Safety and tolerability of transcranial magnetic and direct current stimulation in children: Prospective single center evidence from 3.5 million stimulations.

BACKGROUND: Non-invasive brain stimulation is being increasingly used to interrogate neurophysiology and modulate brain function. Despite the high scientific and therapeutic potential of non-invasive brain stimulation, experience in the developing brain has been limited. OBJECTIVE: To determine the safety and tolerability of non-invasive neurostimulation in children across diverse modalities of stimulation and pediatric populations. METHODS: A non-invasive brain stimulation program was established in 2008 at our pediatric, academic institution. Multi-disciplinary neurophysiological studies included single- and paired-pulse Transcranial Magnetic Stimulation (TMS) methods. Motor mapping employed robotic TMS. Interventional trials included repetitive TMS (rTMS) and transcranial direct current stimulation (tDCS). Standardized safety and tolerability measures were completed prospectively by all participants. RESULTS: Over 10 years, 384 children underwent brain stimulation (median 13 years, range 0.8-18.0). Populations included typical development (n = 118), perinatal stroke/cerebral palsy (n = 101), mild traumatic brain injury (n = 121) neuropsychiatric disorders (n = 37), and other (n = 7). No serious adverse events occurred. Drop-outs were rare (<1%). No seizures were reported despite >100 participants having brain injuries and/or epilepsy. Tolerability between single and paired-pulse TMS (542340 stimulations) and rTMS (3.0 million stimulations) was comparable and favourable. TMS-related headache was more common in perinatal stroke (40%) than healthy participants (13%) but was mild and self-limiting. Tolerability improved over time with side-effect frequency decreasing by >50%. Robotic TMS motor mapping was well-tolerated though neck pain was more common than with manual TMS (33% vs 3%). Across 612 tDCS sessions including 92 children, tolerability was favourable with mild itching/tingling reported in 37%. CONCLUSIONS: Standard non-invasive brain stimulation paradigms are safe and well-tolerated in children and should be considered minimal risk. Advancement of applications in the developing brain are warranted. A new and improved pediatric NIBS safety and tolerability form is included.

Hyperexcitability of brain stem pathways in cerebral palsy.

Individuals with cerebral palsy (CP) experience impairments in the control of head and neck movements, suggesting dysfunction in brain stem circuitry. To examine if brain stem circuitry is altered in CP, we compared reflexes evoked in the sternocleidomastoid (SCM) muscle by trigeminal nerve stimulation in adults with CP and in age/sex-matched controls. Increasing the intensity of trigeminal nerve stimulation produced progressive increases in the long-latency suppression of ongoing SCM electromyography in controls. In contrast, participants with CP showed progressively increased facilitation around the same reflex window, suggesting heightened excitability of brain stem pathways. We also examined if there was altered activation of cortico-brain stem pathways in response to prenatal injury of the brain. Motor-evoked potentials (MEPs) in the SCM that were conditioned by a prior trigeminal afferent stimulation were more facilitated in CP compared with controls, especially in ipsilateral MEPs that are likely mediated by corticoreticulospinal pathways. In some participants with CP, but not in controls, a combined trigeminal nerve and cortical stimulation near threshold intensities produced large, long-lasting responses in both the SCM and biceps brachii muscles. We propose that the enhanced excitatory responses evoked from trigeminal and cortical inputs in CP are produced by heightened excitability of brain stem circuits, resulting in the augmented activation of reticulospinal pathways. Enhanced activation of reticulospinal pathways in response to early injury of the corticospinal tract may provide a compensated activation of the spinal cord or, alternatively, contribute to impairments in the precise control of head and neck functions. NEW & NOTEWORTHY This is the first study to show that in adults with spastic cerebral palsy, activation of brain stem circuits by cortical and/or trigeminal afferents produces excitatory responses in anterior neck muscles compared with inhibitory responses in age/sex-matched controls. This may reflect a more excitable reticulospinal tract in response to early brain injury to provide a compensated activation of postural muscles. On the other hand, a hyperexcitable brain stem may contribute to impairments in the precise control of head and neck functions.

Spinal inhibition and motor function in adults with spastic cerebral palsy.

KEY POINTS: Abnormal activation of motoneurons in the spinal cord by sensory pathways is thought to contribute to impaired movement control and spasticity in individuals with cerebral palsy. Here we use single motor unit recordings to show how individual motoneurons in the spinal cord respond to sensory inputs in a group of participants with cerebral palsy having different degrees of motor dysfunction. In participants who had problems walking independently and required assistive devices such as wheelchairs, sensory pathways only excited motoneurons in the spinal cord. In contrast, in participants with cerebral palsy who walked independently for long distances, sensory inputs both inhibited and excited motoneurons in the spinal cord, similar to what we found in uninjured control participants. These findings demonstrate that in individuals with severe cerebral palsy, inhibitory control of motoneurons from sensory pathways is reduced and may contribute to motor dysfunction and spasticity. ABSTRACT: Reduced inhibition of spinal motoneurons by sensory pathways may contribute to heightened reflex activity, spasticity and impaired motor function in individuals with cerebral palsy (CP). To measure if the activation of inhibitory post-synaptic potentials (IPSPs) by sensory inputs is reduced in CP, the tonic discharge rate of single motor units from the soleus muscle was plotted time-locked to the occurrence of a sensory stimulation to produce peri-stimulus frequencygrams (PSFs). Stimulation to the medial arch of the foot was used to activate cutaneomuscular afferents in 17 adults with bilateral spastic CP and 15 neurologically intact (NI) peers. Evidence of IPSP activation from the PSF profiles, namely a marked pause or reduction in motor unit firing rates at the onset of the cutaneomuscular reflex, was found in all NI participants but in only half of participants with CP. In the other half of the participants with CP, stimulation of cutaneomuscular afferents produced a PSF profile indicative of a pure excitatory post-synaptic potential, with firing rates increasing above the mean pre-stimulus rate for 300 ms or more. The amplitude of motoneuron inhibition during the period of IPSP activation, as measured from the surface EMG, was less in participants with poor motor function as evaluated with the Gross Motor Functional Classification System (r = 0.72, P < 0.001) and the Functional Mobility Scale (r = -0.82, P < 0.001). These findings demonstrate that in individuals with CP, reduced activation of motoneuron IPSPs by sensory inputs is associated with reduced motor function and may contribute to enhanced reflexes and spasticity in CP.

Jaw tremor as a physiological biomarker of bruxism.

OBJECTIVE: To determine if sleep bruxism is associated with abnormal physiological tremor of the jaw during a visually-guided bite force control task. METHODS: Healthy participants and patients with sleep bruxism were given visual feedback of their bite force and asked to trace triangular target trajectories (duration=20s, peak force <35% maximum voluntary force). Bite force control was quantified in terms of the power spectra of force fluctuations, masseter EMG activity, and force-to-EMG coherence. RESULTS: Patients had greater jaw force tremor at ∼8 Hz relative to controls, along with increased masseter EMG activity and force-to-EMG coherence in the same frequency range. Patients also showed lower force-to-EMG coherence at low frequencies (<3 Hz), but greater coherence at high frequencies (20-40 Hz). Finally, patients had greater 6-10 Hz force tremor during periods of descending vs. ascending force, while controls showed no difference in tremor with respect to force dynamics. CONCLUSION: Patients with bruxism have abnormal jaw tremor when engaged in a visually-guided bite force task. SIGNIFICANCE: Measurement of jaw tremor may aid in the detection/evaluation of bruxism. In light of previous literature, our results also suggest that bruxism is marked by abnormal or mishandled peripheral feedback from the teeth.

Adrenergic receptors modulate motoneuron excitability, sensory synaptic transmission and muscle spasms after chronic spinal cord injury.

The brain stem provides most of the noradrenaline (NA) present in the spinal cord, which functions to both increase spinal motoneuron excitability and inhibit sensory afferent transmission to motoneurons (excitatory postsynaptic potentials; EPSPs). NA increases motoneuron excitability by facilitating calcium-mediated persistent inward currents (Ca PICs) that are crucial for sustained motoneuron firing. Spinal cord transection eliminates most NA and accordingly causes an immediate loss of PICs and emergence of exaggerated EPSPs. However, with time PICs recover, and thus the exaggerated EPSPs can then readily trigger these PICs, which in turn produce muscle spasms. Here we examined the contribution of adrenergic receptors to spasms in chronic spinal rats. Selective activation of the α(1A) adrenergic receptor with the agonists methoxamine or A61603 facilitated Ca PIC and spasm activity, recorded both in vivo and in vitro. In contrast, the α(2) receptor agonists clonidine and UK14303 did not facilitate Ca PICs, but did decrease the EPSPs that trigger spasms. Moreover, in the absence of agonists, spasms recorded in vivo were inhibited by the α(1) receptor antagonists WB4010, prazosin, and REC15/2739, and increased by the α(2) receptor antagonist RX821001, suggesting that both adrenergic receptors were endogenously active. In contrast, spasm activity recorded in the isolated in vitro cord was inhibited only by the α(1) antagonists that block constitutive receptor activity (activity in the absence of NA; inverse agonists, WB4010 and prazosin) and not by the neutral antagonist REC15/2739, which only blocks conventional NA-mediated receptor activity. RX821001 had no effect in vitro even though it is an α(2) receptor inverse agonist. Our results suggest that after chronic spinal cord injury Ca PICs and spasms are facilitated, in part, by constitutive activity in α(1) adrenergic receptors. Additionally, peripherally derived NA (or similar ligand) activates both α(1) and α(2) adrenergic receptors, controlling PICs and EPSPs, respectively.

Reduced functional recovery by delaying motor training after spinal cord injury.

The purpose of this study was to examine if a delay in rehabilitative motor training after spinal cord injury affects functional motor recovery. We studied a skilled motor task in which rats traversed a raised horizontal ladder and we quantified errors in accurate stepping, i.e., foot slips between rungs. After lesions to the dorsal quadrant of the thoracic (T8) spinal cord that aimed to unilaterally sever the corticospinal and rubrospinal tracts, rats were re-trained to walk across the ladder, either immediately after injury or after a 3-mo delay. Before training, the error rate in accurate stepping of the affected hindlimb was similar in the immediately (69.4 +/- 5.3%) and delay (62.7 +/- 4.1%; means +/- SE)-trained animals (not significantly different), suggesting that accurate stepping did not improve spontaneously if rats were not exposed to the ladder. After a 3-wk course of training (30 runs across the ladder per day, 5 day/wk), improvements in accurate stepping performance were greater if training was implemented immediately after injury. On average, immediately trained animals improved stepping performance by 61.5 +/- 28.2%, whereas the delay trained group improved by only 34.9 +/- 28.8% (significantly different). The degree of damage to the corticospinal and rubrospinal tracts was very similar in the two groups of animals, indicating that differences in lesion size did not contribute to the differences in performance improvement. Animals with large lesions to the corticospinal and rubrospinal tracts (>70%) displayed poor recovery from training (especially for delay-trained animals), suggesting that these two pathways were important in mediating improvements in accurate stepping. In addition, recovery of stepping-like reflexes appeared not to contribute to the recovery of accurate stepping given that the time course of reflex recovery was not related to the time course of recovery of accurate stepping. We conclude that training of a skilled motor task that relies on descending control is more beneficial when initiated immediately after a partial spinal cord injury.

Spastic long-lasting reflexes in the awake rat after sacral spinal cord injury.

Following chronic sacral spinal cord transection in rats the affected tail muscles exhibit marked spasticity, with characteristic long-lasting tail spasms evoked by mild stimulation. The purpose of the present paper was to characterize the long-lasting reflex seen in tail muscles in response to electrical stimulation of the tail nerves in the awake spastic rat, including its development with time and relation to spasticity. Before and after sacral spinal transection, surface electrodes were placed on the tail for electrical stimulation of the caudal nerve trunk (mixed nerve) and for recording EMG from segmental tail muscles. In normal and acute spinal rats caudal nerve trunk stimulation evoked little or no EMG reflex. By 2 wk after injury, the same stimulation evoked long-lasting reflexes that were 1) very low threshold, 2) evoked from rest without prior EMG activity, 3) of polysynaptic latency with >6 ms central delay, 4) about 2 s long, and 5) enhanced by repeated stimulation (windup). These reflexes produced powerful whole tail contractions (spasms) and developed gradually over the weeks after the injury (< or =52 wk tested), in close parallel to the development of spasticity. Pure low-threshold cutaneous stimulation, from electrical stimulation of the tip of the tail, also evoked long-lasting spastic reflexes, not seen in acute spinal or normal rats. In acute spinal rats a strong C-fiber stimulation of the tip of the tail (20 x T) could evoke a weak EMG response lasting about 1 s. Interestingly, when this C-fiber stimulation was used as a conditioning stimulation to depolarize the motoneuron pool in acute spinal rats, a subsequent low-threshold stimulation of the caudal nerve trunk evoked a 300-500 ms long reflex, similar to the onset of the long-lasting reflex in chronic spinal rats. A similar conditioned reflex was not seen in normal rats. Thus there is an unusually long low-threshold polysynaptic input to the motoneurons (pEPSP) that is normally inhibited by descending control. This pEPSP is released from inhibition immediately after injury but does not produce a long-lasting reflex because of a lack of motoneuron excitability. With chronic injury the motoneuron excitability is increased markedly, and the pEPSP then triggers sustained motoneuron discharges associated with long-lasting reflexes and muscle spasms.

Evidence for plateau potentials in tail motoneurons of awake chronic spinal rats with spasticity.

Motor units of segmental tail muscles were recorded in awake rats following acute (1-2 days) and chronic (>30 days) sacral spinal cord transection to determine whether plateau potentials contributed to sustained motor-unit discharges after injury. This study was motivated by a companion in vitro study that indicated that after chronic spinal cord injury, the tail motoneurons of the sacrocaudal spinal cord exhibit persistent inward currents (I(PIC)) that cause intrinsically sustained depolarizations (plateau potentials) and firing (self-sustained firing). Importantly, in this companion study, the plateaus were fully activated at recruitment and subsequently helped sustain the firing without causing abrupt nonlinearities in firing. That is, after recruitment and plateau activation, the firing rate was modulated relatively linearly with injected current and therefore provided a good approximation of the input to the motoneuron despite the plateau. Thus in the present study, pairs of motor units were recorded simultaneously from the same muscle, and the firing rate (F) of the lowest-threshold unit (control unit) was used as an estimate of the synaptic input to both units. We then examined whether firing of the higher-threshold unit (test unit) was intrinsically maintained by a plateau, by determining whether more synaptic input was required to recruit the test unit than to maintain its firing. The difference in the estimated synaptic input at recruitment and de-recruitment of the test unit (i.e., change in control unit rate, DeltaF) was taken as an estimate of the plateau current (I(PIC)) that intrinsically sustained the firing. Slowly graded manual skin stimulation was used to recruit and then de-recruit the units. The test unit was recruited when the control unit rate was on average 17.8 and 18.9 Hz in acute and chronic spinal rats, respectively. In chronic spinal rats, the test unit was de-recruited when the control unit rate (re: estimated synaptic input) was significantly reduced, compared with at recruitment (DeltaF = -5.5 Hz), and thus a plateau participated in maintaining the firing. In the lowest-threshold motor units, even a brief stimulation triggered very long-lasting firing (seconds to hours; self-sustained firing). Higher-threshold units required continuous stimulation (or a spontaneous spasm) to cause firing, but again more synaptic input was needed to recruit the unit than to maintain its firing (i.e., plateau present). In contrast, in acute spinal rats, the stimulation did not usually trigger sustained motor-unit firing that could be attributed to plateaus because DeltaF was not significantly different from zero. These results indicate that plateaus play an important role in sustaining motor-unit firing in awake chronic spinal rats and thus contribute to the hyperreflexia and hypertonus associated with chronic injury.

Activity of hindlimb motor units during locomotion in the conscious rat.

This paper compares the activity of hindlimb motor units from muscles mainly composed of fast-twitch muscle fibers (medial and lateral gastrocnemius: MG/LG, tibialis anterior: TA) to motor units from a muscle mainly composed of slow-twitch muscle fibers (soleus: SOL) during unrestrained walking in the conscious rat. Several differences in the activation profiles of motor units from these two groups of muscles were observed. For example, motor units from fast muscles (e.g., MG/LG and TA) fired at very high mean frequencies of discharge, ranging from 60 to 100 Hz, and almost always were recruited with initial doublets or triplets, i.e., initial frequencies >/=100 Hz. In contrast, the majority of SOL units fired at much lower mean rates of discharge, approximately 30 Hz, and had initial frequencies of only 30-60 Hz (i.e., there were no initial doublets/triplets >/=100 Hz). Thus the presence of initial doublet or triplets was dependent on the intrinsic properties of the motor unit, i.e., faster units were recruited with a doublet/triplet more often than slower units. Moreover, in contrast to units from the slow SOL muscle, the activity of single motor units from the fast MG/LG muscle, especially units recruited midway or near the end of a locomotor burst, was unrelated to the activity of the remainder of the motoneuron pool, as measured by the corresponding gross-electromyographic (EMG) signal. This dissociation of activity was suggested to arise from a compartmentalized recruitment of the MG/LG motoneuron pool by the rhythm-generating networks of the spinal cord. In contrast, when comparing the rate modulation of simultaneously recorded motor units within a single LG muscle compartment, the frequency profiles of unit pairs were modulated in a parallel fashion. This suggested that the parent motoneurons were responsive to changes in synaptic inputs during unrestrained walking, unlike the poor rate modulation that occurs during locomotion induced from brain stem stimulation. In summary, data from this study provide evidence that the firing behavior of motor units during unrestrained walking is influenced by both the intrinsic properties of the parent motoneuron and by synaptic inputs from the locomotor networks of the spinal cord. In addition, it also provides the first extensive description of motor-unit activity from different muscles during unrestrained walking in the conscious rat.

Activation patterns of hindlimb motor units in the awake rat and their relation to motoneuron intrinsic properties.

The activity of hindlimb motor units from the lateral gastrocnemius and tibialis anterior muscles in the awake rat was compared during locomotion and during slow, sinusoidal muscle stretch. The majority of units were activated with high initial frequencies and often commenced firing with an initial doublet or triplet, even when activated by slow muscle stretch. The high firing rates at recruitment occurred without jumps in the firing rates of other concurrently activated units, the firing rate profiles of which were used as a measure of the net synaptic drive onto the motoneuronal pool. This suggested that the sharp recruitment jumps were not due to an abrupt increase in synaptic drive but rather due to intrinsic properties of the motoneuron. In addition, motor units that were activated phasically by the muscle stretch fired more action potentials during muscle shortening than during muscle lengthening, resulting in rightwardly skewed, asymmetrical firing profiles. In contrast, when the same units fired tonically during the imposed muscle stretch, the frequency profiles were modulated symmetrically and no nonlinearities were observed. Tonically firing units were modulated symmetrically throughout a wide range of firing frequencies, and discrete jumps in rate (i.e., bistable firing) were not observed. The sharp recruitment jumps during locomotion and muscle stretch are proposed to have resulted from the additional depolarization produced by the activation of plateau potentials at recruitment. Likewise, the sustained activation of plateaus subsequent to recruitment may have produced the prolonged firing of the motor units during sinusoidal muscle stretch.

Spasticity in rats with sacral spinal cord injury.

We have investigated sacral spinal cord lesions in rats with the goal of developing a rat model of muscular spasticity that is minimally disruptive, not interfering with bladder, bowel, or hindlimb locomotor function. Spinal transections were made at the S2 sacral level and, thus, only affected the tail musculature. After spinal transection, the muscles of the tail were inactive for 2 weeks. Following this initial period, hypertonia, hyperreflexia, and clonus developed in the tail, and grew more pronounced with time. These changes were assessed in the awake rat, since the tail is readily accessible and easy to manipulate. Muscle stretch or cutaneous stimulation of the tail produced muscle spasms and marked increases in muscle tone, as measured with force and electromyographic recordings. When the tail was unconstrained, spontaneous or reflex induced flexor and extensor spasms coiled the tail. Movement during the spasms often triggered clonus in the end of the tail. The tail hair and skin were extremely hyperreflexive to light touch, withdrawing quickly at contact, and at times clonus could be entrained by repeated contact of the tail on a surface. Segmental tail muscle reflexes, e.g., Hoffman reflexes (H-reflexes), were measured before and after spinalization, and increased significantly 2 weeks after transection. These results suggest that sacral spinal rats develop symptoms of spasticity in tail muscles with similar characteristics to those seen in limb muscles of humans with spinal cord injury, and thus provide a convenient preparation for studying this condition.

Synaptic activation of plateaus in hindlimb motoneurons of decerebrate cats.

Intracellular recordings were made from hindlimb motoneurons in decerebrate cats to study how synaptic inputs could affect the threshold at which plateau potentials are activated with current injections through the recording microelectrode in the cell body. This study was prompted by recent evidence that the noninactivating inward currents that regeneratively produce the plateau potentials arise (partly) from dendritic conductances, which may be relatively more accessible to synaptic input than to current injected into the soma. Initially, cells were studied by injecting a slow triangular current ramp intracellularly to determine the threshold for activation of the plateau. In cells where the sodium spikes were blocked with intracellular QX314, plateau activation was readily seen as a sudden jump in membrane potential, which was not directly reversed as the current was decreased. With normal spiking, the plateau activation (the noninactivating inward current) was reflected by a steep and sustained jump in firing rate that was not directly reversed as the current was decreased. Importantly, the threshold for plateau activation (at 34 Hz on average) was significantly above the recruitment level (13 Hz on average). When tonic synaptic excitation [excitatory postsynaptic potentials (EPSPs)] was provided either by stretching the triceps surae muscle or by stimulating its nerve at a high frequency, the threshold for plateau activation by intracellular current injection was significantly lowered (by 12 Hz or 5.8 mV on average, without and with QX314, respectively). Conversely, tonic synaptic inhibition [inhibitory postsynaptic potentials (IPSPs)], provided by appropriate nerve stimulation, significantly raised the plateau threshold (by 19 Hz or 7.6 mV on average). These effects were graded with the intensity of tonic EPSPs and IPSPs. Strong enough EPSPs brought the plateau threshold down sufficiently that it was activated by the intracellular current soon after recruitment. A further increase in tonic EPSPs recruited the cell directly, and in this case the plateau was activated at or before recruitment. The finding that synaptic excitation can produce plateau activation below the recruitment level is of importance for the interpretation of its function. With this low-threshold activation, the plateau potentials are likely important in securing an effective recruitment to frequencies that produce significant force generation and would subsequently have no further affect on the frequency modulation, other than to provide a steady depolarizing bias that would help to sustain firing (cf. self-sustained firing). Additional jumps in frequency after recruitment (i.e., bistable firing) would not be expected.

Short-term plasticity in hindlimb motoneurons of decerebrate cats.

Cat hindlimb motoneurons possess noninactivating voltage-gated inward currents that can, under appropriate conditions, regeneratively produce sustained increments in depolarization and firing of the cell (i.e., plateau potentials). Recent studies in turtle dorsal horn neurons and motoneurons indicate that facilitation of plateaus occurs with repeated plateau activation (decreased threshold and increased duration; this phenomenon is referred to as warm-up). The purpose of the present study was to study warm-up in cat motoneurons. Initially, cells were studied by injecting a slow triangular current ramp intracellularly to determine the threshold for activation of the plateau. In cells where the sodium spikes were blocked with intracellular QX314, plateau activation was readily seen as a sudden jump in membrane potential, which was not directly reversed as the current was decreased (cf. hysteresis). With normal spiking, the plateau activation (the noninactivating inward current) was reflected by a steep and sustained jump in firing rate, which was not directly reversed as the current was decreased (hysteresis). Repetitive plateau activation significantly lowered the plateau activation threshold in 83% of cells (by on average 5 mV and 11 Hz with and without QX314, respectively). This interaction between successive plateaus (warm-up) occurred when tested with 3- to 6-s intervals; no interaction occurred at times >20 s. Plateaus initiated by synaptic activation from muscle stretch were also facilitated by repetition. Repeated slow muscle stretches that produced small phasic responses when a cell was hyperpolarized with intracellular current bias produced a larger and more prolonged responses (plateau) when the bias was removed, and the amplitude and duration of this response grew with repetition. The effects of warm-up seen with intracellular recordings during muscle stretch could also be recorded extracellularly with gross electromyographic (EMG) recordings. That is, the same repetitive stretch as above produced a progressively larger and more prolonged EMG response. Warm-up may be a functionally important form of short-term plasticity in motoneurons that secures efficient motor output once a threshold level is reached for a significant period. Finally, the finding that warm-up can be readily observed with gross EMG recordings will be useful in future studies of plateaus in awake animals and humans.

Self-sustained firing of human motor units.

Motoneurons of invertebrates and vertebrates can continue to fire repetitively after being activated by a brief, excitatory synaptic input (self-sustained firing). This firing behavior is due to the activation of intrinsic, voltage-gated currents which produce sustained regenerative depolarizations (plateau potentials) of the cell. Examination of these intrinsic cellular properties has been performed in reduced animal preparations and it is unknown if such self-sustained firing occurs in motoneurons of the intact human. In this paper, we present evidence of this in the human by using a technique of dual motor unit recordings. Subjects were instructed to maintain a constant dorsiflexion effort, and the common synaptic input (e.g. descending drive) onto the tibialis anterior (TA) motoneuron pool was monitored by recording the firing frequency of a low threshold 'control' unit. Once the firing rate of the control unit was constant, vibration of the TA tendon recruited a second 'test' unit which continued to fire after the vibration (i.e. synaptic input) was removed, even though the firing rate of the control unit (and thus, the common drive) remained the same or decreased. Self-sustained firing of motoneurons such as this may reduce the need for prolonged synaptic input when constant muscle activation is required (e.g. for postural tone).

Models of ensemble firing of muscle spindle afferents recorded during normal locomotion in cats.

1. The aim of this work was to compare the ability of several mathematical models to predict the firing characteristics of muscle spindle primary afferents recorded chronically during normal stepping in cats. 2. Ensemble firing profiles of nine hamstring spindle primary (presumed group Ia) afferents were compiled from stored data from 132 step cycles. Three sets of profiles corresponding to slow, medium and fast steps were generated by averaging groups of step cycles aligned to peak muscle length in each cycle. 3. Five models obtained from the literature were compared. Each of these models was used to predict the spindle firing profiles from the averaged muscle length signals. The models were also used in the reverse direction, namely to predict muscle length from the firing profiles. A sixth model incorporating some key aspects of the other models was also included in the comparisons. 4. Five of the models predicted spindle firing well, with root mean square (r.m.s.) errors lower than 14 % of the modulation depth of the target profiles. The key variable in achieving good predictions was muscle velocity, the best fits being obtained with power-law functions of velocity, with an exponent of 0.5 or 0.6 (i.e. spindle firing rate is approximately proportional to the square root of muscle velocity). The fits were slightly improved by adding small components of EMG signal to mimic fusimotor action linked to muscle activation. The modest relative size of EMG-linked fusimotor action may be related to the fact that hamstring muscles are not strongly recruited in stepping. 5. Length was predicted very accurately from firing profiles with the inverse of the above models, indicating that the nervous system could in principle process spindle firing in a relatively simple way to give accurate information on muscle length. 6. The responses of the models to standard ramp-and-hold displacements at 10 mm s-1 were also studied (i.e. velocities that were an order of magnitude lower than that during stepping). In these cases components of spindle primary response related to length as well as velocity were needed for good fits. Because these length-related components detracted from rather than improved predictions of the step cycle data, an attenuation of length dependence at high muscle velocities emerged as a possibility. 7. We conclude that in this study we have identified models and parameters that may be used to predict spindle afferent firing from the time course of muscle length in the cat step cycle.

Ensemble firing of muscle afferents recorded during normal locomotion in cats.

1. The main purpose of this study was to collate population data on the firing characteristics of muscle afferents recorded chronically during normal stepping in cats. 2. Ensemble firing profiles of forty-seven muscle spindle and tendon organ afferents were compiled from stored data. The relationships between the firing profiles and the displacement and force signals were analysed with the help of mathematical models of the response characteristics of spindle primary and secondary afferents and tendon organs. 3. Whereas the firing of hamstring spindle afferents could be predicted with reasonable accuracy from the length and velocity signals alone, the firing profiles of triceps surae spindle afferents deviated from the predicted profiles, particularly during electromyogram (EMG) activity. This indicated that the components of fusimotor action linked to extrafusal muscle activity were significant in triceps surae, possibly because this muscle is more strongly recruited in the cat step cycle. 4. From the limited data available, it was not possible to identify the 'best' or most general mathematical function to predict spindle secondary firing. In the two triceps surae spindle secondary units studied, firing was well predicted by using the simplest possible model, rate proportional to displacement, whereas in the hamstring spindle secondary data, a more complex linear transfer function was needed. The results of modelling the spindle secondary data were consistent with a modest amount of phasic, static fusimotor action linked to EMG activity. 5. The averaged ensemble of tendon organ afferent activity from the triceps surae gave predictions of whole-muscle force that agreed well with separate triceps force measurements in normal cat locomotion. This supports the idea that ensembles of tendon organ afferents signal whole-muscle force. 6. Our overall conclusion is that to a first approximation, large muscle afferents in the cat hindlimb signal muscle velocity, muscle length and muscle force, at least in movements of the speed and amplitude seen in locomotion.

Catching a ball: contributions of intrinsic muscle stiffness, reflexes, and higher order responses.

In three sets of experiments in nine normal subjects and a patient with a percutaneous wrist-stabilizing splint, we quantified the open-loop gain (OLG) of the stretch reflex acting about the elbow. The subjects exerted a steady mean flexing force and were instructed not to intervene (i.e., not to resist actively) when force or displacement perturbations were imposed on the forearm. The method was either to reconstruct transmission around the entire loop in a two-part experiment, or to use the attenuation of external perturbations in normal and electrically stimulated muscle to compute gain. Across all experiments, the mean magnitude of stretch reflex OLG was close to unity in the frequency range 1-2 Hz, and declined at higher frequencies, as required to ensure stability, given that the phase lag approached 180 degrees at 5 Hz. Inherent muscle stiffness was approximately equal to reflex stiffness. In functional terms, an OLG of 1 means that the yield caused by a force perturbation is approximately halved by reflex action (prevailing inherent muscle stiffness is doubled). Automatic scaling of reflex transmission at Ia/alpha-motoneuronal synapses ensures that the OLG remains close to unity as inherent stiffness increases. Trials in the patient with the wrist fixator gave similar results, indicating that the reflexes were proprioceptive ly mediated. In a fourth experiment in which the task was to catch a heavy ball, we compared the efficacy of inherent muscle stiffness and reflexes alone, with the subject's intentional reactions, which included predictive and voluntary components of response. The latter were far more effective in maintaining the position of the hand after the ball was caught than inherent and reflex stiffnesses alone. We conclude that stability requirements limit the extent to which stretch reflexes can augment inherent muscle stiffness. When inherent muscle stiffness is low, such as in our ball-catching task, the reflex stiffness is also low, and predictive and pre-programmed reactions predominate in load compensation, thus shifting the emphasis from automatic servo or equilibrium-point behaviour to higher order control.

Corrective responses to loss of ground support during walking. I. Intact cats.

1. In the cat step cycle the electromyographic (EMG) activity in ankle extensor muscles commences approximately 70 ms before foot contact. There is a sharp peak between 10 and 25 ms after contact and the EMG then declines for the remainder of the stance phase. It has been posited that the abrupt transition in EMG after contact is the consequence of reflexes elicited by the large barrage of afferent input that signals foot touchdown. However, it is also possible that the basic profile might be generated within the CNS, with little modification by afferent input. 2. These ideas were tested in 11 normal cats. We compared EMG responses and hindlimb kinematics in steps with normal ground support and steps in which an actuator-controlled trap door unexpectedly opened, withdrawing ground support just before foot contact. 3. In the absence of ground support the transition in EMG activity was still present. The averaged EMG pattern was similar for at least 30 ms after the foot passed through the plane of the floor. We conclude that the basic extensor activation profile in this part of the cycle is generated centrally and is not substantially altered by afferent input. 4. Between 35 and 200 ms after contact the stance phase was aborted and the foot was lifted smartly out of the hole. This reaction varied both in latency and kinematic detail, suggesting a fairly complex corrective response.(ABSTRACT TRUNCATED AT 250 WORDS)

Corrective responses to loss of ground support during walking. II. Comparison of intact and chronic spinal cats.

1. The preceding study described a corrective response in cats when one hind leg steps into a hole. In this investigation we examine the extent to which this behavior is organized at the spinal level by comparing the responses elicited in intact and chronic spinal cats. 2. Adult cats were trained to step bipedally with their hind legs on a treadmill. After training, the responses to stepping into a hole cut in the treadmill belt were monitored with a video recorder and by recording electromyograms from muscles in both hind legs. The responses to stepping into the hole were also recorded in chronic spinal cats that had recovered the ability to step with their hind legs a few weeks after spinalization. 3. The behavioral responses in the two groups of animals differed in two respects. First, the latency of the onset of the flexion movement to remove the foot from the hole was shorter in intact animals (70-150 ms in intact vs. 130-350 ms in spinal animals). Second, the flexion movement in the intact animals was stronger. The exaggerated flexion movement in intact animals lifted the paw well clear of the hole and allowed support to be regained on the treadmill belt. The weaker flexion movement in spinal animals was usually insufficient to lift the paw completely from the hole. 4. Differences in the motor patterns recorded from flexor muscles during the corrective response in intact and spinal animals correspond with the differences in the kinematics. First, the onset of flexor activity after the foot entered the hole was delayed by approximately 100 ms in spinal animals relative to intact animals. Second, in intact animals the magnitudes of flexor bursts were increased relative to the flexor bursts associated with the swing phase during stepping, whereas in spinal animals flexor bursts during the corrective response resembled those occurring during swing. 5. Similarities in the duration and the timing of bursts in different flexor muscles in intact and spinal animals during the corrective response and during swing indicated that the corrective response involves activation of the spinal system that normally produces swing phase flexor activity. We conclude that activation of this system is facilitated by input from supraspinal structures during the corrective response in intact animals. 6. In all intact animals and three of five spinal animals, support of the hindquarters when the foot entered the hole was maintained by the contralateral leg.(ABSTRACT TRUNCATED AT 400 WORDS)

Cerebellar ataxia and muscle spindle sensitivity.

1. The cerebellum has long been known to participate in movement control. One of the enduring theories of cerebellar function is that it "tunes" and coordinates sensorimotor traffic in other parts of the CNS. In particular, it has been implicated in the control of the sensitivity of muscle spindle stretch receptors through the fusimotor system. 2. The stretch sensitivity of spindle primary endings can be varied approximately over a 10-fold range by fusimotor efferent action. For many years it has been believed that cerebellar dysfunction is associated with reduced drive to the fusimotor system and that this in turn causes hypotonia by reducing the reflex excitation of alpha-motoneurons by spindle afferents. 3. The data on which this hypothesis is based were obtained in anesthetized or decerebrate animals. Little direct information is available on animals or humans performing voluntary movements and exhibiting ataxia or other cerebellar symptoms. 4. We tested the hypothesis by recording from nine muscle spindle afferents in behaving cats before and during reversible inactivation of cerebellar interpositus and dentate nuclei. In normal cats fusimotor action varies with motor task, greatly altering spindle stretch sensitivity. We investigated whether this same range of task-related sensitivity manifested itself during ataxia. 5. We found that the full range of spindle sensitivity was still present during ataxia. We therefore conclude that the cerebellar nuclei studied are not primarily responsible for fusimotor control, nor is the ataxia primarily caused by disordered proprioceptive sensitivity.