carroll2016

1
Invited Mini Review 1
2
Recovery of central and peripheral 3 neuromuscular fatigue after exercise 4
5
TJ Carroll 1, JL Taylor 2, SC Gandevia 2 6
7
1 Centre for Sensorimotor Performance, School of Human Movement and 8
Nutrition Sciences, University of Queensland 9
10
2 Neuroscience Research Australia and University of New South Wales 11
12
13
Running title: Recovery of Central and Peripheral Fatigue 14
15
16
Key words: recovery, central fatigue, muscle fatigue 17
Manuscript details: 18
19
Address for correspondence: 20
A/Prof Timothy Carroll 21
School of Human Movement and Nutrition Sciences 22
The University of Queensland 23
St Lucia, Qld, 4072, Australia 24
Ph: [phone]
Fax: [phone]
Email: [email protected]27
Articles in PresS. J Appl Physiol (December 8, 2016). doi:10.1152/japplphysiol.00775.2016
Copyright © 2016 by the American Physiological Society.

2
Abstract 28
29
Sustained physical exercise leads to a reduced capacity to produce voluntary force that typically outlasts 30
the exercise bout. This “fatigue” can be due both to impaired muscle function, termed “peripheral 31
fatigue”, and a reduction in the capacity of the central nervous system to activate muscles, termed 32
“central fatigue”. In this mini-review we consider the factors that determine the recovery of voluntary 33
force generating capacity after various types of exercise. After brief, high-intensity exercise there is 34
typically a rapid restitution of force that is due to recovery of central fatigue (typically within 2 min) and 35
aspects of peripheral fatigue associated with excitation-contraction coupling and re-perfusion of muscles 36
(typically within 3-5 min). Complete recovery of muscle function may be incomplete for some hours, 37
however, due to prolonged impairment in intracellular Ca2+ release or sensitivity. After low-intensity 38
exercise of long duration, voluntary force typically shows rapid, partial, recovery within the first few 39
minutes, due largely to recovery of the central, neural component. However, the ability to voluntarily 40
activate muscles may not recover completely within 30 minutes after exercise. Recovery of peripheral 41
fatigue contributes comparatively little to the fast initial force restitution, and is typically incomplete for 42
at least 20-30 minutes. Work remains to identify what factors underlie the prolonged central fatigue that 43
usually accompanies long-duration single joint and locomotor exercise, and to document how the time-44
course of neuromuscular recovery is affected by exercise intensity and duration in locomotor exercise. 45
Such information could be useful to enhance rehabilitation and sports performance. 46
47
48
Keywords 49
50
central fatigue, endurance, exercise, muscle fatigue, recovery 51
52

3
Introduction 53
Sustained physical exercise leads inexorably to a reduced capacity to produce voluntary force. Although 54
multiple processes contribute to this “muscle fatigue”, it is ultimately manifest as impaired muscle 55
function, and/or a reduction in the capacity of the central nervous system to activate muscles. The term 56
“peripheral fatigue” is typically used to describe force reductions due to processes distal to the 57
neuromuscular junction, whereas those due to processes within motoneurons and the central nervous 58
system are commonly known as “central fatigue”. The physiology of fatigue has been studied for well 59
over a century (see (28) for a comprehensive historical review), and recent reviews have summarized 60
current understanding of various aspects of fatigue (2, 13, 23, 62, 84, 92, 93). As part of this 61
Highlighted Topic series, we consider the factors that determine the recovery of voluntary force 62
generating capacity after various types of exercise. 63
64
In attempting to document the mechanisms of fatigue and recovery, an important consideration is that 65
sustained exercise affects physiological processes throughout the neuromuscular system. Critically, 66
alterations in these underlying processes may either contribute to, or compensate for, fatigue (see 67
sections below for examples and details). Although some progress has been made in documenting inter-68
relationships between exercise characteristics, physiological responses and impaired force generating 69
capacity, much remains to be learned. Given this, and the constraints of the mini-review format, we 70
consider primarily recovery of neuromuscular performance in terms of voluntary or artificially-evoked 71
forces measured in intact humans. We attempt to link these functional measures of fatigue with the 72
likely underlying processes where possible, and highlight areas in which a lack of available evidence 73
prevents this. As a consequence of our focus on work that we believe provides the clearest inferences 74
regarding the mechanisms of recovery, a number of interesting and important issues are omitted. For 75
example, we do not consider the effects of low-intensity exercise, nutrition or other interventions on the 76
recovery process. Some of these issues are dealt with in other papers in the Highlighted Topic series. 77
78
Due to the task-dependent nature of fatigue (23), the review is structured according to sections that each 79
consider the classes of exercise in which recovery has been documented. We initially consider maximal 80
voluntary contractions at a single joint, because central fatigue is most easily studied in this type of task. 81
In particular, we extrapolate recent insights into the mechanisms of central fatigue in these tasks to the 82

4
post-exercise recovery period. We then consider what general implications can be drawn about recovery 83
from apparent differences in fatigue and recovery between maximal and sustained sub-maximal 84
contractions at a single joint. Finally, we consider recovery from everyday exercise, such as running and 85
cycling, which involve large muscle masses, and consequently challenge systemic homeostasis. 86
87
Recovery from maximal contractions 88
A reduction in the maximum force that a person can produce during a voluntary isometric contraction 89
(MVC) provides the most straightforward demonstration of fatigue. Accordingly, tasks involving a 90
sustained MVC provide a convenient model to study fatigue and recovery, because there is a continuous 91
measure of fatigue during the protocol (i.e. instantaneous MVC force), and because recovery can easily 92
be tracked with the same apparatus used to induce fatigue (i.e. without the requirement to re-position the 93
subject or switch between tasks). Isometric conditions are also convenient for measurement of forces 94
evoked artificially by electrical or magnetic stimulation of motor nerves, descending tracts or the motor 95
cortex. Evoked forces at rest can provide information about fatigue and recovery of the muscle fibers, 96
whereas force responses to stimulation that is “superimposed” upon voluntary contractions can reveal 97
the extent to which voluntary neural drive is sufficient to generate the maximum force of which the 98
muscles are capable. 99
100
It is important to acknowledge that, although measures obtained during MVC provide valid and easily-101
interpreted information about neuromuscular function, such measures may not be ideally sensitive to 102
some physiological changes that are important for exercise performance. Examples that demonstrate this 103
point include observations of exacerbated force declines during low-frequency motor unit firing (16, 22, 104
83, 95, 102), and observations that some interventions such as hyperthermia (98) or prior locomotor 105
exercise (83) have much greater effects on sustained than brief MVC performance. Despite these 106
limitations, in many cases measurements obtained during MVC provide the best available evidence 107
regarding muscle force generating capacity and the capacity of the CNS to drive muscles, and the 108
current review will focus extensively on work that exploits these measures. 109
110

5
Maximal voluntary force declines rapidly and progressively during a sustained MVC, typically falling to 111
below 50% of baseline within 1-2 min. There is also a rapid but partial recovery of voluntary force over 112
the first few minutes after cessation of this type of exercise, with the largest component occurring within 113
15-30 s. This suggests that re-perfusion of the exercising muscles is a key factor in initial recovery, a 114
conclusion supported by the observation that recovery is delayed if the muscles are held ischemic. 115
Further recovery of MVC force is much slower, and may reach only ~80% by 4-5 minutes post-exercise 116
(see Fig. 1a and (29, 50, 98)). 117
118
The size of superimposed twitches evoked by stimulation of motor nerves or the motor cortex also 119
increases within 15-30s of sustained MVC, indicating that part of the voluntary force reduction is due to 120
sub-optimal output from the motor cortex (29, 40, 41, 48, 50, 51, 91, 98). This failure of voluntary 121
activation has been estimated to account for ~25% (100) of the total force reduction during sustained 122
maximal contractions, but voluntary activation usually completely recovers to pre-fatigue levels within 123
~30 s of exercise termination (see Fig. 1c and (29, 48, 50, 51, 98)). The dissociation in the time-course 124
of recovery between MVC and voluntary activation implies that the sustained impairments in voluntary 125
force production originate predominantly within the muscle fibers. Further support for this conclusion 126
derives from observations of prolonged, incomplete recovery of electrically-evoked twitches and tetani 127
following repeated isometric contractions to the limit of tolerance (22). 128
129
Mechanisms of recovery 130
A detailed coverage of what is currently known about the physiological processes that accompany 131
sustained exercise is beyond the scope of this paper, but see Taylor et al (92) and Allen et al (2) for 132
fuller accounts of central and peripheral fatigue mechanisms, respectively. Here, we provide a brief 133
overview (see Fig. 2 for a summary), and emphasize that the time-courses of change in these processes 134
need not reflect that of the functional recovery in voluntary force. This is because physiological 135
responses to sustained exercise may either contribute to, or compensate for fatigue, and recovery of 136
voluntary force is ultimately determined by the interplay of such underlying processes. For example, 137
during sustained maximal contractions, both the excitatory and the inhibitory (silent period) responses 138
of motor cortex output cells to transcranial magnetic stimulation increase. These changes suggest extra 139
cortical excitability, which should improve motor output, but also extra cortical inhibition, which might 140