short latency El and I1 components are identifiable in some A Lying down El Extensor digitorum brevis Standing up B Tibialis anterior E3 2% C Erector spinae 5 1 L 5 I 1 Figure 2. Cutaneomuscular reflex responses recorded during voluntary EMG activity while lying down and postural activity while standing up Averaged rectified EMG recordings from extensor digitorum brevis (A), tibialis anterior (B) and erector spinae (C) following stimulation of the digital nerves of the second toe at 3 s-' single electical pulses (pulse width -1 ms) at a strength twice threshold for perception. Reflex recorded while subject produced a sustained voluntary contraction while lying down or when muscles were active posturally while standing up. Same subject in each case. 5 sweeps. Stimulation at time zero. Vertical calibration bar indicates 2% modulation of background EMG. El, I1, and E3 indicate components of the CMR.

4 24 2 Gibbs, L. M. Harrison and J A. Stephens J Phy8iol Table 1. Task dependency of cutaneomuscular reflexes recorded from leg and trunk muscles in man El I1 Muscle Condition n/1 Lat range Amp mean n/1 Lat range Amp mean (ins) (S.D.) (ins) (S.D.) EDB Ly St TA Ly St 1 43 Sol Quad 55(58) 42(22) 45(47) n/1o Lat range Amp mean (ms) (S.D) (26) 46(28)* (7) (5) (8) 13(8)** Ly St Ly St _ ES Ly St ~~~~ ~~~~ 17(3) 7(2)** (15) Amplitudes and latencies of cutaneomuscular reflex responses measured from the average rectified EMG recorded from EDB (extensor digitorum brevis), TA (tibialis anterior), Sol (soleus), Quad (quadriceps femoris) and ES (erector spinae) elicited by twice threshold perception, single electrical pulses applied to the second toe during voluntary contraction of these muscles while lying down and postural contractions while standing up. Ly, voluntary contraction while lying down; St, postural contraction while standing up; Lat, latency; Amp, amplitude as a percentage of background EMG. * P < 5, ** P < 1, paired t tests comparing Ly and St amplitudes. n, number of subjects. El, I1 and indicate components of the CMR. of the distal muscles but not in the proximal muscles. The longer latency component is more widely distributed. An E3 component can be identified in Fig. 1B and D. The origin of the rather variable long latency components that occur after the component is uncertain (Garnet & Stephens, 198) and such components have not been considered further in this study. Figure 2 shows CMR responses recorded from EDB, TA and ES when these muscles were voluntarily contracted while lying down and when these muscles were active posturally during standing. The same background level of EMG was maintained in each case. When EDB is contracted voluntarily the El and components are prominent (Fig. 2A). When active posturally, the response is smaller and the El component, in particular, is much reduced. A similar task-effect can be observed for the response recorded from TA muscle (Fig. 2B). The response recorded from the ES muscle when contracted voluntarily is absent when this muscle was active posturally (Fig. 2C). The CMR response that could be recorded from the ES muscle in five of the subjects always showed a marked task-dependent effect; a response was only apparent during voluntary contraction of this muscle while lying. No CMR responses were recorded from muscles contralateral to the digital nerve stimulus in any subject, either when contracted voluntarily while prone or supine or when active posturally while standing. Table 1 summarizes the CMR responses recorded from ipsilateral muscles in the ten subjects. During voluntary contraction while lying, prominent short latency El components were found in almost all subjects (9/1) in the intrinsic foot muscle EDB which acts on the toes. In only a few cases was this component found in TA (3/1) and sol (1/1), muscles which act at the ankle. The short latency decrease in EMG, the Il component, was found in almost all subjects in TA (9/1) but only in two out of ten in Sol. It should be noted that the recorded configuration of the CMR is the result of summation at the motoneurones of excitatory and inhibitory reflexes. Thus the absence of an Il component in the average rectified EMG signal recorded from EDB cannot be taken to indicate an absence of an inhibitory reflex component contributing to the total reflex drive to this muscle: its presence may be hidden because its inhibitory effect may not have been strong enough to counteract the stronger excitatory synaptic effects of the concurrent excitatory component El. Thus the distribution of the El component elicited by electrical stimulation of the digital nerves of a toe has a more distal restriction than the Il component and was prominent in a muscle which acts directly on the toe being stimulated to cause dorsiflexion. When observed, the Il component was small and, for most subjects, restricted to TA, an ankle flexor. The component was found in EDB and TA in all ten subjects and in eight out of ten in Sol. In one subject a small response was found in Quad and in five out of ten in ES. Thus of the three components of the CMR, this component had the greatest proximo-distal distribution. When reflex responses were present, a clear task dependency was apparent in the CMR response recorded from lower limb and trunk muscles in all subjects. The

5 J Physiol Lower limb cutaneous reflexes in man 241 amplitudes of the components, recorded from distal muscles, were significantly larger when these muscles were voluntarily contracted while lying down than when these same muscles were active posturally when standing up (EDB, P < 5; TA, P < 1; Sol, P <1; paired t tests). There was a tendency for the amplitudes of the shorter latency El and I1 components also to be larger when recorded from distal muscles during voluntary, as compared with postural, activation; although this difference was not statistically significant. The component was recorded in some subjects from proximal muscles during voluntary contraction while lying down but was invariably absent when these muscles were active posturally during standing up. There was no significant difference between the latencies of any of the CMR components when the responses recorded from lower limb and trunk muscles contracted voluntarily while lying down were compared with the responses recorded from the same muscles active posturally during standing up (paired t test, P > 5). DISCUSSION Distribution of the CMR response Following modest, non-noxious electrical stimulation of the digital nerves of the right second toe, CMR responses were found in a variety of ipsilateral lower limb muscles and also in the ipsilateral erector spinae muscle in some subjects. No contralateral responses were observed. This distribution of the CMR response indicates that non-noxious cutaneous input arising from the toes has a widespread modulating effect on motor activity in the lower limb. CMR responses were more commonly observed in distal, rather than proximal muscles, reflecting the variation in the strength of reflex connections from cutaneous afferents of the toe to the motoneurones innervating the relevant muscles. The size of the CMR response also varied between subjects, with some subjects showing no appreciable responses in the more proximal muscles, indicating the diversity in the strength of this reflex between individual subjects. More widely distributed responses in the lower limbs have been demonstrated by other workers following cutaneous stimulation at a higher intensity than employed in the current study. In general, the more intense the cutaneous stimulation the greater the probability of this stimulus producing a reflex response. In particular, we have found that the size of the components increases with increasing intensity of stimulation and when stimuli are perceived as unpleasant or painful, then a late inhibitory component is seen which may encroach upon the earlier components (J. Gibbs, L. M. Harrison & J. A. Stephens, unpublished observations). In their earlier study, Burke, Dickson & Skuse (1991) used short trains of five pulses (interpulse interval, 3 ins) to stimulate the sural nerve. Stimulus intensity was set to produce intense paraesthesiae, short of pain, at a level 2-4 times perceptual threshold. Responses were recorded from contralateral lower limb muscles in four out of seven subjects in addition to the ipsilateral responses seen in all subjects. In contrast, we never recorded reflex responses contralateral to the stimulus even when we employed stimulus strengths that were considered by the subject to be unpleasant. In the present study, the shorter latency El and I1 components were largely confined to muscles acting on the foot or toes. These reflex components have a short central delay and are believed to be mediated by oligosynaptic spinal pathways (Jenner & Stephens, 1982; Rowlandson & Stephens, 1985). Taking the data as a whole, the predominantly observed reflex effects of the El and I1 components were excitation of a toe dorsi-flexor muscle (EDB) and inhibition of the ankle dorsi-flexor TA. The longer latency component in EDB and TA requires the functional integrity of the dorsal columns, motor cortex and descending supraspinal pathways, most likely involving activity in the corticospinal tract (Jenner & Stephens, 1982; Choa & Stephens, 1982; Rowlandson & Stephens, 1985a, b). The responses in Sol are of the same latency as in TA and are likely to be of the same origin. The latency of the more proximal ES responses were commensurately shorter than those in TA and Sol and may result from activity in the same pathways. It is interesting to note that in some subjects the component could be recorded from TA whilst supine or leaning back and from Sol whilst prone or leaning forwards. These muscles are not normally co-activated and therefore any reflex effects would probably be very small and of no functional importance in the non-active muscle. The component is also more widespread than the earlier El and/or I1 components, which are largely confined to foot and toe muscles. Activity in the reflex pathway thus seems to produce a widespread reflex response in cocontracting synergistic muscles. In the present study the component produced on average between 15 and 55% modulation of background muscle activity. These values are similar to those we found earlier for finger muscles (Evans, Harrison & Stephens, 1989). In both upper and lower limb, and trunk muscles, the component of cutaneous reflexes elicited by stimulation of cutaneous afferents from a single digit has a similar powerful modulating effect on on-going muscle activity, indicating an important general functional role for these reflexes. Task-dependent changes in the CMR response In an earlier study we have shown task-dependent changes in the configuration of CMRs recorded from hand and forearm finger muscles following electrical stimulation of the index finger (Evans, Harrison & Stephens, 1989). For precision tasks, such as flexing or extending a single finger, which are believed to involve considerable motor cortical

6 242 J Gibbs, L. M. Harrison and J A. Stephens J Physiol activity, responses were prominent. In contrast, for non-precision tasks, such as when the fingers are used to hold an object in a power grip, responses were small or absent. The results of the present study show again that the component of the CMR is more pronounced when elicited during a voluntary contraction, when presumably there is more cortical involvement, rather than during a supporting or postural task. In their study of somatosensory evoked potentials following electrical stimulation of the digital nerves of the index finger, Harrison, Mayston & Stephens (1994) found that the amplitude of the N2-P25 component was smaller when elicited during a nonprecision task such as a power grip than during a precision task such as abduction of the index finger, indicating that central gating of the afferent volley may also be involved. The same may be true in the lower limb. Task-dependent changes in the responses to cutaneous afferent stimulation in the lower limb have also been described by Burke et al. (1991) who recorded from a variety of muscles during different postural and nonpostural tasks while stimulating the sural nerve. Their stimulus parameters would have produced a more widespread reflex effect than did the single shocks to the digital nerves of the second toe which were used in the present study. In TA, for example, the response recorded by Burke et al. (1991) consists predominantly of an inhibition, reaching a maximum at a mean latency of 93 ms, followed by a period of facilitation. Two small, earlier components were also noted at latencies equivalent to that of El and components described in the present study. The -equivalent component was present as an inflexion on the down slope of the much more prominent inhibitory component. This -equivalent component was small or absent either during sitting or maintenance of stable bipedal stance, but more prominent in the voluntarily ipsilaterally flexed leg during unipedal stance on the unstimulated leg. This change in size of the component is consistent with the finding in the present study that the component was significantly smaller when recorded from muscles active posturally than when recorded from muscles contracted voluntarily. The present study indicates that a greater role is played by cutaneous afferent input from the toes in modulating motor output when the lower limbs are not engaged in maintaining an upright posture. Thus the generalization emerges that in arm, leg and trunk muscles the component of cutaneomuscular reflexes is more pronounced during tasks likely to be associated with a high level of activity in the motor cortex. BURKE, D., DICKSON, H. G. & SKUSE, N. F. (1991). Task-dependent changes in the responses to low-threshold cutaneous afferent volleys in the human lower limb. Journal of Physiology 432, CACCIA, M. R., MCCOMAS, A. J., UPTON, A. R. M. & BLOGG, T. (1973). Cutaneous reflexes in small muscles of the hand. Journal of Neurology, Neurosurgery and Psychiatry 36, CHOA, B. H. G. & STEPHENS, J. A. (1982). Cutaneous reflex responses and central nervous lesions studied in the lower limb in man. Journal of Physiology 328, 23-24P. EVANS, A. L. (1992). Development and function of cutaneomuscular reflexes and their pathophysiology in cerebral palsy. MD thesis, University of Cambridge. EVANS, A. L., HARRISON, L. M. & STEPHENS, J. A. (1989). Taskdependent changes in cutaneous reflexes recorded from various muscles controlling finger movement in man. Journal of Physiology 418, EVANS, A. L., HARRISON, L. M. & STEPHENS, J. A. (199). Maturation of the cutaneomuscular reflex recorded from the first dorsal interosseous muscle in man. Journal of Physiology 428, GARNETT, R. & STEPHENS, J. A. (198). The reflex responses of single motor units in human first dorsal interosseous muscle following cutaneous afferent stimulation. Journal of Physiology 33, GIBBS, J., SHIERS, P., MALLIK, A. K., HARRISON, L. M. & STEPHENS, J. A. (1993). Cutaneomuscular reflexes recorded from leg and trunk muscles in man. Journal of Physiology 467, 197P. HARRISON, L. M., MAYSTON, M. J. & STEPHENS, J. A. (1994). Central mechanisms underlying task-dependence of cutaneous reflexes in man. Journal of Physiology 476.P, 18P. ISSLER, H. & STEPHENS, J. A. (1983). The maturation of cutaneous reflexes studied in the upper limb in man. Journal of Physiology 335, JENNER, J. R. (1981). Cutaneous reflexes in man studied in health and disease. MD thesis, University of London. JENNER, J. R. & STEPHENS, J. A. (1982). Cutaneous reflex responses and their central nervous pathways studied in man. Journal of Physiology 333, ROWLANDSON, P. H. & STEPHENS, J. A. (1985a). Maturation of cutaneous reflex responses recorded in the lower limb in man. Developmental Medicine and Child Neurology 27, ROWLANDSON, P. H. & STEPHENS, J. A. (1985b). Cutaneous reflex responses recorded in children with various neurological disorders. Developmental Medicine and Child Neurology 27, Acknowledgements This work was supported by the Scholl Trust. Received 15 August 1994; accepted 7 February 1995.