The crossover concept in exercise metabolism
This is an excerpt from Exercise Physiology-6th Edition by George A. Brooks,Thomas D. Fahey,Kenneth M. Baldwin.
A major problem in the field of exercise biochemistry has been lack of appreciation for how the glucose–fatty acid cycle operates in an intact, functioning person. In the past, the bias has been that because endurance training increases the capacity to use fat, as opposed to carbohydrate, and because the glucose–fatty acid cycle explains substrate interactions, the glucose–fatty acid cycle must be important for explaining fuel use during exercise. In contrast, evaluation of data obtained on exercising humans requires adoption of another conceptual model, which is termed the Crossover Concept.
CROSSOVER CONCEPT
That power output is the most important factor in determining the fuels used during exercise was recognized by Jacques Mercier of Montpellier, France, and George Brooks (1994) of Berkeley, California, who recognized and articulated the Crossover Concept (figure 7.16). Other factors, such as diet, training status, gender, menstrual cycle phase, and age are of secondary importance in determining the fuels used by working muscles and the remainder of the body during exercise. The Crossover Concept allows that in the postabsorptive person, resting muscles and the remainder of the body utilize lipids predominantly as fuels. However, as exercise starts and intensity progresses from mild through moderate to hard intensities, the fuel mix switches (“crosses over”) from lipid to carbohydrate. The evidence supporting the concept follows.
Results of classic indirect calorimetry studies indicate that in a resting postabsorptive person, most energy (≈60%) is from lipid oxidation. Most of the remainder is from carbohydrate oxidation (≈35%), and approximately 5% is from proteins (Bergman and Brooks 1999). However, as soon as exercise starts, even though the same amount or more of lipid is used, the relative contribution of lipid to total energy release declines, and that of carbohydrate rises (figure 7.16). Then, as exercise power output increases from mild to moderate and then to hard intensity, the relative contribution of lipids declines, while that of carbohydrate increases. Thus, hard-intensity exercise is accomplished by “crossover” to dependence on carbohydrate oxidation and utilization. Perhaps most impressive is that the Crossover Concept applies to the heart. In healthy persons at rest, circulating fatty acids provide most (≈80%) of energy for the beating heart. However, during exercise, lactate released into the systemic circulation from working muscles is taken up by the heart and becomes the major cardiac energy source (Brooks 2021).

Adapted from Brooks and Mercier (1994).
When contractions start, GLUT-4 is translocated to the sarcolemma, and phosphorylase is activated by elevated cytosolic Ca2+ and Pi. These events, in combination with high enzymatic rates (Vmax) and an abundance of glycolytic enzymes in the muscle, give rise to rapid glycolysis, leading to pyruvate and lactate formation. Arterial FFA concentration falls, due in part to the inhibition of lipolysis by lactate anion binding to a receptor [hydroxycarboxylic acid receptor 1 (HCAR1) or lactate receptor 1 (LR1)], inactivating adenylcyclase, thus decreasing cyclic AMP (cAMP), which works through the cAMP response element-binding protein (CREB) to inhibit muscle hormone-sensitive lipase (L-HSL). This causes circulating FFA levels to fall.
More importantly, when contractions start and glycolysis is stimulated, pyruvate formation leads to malonyl-CoA formation, thus inhibiting CPT1 and mitochondrial FFA uptake. As well, acetyl-CoA from pyruvate inhibits β-ketothiolase, the terminal enzyme in β-oxidation, thus limiting lipid oxidation. As a consequence, in hard-intensity exercise, carbohydrate is the predominant energy source, regardless of training state. Simultaneously, as shown by Inigo San-Millán and colleagues (2022), lactate produced from glycolysis binds to cardiolipin, which blocks CPT2 and also limits mitochondrial activated FFA uptake.
Endurance training has the effect of displacing to a higher absolute level the point where crossover occurs. By suppressing epinephrine secretion, by increasing capacity for lactate clearance, and by increasing sensitivity of respiratory control through an increase in mitochondrial mass, training shifts the crossover point to higher absolute and relative power outputs. However, because athletes usually train and compete at hard or greater relative intensities, they invariably cross over to the region where most energy is from carbohydrate.
Even though athletes utilize relatively more energy from carbohydrate than lipid, lipid utilization in the athlete remains an important means of suppressing muscle glycogenolysis, thereby prolonging the time to muscle glycogen depletion and fatigue. Moreover, during recovery from hard training and competition, utilization of lipid is essential to allow normal functioning of the individual and repletion of muscle glycogen. Thus, under the larger umbrella of the Crossover Concept, the glucose–fatty acid cycle is understood to function during mild to moderate intensity exercise, especially in highly trained persons, but mainly to function after exercise that depletes muscle glycogen.
Evidence for the Crossover Concept has been provided by Anne L. Friedlander, G.A. Brooks, and associates. They used stable, nonradioactive tracers (Friedlander et al. 1997, 1998a, 1998b, 1999) and classical mass balance (arterial-venous) difference and muscle biopsy measurements (Bergman et al. 1999a, 1999b) to study glucose, glycogen, glycerol, fatty acid, and IMTG metabolism during exercise. Results not only support the Crossover Concept but also show that working muscles and the rest of the body coordinate their use of substrates by shifting their substrate utilization patterns during exercise.
Figure 7.17 shows blood glycerol and FFA concentrations in subjects during rest and moderate and hard exercise, both before and after 10 wk of endurance training. When exercise starts, blood glycerol concentration rises as a function of relative power output. Thus, in untrained subjects, the rise in glycerol concentration is greater during exercise at 65% than at 45% V̇O2max. Training does increase glycerol somewhat for a given absolute, but not relative, power output. Comparison of figures 7.17a and 7.17b indicates that blood glycerol concentration is a good marker of its appearance (production) and, therefore, the rate of lipolysis in adipose tissue. This is because glycerol cannot be reesterified back to triglyceride in peripheral tissues, such as adipose and striated muscle, and because in humans adapted to a balanced diet glycerol is a poor gluconeogenic precursor that is removed slowly by the liver for that purpose. Therefore, when glycerol, a water-soluble metabolite, is released from adipose, glycerol accumulates in the blood.
Compared to that for glycerol, the pattern of FFA concentration response and use shows some similarities as well as differences (figure 7.17b). These reflect mainly the greater uptake of FFA by working muscle, but also the limited ability of plasma albumin to access FFAs released in adipose. Typically, when exercise starts, plasma FFA concentration falls because uptake by working muscles suddenly rises. Thereafter, concentration of FFAs rises in the plasma, but the rise depends on variations in the relationship between release from adipose depots and uptake by working muscles and other tissues. Consequently, plasma FFA concentration is not a good indicator of either lipolysis or FFA disposal.
![Figure 7.17 (a) Blood glycerol and (b) free fatty acid [FFA] concentrations in female subjects, during rest and exercise before and after training. Subjects were studied at 45% and 65% of V̇O2max before training and after training at the same absolute power output (ABT) that elicited 65% of V̇O2max before training and 65% of the posttraining V̇O2max (i.e., the same relative intensity [RLT]). Typically, glycerol rises during exercise, whereas FFAs fall and then rise. Values are mean ± SEM of the last 15 and 30 min for rest and exercise, respectively; n = 17; * significantly different from 45UT; # significantly different from ABT (p [less than] .05). Note the absence of statistical difference symbols for FFA because of complex regulation including disposal by oxidation and esterification; see figure 7.18. Reprinted from A.L. Friedlander, G.A. Casazza, M.A. Horning, T.F. Budinger, and G.A. Brooks, “Effects of Exercise Intensity and Training on Lipid Metabolism in Young Women,” American Journal of Physiology 275, no. 5 (1998a): E853-E863.](https://cdn.shopify.com/s/files/1/1281/4359/files/E9786_117art1.jpg)
Reprinted from A.L. Friedlander, G.A. Casazza, M.A. Horning, T.F. Budinger, and G.A. Brooks, “Effects of Exercise Intensity and Training on Lipid Metabolism in Young Women,” American Journal of Physiology 275, no. 5 (1998a): E853-E863.
Results in figure 7.18 were derived from use of glycerol and fatty acid isotopic tracers to provide measures of blood flux [appearance (Ra) and disposal (Rd) rates] in the same subjects as in figure 7.17. In the case of glycerol (figure 7.18a), Ra gives the rate of adipose lipolysis. In the case of fatty acids (figure 7.18b), results depend on FFA release from lipolysis minus disposal by oxidation or reesterification. Hence, data in figure 7.18 reveal several things about the effects of training on FFA mobilization and use. Prior to training, FFA rate of appearance in the blood is higher at a low power output (45% V̇O2max) than a higher power output (65% V̇O2max). However, training increases whole-body FFA Ra and Rd at given absolute and relative power outputs. Thus, in figure 7.18b, the highest FFA Ra is seen at the relatively high power output eliciting 65% V̇O2max. However, even with the training-induced rise in FFA use, the amount of fat use by working human muscle is small.
![Figure 7.18 The effect of exercise intensity and training on the plasma FFA rate of (a) appearance and (b) disappearance. Subjects were studied at 45% and 65% of V̇O2max before training, and then after training at the same absolute power output (ABT) that elicited 65% of the pretraining and 65% of the posttraining V̇O2max—i.e., at the same relative exercise intensity (RLT). Values are mean ± SEM of the last 15 and 30 min for rest and exercise, respectively; n = 8; Δ significantly different from rest; * significantly different from 45UT; # significantly different from ABT (p [less than] .05). Reprinted from A.L. Friedlander, G.A. Casazza, M.A. Horning, T.F. Budinger, and G.A. Brooks, “Effects of Exercise Intensity and Training on Lipid Metabolism in Young Women,” American Journal of Physiology 275, no. 5 (1998a): E853-E863.](https://cdn.shopify.com/s/files/1/1281/4359/files/E9786_117art2.jpg)
Reprinted from A.L. Friedlander, G.A. Casazza, M.A. Horning, T.F. Budinger, and G.A. Brooks, “Effects of Exercise Intensity and Training on Lipid Metabolism in Young Women,” American Journal of Physiology 275, no. 5 (1998a): E853-E863.
Figure 7.19 comes from a study in which working muscle FFA use was assessed by measurements of arterial-venous (a-v) concentration differences measured across legs of resting and exercising men. The experimental design was the same as the studies using stable isotopes (figures 7.17 and 7.18), and the results show limited FFA uptake by working muscles and only small training effects. The same conclusion obtained by different methods increases veracity of the conclusions. Training increases whole-body (tracer-measured) FFA disposal (figure 7.18b) and working-muscle FFA uptake (figure 7.19). In both cases, however, the change due to training is small.
![Figure 7.19 Arterial-venous differences of FFA measured by femoral catheterizations of men during rest and exercise, both before and after training. Results show resting muscle FFA release, and working muscle FFA uptake, but (a-v) differences were small; only rest-to-exercise transitions were significantly different. Values are mean ± SEM of the last 15 and 30 min for rest and exercise, respectively; n = 8; Δ significantly different from rest; p [less than] .05. Reprinted from B.C. Bergman, G.E. Butterfield, E.E Wolfel, et al., “Muscle Net Glucose Uptake and Glucose Kinetics After Endurance Training in Men,” American Journal of Physiology 277, no. 1 (1999b): E81-E92.](https://cdn.shopify.com/s/files/1/1281/4359/files/E9786_118art.jpg)
Reprinted from B.C. Bergman, G.E. Butterfield, E.E Wolfel, et al., “Muscle Net Glucose Uptake and Glucose Kinetics After Endurance Training in Men,” American Journal of Physiology 277, no. 1 (1999b): E81-E92.
SHOP

Get the latest insights with regular newsletters, plus periodic product information and special insider offers.
JOIN NOW