Volume 85, Issue 15 p. 3334-3339
Mini-Review
Free Access

The brain at work: A cerebral metabolic manifestation of central fatigue?

Mads K. Dalsgaard

Corresponding Author

Mads K. Dalsgaard

Department of Anaesthesia, The Copenhagen Muscle Research Centre, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark

Department of Anaesthesia, Rigshospitalet 2041, Blegdamsvej 9, DK-2100 Copenhagen, DenmarkSearch for more papers by this author
Niels H. Secher

Niels H. Secher

Department of Anaesthesia, The Copenhagen Muscle Research Centre, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark

Search for more papers by this author
First published: 29 March 2007
Citations: 38

Abstract

Central fatigue refers to circumstances in which strength appears to be limited by the ability of the central nervous system to recruit motoneurons. Central fatigue manifests when the effort to contract skeletal muscles is intense and, thus, is aggravated when exercise is performed under stress, whereas it becomes attenuated following training. Central fatigue has not been explained, but the cerebral metabolic response to intense exercise, as to other modalities of cerebral activation, is a reduction in its “metabolic ratio” (MR), i.e., the brain's uptake of oxygen relative to that of carbohydrate. At rest the MR is close to 6 but during intense whole-body exercise it decreases to less than 3, with the uptake of lactate becoming as important as that of glucose. It remains debated what underlies this apparent inability of the brain to oxidize the carbohydrate taken up, but it may approach ∼10 mmol glucose equivalents. In the case of exercise, a concomitant uptake of ammonium for formation of amino acids may account for only ∼10% of this “extra” carbohydrate taken up. Also, accumulation of intermediates in metabolic pathways and compartmentalization of metabolism between astrocytes and neurons are avenues that have to be explored. Depletion of glycogen stores and subsequent supercompensation during periods of low neuronal activity may not only play a role but also link brain metabolism to its function. © 2007 Wiley-Liss, Inc.

Fatigue manifests after a determined mental or physical effort and is an important symptom in somatic and neuropsychiatric diseases (Chaudhuri and Behan, 2004). Fatigue established during maximal exercise is predictable, and insight into exercise-induced fatigue may be extrapolated to the fatigue associated with diseases. In patients suffering from ischemic heart disease, work capacity is limited by oxygen delivery to the heart, causing angina pectoris, insofar as, in general, physical activity hinges upon sufficient oxygen delivery to contracting muscles rather than to the capacity of the muscles for aerobic metabolism (Secher and Volianitis, 2006). Thus, provision of energy to contracting muscles is supplemented with anaerobic metabolism. Whether exhaustion is reached depends on factors within the central nervous system balancing motivation to continue exercise with the physiological ability to recruit the motor neurons as exemplified by neuronal inhibition elicited by pain. This Mini-Review addresses circumstances under which central fatigue is provoked and speculates on a link between such attenuation of brain efficiency and its metabolic response to activation.

CENTRAL FATIGUE

Central fatigue was demonstrated by the Italian physiologist A. Mosso (1904), who showed that fatigue becomes pronounced after a demanding mental task, such as giving a lecture. Since then, central fatigue has been documented in wide range of situations, and, conversely, the enhanced performance associated with so-called diverting activities may be seen as alleviating that type of fatigue. The influence of diverting activities on muscle fatigue (Setchemow's phenomenon) refers to the observation that performance is enhanced when exercise with one muscle group is supplemented by activity with another muscle group (Asmussen and Mazin, 1978a). Similarly, if exercise is continued until exhaustion with the eyes closed, opening of the eyes enhances strength, and work can be continued for some time (Asmussen and Mazin, 1978b).

Another delineation of a limitation to recruitment of muscles is the varying strength established when contractions are performed with either one or both legs. For instance, the force developed during simultaneous contraction of the legs is less than the sum of strength developed during contractions of one leg at a time, and this “leg strength paradox” is modulated by training (Secher, 1975). Additionally, training enhances the ability to resist fatigue, and this effect is so specific that it is ascribed to the increased ability to recruit motoneurons. First, the training effect of repeated one-legged contractions is to postpone the time when fatigue sets in during one-legged contractions, whereas performance with both legs remains unchanged (Rube and Secher, 1991). Conversely, two-legged training enhances endurance for both legs but not for one-leg contractions. In both cases the electromyographic activity over the muscles decreases in parallel with force. Central fatigue also presents after maximal dynamic exercise in which electrical stimulation of the motor nerve yields a greater power than evoked by voluntary contractions (Nybo and Secher, 2004). Finally, the consistent finding that an ∼25% increase in strength takes place without hypertrophy of the muscle fibers confirms that full recruitment of muscle fibers requires a central adaptation or a learning process.

With varying ability of the central nervous system to recruit the motoneurons, the pattern of muscle contractions developed during central fatigue is of interest. During partial neuromuscular blockade, two types of contractions can be delineated. With the use of an agonist drug to acetylcholine (decamethonium or suxamethonium), muscle contractions become slow but enduring. Conversely, a nondepolarizing neuromuscular blocking agent (e.g., tubocurarine) provokes contractions that maintain a high rate of rise of tension, but the developed force fades off rapidly. Because the two types of neuromuscular blocking agents affect fast and slow twitch fibers, respectively (Zaimis, 1953), the contraction pattern manifested during partial neuromuscular blockade provides some insight into the characteristics of the two main fiber types in humans. From that perspective, it seems that central fatigue affects the recruitment of slow rather than fast twitch muscle fibers, in that the contraction maintains its rate of rise of tension and, at the same time, looses its endurance (Fig. 1).

Details are in the caption following the image

Registration of force during repeated maximal voluntary handgrip contractions. As force decreases, also the ability to maintain the contraction becomes affected, although there is little reduction in the rate of rise of tension. Thus, at exhaustion (A), the contraction has a duration of ∼0.3 sec, while the normal maximal contraction (E) lasts for some ∼2 sec. Intermediate conditions of increasing fatigue are also depicted (B–D).

The central nervous system mechanisms responsible for central fatigue remain elusive. One consideration relevant to central fatigue is that stimulation of cortical areas provokes facilitation of reflexes over fast muscles and, conversely, inhibits reflexes involving slow muscles. In the following, the cerebral metabolic response to exercise will be addressed from the perspective that central fatigue may be provoked by a limited provision of oxygen and substrate to relevant areas of the brain.

CEREBRAL ENERGY METABOLISM

Although demonstration of central fatigue during different human activities may seem trivial, it has been a challenge to identify changes within the brain that correlate with hampered recruitment of the muscles. Given that the traditional markers of (global) cerebral activity, i.e., cerebral blood flow (CBF) and cerebral metabolic rate of oxygen (CMRurn:x-wiley:03604012:media:JNR21274:tex2gif-inf-1), remain unaffected during activation, including exercise, favorably, acute changes in brain metabolism are expressed in the cerebral metabolic ratio (MR; Dalsgaard, 2006). The MR relates the brain oxygen uptake to that of carbohydrate and thereby minimizes uncertainty in relation to the influence of CBF on evaluation of metabolism, although it is essential that the brain region of interest drains to the jugular vein from which blood is sampled. Cerebral metabolism depends on oxidation of carbohydrate; the brain uptake of other energy sources, e.g. ketone bodies, amino acids, and free fatty acids, is of little quantitative importance (Dalsgaard et al., 2002). Thus, at rest, MR is close to 6, but, during brain activation, MR decreases as demonstrated for the human visual cortex during exposure to light (Fox et al., 1988) and for the brain as a whole during mental stress (Madsen et al., 1995). In terms of reduction in MR, physical exercise represents an extreme, with the largest reduction in MR established during exhaustive whole-body exercise (Dalsgaard, 2006). For intense exercise-induced activation of the brain, MR decreases to values lower than 3, indicating that 50% of the carbohydrate taken up or some ∼10 mmol glucose-equivalents is not oxidized (Dalsgaard et al., 2004a).

Strenuous exercise differs from other paradigms of cerebral activation by a marked increase in arterial lactate, e.g., from ∼1 mM at rest to 15 mM or more. Whereas at rest the brain releases small amounts of lactate, when exercise increases blood lactate substantially there is a proportional cerebral lactate uptake (Dalsgaard, 2006). Seemingly, this lactate taken up by the brain is metabolized, insofar as it does not accumulate within the cerebrospinal fluid or the brain tissue (Dalsgaard et al., 2004b), so lactate is included in the MR as oxygen/(glucose + 0.5 lactate).

During exercise the molar amount of lactate taken up by the brain may approach that of glucose, so lactate becomes a quantitatively important substrate as confirmed by preliminary data on 13C-labelled lactate infusion in human subjects. Both neurons and astrocytes can metabolize lactate, and in humans lactate infusion lowers glucose consumption by the brain as determined by positron emission tomography, both at rest (Smith et al., 2003) and following exercise (Kemppainen et al., 2005). Breakdown of MR during exercise, however, does not depend on the lactate taken up by the brain. During prolonged exercise, the increase in blood lactate is negligible and, accordingly, there is no lactate uptake by the brain, but MR decreases at exhaustion (Nybo et al., 2003). In other words, lactate may contribute 30–50% to the substrate uptake of the activated brain, but this lactate uptake is not a prerequisite for the reduction in MR at exhaustion.

The “extra” carbohydrate taken up by the brain could support increased turnover of neurotransmitters. Shuttling of glutamate and glutamine between neurons and astrocytes, with its inherent need for accelerated metabolism, is likely to cause some build-up in the pools of these amino acids as well as metabolic intermediates. However, after cerebral activation in the rat, such amino acid accumulation seems less important (Dienel et al., 2002), and, even assuming that all the cerebral uptake of ammonium in exercising subjects is channelled to formation of amino acids, it accounts for less than 10% of the carbohydrate that is not oxidized (Nybo et al., 2005).

Alternatively, a transient decrease in MR during cerebral activation may be counterbalanced at a later stage. Although data on brain substrate uptake obtained during recovery suggest that the MR is restored soon after exercise, a detailed analysis beyond the initial 1 hr of recovery is pending. Of interest are periods of low cerebral activity in which the otherwise remarkably stable CBF and CMRurn:x-wiley:03604012:media:JNR21274:tex2gif-inf-3 decrease, e.g., by 25% or more, during deep sleep (Madsen and Vorstrup, 1991). An increase in MR during attenuation of cerebral activity would be inferred from the decrease observed during cerebral activation, and, accordingly, during surgical anesthesia in humans whole-brain MR becomes elevated to ∼6.5 (Fig. 2). Because the brain cannot store oxygen, such an elevated MR suggests that endogenous energy sources are oxidized. This raises the possibility that periods of physiologically low brain involvement, such as sleep, allow the brain to restore its carbon balance. Part of the restorative process may be to permit glycogen resynthesis as opposed to decreasing brain glycogen levels in the sleep-deprived rat (Kong et al., 2002).

Details are in the caption following the image

The cerebral metabolic ratio (open circles, brain uptake of oxygen relative to that of carbohydrate, oxygen/[glucose + 1/2 lactate]; solid circles, brain uptake oxygen/glucose) during general anesthesia, at rest and during various types of brain activation including several types of exercise. Data represent mean ± SE. Different from rest: *P < 0.05; †P < 0.01. Data for anesthesia by C. Selmer (unpublished). Modified from Dalsgaard (2006).

BRAIN GLYCOGEN METABOLISM

Brain glycogen is also interesting insofar as it, along with MR, decreases during cerebral activity (Swanson et al., 1992), and its level in humans is significant (Dalsgaard et al., 2006; Oz et al., 2006). Under anesthesia for surgical treatment of otherwise intractable epilepsy, the human brain glycogen concentration in vivo is ∼6 mM (glucosyl units) in apparently normal gray and white matter and as high as 13 mM in the pathological hippocampus (Dalsgaard et al., 2006). Although significant, the brain glycogen concentration constitutes only ∼10% of that in skeletal muscle, and as an energy reservoir brain glycogen seems of limited use. On the other hand, because brain glycogen is confined primarily to the astrocytes (Cataldo and Broadwell, 1986), their intracellular concentration may approach that of skeletal muscle underpinning its potential importance. Additionally, glycogen could be important for fast-accessible glucose phosphate, of which the brain possesses limited quantities. At least in a model of neuron–glia interaction, the mouse optic nerve preparation, it seems that glycogen becomes of value especially when neuronal activity is intense (Brown et al., 2003). In contrast, glycogen turnover seems slow in visually stimulated subjects as determined by 13C nuclear magnetic resonance (NMR) spectroscopy, also estimating glycogen concentration to a lower value of ∼3.5 μmol/g (Oz et al., 2006). Moreover, the most peripheral part of the astrocyte (and perhaps that of the neuron) may rely on anaerobic metabolism as potentially supported by glycogen breakdown.

As opposed to the natural flow of thoughts, exercise and test situations of cerebral activation confine stimulation to a specific brain region and may as such lead to a metabolic “drenage” of that region, e.g., by depletion of glycogen (Dalsgaard, 2006). Thus, the release of interleukin-6 from the brain during exhaustive exercise is taken to reflect an energy crisis, and even CMRurn:x-wiley:03604012:media:JNR21274:tex2gif-inf-5 may decrease during prolonged exercise (Nybo and Secher, 2004). By provoking similar crises during exercise in hypoxia, isometric contractions become more affected than fast contractions (Rasmussen et al., 2006), suggesting that recruitment of slow twitch fibers is more vulnerable to oxygen availability than that of fast twitch muscle fibers.

To the glucose taken up but not oxidized during cerebral activation is added the glucose liberated from glycogen breakdown; i.e., glycogen metabolism does not explain the immediate decline in MR. However, the duration of glycogen resynthesis is protracted (Dienel et al., 2002). If glycogen deposits were replenished to levels even higher than those before activation, i.e., supercompensation, the carbon balance could be reestablished over time.

SUMMARY

Exercise-induced brain activation decreases the cerebral uptake ratio of oxygen vs. carbohydrate, most profoundly so at the time of exhaustion, when 30–50% of the glucose plus lactate taken up is not oxidized (Fig. 3). Although unexplained, this surplus cerebral uptake of carbohydrate may reflect an enhanced need for fast-accessible glucose as supported by glycogenolysis. The astrocytes may contribute to compartmentalization of energy metabolism and to the simultaneously declining metabolic ratio and glycogen level. For brain regions continuously stimulated as during exhaustive whole-body exercise, the metabolic need could exceed provision of energy and, eventually, lead to glycogen depletion and as such play a role in central fatigue. Although the mechanism(s) remains undefined, from this perspective a low cerebral metabolic ratio may represent an impending metabolic crisis.

Details are in the caption following the image

Brain energy metabolism. Under normal conditions glucose (A) is the principal energy source for the brain, possibly (B) with the majority taken up in the glial compartment while the largest fraction of energy is expended by neurons. Whether glucose is shuttled directly to oxidation in the neuron or passes through glycolysis in the astrocyte and released as lactate (C; Pellerin and Magistretti, 1994) does not affect the MR. Glycogenolysis, which may be required in processes of the astrocyte (D), adds to the amount of glucose equivalents that seems not to become oxidized. Lactate taken up by the brain during exercise (E) may support neurons and astrocytes as both cell types metabolize lactate. Modified from Dalsgaard (2006).

    The full text of this article hosted at iucr.org is unavailable due to technical difficulties.