Exercise and Muscle Glycogen Metabolism

Abstract

Muscle glycogen is an important fuel source for contracting skeletal muscle, and it is well documented that exercise performance is impaired when the muscle’s stores of glycogen are exhausted. The role of carbohydrate (CHO) availability on exercise performance has been known for more than a century, while the specific role of muscle glycogen for muscle function has been known for half a century. Nonetheless, the precise cellular and molecular mechanisms by which glycogen availability regulates cell function and contractile-induced fatigue are unresolved. Alterations of pre-exercise muscle glycogen reserves by dietary and exercise manipulations or modifying the degree of dependency on endogenous glycogen during exercise have collectively established a close relationship between muscle glycogen and the resistance to fatigue. It is also apparent that glycogen availability regulates rates of muscle glycogenolysis and resynthesis, muscle glucose uptake, key steps in excitation-contraction coupling, and exercise-induced cell signaling regulating training adaptation. The present review provides both a historical and contemporary overview of the effects of exercise on muscle glycogen metabolism, addressing factors affecting glycogen use during exercise as well as the evolving concepts of how glycogen and glycolysis are integrated with cell function, skeletal muscle fatigue, and training adaptation.

5.1 Introduction

It is remarkable how skeletal muscle fibers can adapt acutely to provide the necessary production of energy during exercise, where a several-fold elevated energy turnover can be sustained for hours or a more than a hundred-fold increase can be executed for minutes. This ability to balance the energy turnover during various types of exercise is achieved by an integration of different energy pathways and by an efficient regulatory system, ensuring that ATP resynthesis is closely matched to the ATP demand of exercise. By the end of the 1930s, it had already been established that both fat and carbohydrate (CHO) could be used as fuel sources for aerobic metabolism during exercise and that fuel used during exercise can be modified by dietary manipulation (Zuntz 1896; Frentzel and Reach 1901; Krogh and Lindhard 1920; Edwards et al. 1934; Christensen and Hansen 1939). It was also understood that CHO was the predominant metabolic substrate when exercising at high intensities and that a relation existed between increasing CHO utilization with increased intensity. Although these early studies had documented that CHO is a major substrate during exercise and that the diet plays an important role in endurance capacity, it was the introduction of the needle biopsy technique in the 1960s that initially demonstrated that work time to exhaustion is highly correlated with muscle glycogen concentration (Bergström et al. 1967). Furthermore, it was also established that muscle glycogen content can be easily altered by isocaloric diets with varied CHO content and that at exhaustion there was a near depletion of muscle glycogen. These seminal studies on the important role of muscle glycogen on prolonged submaximal exercise performance have since been confirmed and extended numerous times. It is also well established that endurance training increases the basal stores of muscle glycogen, as demonstrated in rodent muscle already in the 1930s (Palladin 1945) and later in humans (Taylor et al. 1992; Gollnick et al. 1974). Additionally, trained humans have higher muscle glycogen content and a lower carbohydrate utilization for a given absolute submaximal exercise intensity when compared to untrained subjects (Karlsson et al. 1974; Saltin et al. 1976).

Based on the fundamental findings that carbohydrate combustion is more efficient than fat combustion (~6.2 ATP per O₂ and ~ 5.6 ATP per O_2, respectively) (Krogh and Lindhard 1920; McGilvery 1975) and that the degradation of glycogen to lactate or CO₂ and H₂O in muscles provides a more rapid energy production than that provided by the utilization of fatty acids (~1, 0.5 and 0.24 μmoles ATP g⁻¹ s⁻¹, respectively) (Margaria et al. 1964), a large body of studies have been undertaken from around the mid-1960s to today to investigate which factors influence glycogen utilization during exercise and how this affects the function of skeletal muscle fibers and athletic performance. As discussed above, a fundamental observation is that exercise performance is impaired when the muscle’s stores of glycogen are exhausted. During exercise, glycogen is utilized and can be depleted to very low levels often reaching one-fifth to one-sixth of the pre-exercise level (Gollnick et al. 1974). This is observed in humans, who are unable to withstand exercise at or above moderate intensity for a prolonged time when the stores are depleted to very low levels (<150 mmol kg⁻¹ dw) compared to pre-exercise levels of 500–900 mmol kg⁻¹ dw (Bergström et al. 1967; Hermansen et al. 1967) even with carbohydrate supplementation during the exercise (Coyle et al. 1986; Rauch et al. 1995). Therefore, the understanding of factors affecting the graded utilization of glycogen during exercise is key to avoid unforeseen glycogen-dependent muscle fatigue. Indeed, the precise mechanisms underpinning the role of glycogen in muscle function and performance are far from understood.

With this in mind, the present chapter provides both a historical and contemporary overview of the regulation of muscle glycogen metabolism during exercise in humans. After an initial discussion of glycogen storage and regulatory processes of glycogenolysis, a critical review of the factors that modulate glycogen utilization during exercise (i.e., intensity, duration, training status, modality, fiber type, sex, subcellular location, and environmental factors) is then presented. Subsequently, the evolving concepts of how glycogen and glycolysis are integrated with cell function and skeletal muscle fatigue are discussed from both a biochemical and physiological context. Finally, we close by outlining how glycogen utilization may serve as a signal to regulate cell signaling processes associated with modulating the endurance phenotype (i.e., training adaptations).

5.2 Glycogen Storage and Regulation

5.2.1 Biochemistry of the Glycogen Particle and Its Turnover

Glycogen is a unique molecule among several glucose polymers found in nature with structural and energy storage functions. Polymers of glucose with structural function include chitin (polymer of n-acetylglucosamine, a derivative of glucose), predominantly in arthropods and fungi, and cellulose in plants, algae, and oomycetes, which represents the most abundant polymer on earth (Klemm et al. 2005). For energy storage, the main polymers of glucose are starch in plants and glycogen which is by far the most widespread form of storage, found in archaea, bacteria, and eukaryotes (from protozoa and fungi to mammals) (Ball et al. 2011). Synthesis of glycogen through digestion of other glucose polymers is possible in humans for which starch represents a digestible form, while other mammals (e.g., ruminants and ungulates) can also digest cellulose by endosymbiont bacteria in the gut. In this context, two facts make glycogen standout: (1) it is a highly evolutionarily conserved molecule in prokaryotes and from unicellular eukaryotes to mammals, and (2) it represents the largest storage form of energy for high metabolic power output processes in mammalian cells, for which different mechanisms have evolved to maximize its synthesis and storage. These facts attest its key role in energy storage and place it in the centerstage of cellular energy production in skeletal muscle for exercise.

The glycogen polysaccharide is termed as a glycogen molecule, granule, or particle (Fig. 5.1). Due to its physical association with several proteins, it has also been termed glycosome to indicate its more organelle-like structure (Rybicka 1996). When glucose enters the cell, it is phosphorylated to glucose 6-phosphate by hexokinase I and can either be metabolized in the glycolytic pathway or be added to existing glycogen particles by the action of phosphoglucomutase, UDP-glucose pyrophosphorylase, glycogen synthase, and branching enzyme (Roach et al. 2012). While the first two steps activate the glucose, the glycogen synthase catalyzes the reaction, where a ɑ-1,4-glycosidic linkage connects the new glycosyl units with the nonreducing end of a chain of glycosyl units. The chain grows to a size of 11–13 glycosyl units. The branching enzyme transfers a set of glycosyl units from a chain to another position of a chain by creating ɑ-1,6-glycosidic linkages. Since each chain branches out two times with new chains organized in concentric tiers, the number of glycosyl units in the glycogen particle increases exponentially from around 200 in a small particle of 4 tiers (diameter of 11 nm) to around 55,000 in the largest particle of 12 tiers (diameter of 42 nm). In resting skeletal muscles, the typical (around 80% of all particles observed) diameter of glycogen particles is 20–32 nm (Marchand et al. 2002; Nielsen et al. 2010; Hokken et al. 2020), corresponding to 1000 to 9000 glycosyl units per particle (Melendez-Hevia et al. 1993). This structure of the glycogen particle is probably optimized by evolution to increase solubility and decrease the osmotic effect (Melendez-Hevia et al. 1993). Despite this, the relatively high numerical density of glycogen particles equals an osmotic effect of around 3 grams of water per 1 gram of glycogen (Olsson and Saltin 1970), which can accumulate up to 2 kilograms of glycogen-dependent stored water in humans during glycogen-loaded conditions (Olsson and Saltin 1970; Shiose et al. 2016).

Fig. 5.1

Schematic diagram of a glycogen particle (with four concentric tiers of chains of glycosyl units) and the turnover of glycosyl units to lactate generating a high metabolic power or to CO₂ and H₂O ensuring a high metabolic capacity. G glycogenin, GP glycogen phosphorylase, G1P glucose-1-phosphate, G6P glucose-6-phosphate, TCA tricarboxylic acid cycle, ETC electron transport chain

Although the true numerical density of glycogen particles cannot be investigated with the currently available techniques, measures of the size of the observable glycogen particles indicate that the utilization of glycogen during exercise is mainly attributable to a decrease in the average particle size (Marchand et al. 2007; Gejl et al. 2017a, b; Hokken et al. 2020); on the other hand, with training and diet interventions, the increase in resting glycogen content is ascribed to an increase in particle number (Nielsen et al. 2010; Jensen et al. 2021). Thus, in resting muscles, there may be a preferable size of particles, which could be the result of a trade-off between storage efficiency (larger particles store more glycosyl units per volume) and metabolic power (smaller particles possess a higher glycogenolytic rate) as also suggested by Shearer and Graham (2004). An increase in glycogen particle number has been recognized to require a de novo synthesis of the self-glycosylating protein backbone glycogenin (Roach et al. 2012), but recent studies have suggested an increase in glycogen particle number based on a mechanism, which is independent on glycogenin (Testoni et al. 2017; Visuttijai et al. 2020).

The degradation of glycogen is conducted by the action of glycogen phosphorylase, which cuts the ɑ-1,4-glycosidic linkages and liberates the glycosyl units to the glycolytic pathway. Interestingly, given the branched structure of the glycogen particle, around one third of the glycosyl units are in the outermost tier (Melendez-Hevia et al. 1993) and readily available for glycogen phosphorylase securing a fast mobilization of energy. A more pronounced degradation of the particle requires a coordinated action of glycogen debranching enzyme cutting the ɑ-1,6-glycosidic linkages. As discussed, the breakdown of muscle glycogen to produce glucose 1-phosphate is thus under the control of glycogen phosphorylase, and this reaction requires both glycogen and P_i as substrates. Phosphorylase, in turn, exists as a more active a form (which is under the control of phosphorylation by phosphorylase kinase) and also as a more inactive b form (which exists in a dephosphorylated form due to the action of protein phosphatase 1). Phosphorylase can be transformed via covalent modification (i.e., phosphorylation by phosphorylase kinase) as mediated through epinephrine (Roach et al. 2012). Additionally, Ca²⁺ is a potent positive allosteric regulator of phosphorylase kinase through binding to the calmodulin subunit (Jensen and Richter 2012), though glycolytic flux is not controlled by Ca²⁺ directly, but by factors related to energy state (Ørtenblad et al. 2009). During contractile activity, the increased accumulation of Pi as a result of increased ATP hydrolysis can increase the rate of glycogenolysis as it provides increased substrate required for the reaction. Furthermore, greater accumulations of free ADP and AMP can also subsequently fine-tune the activity of phosphorylase a through allosteric regulation (Howlett et al. 1998). When taken collectively, the regulation of glycogen phosphorylase is dependent on hormonal control, substrate availability, and local allosteric regulation, the precise contribution of which is dependent on the specific exercise challenge.

5.2.2 Measurement of Muscle Glycogen

The golden standard for determination of muscle glycogen is the biochemical technique where the glycosyl units are liberated by acid hydrolysis and the amount determined spectrophotometrically (Passonneau and Lowry 1993). This is performed in skeletal muscle biopsies in small specimens of ~10 mg (wet weight), and with small adjustments can also be applied to single fibers (Hintz et al. 1982). The latter approach makes it possible to directly combine glycogen content with other biochemical or physiological measures and to distinguish between different fiber types. Noninvasively, nuclear magnetic resonance (NMR) has been optimized to detect muscle glycogen signals (Taylor et al. 1992) and can serve as an alternative to the biochemical assessment when biopsies are unsuitable for the target group.

To investigate fiber-type-specific glycogen, a semi-quantitative histochemical periodic acid-Schiff (PAS) staining has been employed in a large body of studies. Here, the glycogen content is assessed by staining intensity from empty to full using a scale of 4–5 steps and compared with the myosin ATPase characteristics (Pearse 1961). The PAS staining is based on a reaction of periodic acid with all sugars in the muscle, and it is therefore not specific for glycogen. It has been suggested to be replaced by an antibody-based technique (Nakamura-Tsuruta et al. 2012; Skurat et al. 2017), which can also be used by dot blotting (Albers et al. 2015).

With the binding properties of glycogen particles to a reduced form of osmium, a protocol for staining glycogen in transmission electron microscopy has also been developed (de Bruijn 1973; Marchand et al. 2002). With this approach, the subcellular distribution of glycogen can be envisaged along with information on fiber types (based on Z-discs and M-band appearances) and other ultrastructural parameters (Sjöström et al. 1982b).

5.2.3 Inter-fiber Variability and Subcellular Differences

While most studies on glycogen content are conducted using homogenates of small pieces of muscle, a string of studies have shown large inter-fiber variability ranging from around 100 to 1000 mmol kg⁻¹ dw (Essén and Henriksson 1974). With discrimination between fiber types based on myosin ATPase activity, it has repeatedly been shown that type 2 fibers contain about 10–30% more glycogen than type 1 fibers in human skeletal muscles (e.g., Essén and Henriksson 1974; Ball-Burnett et al. 1991; Greenhaff et al. 1993). At the subcellular level, glycogen particles are dispersed heterogeneously throughout the myoplasm with one large pool located in the intermyofibrillar space close to sarcoplasmic reticulum and mitochondria and two small pools located in the intramyofibrillar and subsarcolemmal spaces, respectively (Sjöström et al. 1982a; Fridén et al. 1985, 1989; Marchand et al. 2002, see Fig. 5.4, Sect. 5.3). Although all the pools are utilized during exercise taxing the endogenous glycogen stores (see later), their content seems not to be related in resting muscles (Nielsen et al. 2010), suggesting that local independent factors are involved in their regulation.

5.3 Utilization of Glycogen During Exercise

As alluded to earlier, a series of studies from Scandinavian researchers during the late 1960s and early 1970s collectively demonstrated the role of skeletal muscle glycogen as a key fuel source for prolonged exercise capacity, particularly when completed at higher exercise intensities (Bergström and Hultman 1966a, b; Bergström et al. 1967; Saltin and Karlsson 1971). For example, pioneering work from Bergström et al. (1967) highlighted the role of dietary carbohydrate in elevating muscle glycogen reserves and subsequently demonstrated that high pre-exercise muscle glycogen concentration (~890 mmol.kg⁻¹ dw) extends exercise capacity at 75% of VO_2max by a remarkable 320% (from 59 to 189 min) when compared to low (~170 mmol.kg⁻¹ dw) muscle glycogen stores. It was later demonstrated that increasing exercise intensity results in an exponential increase in the rate of muscle glycogen utilization (Saltin and Karlsson 1971).

Further association between starting skeletal muscle glycogen and prolonged aerobic exercise performance (Karlsson and Saltin 1971) solidified the central role of skeletal muscle glycogen as key substrate for intense exercise of prolonged duration. These seminal studies paved the way for the next 50 years of research examining the dietary and exercise-related factors that can subsequently affect the pattern of glycogen utilization during exercise. Such factors have been the subject of a recent meta-analysis (Areta and Hopkins 2018) and are portrayed conceptually in Fig. 5.2. The text to follow provides a critical overview of factors that can affect glycogen utilization during exercise.

Fig. 5.2

Conceptual representation of skeletal muscle glycogen utilization dynamics at different intensities and durations, based on mathematical modelling of data extracted from a large sample of published literature (Areta and Hopkins 2018). Note that the starting muscle glycogen has been set constant at 600 mmol^. kg⁻¹ dw for exercise of intensities ranging from equivalent to 40 to 120% of VO_2max, ranging from 5 to 120 min duration

5.3.1 Duration and Intensity

Although the duration of exercise that can be maintained is inevitably dependent on exercise intensity, the effect of intensity and duration on muscle glycogen utilization can be analyzed independently. Glycogen utilization increases in accordance with increased exercise duration though it is noteworthy that the rates of utilization vary at different stages during exercise. For example, during the first ~20 min of exercise, the rates of glycogen utilization seem to be the highest, a factor that is likely related to higher activity of glycogenolytic enzymes and lower availability of other metabolic substrates (Green et al. 1995; Chesley et al. 1996; Dyck et al. 1996; Graham et al. 2001), over and above the fact that higher glycogen concentration per se is associated with increased glycogen utilization (see below substrate availability and glycogen utilization). Thereafter, rates of utilization during prolonged exercise seem to be rather constant until they reach low levels (~200 mmol^. kg⁻¹ dw) at which point the rates are then reduced significantly (Coyle et al. 1986; Bosch et al. 1993, 1996). Fatigue during prolonged submaximal exercise (i.e., 60–70% of VO_2max) typically occurs after a duration of 2 h with glycogen at exhaustion within the range of ~100 to 200 mmol^.kg⁻¹ dw (Areta and Hopkins 2018). Indeed, early studies on muscle glycogen and athletic performance suggested that the inability to maintain high rates of work would coincide with theoretical muscle glycogen content of 70–120 mmol^.kg⁻¹ dw during an endurance competition of ~135 min (Karlsson and Saltin 1971). With this in mind, it is most likely that only competitive events lasting >90 min would benefit from CHO loading strategies that aim to super compensate pre-exercise muscle glycogen stores, where performance has been suggested to improve by ~2–3% (Hawley et al. 1997a, b).

During intense exercise, ATP provision is achieved principally by the oxidation of carbohydrate, and muscle glycogen utilization increases exponentially with exercise intensity. At intensities ranging from 75% VO_2max to near maximal workloads, glycogen is the main energy substrate (Saltin 1973; Gaitanos et al. 1993; Hultman and Greenhaff 1999). Thus, even a single 6 s or 30 s all-out sprint can reduce muscle glycogen by15 or 20–30%, respectively (Gaitanos et al. 1993; Bogdanis et al. 1996;Parolin et al. 1999). Such high glycogen utilization is achieved by estimated glycogenolytic rates of around 4.5 mmol glucosyl units^.kg^-1.dw^.s⁻¹ (Gaitanos et al. 1993; Parolin et al. 1999). During longer durations of high-intensity exercise, the glycogenolytic rate is known to decreases though it is noteworthy that muscle glycogen stores in both arms and legs are reduced by 20–25% during 4 min of high-intensity all-out cross-country sprint skiing (Gejl et al. 2014).

The first study to systematically investigate the effect of intensity demonstrated a clear exponential increase in muscle glycogen use (Fig. 5.3), with a disproportionate increase in rates at intensities close to VO_2max and above (Saltin and Karlsson 1971). More recent studies addressing the contribution of different energy substrates using metabolic tracers show a clear disproportionate increase in reliance on skeletal muscle glycogen with increasing intensities when compared to all other fuel sources including intracellular lipids and plasma lipids and glucose (Romijn et al. 1993; van Loon et al. 2001). Corroborating these findings, a meta-analytic evaluation of the effect of exercise intensity on muscle glycogen use demonstrates moderate and large effects with increasing exercise intensity, albeit with substantial variation between studies (Areta and Hopkins 2018). The variation in glycogen use observed in different studies and individuals at the same intensity may be related to the fact that % VO_2max is typically used as the “default” intensity parameter, and this parameter may not necessarily match the metabolic and substrate demands of the effort at the muscular level. Indeed, muscular oxidative capacity can vary greatly between individuals with the same VO_2max (Holloszy 1973; Holloszy and Coyle 1984), and lactate threshold represents a more suitable method for which to match intensity within and between studies (Coyle et al. 1988; Poole et al. 2020).

Fig. 5.3

Relationship between exercise intensity and rates of skeletal muscle glycogen utilization. Linear increases in exercise intensity (as % of VO_2max equivalent) result in exponential increases in skeletal muscle glycogen utilization, as shown by early research (Figure redrawn from Saltin and Karlsson 1971)

Given that phosphorylase can be transformed via covalent modification (i.e., phosphorylation by phosphorylase kinase) mediated through epinephrine, it would be reasonable to expect that greater phosphorylase transformation from b to a may be one mechanism to explain increased glycogenolysis that is evident with increasing exercise intensity. This would also be logical given that sarcoplasmic Ca²⁺ levels would be increased with high-intensity exercise (given the need for more rapid cross-bridge cycling) and that Ca²⁺ is a potent positive allosteric regulator of phosphorylase kinase through binding to the calmodulin subunit. However, the percentage of phosphorylase in the more active a form does not appear to be increased with exercise intensity and in actual fact is decreased after only 10 min of high-intensity exercise, which may be related to the reduced pH associated with intense exercise (Howlett et al. 1998). Whereas this mechanism of transformation (mediated by Ca²⁺ signaling) may be in operation within seconds of the onset of contraction (Parolin et al. 1999), it appears that post-transformational mechanisms are in operation during more prolonged periods of high-intensity exercise given that glycogenolysis still occurs despite reduced transformation. In this regard, vital signals related to the energy status of the cell play a more prominent role. Indeed, as exercise intensity progresses from moderate- to high-intensity exercise, the rate of ATP hydrolysis increases so much so that there is a greater accumulation of ADP, AMP, and Pi, thus providing allosteric and substrate level control (Howlett et al. 1998).

One final consideration is that the rapid decrease in muscle glycogen at intensities close to and above VO_2max most likely represents mere glycogenolysis, rather than oxidation of its glucose units. For example, all-out exercise of ~2 min at an intensity well above the VO_2max decreases muscle glycogen by about 25% (Medbø 1993; Medbø et al. 2006), but the net oxidation of the glycogen utilized is estimated to be ~4–13% (Medbø 1993). In this scenario, most of it ends as lactate, of which half is released to the bloodstream and the other half is utilized for rapid muscle glycogen resynthesis within the muscle (Medbø et al. 2006). Bangsbo et al. (1991) estimated that between 13 and 27% of lactate released from high-intensity exercise was converted back to glycogen (Bangsbo et al. 1991). From a practical perspective, it is worth noting that despite the disproportionate increase of skeletal muscle glycogen with increased intensity, prolonged steady-state exercise results in increased absolute muscle glycogen utilization when compared with shorter high-intensity intervals (Impey et al. 2020).

5.3.2 Substrate Availability

Although exercise intensity and duration are key drivers of glycogen use, high or low glycogen availability can result in higher or lower glycogenolysis, respectively, despite divergent exercise intensities (Arkinstall et al. 2004). Due to the significant effects of nutrient availability on glycogen utilization dynamics and exercise performance, manipulation of substrate availability and skeletal muscle glycogen has therefore been a major area of focus in research. Indeed, given the relationship between high muscle glycogen and increased work capacity, a number of dietary practices manipulating substrate availability have been tested to maximize muscle glycogen stores or to minimize its use during exercise in order to further increase work capacity. The most noteworthy practices manipulating substrate availability are the increase of muscle glycogen itself, the increase in availability and oxidative capacity of lipids (through diet and intravenous provision of lipids or lipolytic agents), and the ingestion of glucose during exercise.

5.3.2.1 Increased Muscle Glycogen Stores and Glycogen Use

Skeletal muscle glycogen content is directly related to dietary carbohydrate as shown by seminal studies (Bergström et al. 1967; Gollnick et al. 1972), a finding that has systematically been corroborated in the literature (Areta and Hopkins 2018). Depleting muscle glycogen through exhaustive exercise followed by high carbohydrate intake has been shown to subsequently overshoot skeletal muscle glycogen reserves above normal resting values in the previously contracted muscle (Bergström and Hultman 1966a, b), a phenomenon that was termed as “glycogen supercompensation.” The precise molecular mechanisms underpinning the supercompensation effect remain an active area of research (Hingst et al. 2018). Since the original work documenting this effect, it is now accepted that prior depletion of muscle glycogen is not necessary for individuals who regularly practice aerobic-type exercise and that a taper of training load in conjunction with elevated dietary carbohydrate intake of ~10–12 g/kg of body mass/day can maximize skeletal muscle glycogen concentration within 24–48 h (Burke et al. 2017). Regardless, higher resting muscle glycogen leads to higher rates of muscle glycogen use during exercise, likely a reflection of substrate regulation of glycogen phosphorylase (Hargreaves et al. 1995). Indeed, when intensity and duration of exercise are kept constant, commencing exercise with high muscle glycogen (e.g., > 600 mmol^.kg⁻¹ dw) leads to higher muscle glycogen use compared to normal (e.g., ~450 mmol^.kg⁻¹ dw) or low (e.g., <300 mmol^.kg⁻¹ dw) starting muscle glycogen (Galbo et al. 1979; Hargreaves et al. 1995; Areta and Hopkins 2018). Accordingly, when examining performance testing data of prolonged endurance exercise of a set distance and workload, exercise is often finished with similar muscle glycogen levels despite significant differences in starting muscle glycogen (Karlsson and Saltin 1971; Sherman et al. 1981; Rauch et al. 1995; Tomcik et al. 2018).

5.3.2.2 Increased Lipid Availability and Glycogen Sparing

Given the comparatively unlimited storage of energy in the form of fat, increasing the use of lipids as a fuel source during exercise to “spare” muscle glycogen with the goal of improving physical performance has been a topic of intense research. Intravenous lipid infusion, lipolytic agents, high-fat meals, and high-fat diets have been all been used to test this hypothesis. Interventions using intralipid infusion and heparin have been consistently shown to spare muscle glycogen between ~15 and 45% at moderate and high intensities lasting 15–60 min (Vukovich et al. 1993; Dyck et al. 1993, 1996; Odland et al. 1998; Hawley 2002). The effects of a high-fat meal consumed prior to exercise, however, have been less consistent with studies ranging from 40% glycogen sparing to no glycogen sparing at all (Costill et al. 1977; Vukovich et al. 1993; Hawley 2002;). The possibility of sparing skeletal muscle glycogen via dietary fat intake has been expanded through the use of high-fat diets during days and weeks.

The use of high-fat diets consumed over durations greater than 3 days has been termed “fat adaptation” and is effective in increasing circulating free fatty acids (FFA), FFA uptake by muscle, and increased fat oxidation during exercise (Stellingwerff et al. 2006). The increased rates of fat oxidation, however, occur in parallel with decreased skeletal muscle glycogen due to reduced dietary carbohydrate intake (Hammond et al. 2019; Areta et al. 2020) and hence, as detailed in the previous section, lower muscle glycogen results in lower glycogen utilization. Nonetheless, 24 h of restoration of carbohydrate intake after days of fat adaptation permits skeletal muscle glycogen storage while still retaining the increased capacity to utilize higher amounts of fat during exercise and spare glycogen to some extent (Burke et al. 2000; Stellingwerff et al. 2006). However, fat adaptation protocols also decrease pyruvate dehydrogenase (PDH) activity (Stellingwerff et al. 2006), the enzyme that represents a rate-limiting step in the flux of glucose into mitochondrial ATP, and thereby this condition impairs the capacity of generating metabolic power through the oxidation of glycogen. Therefore, in the majority of the studies to date, fat adaptation has not been able to enhance physical performance in endurance exercise, likely due to impairment of the capacity for high-intensity exercise (Burke 2006). However, over prolonged exercise (>5 h), fat adaptation with carbohydrate restoration is potentially beneficial for performance (Rowlands and Hopkins 2002). In conclusion, increasing fat availability acutely and chronically has the capacity to reduce muscle glycogen use during exercise, likely due to the independent and combined effects of increasing fat utilization and decreasing the capacity to oxidize glucose. Nonetheless, these strategies do not seem conducive to promoting exercise performance.

5.3.2.3 CHO Supplementation During Exercise and Glycogen Sparing

The fact that carbohydrate consumption during exercise increases carbohydrate oxidation led to the idea that muscle glycogen would be spared as a fuel source at its expense (Coyle et al. 1986). This intuitive idea, however, has not been supported by research findings. When considering whole skeletal muscle glycogen, a meta-analytic evaluation of CHO supplementation during exercise (n = 24 studies) determined there was no glycogen-sparing effect (Areta and Hopkins 2018). It is important to highlight that this meta-analysis only evaluated the presence or absence of CHO consumption. Indeed, although the effects of consuming 0, 15, 30, and 60 g/h of CHO showed no differences in glycogen use during exercise (Smith et al. 2010), a more nuanced analysis should perhaps determine whether low, medium, and high amounts of carbohydrate (e.g., <30, ~60, and ≥ 90 g/h respectively) across different exercise durations can affect glycogen utilization. No unequivocal evidence exists to support sparing with specific fibers, with reports of no sparing in either fiber type (Coyle et al. 1986; Mitchell et al. 1989), sparing in type 2a fibers (De Bock et al. 2006), and sparing in type 1 fibers (Yaspelkis et al. 1993; Tsintzas et al. 1995, 1996; Tsintzas et al. 2001). Such differences between studies may be related to differences in CHO availability prior to exercise, the CHO feeding strategy, and contrasting exercise protocols related to duration, modality, and intensity. In a recent study, Fell et al. (2021) demonstrated that consuming 45 or 90 g/h of CHO (in the form of solids, gels, and fluids) does not spare glycogen use in type 1 or 2 fibers when compared with no CHO ingestion, as assessed during 3 h of steady-state cycling conducted at lactate threshold. Importantly, exercise was commenced after a 36 h CHO loading protocol of 12 g/kg as well as consumption of a pre-exercise meal of 2 g/kg, thus replicating nutritional practices of elite road cyclists. Despite the apparent consensus that CHO feeding does not exert glycogen sparing, future studies should also examine the effects of CHO feeding on utilization within the subcellular glycogen storage pools of both type 1 and 2 fibers.

5.3.3 Training Status

For a given absolute exercise intensity, longitudinal studies demonstrate that glycogen utilization is reduced with exercise training (Karlsson et al. 1974), an effect that is confined locally to the actual muscles that were trained (Saltin et al. 1976). The reduced glycogenolysis observed after training is not due to any change in phosphorylation transformation but rather allosteric mechanisms (Chesley et al. 1996). Indeed, exercise in the trained state is associated with reduced content of ADP, AMP, and Pi thereby providing a mechanism leading to reduced phosphorylase activity. Importantly, the reduced rates of glycogenolysis that are evident with training are also apparent despite the fact that training induces elevations in resting glycogen stores and that higher basal glycogen is normally associated with increased glycogen utilization. As such, local allosteric control exerts a more pronounced regulatory role than substrate-level regulation. On this basis, it is now accepted that glycogen utilization during exercise is inversely related to training status and that exercise in the trained state, at same absolute intensity, requires comparatively less muscle glycogen. However, despite this well-documented finding, a meta-analytic evaluation of the effects of training status on glycogen utilization has shown only a trivial or small effect in relation to increases of VO_2max of 10 ml/kg/BM (Areta and Hopkins 2018).

This latter finding may be attributable to the fact that VO_2max is not an ideal parameter of training status when assessing changes in local muscle metabolism, despite it being considered as the main parameter for maximal aerobic capacity. Rather, a parameter of training status (and indeed exercise intensity) that is more reflective of skeletal muscle oxidative capacity is likely a more suitable approach to assessing the changes in glycogen utilization associated with training. For example, as little as 3 days of prolonged (2 h) aerobic-type training induces glycogen sparing by ~60% when assessed during exercise undertaken at 65% VO_2peak in healthy individuals (Green et al. 1995). As such, parameters of training status that are more indicative of the oxidative capacity of skeletal muscle (i.e., lactate threshold) are likely a better parameter for classification of status when determining the effects of training on skeletal muscle glycogen utilization. Indeed, individuals with comparable VO_2max but who possess a “high” or “low” lactate threshold present with distinct differences in glycogen utilization when cycling for 30 min at 80% VO_2max, i.e., subjects with a low threshold utilize more than twice as much muscle glycogen during 30 min exercise (Coyle et al. 1988).

In summary, substantial evidence demonstrate that a more trained skeletal muscle (i.e., with higher oxidative capacity) is less reliant on skeletal muscle glycogen use during exercise. However, there is a large variation in rates of muscle glycogen use in individuals when exercising at a percentage of their VO_2max. The use of a parameter of intensity and training status representative of skeletal muscle oxidative capacity will be important to further characterize the estimation of muscle glycogen use during exercise.

5.3.4 Exercise Mode

5.3.4.1 Running vs. Cycling

When comparing glycogen utilization within the vastus lateralis muscle, it has been consistently demonstrated that cycling induces greater absolute utilization than when running at a matched relative exercise intensity and duration. For example, Arkinstall et al. (2004) compared glycogen utilization in the vastus lateralis of moderately trained males during 60 minutes of cycling and running at lactate threshold and observed an absolute utilization of approximately 220 and 120 mmol^.kg⁻¹ dw, respectively. Accordingly, a recent meta-analysis demonstrated that absolute glycogen utilization is “small but very likely” reduced during matched protocols of running versus cycling (i.e., relative exercise intensity and duration), where the expected reduction in running equates to 70 mmol^.kg⁻¹ dw (Areta and Hopkins 2018). Such differences between modalities are, of course, reflective of greater recruitment of the vastus lateralis muscle during cycling when compared with running. The role of muscle recruitment in modulating glycogen utilization is also evident during running where both the absolute and rate of glycogen utilization are higher in the gastrocnemius muscle when compared with the vastus lateralis (Costill et al. 1974). This pattern of greater glycogen utilization within the gastrocnemius versus vastus lateralis when running is evident in both males and females during moderate- and high-intensity exercise protocols (Impey et al. 2020).

5.3.4.2 Resistance Exercise

Resistance exercise sessions typically consist of several sets (interspaced by pauses of 1–4 min) of repeated near maximal force productions. Here, the short duration and the high intensity of each force production necessitate a predominant use of the rapid energy systems taxing glycogen and creatine phosphate as substrates. During recovery periods between sets, however, other substrates such as intramuscular triglycerides can also contribute to the overall energy provision (Koopman et al. 2006). The most important factor for the utilization of glycogen during resistance exercise seems to be the total work performed rather than the exercise intensity (Robergs et al. 1991). Typical high-volume resistance exercise sessions decrease skeletal muscle glycogen by about 20–40% (100–250 mmol^.kg⁻¹ dw) (Tesch et al. 1986; Essén-Gustavsson and Tesch 1990; Robergs et al. 1991; Pascoe et al. 1993; MacDougal et al. 1999; Haff et al. 2003; Harber et al. 2010; Samuelson et al. 2016), but, in contrast to endurance exercise, the decrease is larger in type 2 fibers than in type 1 fibers (Tesch et al. 1998; Koopman et al. 2006; Morton et al. 2019).

5.3.5 Lower vs Upper Body

Vastus lateralis of the quadriceps has been the muscle predominantly sampled in research on skeletal muscle glycogen and used as a model that is normally extrapolated to all other muscles. However, several lines of evidence point to different intrinsic regulations of metabolism in the upper and lower body. These differences include a higher relative fat oxidation in the leg compared with arm exercise (Helge et al. 2008; Larsen et al. 2009), and a higher mitochondrial oxygen flux is present in the vastus lateralis compared to the arm (deltoid) muscle measured by high-resolution respirometry (Larsen et al. 2009). Further, it is well recognized that lactate release is higher during arm compared with leg exercise of comparable intensity (Jensen-Urstad and Ahlborg 1992; Jensen-Urstad et al. 1993). Thus, Ahlborg and Jensen-Urstad (1991) had two groups performing arm cranking and leg cycling, respectively, and demonstrated a higher relative carbohydrate utilization during arm cranking due to both a higher muscle glycogenolysis and a higher glucose uptake, while arms also had a higher lactate release (Ahlborg and Jensen-Urstad 1991). In a later study, the glucose uptake was not different between limbs when expressed per muscle volume, whereas the lactate release was noticeably higher in the arm than in the leg (Jensen-Urstad et al. 1993). Overall, this suggests that exercising arm muscle displays a different metabolic response compared with leg muscle, with a higher glycogen utilization and lactate release compared to leg muscle during same relative exercise intensity, implying a possible qualitative difference between muscles from the arm vs. the leg. However, another explanation for the differences described above may simply be that the upper body musculature, including the arm muscle, resides in less trained musculature than that of the lower body, as in the leg muscles (see Sect. 5.3.3). Still, when arm-trained athletes are exercising with either arm or leg exercise at the same relative intensity, there was a markedly higher lactate release during arm exercise (Jensen-Urstad and Ahlborg 1992). Also, in highly trained cross-country skiers, a net lactate release from arms and uptake by legs has been demonstrated during exercise involving both the upper and lower body (van Hall et al. 2003). In line with this, direct comparisons of the highly trained arm and leg muscles of elite cross-country skiers reveal that despite same mitochondrial volume percentage and citrate synthase activity in the legs and arms, the muscles exhibited clear difference in their enzyme-linked ability to oxidize fatty acids (HAD capacity) and a fourfold higher intramyocellular lipid volume contents in leg muscles (Ørtenblad et al. 2018; Koh et al. 2018). These data point to a clear limb difference in metabolism between the leg and the arm, which cannot be explained by training status or different fiber-type distributions. Taken together, the current data suggests that when exercising at the same relative intensity, arm muscle has a higher muscle glycogen use and lactate release as compared to the leg. This is also apparent in subjects with trained both upper and lower body.

5.3.6 Temperature

Exercise in the heat results in exaggerated fatigue concomitant with major alterations in several physiological and metabolic factors. The majority of the research conducted on the effects of heat stress on energy metabolism during exercise has demonstrated a shift toward increased carbohydrate use. During exercise in the heat, the rate of muscle glycogen degradation is significantly increased (Fink et al. 1975; Febbraio et al. 1994) with an increase in both carbohydrate oxidation and lactate accumulation at a given exercise intensity, while the muscle glucose uptake and utilization appear to be unaltered during exercise in the heat, despite hyperglycemia and an augmented liver glucose output. Thus, the increase in carbohydrate utilization is largely explained by an increased muscle glycogenolysis observed via both aerobic and anaerobic energy turnover. The mechanisms thought to be responsible for the enhanced muscle glycogenolysis likely are due to increased sympatho-adrenal response and increased muscle temperature (Hargreaves et al. 1996). Although exercise in the heat increases the intramuscular glycogen utilization (Fink et al. 1975; Febbraio et al. 1994; Hargreaves et al. 1996), depletion seems not to be the cause of fatigue during exercise in the heat. Thus, a general observation is that muscle glycogen stores are not critically low at fatigue and exhaustion, i.e., <250–300 mmol^.kg⁻¹ dw, suggesting that exercise in the heat is terminated before available glycogen stores have been limiting. Further, the total amount of carbohydrate oxidized during exercise in the heat is relatively low as exercise time is shorter than exercising in moderate or low temperatures (Fink et al. 1975; Febbraio et al. 1994; Galloway and Maughan 1997). Still, diet-induced increase in muscle glycogen before exercise is associated with enhanced exercise capacity in the heat, and carbohydrate ingestion during exercise increases exercise capacity in the heat (Pitsiladis and Maughan 1999; Carter et al. 2003). These outcomes cannot be explained by direct effects on either hyperthermia or substrate depletion but may exert an ergogenic effect related to factors peripheral to the muscle. Taken together, exercising in the heat results in an increase in intramuscular glycogen utilization; however, fatigue in these circumstances appears to be related to factors other than muscle glycogen per se.

5.3.7 Altitude

As the contribution of carbohydrate is determined by the exercise intensity, the hypoxic exposure experienced at altitude causes, in most studies, an increase in relative carbohydrate utilization, when exercising at the same absolute intensity as at sea level (Young et al. 1991; Brooks et al. 1991; Brooks 1992; Roberts et al. 1996a, b; Lundby et al. 2004; Katayama et al. 2009), but not all (Braun et al. 2000; O'Hara et al. 2017; Matu et al. 2018). A more direct comparison of possible effects on substrate utilization during exercise at hypoxic conditions can be gained by comparing exercise at sea level and at altitude, matched for the same relative intensities. A meta-analysis of the effects of exposure to hypoxia during exercise matched for relative intensities, compared with normoxia, demonstrated no consistent change in the relative contribution of carbohydrate to the total energy yield (Griffiths et al. 2019). This has been evidently demonstrated in a study where subjects exercised at both the same absolute and relative intensity as at sea level and under acute hypoxia, before and after 4 weeks exposure to 4100 m altitude (Lundby and Van Hall 2002). Submaximal substrate utilization was unchanged with acute and chronic hypoxia when exercising at the same relative intensity, while the carbohydrate utilization was increased when exercising at the same absolute intensity. Further, 4 weeks of acclimatization to altitude did not affect substrate utilization, also confirming data from rodents (McClelland et al. 1998). Overall, these studies demonstrate that relative work intensity is the main factor determining fuel selection during exercise and prolonged hypoxia does not cause a significant shift in fuel selection. However, little is known about the relative contributions of muscle glycogen during hypoxia.

5.3.8 Sex Differences

Sex-based differences in substrate metabolism during endurance exercise are well documented in that it is now recognized that females have reduced reliance on whole-body CHO metabolism (typically reflective of reduced liver glycogen metabolism) to support energy production (Devries 2016). Although not always consistent, there is also some evidence that females use less glycogen than males. For example, when assessed during the luteal phase, females use less glycogen in the vastus lateralis muscle (25%) compared with males, as assessed during 90 minutes of cycling at 65% VO_2peak (Devries et al. 2006). When tested in the mid-follicular phase, it has also been reported that females use less glycogen (25%) in the vastus lateralis muscle than males when running a set distance of 15.5 km on a treadmill at 65% VO_2peak (Tarnopolsky et al. 1990). When completing a 16 km road run undertaken at lactate threshold in the mid-follicular phase, we also recently reported that females use less glycogen in both the vastus lateralis (30%) and gastrocnemius muscles (20%) when compared with males (Impey et al. 2020). It is, of course, difficult to offer definitive mechanisms underpinning such differences in local muscle metabolism owing to the challenge of matching resting glycogen concentration, training status, total work done, or distance covered between sexes (i.e., thus reflective of exercise intensity and duration). Additionally, variations in phase of the menstrual cycle studied as well as the use of contraceptives also make it difficult to compare between studies. On the basis that studies examining the impacts of sex hormones (e.g., estrogen and progesterone) have generally indicated they exert minimal regulatory effects on muscle glycogen utilization (Devries 2016), it is possible that the aforementioned factors may indeed play a more subtle but influential role. A recent meta-analysis demonstrated that sex-based differences in glycogen utilization are indicative of a “likely small” reduction on absolute glycogen utilization of approximately 30 mmol^.kg⁻¹ dw (Areta and Hopkins 2018). From a practical perspective, we therefore suggest that such magnitudes of differences in glycogen utilization are unlikely to require sex-specific nutritional strategies and that both males and females should simply ensure they commence their training or competitive scenario with sufficient glycogen stores to meet the subsequent metabolic demand.

5.3.9 Subcellular Compartmentalization

In the above sections, the utilization of glycogen has been described based on mixed muscle homogenates or histochemical-defined specific fiber types. This view assumes a uniform utilization of glycogen within the muscle fibers, i.e., with no spatial compartmentalization of metabolic reactions and no existence of local gradients of ions and metabolites. A wealth of studies have shown that this assumption is not valid and, in contrast, demonstrated that the muscle fibers’ interior is extremely crowded (high concentrations of, e.g., proteins, metabolites, and ions) with limited free diffusion, but with a highly developed compartmentalized network of enzymatic reactions (e.g., Srere 1967; Sear 2019). Within this crowded interior, glycogen particles are dispersed in distinct compartments. The definition of these compartments is man-made and may be limited by the typical 2D portray of a muscle fiber. However, the studies from independent research groups (Sjöström et al. 1982a; Marchand et al. 2002; Nielsen et al. 2011) have all defined three clearly separated compartments (Fig. 5.4): (1) the intermyofibrillar space (between the myofibrils); (2) the intramyofibrillar space (within the myofibrils); and (3) the subsarcolemmal space (just beneath the surface membrane). In addition, some studies have addressed glycogen particles in physical interaction with the sarcoplasmic reticulum (Wanson and Drochmans 1968; Goldstein et al. 1985; Tammineni et al. 2020) or have discriminated between intramyofibrillar glycogen located either in the I- or A-band of the sarcomere (Fridén et al. 1985, 1989). In the subsarcolemmal space, glycogen particles have been described as perinuclear (Caulfield and Klionsky 1959), nuclear (Sun et al. 2019), or lysosomal (Viragh et al. 1982). Since quantitative data is available only for the three clearly separated compartments, this will form the basis for the discussion below.

Fig. 5.4

Representative transmission electron micrographs showing the typical pattern of glycogen storage (glycogen particles are the black dots) at pre- (a) and post-prolonged exhaustive exercise (b). Mi, mitochondria. The glycogen particles located within the sarcomere often close to the Z-disc and as strings between the filaments are termed intramyofibrillar glycogen (thick arrow). The particles located close to mitochondria (Mi) between the myofibrils are termed intermyofibrillar glycogen (thin arrows). The images are collected as a part of the project described in Nielsen et al. 2011

Of the three compartments, intramyofibrillar glycogen are utilized relatively most during various types of exercise (Marchand et al. 2007; Nielsen et al. 2011; Gejl et al. 2017a, b, c; Jensen et al. 2020b; Hokken et al. 2020, Fig. 5.4). In a recent study where participants cycled at 75% of ̇VO₂max, exhaustion was associated with mixed muscle homogenate glycogen concentration well above zero (120 mmol kg dw), but with intramyofibrillar glycogen levels close to zero in about 60% of the type 1 fibers (Jensen et al. 2020b). In comparison, subsarcolemmal and intermyofibrillar glycogen levels were close to zero in only 40 and 10% of the type 1 fibers, respectively, clearly suggesting a link between the depletion of intramyofibrillar glycogen and exhaustion. However, in a few exceptions, intramyofibrillar glycogen was not found to be preferentially utilized, which forms some basis for the understanding of compartmentalized glycogen utilization. Firstly, with repeated high-intensity exercise (4 × 4 min sprint skiing), only intermyofibrillar glycogen was utilized during the fourth exercise bout (Gejl et al. 2017a, b, c), indicating that the energy production by degradation of intramyofibrillar glycogen can be replaced by another energy source if the exercise is repeated. This could be from intermyofibrillar glycogen degradation, which in absolute terms was utilized more during the fourth exercise bout, or from phosphocreatine, which is also localized within the myofibrils (Wallimann and Eppenberger 1985) and known to super-compensate in response to repeated exercise (Sahlin et al. 1997). Secondly, if subsarcolemmal glycogen is super-compensated, its utilization rate is increased concomitant with a reduced utilization of intramyofibrillar glycogen (Jensen et al. 2020b). This implies the existence of a mechanism linking subsarcolemmal glycogen with intramyofibrillar glycogen utilization. Thirdly, after resistance exercise, intermyofibrillar glycogen was preferentially utilized in type 1 fibers, which contrasts to type 2 fibers, where glycogen was utilized from all compartments (Hokken et al. 2020). The resistance exercise was characterized by work periods interspaced by 2–4 min pauses, which may facilitate some resynthesis of glycogen. With a low rate of glycogen utilization in type 1 fibers, a small preferential resynthesis of intramyofibrillar glycogen during rest periods may almost equal the degradation during work, which underscores that net results (pre-post) from intermittent exercises should be carefully interpreted.

5.4 Glycogen Depletion and Fatigue

Most of the studies investigating the role of glycogen in muscle fatigue are based on associations or correlational findings, but the causative effect and mechanisms explaining glycogen depletion-induced fatigue are not clear. To definitively address this experimentally, these human studies with correlative findings should be combined with mechanistic in vitro studies, but while glycogen can be removed enzymatically by amylase, it cannot be instantly added exogenously. Due to these inherent limitations in research designs, our understanding of the role of glycogen in muscle fatigue must include careful interpretations of the available data.

5.4.1 Correlations with Performance

Pioneering research by A.V. Hill suggested that glycogen was the sole fuel for muscle work and that lactate was necessary to activate the muscle (e.g., Hill 1913). Today, it is widely known that ATP generated through several metabolic pathways is the fuel for muscle work and that Ca²⁺ ions ultimately mediate switching off the brake on actin filaments and facilitate cross-bridge formation, which is the main feat of ATP consumption of contracting skeletal muscle. Although Hill unintendedly exaggerated the role of glycogen in skeletal muscle, much evidence still points toward glycogen as a key fuel source mediating these processes. In the following sections, we review the literature examining the role of glycogen for performance during prolonged, short-term, and resistance exercise.

5.4.1.1 Prolonged Exercise (60–180 Min)

When performance is evaluated as maximal work conducted at a fixed protocol of working intensity (time to exhaustion), most (Bergström et al. 1967; Galbo et al. 1979; Lamb et al. 1994; Walker et al. 2000; McInerney et al. 2005; Duhamel et al. 2006c; Alghannam et al. 2016; Jensen et al. 2020b), but not all (Madsen et al. 1990), studies demonstrate improved performance with high pre-exercise glycogen stores. Closed-end tests (e.g., time trials) show small effects (Karlsson and Saltin 1971; Widrick et al. 1993; Rauch et al. 1995) or no effect (Sherman et al. 1981; Hawley et al. 1997a, b; Burke et al. 2000; Tomcik et al. 2018) of high pre-exercise glycogen stores on performance. Independent of test protocol, the characteristics of the studies showing a positive effect of elevated glycogen levels are that the participants either worked for 90–180 min and/or had large differences (>50%) in pre-exercise glycogen levels between the experimental conditions.

5.4.1.2 Short-Term Exercise (<15 Min)

If the pre-exercise glycogen level is well above the utilization level, the current consensus is that there are no effects of above-normal levels of glycogen on short-term exercise performance. This is based on tests of time to exhaustion at 125% (Vandenberghe et al. 1995) and 85% (Lambert et al. 1994) of VO_2max as well as a 75 sec all-out time trial (Hargreaves et al. 1997). However, if the pre-exercise glycogen level is very low, both repeated 6 sec all-out sprint (Balsom et al. 1999) and repeated one-legged intense exercise (Bangsbo et al. 1992) performance are impaired. Collectively, muscle glycogen seems to be important for both continuous high-intensity exercise tolerance (>60 s duration) and single or repeated sprint performance (<60 s duration), only if a substantial degree of depletion is achieved, whereas loading the stores above normal levels imposes no consistent additional benefit (Vigh-Larsen et al. 2021).

5.4.1.3 Resistance Exercise

Several studies have found that a carbohydrate-restricted diet is not associated with reduced strength and power output (i.e., Mitchell et al. 1997; Sawyer et al. 2013), suggesting that resistance exercise performance is not related to the muscle glycogen levels. This is in accordance with one study, where a very low pre-exercise glycogen level did not affect maximal voluntary isometric contraction or peak force during 50 repetitions (Symons and Jacobs 1989). While this is corroborated by the finding that only around 40% of the glycogen stores in the active muscles are utilized during typical resistance exercise protocols (Tesch et al. 1986; Essén-Gustavsson and Tesch 1990; Robergs et al. 1991; MacDougal et al. 1999), a recent study demonstrated, however, that a local pool (intramyofibrillar) of glycogen in type 2 fibers is very low after resistance exercise (Hokken et al. 2020), suggesting that fatigue development of some type 2 fibers may limit resistance exercise performance.

5.4.2 A Causal Link to Fatigue?

The causal link between glycogen depletion and impaired performance is most likely multifactorial, but studies on isolated muscles from rodents (Chin and Allen 1997) and amphibians (Stephenson et al. 1999) clearly suggest a local factor within the muscles. To our knowledge, till date, only one paper has questioned this finding. Here, mice were lacking glycogen due to a whole-body gys1 (glycogen synthase) deletion and showed surprisingly normal exercise capacity (Pederson et al. 2005). However, only around 10% of the mice survived birth suggesting a survival phenotype, which may not be comparable to a wild-type phenotype, and a later study using inducible skeletal muscle-specific gys1 deletion showed considerable reduced exercise capacity (Xirouchaki et al. 2016). Thus, the lack of glycogen inevitable leads to diminished muscle function. Nonetheless, the mechanisms explaining why and how glycogen and glycolytic rate are integrated with cell function are far from understood.

According to one’s intuition and the recognized “energy crisis” theory, the association between low muscle glycogen levels and impaired contractile function is that low glycogen causes a slowed glycogenolytic and glycolytic flux, compromising the required rate of ATP regeneration for the sustained muscle function during a given intensity (Green 1991; Sahlin et al. 1998). Consequently, the muscle is unable to maintain an adequate ATP supply to one or more of the processes involved in E-C coupling, leading to impaired muscle function, i.e., fatigue. This is supported by observations of whole muscle cell PCr decreases along with an increase in free ADP and IMP following prolonged glycogen depleting exercise (Norman et al. 1988; Sahlin et al. 1997). However, this theory is challenged by both in vitro and in vivo studies demonstrating a strong association between low glycogen and decreased muscle function even after recovery periods, where ATP levels would be normal (Bangsbo et al. 1992; Chin and Allen 1997). Moreover, muscular fatigue is also observed even when glycogen is far from depleted in different conditions where glycogen is decreased prior to the start of exercise (Duhamel et al. 2006a, b; Ørtenblad et al. 2011). Also, low glycogen affects muscle function under experimental circumstances in vitro where global ATP and PCr levels can be maintained at near resting levels (Kabbara et al. 2000; Helander et al. 2002; Nielsen et al. 2009). Together, these series of experiments do not provide experimental support for the energy crisis hypothesis.

It is noteworthy, however, that the ATP concentration inside cells may not be uniform at a subcellular level (Jones 1986). The highly organized muscle cell forms many compartments and hence microenvironments with high ATPase activity, and restricted diffusional access of metabolites and observations on whole muscle experiments or fiber preparations does not rule out a role of glycogen in maintaining a subcellular compartment energy status. Such a functional compartmentalization of glycolytic metabolism is known in a variety of tissues, including skeletal muscle. In this way, the model that has evolved is that glycolytic-derived ATP regulates key steps in the muscle excitation-contraction (E-C) coupling by delivering ATP in the microenvironment of the triad junction (Han et al. 1992; Korge and Campbell 1995). This is particularly significant in the muscle triad junction between the transverse tubular system and the sarcoplasmic reticulum (SR), with a diffusional restricted space around 12 nm wide and with a high metabolic activity (Dulhunty 1984). Consistent with the notion of compartmentalized glycolysis, most of the glycolytic enzymes are associated with membranes of intracellular compartments such as the SR (Wanson and Drochmans 1972; Entman et al. 1980; Xu and Becker 1998). Furthermore, glycogen is stored in particles located in specific regions of the muscle fiber, and variable utilization of these depots occurs during exercise, with the depot localized near the triad region being repeatedly associated with muscle function and whole-body exhaustion (see Sect. 5.3.9). Physiologically, this organization places muscular energy stores in close proximity to their site of utilization and provides support for the emerging concept for functionally compartmentalized energetic networks, ensuring an efficient energy transfer and signal transduction between energy production and utilization in different cellular compartments (Korge and Campbell 1995; Saks et al. 2008; Nielsen and Ørtenblad 2013). The following sections will focus on the experimental evidence of how glycogenolytically/glycolytically derived products preferentially regulate key steps in the muscle E-C coupling, i.e., SR Ca²⁺ regulation and muscle excitability, thus providing an explanation of the observed association between muscle glycogen contents and fatigue.

5.4.2.1 SR Ca²⁺ Regulation

Contraction of skeletal muscle is governed by the series of events in the E-C coupling, in which the Ca²⁺ release and uptake from the sarcoplasmic reticulum (SR) are an integral part through initiation and termination of the cross-bridge cycling. The SR Ca²⁺ release is triggered through an action potential (AP) activation of the voltage-sensor molecules in the t-system membrane, which open the SR Ca²⁺ release channels (RyR), leading to a rise in intracellular free concentration ([Ca²⁺]_i) and generation of force by the contractile apparatus (Melzer et al. 1995; Stephenson 1996).

Both direct and indirect evidence point to a modulating role of glycogen availability on SR Ca²⁺ handling, as demonstrated in animal (Chin and Allen 1997; Stephenson et al. 1999; Kabbara et al. 2000; Barnes et al. 2001; Helander et al. 2002; Nielsen et al. 2009) and human models (Gejl et al. 2014; Duhamel et al. 2006b; Ørtenblad et al. 2011). Using both single fibers and muscle bundles, Chin and Allen (1997) elegantly demonstrated that muscle force and [Ca²⁺]_i are associated with muscle glycogen content. Thus, through the manipulation of glucose availability in the recovery phase after fatiguing contractions, it was shown that a reduced resting level of glycogen was associated with a faster decrease in tetanic [Ca²⁺]_i and force during subsequent contractions. These results have subsequently been confirmed (Kabbara et al. 2000; Helander et al. 2002; Nielsen et al. 2014), and together, experiments on rodent muscle suggest that the change in tetanic [Ca²⁺]_i associated with fatigue and recovery has a component that is glycogen dependent.

The mechanisms linking low muscle glycogen with decreased [Ca²⁺]_i has further been elucidated by direct measures of SR vesicle Ca²⁺ release rate (Duhamel et al. 2006a, b, c; Ørtenblad et al. 2011). These studies on SR vesicles from the human muscle, where glycogen levels have been modulated by the diet either before or during the recovery phase after exercise, demonstrate a clear association between the SR vesicle Ca²⁺ release rate and muscle glycogen levels (Duhamel et al. 2006a, b; Ørtenblad et al. 2011; Gejl et al. 2014; Watanabe and Wada 2019). Importantly, there seems to be a critical level of muscle glycogen at around 250–300 mmol^.kg⁻¹ dw below which the SR Ca²⁺ release rate is impaired (Duhamel et al. 2006a; Ørtenblad et al. 2011; Gejl et al. 2014). Such data explain why minor decreases in muscle glycogen do not cause significant impairments in SR Ca²⁺ release rate and why exhaustive exercise starting with a low or high muscle glycogen store decreases or improves the endurance performance, respectively, possibly affected by SR Ca²⁺ regulation (Ørtenblad and Nielsen 2015; Ørtenblad et al. 2013).

The use of the mechanically skinned fiber preparation has provided a unique means to investigate the possible interactions between glycogen and basic muscle function (Lamb and Stephenson 2018). Indeed, the selective removal of the sarcolemma allows for the study of muscle function and the effects of glycogen content per se while maintaining the cellular architecture and control of the intracellular milieu, i.e., keeping PCr and ATP high and constant. Studies using the mechanically skinned fiber have provided experimental evidence that low glycogen content in the muscle fiber is associated with force depression during repeated voltage sensor activated contractions in most studies (Stephenson et al. 1999; Barnes et al. 2001; Watanabe and Wada 2019), but not all (Goodman et al. 2005), as well as AP-induced contractions (Nielsen et al. 2009: Jensen et al. 2020a).

Taken together, emerging evidence suggests that low muscle glycogen affects the SR Ca²⁺ release and in turn [Ca²⁺]_i and muscle function despite global ATP being held constant. This conclusion supports the concept of a functionally compartmentalized energetic network regulating key steps in the muscle E-C coupling. In support hereof are the observations (from both mechanically skinned fiber preparation, intact mouse fibers, and human SR vesicles) that the specific pool of intramyofibrillar glycogen within the myofibrils is associated with SR Ca²⁺ release and [Ca²⁺]_i (Nielsen et al. 2009; Ørtenblad et al. 2011; Nielsen et al. 2014). At present, little is known about the precise mechanism(s) which links glycogen levels in the muscle with SR Ca²⁺ release rate.

In skeletal muscle, Ca²⁺ is released from the SR Ca²⁺ stores via the specific Ca²⁺ channels (RyR1), located at the junctional SR of the triad, which ensures efficient Ca²⁺ release to the contractile proteins. In relation to the role of glycogen affecting the SR Ca²⁺ release, low glycogen may in the compartmentalized cell lead to changes in metabolic status, especially in triad region with the RyR localization. This may lead to increase in free [Mg²⁺] and decrease in free [ATP], which are strong regulators of the RyR1, also in the physiological range of changes (Lamb and Stephenson 1994; Blazev and Lamb 1999). Furthermore, glycolytic intermediates as fructose 1,6-bisphosphate have been demonstrated to increase the open probability of the RyR channels (Han et al. 1992). However, low glycogen has also been demonstrated to modulate the Ca²⁺ release rate in isolated vesicles without restricted metabolic space and during resting metabolic conditions. This may indicate a crucial role of the metabolic machinery associated with the SR. Also, the RyR protein contains numerous phosphorylation sites, which may be affected by PKA and Ca²⁺-calmodulin-dependent kinase II in fast twitch fibers (Fill and Copello 2002). However, the impact of phosphorylation and/or dephosphorylation on single RyR channel behavior and the role of glycogen and energy status are at present not fully unraveled.

5.4.2.2 Muscle Excitability and Na, K-Pump

With repeated intense activation, a change in the electrochemical gradients for K⁺ can cause a substantial membrane depolarization leading to failure of excitation and SR Ca²⁺ release and an ensuing decrease in force responses (Sejersted and Sjøgaard 2000; Clausen and Nielsen 2007), although several mechanisms interact during exercise to counteract the depressive effects of elevated extracellular levels of K⁺ (Pedersen et al. 2003; de Paoli et al. 2007, 2010). There is a reasonably well-established relationship between glycolytic-derived ATP and Na, K-pump activity, and evidence exists supporting that glycolysis and the Na, K-pump are functionally coupled. This seems to be an evolutionary conserved metabolic coupling and has been observed in several tissue types, including the mammalian erythrocytes (Schrier 1966; Mercer and Dunham 1981; Kennedy et al. 1986), axons (Caldwell et al. 1960), cardiac myocytes (Philipson and Nishimoto 1983; Hasin and Barry 1984; MacLeod 1989), and skeletal muscle (Clausen 1965; James et al. 1999; Okamoto et al. 2001; Jensen et al. 2020a). In line with this, many tissues generate pyruvate and lactate under aerobic conditions (aerobic glycolysis) in a process linking glycolytic ATP supply to the activity of the Na, K-pump (Brooks 1986; Dhar-Chowdhury et al. 2007). Indeed, aerobic glycolysis and glycogenolysis occur in resting, well-oxygenated skeletal muscles and is closely linked to stimulation of active membrane Na, K-pump transport due to epinephrine release (James et al. 1996, 1999a, b; Bundgaard et al. 2003; Levy et al. 2005). The role of glycolysis and oxidative phosphorylation in providing fuel to the Na, K-pump in the skeletal muscle was investigated using the Na, K-pump inhibitor ouabain in resting rat extensor digitorum longus muscles, demonstrating that Na, K-pump activity is only impaired when the glycolysis is inhibited (Okamoto et al. 2001). A tight coupling between the glycogenolytic rate and Na, K-pump activity is further demonstrated by the observation that intracellular Na⁺ decreases if glycogen breakdown is stimulated with epinephrine at rest, while ouabain significantly attenuates glycogen utilization (James et al. 1999b). The data suggest that in the skeletal muscle, glycolysis is the predominant source of the fuel for the Na, K-pump. Further, a direct link between muscle glycogenolysis and Na, K-pump activity was demonstrated by the observation of decreased glycogen utilization in resting muscle when the muscle Na, K-pump activity was blocked with ouabain (Clausen 1965). Moreover, lactic acid production was increased in proportion to activation of the Na, K-pump. Taken together, these data indicate a clear association of glycogenolytic/glycolytically derived ATP on active cation transport across the muscle membranes. Indeed, a direct link between energy state and excitability of the muscle was demonstrated by blocking cross-bridge cycling and SR Ca²⁺ release with the cross-bridge cycling blocker BTS (N-benzyl-p-toluene sulphonamide) and dantrolene, respectively, thereby conserving energy during repeated electrical stimulations, which in turn improved the ability of muscles to maintain excitability during high-frequency stimulation (Macdonald et al. 2007).

The effects of glycogenolytically derived ATP on muscle excitability are further substantiated in mechanically skinned fibers with a high and constant global ATP, demonstrating that glycogenolytically derived energy is associated with fiber contractile endurance irrespective of global ATP levels. Thus, enzymatically lowering glycogen by 70% led to a reduction in both voltage sensors-activated and AP-induced forces in the skinned fibers, with a larger decrease in the AP force by the glycogen lowering. These data suggest that low glycogen and glycogenolytic rate affects t-system polarization and excitability, as the voltage sensor inactivation is displaced to markedly more positive Vm values compared with the AP inactivation (Ørtenblad and Stephenson 2003; Nielsen et al. 2004) and thereby force production will be less affected by voltage sensor compared to AP activation. For a more direct estimate of the Na, K-pump function, one can quantify the membrane ability to respond to two closely spaced AP. The repriming time until the second pulse generates an AP is in turn dependent on the Na, K-pump function. A depolarization of the t-system increases the repriming time as expected; however, the addition of phosphoenolpyruvate, which increases glycolytic ATP resynthesis, decreases t-system repriming period (Dutka and Lamb 2007a, b). Also, when glycogen is enzymatically lowered by glucoamylase treatment, the repriming period increases (Watanabe and Wada 2019; Jensen et al. 2020a). This was confirmed by the use of glycogen phosphorylase inhibitors and glycogen lowering treatment in mechanically skinned fibers, which invariably prolonged repriming time, strongly indicating an attenuated Na, K-pump activity (Jensen et al. 2020a). Collectively, substantial evidence indicates a tight coupling between glycogenolytic-glycolytic-derived ATP production on Na-K-pump function, with low glycogen or inhibited GP in turn leading to an attenuated t-system excitability and ultimately force production.

5.4.2.3 Insights from McArdle Patients

Insights into the role and importance of muscle glycogen can be gained from McArdle patients (glycogen storage disease type V) who lack glycogen phosphorylase and thereby not able to catalyze the breakdown of glycogen into glucose-1-phosphate in muscle fibers (McArdle 1951; Santalla et al. 2014). McArdle patients’ have low peak work capacities and exercise intolerance and a suppressed plasma lactate concentration during exercise (De Stefano et al. 1996; Ørngreen et al. 2015). The exercise intolerance is especially evident during the onset of exercise, during high-intensity exercise and high force output contractions with smaller muscle groups (Lucia et al. 2008; Santalla et al. 2014). These patients may provide a unique paradigm to gain insight to the role of muscle glycogen per se in muscle function and exercise tolerance. Interestingly, a higher surface electromyography signal is measured during submaximal contractions in McArdle patients, which is indicative of a need to activate larger muscle mass for a given force output, suggesting reduced muscle excitability (Santalla et al. 2014). Interestingly, McArdle patients have less Na⁺-K⁺-ATPases as compared to controls (Haller et al. 1998). Taken together, insights from McArdle patients are in line with the idea of glycolytic/glycogenolytic-derived ATP as required for muscle E-C coupling, possibly by a tight coupling between glycogenolytic/glycolytic rate and Na⁺-K⁺-ATPase activity.

5.5 Muscle Glycogen as a Regulator of Skeletal Muscle Training Adaptations

In addition to the well-documented role as a metabolic substrate for ATP provision, it is becoming increasingly accepted that the glycogen granule can also act as a regulator of training adaptations (Impey et al. 2018; Impey et al. 2016). Accordingly, the concept of deliberately training with reduced pre-exercise muscle glycogen availability in an attempt to enhance the activation of the molecular signaling pathways that regulate mitochondrial biogenesis (the so-called train low paradigm) has received significant research in the previous decade (Burke et al. 2018). In this regard, findings from acute exercise studies demonstrate that commencing exercise with “reduced” pre-exercise muscle glycogen concentration upregulates cell signaling pathways with putative roles in the regulation of both the nuclear and mitochondrial genomes (Pilegaard et al. 2002; Wojtaszewski et al. 2003; Yeo et al. 2010; Bartlett et al. 2013; Psilander et al. 2013). Furthermore, repeated bouts of train-low exercise can subsequently augment many hallmark muscle adaptations inherent to the endurance phenotype (as reviewed in Impey et al. 2018). For example, the strategic periodization of dietary CHO in order to commence exercise with low muscle glycogen (during 3–10 weeks of training) enhances mitochondrial enzyme activity and protein content (Hansen et al. 2005; Morton et al. 2009; Yeo et al. 2008) and whole body and intra-muscular lipid metabolism (Hulston et al. 2010) and in some instances improves exercise capacity (Hansen et al. 2005) and performance (Marquet et al. 2016a, b), though performance enhancing effects are not always evident (Yeo et al. 2008; Hulston et al. 2010; Burke et al. 2017; Gejl et al. 2017a, b; 2018). As such, the train-low paradigm and wider CHO periodization strategies have subsequently gained increased recognition among athletic populations (Stellingwerff 2012; Burke et al. 2018; Impey et al. 2018). It should also be noted that some of the enhanced adaptations associated with “train-low” (at least in the twice per day training model) may be due to performing two consecutive training sessions in close proximity to one another, as opposed to the effects of low pre-exercise muscle glycogen per se (Andrade-Souza et al. 2020).

Skeletal muscle glycogen may exert its regulatory effects upon training adaptation through modulation of the AMP-activated protein kinase (AMPK)-peroxisome proliferator-activated receptor γ coactivator 1α (PGC- 1α) signaling axis. For example, exercise-induced AMPKα2 activity (Wojtaszewski et al. 2003), phosphorylation (Yeo et al. 2010), and nuclear abundance (Steinberg et al. 2006) are all augmented under conditions of reduced pre-exercise muscle glycogen. These effects may be partly mediated through the glycogen-binding domain present on the β-subunit of AMPK (McBride and Hardie 2009; McBride et al. 2009). Commencing acute exercise with reduced muscle glycogen also potentiates the phosphorylation of the tumor suppressor protein, p53 (Bartlett et al. 2013), that in turn may coordinate regulation of the mitochondrial genome, through modulation of mitochondrial transcription factor A (Tfam). It is noteworthy that training with low muscle glycogen also increases epinephrine, stimulates lipolysis, increases circulating free fatty acids (FFAs), and therefore augments both whole-body and intramuscular lipid oxidation (Hearris et al. 2018, 2020; Maunder et al. 2021). In this way, it is possible that FFAs may also regulate the enhanced adaptations associated with training low through acting as stimulatory signaling molecules for PPARδ signaling.

Given that the enhanced training response associated with train-low is potentially mediated by muscle glycogen availability, it is pertinent to consider the absolute glycogen concentrations required to facilitate the response. In relation to research design, high glycogen trials are commonly commenced with muscle glycogen concentrations between 400 and 600 mmol^.kg⁻¹ dw and remain above 300 mmol^.kg⁻¹ dw after exercise (Wojtaszewski et al. 2003; Roepstorff et al. 2005; Bartlett et al. 2013). In such instances, these researchers observed attenuated (Wojtaszewski et al. 2003) or abolished (Roepstorff et al. 2005; Bartlett et al. 2013) activation of cell signaling pathways. For example, we have previously observed that both AMPK and p53-related signaling is reduced when exercise is commenced and finished with high (reducing from ~500 to ~300 mmol.kg⁻¹ dw, respectively) versus low (reducing from ~150 to ~50 mmol.kg⁻¹ dw, respectively) muscle glycogen (Bartlett et al. 2013). A detailed examination of available train low studies (see Impey et al. 2018) indeed demonstrates that adaptations associated with CHO restriction are particularly evident when the absolute pre-exercise muscle glycogen concentration permits exercise-induced depletion to post-exercise concentrations that are ≤200–300 mmol^.kg⁻¹ dw. Accordingly, we also observed that comparable cell signaling responses occur if exhaustive (Hearris et al. 2019) and non-exhaustive (Hearris et al. 2020) exercise is finished with similar absolute post-exercise muscle glycogen concentrations (i.e., 100–300 mmol^.kg⁻¹ dw), despite commencing exercise with graded levels of muscle glycogen concentration (i.e., 300–600 mmol^.kg⁻¹ dw). On this basis, it was suggested that the absolute post-exercise muscle glycogen concentration may be a more influential factor than pre-exercise muscle glycogen concentration in relation to modulation of exercise-induced cell signaling. In keeping with this hypothesis, CHO feeding during exercise attenuates AMPK-mediated signaling but only when glycogen sparing occurs (Akerstrom et al. 2006). Furthermore, restoring post-exercise glycogen levels to >500 mmol^.kg⁻¹ dw attenuates exercise-induced changes in gene expression (Pilegaard et al. 2005). From a practical perspective, such data collectively suggest that distinct differences in post-exercise muscle glycogen concentration would likely be required between low (e.g., <200 mmol^.kg⁻¹ dw) and high (> 300 mmol^.kg⁻¹ dw) glycogen conditions in order to achieve any potential physiological advantage of reduced muscle glycogen availability in relation to augmenting training adaptation. While the optimal approach to CHO periodization remains to be determined, further research is also required to examine the role of the specific subcellular storage pools of glycogen as a regulator of skeletal muscle cell signaling pathways. Moreover, despite the theoretical rationale for carefully scheduled periods of “training-low,” it is noteworthy that definitive evidence supporting that this approach to training induces superior improvements in performance is currently limiting. In this regard, there is also a requirement to utilize experimental testing protocols (e.g., exercise durations of 3–6 h) in which the physiological adaptations associated with training low (i.e., increased oxidative capacity and lipid oxidation) may actually manifest as improved exercise performance.

5.6 Concluding Remarks and Future Directions

Glycogen is an important fuel source for contracting skeletal muscle to sustain high metabolic power output during exercise, and modern research has also shown it to be far more than just an inert energy source. Early research determined a clear direct association between the capacity to withstand intense exercise for prolonged periods and starting muscle glycogen, showing that the onset of fatigue coincides with the depletion of glycogen to very low levels, even with exogenous carbohydrate supplementation. In addition to exercise intensity and duration, baseline muscle glycogen and training status have also been identified as important factors in determining its rate of use, with other factors such as sex substrate availability and environment also having some impact. Glycogen use has shown to be divergent in different fiber types, and more recent research has identified specific subcellular loci of glycogen (namely, intramyofibrillar, intermyofibrillar, and subsarcolemmal), with possible diverse utilization and metabolic function depending on their localization. The mechanisms explaining how glycogen and glycolytic rate are integrated with muscle function are far from understood; however, emerging evidence link glycogen availability with key steps in the excitation-contraction coupling. In addition, it is also apparent that glycogen availability can exert important effects on metabolic regulation and exercise-induced cell signaling, regulating training adaptation. Thus, glycogen has a far more diverse function than just being an energy storage. Despite over 100 years of research in carbohydrate metabolism, fundamental aspects of muscle glycogen metabolism and regulation continue to challenge the scientific community and remain an evergreen area of exciting research. At several levels, we appear to have a complete understanding of the physiology of muscle glycogen, and yet at others, we know very little. Indeed, many aspects of muscle glycogen metabolism in exercise and cell function warrant further knowledge. Table 5.1 lists possible future studies further generating our understanding of muscle glycogen metabolism and regulation, as well as the causal link underlying the association between glycogen availability in the etiology of muscle fatigue and key steps in muscle function and regulator of training adaptations.

Table 5.1 Possible future directions in the study of muscle glycogen metabolism