The fatigued spinal cord

If we contract a muscle for a long period of time we experience fatigue: it feels as if we need more effort to produce the same amount of muscle force. Furthermore, if we continue to contract we will reach a point where it is no longer possible to produce that force. Much of the loss of force is caused by changes in the muscle itself. However, an important part of the force loss is due to a gradual reduction in the central ‘volitional’ drive to the muscle such that we fail to recruit the maximum force that it is capable of delivering. This is termed ‘central fatigue’.

In an excellent series of papers over the past few years, Gandevia and colleagues have probed the mechanisms responsible for the loss of volitional drive during fatigue (Gandevia, 2008). Two techniques have proved particularly useful: transcranial magnetic stimulation (TMS) of the motor cortex and transmastoidal stimulation of the corticospinal tract at the level of the pyramidal decussation. The latter gives information about the excitability of spinal cord whereas responses to TMS are also influenced by the excitability of intrinsic circuits of the cortex which synapse onto the corticospinal neurones. By comparing the effects of muscle contraction on each of these, the experimenters can tease out fatigue related changes in spinal motor circuits from changes in cortical motor circuits.

In a typical experiment, healthy individuals perform maximal voluntary contractions of the elbow flexors. Over a 2 min period, the force exerted typically declines by about 60–70%. Although much of this is the result of peripheral changes in muscle, central fatigue can be demonstrated by examining what happens when a TMS pulse is given to the motor cortex, in a version of the ‘twitch interpolation’ technique devised by Merton (Merton, 1954; Gandevia, 2009). At the start of the experiment a TMS stimulus that may be able to evoke a large muscle twitch at rest recruits a small additional force at the onset of a maximal contraction. This is because the muscle is being driven near maximally and it is not possible for it to exert much extra force. However, after 10–20 s, the same TMS pulse can evoke additional force over and above the voluntary force being exerted. The conclusion is that the volitional effort is failing to activate the muscle sufficiently to recruit its maximum force.

The response to TMS depends not only on the excitability of the motor cortex but also on the responsiveness of the motor circuits in the spinal cord. So where does volitional drive fail? Is it because we can no longer produce as much output from the cortex as in the fresh state, or is it because the spinal cord no longer responds to that output as strongly as it used to do, or is it a combination of the two?

One clue to this comes from looking at the EMG responses (MEP) evoked by the TMS pulse rather than the muscle twitch. We find that the MEP actually increases in amplitude throughout the contraction, even though muscle force declines. It appears as if the excitability of the corticospinal output (i.e. how easy it is to activate output with a TMS pulse) increases during the fatiguing contraction. Now this only tells us about the excitability of the system, and not about how well it can be driven by volitional effort. Nevertheless, it does open the possibility that rather than declining, perhaps our voluntary effort is able to increase corticospinal output during fatigue.

In fact, the results from transmastoidal stimulation give some clues as to whether this is a real possibility. This form of stimulation activates corticospinal axons directly and can be regarded as producing a constant level of input to spinal motor output. It can therefore be used to monitor the excitability of spinal motor output. Unexpectedly, the EMG activity evoked by transmastoidal stimulation (known as CMEPs, or cervicomedullary evoked EMG potentials) declines during a maximal contraction suggesting that, in contrast to motor cortical output, spinal motor output becomes less excitable as fatigue progresses.

This leaves three scenarios to account for the inability of voluntary output to produce maximal muscle force during a fatiguing contraction. In the first instance, voluntary drive might be able to activate cortical output to the same level as when non-fatigued; however, reduced spinal excitability might mean that this cannot recruit the maximum force that can be delivered by the muscle. The second possibility is that volitional effort could increase cortical output above non-fatigued levels, yet still fail to activate muscle maximally because of the drop in spinal motor excitability. Finally, there could be a double problem of reduced output from cortex in addition to reduced excitability of cord.

In this issue of The Journal of Physiology, McNeil et al. (2009) provide a clever way of distinguishing these possibilities. When a TMS pulse is applied during contraction, it evokes an MEP; however, this transient excitation is followed by a period in which there is complete suppression of ongoing EMG lasting 150 ms or so. This is thought to be caused by inhibition of volitional output from cortex and withdrawal of corticospinal input to spinal cord. As expected, a CMEP evoked during this period of silence is smaller than the CMEP evoked during ongoing contraction. The amount of suppression gives some indication of how much the ongoing corticospinal input was facilitating spinal motor output. During fatigue, the CMEP in the silent period is much smaller or even absent compared to that in the silent period of a non-fatigued contraction. This suggests quite strongly that the volitional drive of corticospinal output increases during fatigue in order to compensate for the reduced excitability of the spinal motor output. When this descending facilitation is removed during the silent period, the true extent to which fatigue reduces spinal excitability is exposed.

What is the cause of the reduced excitability of the spinal motor output? As the authors argue, there are several possibilities, including changes in the membrane properties of neurones that discharge rapidly during strong contractions and increased inhibitory feedback from muscle group III and IV receptors. However, the main source is likely to be reduced excitatory input from muscle spindles which are known to reduce their discharge rate during fatigue (Macefield et al. 1991). Whatever the factors, the results imply that this reduction in spinal excitability is an important contributor to central fatigue. Although volitional drive increases the output from cortex, it cannot achieve the levels necessary to overcome the reduced responsiveness of the spinal motor apparatus. The result is that we can no longer obtain the maximum force that the muscle is capable of delivering.