Maintaining an elevated training status impacts FAox potential due to the increase in IMTG, cellular/mitochondrial protein changes, and hormonal regulation. The adaptations that occur due to regular endurance training favor the ability to oxidize fat at higher workloads in addition to increasing over all MFO [39, 40]. Increased fat oxidation has been shown to improve with endurance training, and therefore increases in MFO parallels changes in training status. Bircher and Knechtle, (2004)  demonstrated this concept by comparing sedentary obese subjects with athletes and found that MFO was highly correlated with respiratory capacity, and thus training status.
Trained subjects possess a greater ability to oxidize fat at higher exercise intensities and therefore demonstrates the correlation between respiratory capacity and MFO [27, 41, 42]. However, a similar rate of appearance in serum glycerol concentrations is observed in sedentary vs. trained subjects . These results, however, conflict with results from Lanzi et al. (2014)  who reported that obese subjects had a higher serum FA rate of appearance likely due to increased total adipose tissue mass (kg). Furthermore, sedentary/obese subjects have a reduced cellular transport and fat oxidation capabilities, therefore maintaining higher serum FA concentrations . Despite the reported reduced rate of glycerol appearance for the trained population reported by Lanzie et al., trained women were shown to oxidize fat at twice the rate compared with the obese population .
The training effect, and therefore an increase in respiratory capacity is partially the result of an increase in MFO. Scharhag-Rosenberger et al. (2010)  conducted a prospective study to demonstrate this concept using sedentary subjects who met or exceeded ACSM’s minimum cardiorespiratory exercise recommendations for a period of 1 yr. Maximal fat oxidation (rate) increased over 12 months of training (pre-training 0.26± 0.10; post-training 0.33± 0.12 g/min) and it occurred at a higher exercise intensity (pre-training 35±6% VO2max; post-training 50±14% VO2max). The training status effect on MFO further applies to athletic populations. In moderate vs highly trained subjects, the exercise intensity (%VO2max) that MFO occurred was not significantly different, but MFO was elevated for the highly trained subjects (0.29±10 vs 0.47±.17 g/min, respectively) . Furthermore, mitochondrial enzymes citrate synthase and HAD were found to be significantly increased (49% and 35%) in highly trained vs. moderately trained participants respectively . Increasing HAD directly elevates beta-ox rate while citrate synthase increases the TCA cycle rate . This evidence suggests that lipolysis and systemic FA delivery are not limitations to FAox at higher exercise intensities. Therefore, FA cellular transport proteins (CD36 and CPT-1) [24, 25] and mitochondrial density (HAD) are likely the limitation of FAox during high intensity exercise . Elevating FAox potential by increasing cellular respiration capacity increases FAox at higher exercise intensities which can have a positive influence on aerobic capacity.
Acknowledging the occurrence of large inter-individual differences in MFO, differences in MFO relative to training status are still observed . Lima-Silva et al. (2010)  showed that differences in the lipid oxidative potential may exist in high vs. moderately trained runners referenced above. However, while no statistical differences were observed between groups at the exercise intensity that MFO occurred, there was an increased capacity to oxidize fat in the highly trained subjects. It is worth noting that the increased performance capacity in highly trained runners is most likely attributed to an increased CHO oxidative potential at higher exercise intensities in order to maintain higher steady state running workloads . Subsequently, cellular protein expression, oxidative capacity and therefore training status do have the ability to influence fat oxidation.
Training status further influences maximal fat oxidative potential by increasing endogenous substrate concentrations [19, 20]. Endurance training enhances type I fiber IMTG concentrations as much as three-fold compared with type II fibers. Increased MFO potential due to endurance training is further influenced by IMTG FA-liberating HSL  and LPL proteins , which are responsible for the liberation of intramuscular FAs from the IMTG molecule. However, during exercise, the IMTG pool is constantly being replenished with plasma-derived FAs during exercise [20, 45]. Nonetheless, the reliance of IMTGs during submaximal exercise durations lasting <2 h is essential to maintaining workloads . The exercise duration effect could be due to β-adrenergic receptor saturation, which has been shown to occur during prolonged bouts of exercise [16, 46]. Furthermore, HSL activity has been shown to increase initially within 10-60 min, but returned to resting levels after 120 min of exercise, increasing reliance on serum derived FAs [20, 45]. More research in the area of hormone related FA kinetic limitations is warranted.
The exercise intensity that MFO occurs has been reported to range from 45–75% VO2max [1, 4, 6, 41, 43], with a recent paper highlighting MFO rates from 1121 athletes in various disciplines (American Football, triathlon, golf, soccer, motor sports, cross county, and water sports amongst others) ranging from 23 to 89%VO2max . Factors such as training status, sex, and nutrition  all impact FAox kinetics and thereore the exercise intensity that MFO occurs. Exercise intensity has the most profound effect on MFO based on a combination of events which include FA transport changes [24, 25] and hormone fluctuation, which can increase lipolytic rate . The cellular and hormonal changes that occur during exercise are directly related to exercise intensity which can influence FAox .
Fatty acid oxidation varies relevant to exercise intensity and therefore examining lipid oxidation at specific exercise intensities is warranted. At 25% VO2max, FAox comprises >90% of energy expenditure and more specifically plasma FAs provide the largest energy contribution, where muscle glycogen and IMTG contribute very little . At exercise intensities <65% VO2max muscle glycogen and IMTG oxidation increase considerably to as much as 50% of energy expenditure, depending on exercise duration [15, 48]. Bergomaster et al. (2008)  compared 6wks of sprint interval training (Wingate Tests) to endurance training (~65% VO2max) and found no differences in MFO. These findings suggest that training ≥65% VO2max will not increase MFO potential, which is in disagreement with more recent literature . Previous research suggests that training at higher exercise intensities greatly influences substrate utilization [5, 42, 50]. It is worth noting that Bergomaster et al.  used moderately trained subjects (VO2max = 41.0 ± 2.0 ml/kg/min) where Achten et al. (2004)  and Nordby et al. (2005)  used higher trained subjects (VO2max = 58.4 ± 1.8 and 56.6 ± 1.3 respectively) to formulate their conclusions.
The increased expression of FAox transport and oxidative cell proteins (CD36, CPT-1, HAD, etc.) that results in an increase FAox are a result of exercise intensity [24, 49]. Bergomaster et al. (2008)  suggests a minimum training volume of two weeks is necessary independent of training status for sufficient cellular adaptation to occur. The Lima-Silva et al. (2010)  data, however, show that a heterogeneous sample of highly vs moderately trained subjects (VO2max of 68.4 ± 4.5; 58.6 ± 5.4 ml/kg/min respectively) training for a minimum of 3 yrs. at varying exercise intensities had a 62% difference in fat oxidation rates (0.47 ± 0.17; 0.29 ± 0.10 g/min respectively). Thus, FAox adaptation potential is related to training at higher exercise intensities rather than non-descript chronic exercise adaptation. Additionally, it has also been shown that carnitine concentrations are a direct limitation of FAox (Fig. 1) at higher exercise intensities (>65% VO2max) to both IMTG  and serum FAox , regardless of mitochondrial enzymatic activity in untrained and moderately trained subjects. Interestingly, efforts to mitigate the limitations of free carnitine on MFO at high exercise intensities have been unsuccessful . While exogenous carnitine supplementation increased muscle carnitine by 21% after 4 weeks of supplementation, no differences in performance were observed. While FAox was not measured, increases in muscle carnitine were able to buffer excess acetyl CoA by forming acetylcarnitine and thus increase pyruvate dehydrogenase (PDH) activity (38%) at 80% VO2max .
Exercise intensity may further influence MFO by influencing catecholamine concentrations which have regulatory effects on lipolysis [
], glycogenolysis, as well as gluconeogenesis [
]. Increased epinephrine concentrations that parallel increases in exercise intensity stimulate both glycogenolysis and gluconeogenesis [
]. As exercise intensity increases, so does catecholamine concentrations facilitating a concurrent increase of serum CHO and FAs into the blood [
]. However, the body still favors FAox at exercise intensities
]. When exercise intensity exceeds MFO, FAox (g/min) begins to decline; this process is described as the crossover concept [
] shown in Fig.
The crossover concept. The relative decrease in energy derived from lipid (fat) as exercise intensity increases with a corresponding increase in carbohydrate (CHO). The crossover point describes when the CHO contribution to substrate oxidation supersedes that of fat. MFO: maximal fat oxidation. Adapted from Brooks and Mercier, 1994
The concept of the crossover point represents a theoretical means to understand the effect of exercise intensity on the balance of CHO and FA oxidation  (Fig. 2). More specifically, the crossover concept describes the point that exercise intensity influences when the CHO contribution relevant to energy demand exceeds FAox. The limitations of FAox at higher intensities is due to the vast amount of acetyl-CoA produced by fast glycolysis [24, 38]. The abrupt increase in total acetyl-CoA production at high intensity is due to fast glycolysis flooding the cell with potential energy, which suppresses FA mitochondrial transport potential resulting in decreased FAox (Fig. 1). Notably, the large inter-individual fluctuation of when the crossover point occurs at a given exercise intensity can be attributed in part to training status [39, 40]. Training status has been shown to effect catecholamine release and receptor sensitivity , endogenous substrate concentrations, and cellular transport protein expression; all of which contribute to the variability of when MFO occurs relevant to exercise intensity . Nonetheless, MFO occurs in all populations regardless of training status, nutritional influence, etc., and is decidedly dictated in large part by exercise intensity [5, 6, 42].
Another factor that significantly influences FAox is the duration of exercise [13, 45, 48]. Throughout a prolonged exercise bout, changes in hormonal and endogenous substrate concentrations trigger systematic changes in substrate oxidation [20, 51]. Studies show that endurance training promotes reliance on endogenous fuel sources for up to 120 min of submaximal exercise [47, 51, 52].
Exercise duration has a large effect on the origin of FAs for oxidative purposes. While the initiation of exercise relies heavily on endogenous fuel sources (IMTG and glycogen), reductions in IMTG concentrations have been shown to occur when exercise duration exceeds 90 min . Beyond 90 min of submaximal exercise (≥65% VO2max), IMTG oxidation is mitigated by the increase in serum derived LCFAs [20, 45]. Increases in both epinephrine and plasma LCFA concentrations were observed when exercise exceeded 90 min with a simultaneous reduction in HSL activity. Therefore the increase in serum LCFAs [20, 45] and the saturation of HSL to epinephrine [16, 46] are postulated to inhibit HSL reducing IMTG oxidation when exercise exceeds 90 min . When exercise exceeded 120 min of exercise, IMTG oxidation returned to resting values and was offset by a 46% increase  in serum FA delivery and oxidation . Additional evidence shows that after 12 h of prolonged exercise, IMTG stores are 50–80% of pre-exercise concentrations despite the extreme duration of exercise .
The shift from intramuscular fuel sources to serum derived FAs after 2 h of submaximal exercise parallel changes in blood glucose concentrations. Untrained subjects who completed 3 hrs of knee extensions at 60% of 1RM had a 66% increase in serum glucose concentrations during the second to the third hour of exercise . Trained subjects however experienced a reduction in muscular CHO uptake during the same time frame compared with the untrained. This suggests that the trained subjects were able to maintain FAox (despite substrate origin) during prolonged exercise to stave off CHO usage for high intensity exercise . While the exercise intervention used in this study is not typically classified as endurance exercise, the exercise protocol does clarify the variation in the origin of substrate oxidation over time, and expands on the diverse effects exercise duration has on substrate oxidation.
Training duration has a large influence on FA and CHO oxidation during prolonged submaximal exercise. However, training status has little influence on the origin of FAs during the first 120 min of submaximal exercise. Nonetheless, trained subjects are able to maintain higher workloads with decreased metabolic work (HR) for longer periods compared to untrained individuals based on the ability to maintain FAox for longer durations . Despite the training status effect on FAox, exercise duration will dictate substrate origin during submaximal exercise [20, 45, 51].
Variability in FAox owing to sex exist due to the inherent hormonal differences specific to men and women [53–56]. In a comprehensive study with over 300 men and premenopausal women, the energy contribution of fat was significantly higher in women vs. men at all exercise intensities measured ranging from 41-61% VO2max . Studies have consistently shown that premenopausal women have a significantly greater ability to oxidize fat during exercise [2, 57, 58].
The sex differences in fat oxidation [58, 59] during exercise is attributed to the increased circulation of estrogens [53, 54, 60]. Evidence suggests that estrogen directly stimulates AMPK  and PGC-1α activity , which is thought to increase the downstream FAox transport protein CD36 and beta-oxidative protein HAD . Additionally, beta-oxidative proteins that oxidize LCFA oxidation have been shown to be regulated in part by estrogen [54, 60]. The result of increased beta-oxidative proteins is directly related to increased FAox potential [29, 54]. Interestingly, when men were supplemented with estrogen, increases in FAox were observed along with increased cellular expression of beta-ox proteins within eight days of supplementation .
Circulating estrogen is naturally higher for premenopausal women compared to men. Additionally, fluctuation in estrogen levels is inherent throughout the menstrual cycle [53, 59]. Estrogens are generally higher during the follicular phase of the menstrual cycle compared to the luteal phase . Paradoxically, elevated estrogens during the follicular phase do not affect FAox when compared to the luteal phase [29, 53]. Nevertheless, elevations in endogenous circulating estrogens inherent to premenopausal women increase the expression of cellular proteins responsible for increased FA transport and oxidation compared to men.
Cellular protein expression and the corresponding endogenous vs. systematic substrate oxidation vary according to dietary macronutrient intake [19, 35, 61]. It has been recently shown that high fat diets promote FAox and have performance enhancement capabilities [3, 60]. However, definitive conclusions regarding pre-exercise macronutrient dominant diets and exercise performance improvements are contingent on specific exercise applications  that are directed by exercise duration and intensity [63–65].
Diets that have higher proportions of a specific macronutrient (e.g. fat/CHO) have shown an increased ability to oxidize the primary macronutrient consumed [66–68]. Furthermore, endogenous substrate concentrations increase after acclimating to high fat/high CHO diets [65, 68, 69]. High fat diets increase IMTG concentrations while decreasing glycogen levels within muscle [17, 35]. Alternatively, high CHO diet conditions increase glycogen concentrations while IMTGs decrease . After acclimation, during exercise the body favors oxidation of specific substrates [65, 67] based on long-term (>48 h) cellular adaptation in accordance to macronutrient consumption [3, 35, 69]. However, post-exercise predominant macronutrient (CHO) consumption has been shown to influence cellular protein expression in as little as 2 hrs . The plasticity of cellular changes relevant to chronic adaptation are compromised when macronutrient content is altered [65, 67].
Macronutrient proportion and timing has been shown to have effects on cellular adaptation  as well as the physiological response to exercise [70–72]. High fat diets increase beta-ox potential at rest  and during exercise , however, the limitations of high fat diets (including short term adaptation (5dys)) reside with high intensity exercise [70, 72, 73]. High intensity exercise (>75% VO2max) eclipses the FAox oxidative potential relying on fast glycolysis, or more specifically PDH to produce CHO-derived acetyl-CoA  for ATP re-synthesis [35, 67]. Pyruvate dehydrogenase is the enzyme responsible for oxidizing pyruvate as the final substrate of the glycolytic pathway. The deleterious cellular adaptation of reduced PDH activity due to high fat diets has been found to compromise high intensity exercise performance potential [35, 63, 67].
High fat diets (>68% total daily calorie intake) have had positive effects on lowering RER values [64, 71, 72] during moderate intensity exercise (~64–70% VO2max), and for prolonged exercise durations (~3 hrs)  indicating an increase in FAox. Adapting the body to high fat diets allows the body to increase IMTG storage as well as increase FAox [21, 35]. Contrariwise, PDH activity and therefore CHO oxidation was shown to be compromised [35, 67] along with power output at exercise intensities ≥70% VO2max . However, crossover diet applications where the body was adapted to a high fat diet prior to short term high CHO loading (36-72 h) was shown to maintain IMTG stores  while increasing glycogen stores , partially restore glycolytic enzymes , as well as partially restore CHOox . Increasing MFO (g/min) and the exercise intensity that MFO occurs (% VO2max)  is ideal for long duration exercise performance. However, further research on the initial compromised cellular PDH activity after fat adaptation, and the capacity to restore glycolytic potential after short term CHO adaptation is warranted for prolonged intermittent high intensity exercise (≥70% VO2max) applications.
Alternating pre-exercise macronutrient specificity has the potential to be effective in accommodating the stress of sustained high intensity exercise due to both ideal cellular protein expression, and adequate storage of IMTG and muscle glycogen. The adaptation to high fat diets (>50% Fat of total K/cal)  has been shown to reduce PDH activity  by 59% at rest and 29% at moderate intensity exercise (70% VO2max) . The reduction in PDH activity due to high fat diets is a limiting factor to the necessary CHO oxidation at high intensity exercise despite adequate endogenous energy stores. However, five days of fat adaptation (~67% of total energy intake) with a 24 hr short-term CHO loading period (compared with high CHO diet ~70% total energy intake) maintains IMTG concentrations and partially restores PDH activity (71% of high CHO diet) while maintaining 80% HSL activity . Additionally, no differences were observed in time trial performance (~10 min at 90% VO2peak) between the high CHO and high fat/short term CHO diet adaptation . Maintaining the ability to store and oxidize fat after acclimating to a high fat diet while restoring the ability to oxidize CHO with short-term CHO loading is an ideal physiological state for endurance exercise performance. Furthermore, glycogenolysis is elevated during exercise after CHO loading  indicating an increase in both glycogen storage as well as an increased ability to produce/maintain CHO availability during intense exercise .
Current research asserts that high fat diets favorably enhance FAox at both rest and during exercise [3, 74]. However, exercise intensity dictates substrate utilization regardless of dietary influence, training status, and exercise duration. Because of this, high fat diets are sometimes encouraged during preparatory off-season training when training volumes are high and exercise intensities are low to moderate . However, during sustained high intensity exercise (>70% VO2max) which is common during competition, CHO is the primary substrate relied upon despite short and long term fat acclimation [71, 75]. More research into the short-term macronutrient manipulation effect on endogenous substrate concentrations, plasticity of cellular expression, and preferential substrate oxidation are necessary to ascertain if there is benefit on exercise performance outcomes.