It is generally accepted that optimal adaptation to repeated days of heavy endurance training requires a diet that replenishes muscle glycogen reserves (see the companion paper by M. Saunders on this site for further information regarding the importance of carbohydrate during heavy training). However, some studies have found that when exercise is undertaken with low muscle glycogen content, the transcription of a number of genes involved in training adaptations is enhanced. In fact, exercising with low muscle glycogen stores amplifies the activation of a number of signaling proteins, including the AMP-activated protein kinase (AMPK) and the p38 mitogen-activated protein kinase (MAPK). These two enzymes have direct roles in controlling the expression and activity of several transcription factors involved in mitochondrial biogenesis and other training adaptations (1,2). Thus, some practitioners have suggested that athletes deliberately train in a glycogen-depleted state ("train low") to maximize physiological adaptations to endurance training.
There are many ways to reduce carbohydrate availability for training (1,2):
1) A chronically low carbohydrate diet (carbohydrate intake less than fuel requirements for training).
2) Twice a day training (lowering carbohydrate availability for the second training session by limiting the duration of recovery, and carbohydrate intake between the training sessions).
3) Training after an overnight fast.
4) Prolonged training while withholding carbohydrate intake during the session.
5) Withholding carbohydrate during the first hours of recovery after training.
Most of the "training low" research has used "twice a day" training (starting the second session with low muscle glycogen stores) and withholding carbohydrate during training sessions (1,2).
A 2005 study by Hansen and colleagues sparked intense interest in the "train low" concept (3). Seven untrained men participated in a 10-week program of leg knee extensor exercise and consumed a carbohydrate-rich diet (8 grams/kg per day). Each subject's legs completed the same weekly 5 hour training program but received a different daily schedule: one leg was trained twice-a-day on every second day and the other leg was trained once daily. Thus, one leg exercised with full restoration of glycogen each day ("train high"), while the twice-a-day leg undertook the second session with depleted muscle glycogen stores ("train low"). After 10 weeks, the increase in maximal power was identical for the two legs. However, the training-induced increases in time to fatigue in the "train low" leg were about two-fold greater than the "train high" leg. The "train low" leg also had higher resting glycogen content and citrate synthase activity compared to the "train high" leg. These results suggested that training adaptations may be enhanced by low glycogen availability, thereby improving endurance performance (3).
This study has several limitations (2). The subjects were untrained, so these results may not apply to endurance-trained athletes. The training sessions were also held at a fixed submaximal intensity during the training program whereas athletes typically follow periodised training which incorporate a ‘hard-easy' training sessions and progressive overload. Lastly, the type of training (one-legged knee extensor exercise) and performance trial do not remotely resemble how most competitive athletes train and compete (2). Thus, it is unclear what effects the training low regimen would have on performance in trained athletes. Subsequent studies have addressed some of these issues (4,5).
Yeo and associates recruited endurance-trained male cyclists and triathletes to evaluate the effects of "training low" on training capacity, endurance performance, and substrate metabolism (4). The authors also measured the activity of several muscle enzymes associated with training adaptation – β-hydroxyacyl-CoA-dehydrogenase (an indicator of beta-oxidation) and citrate synthase (an indicator of Kreb's cycle activity). The subjects consumed a carbohydrate-rich diet (8 grams of carbohydrate/kg per day) and completed six training sessions per week for three weeks. Seven subjects trained daily ("train high"), alternating between a steady-state aerobic ride one day (100 minutes at 70% VO2 peak) and a high-intensity interval training session the next day (5 minute intervals at maximum self-selected power output, repeated 8 times with a 1 minute recovery between intervals. Another seven subjects trained twice daily, every other day ("train low"), with the steady-state aerobic workout followed by the interval session an hour later. The subjects' training intensity was measured as the self-selected power outputs achieved in the interval session. Performance was measured 48 hours before and after the first and last training session, using a one hour time trial completed immediately after an hour of steady-state cycling (4).
The training intensity achieved during interval training was significantly lower in the "train low" group for the first two weeks, but by the third week the training intensities were not different between groups. The "train low" group experienced significant increases in resting muscle glycogen concentrations, fat oxidation during steady-state cycling, and activity of the muscle enzymes β-hydroxyacyl-CoA-dehydrogenase and citrate synthase, while the "train high" group did not observe significant changes in these responses (4).
Despite metabolic and muscle enzyme changes indicating an enhanced training adaptation in the "train low" group, there was no obvious benefit to endurance performance (4). Performance during the one hour time trial was significantly higher following training in both groups (12.2% for "train low" and 10.2% for "train high") but there were no significant differences in performance between the two groups.
Hulston and colleagues recruited endurance-trained male cyclists to determine the effects of training with low muscle glycogen on exercise performance, substrate metabolism, and skeletal muscle adaptation (5). The experimental design was similar to that of Yeo and associates. The subjects consumed a carbohydrate-rich diet (8 grams of carbohydrate/kg per day) and completed six training sessions per week for three weeks. Seven subjects trained daily ("train high"), alternating between a steady-state aerobic ride one day (90 minutes at 70% VO2max) and a high-intensity interval training session the next day (5 minute intervals at maximum self-selected power output, repeated 8 times with a 1 minute recovery between intervals. Another seven subjects trained twice every other day ("train low"), with the steady-state aerobic ride followed by the interval session an hour later. Performance was measured 48 hours before and after the first and last training session by a one hour time trial completed after an hour of steady-state cycling (5).
The power output during high-intensity interval training was significantly lower in "train low" (297 W) compared with "train high" (323 W) throughout the training period. Performance during the one hour time trial was significantly higher following training for both groups (10.2% for "train low" and 10.5% for "train high") but there were no significant differences in performance between the two groups (5).
Fat oxidation during steady-state cycling significantly increased following training in "train low" from 26 to 34 μmol/kg/minute, primarily due to an increase in muscle triglyceride oxidation from 16 to 23 μmol/kg/minute. Plasma free fatty acid oxidation was similar before and after training in both groups. Training with low muscle glycogen also increased β-hydroxyacyl-CoA-dehydrogenase protein content (5).
Training with low muscle glycogen reduced training intensity and was no more effective in improving performance than training with high muscle glycogen. However, fat oxidation was increased after training with low muscle glycogen, which may have been due to the enhanced metabolic adaptations in skeletal muscle. Further research is needed to determine whether the increased ability for fat oxidation translates into better performance during longer duration endurance exercise (5).
"Training low" proponents assert that elite endurance athletes need to periodically train in a glycogen-depleted state to fully exploit endurance training responses (6,7). Proponents also claim that "training low" decreases the need for carbohydrate during competition (2). Thus, "training low" could potentially reduce the risk of developing gastrointestinal distress by decreasing the amount of carbohydrate-rich foods and fluids consumed during competition (2).
There is good evidence that undertaking endurance training with low muscle glycogen content amplifies training adaptations (mitochondrial biogenesis) versus when the same training is undertaken with normal or elevated muscle glycogen levels. However, despite creating metabolic and muscle enzyme adaptations that should enhance endurance, there is no clear proof that "training low" improves endurance performance. Hawley and Burke offer several explanations for these equivocal findings (1,2):
1) It is possible that performance is not related to the muscle and metabolic markers that have been measured.
2) The duration of the studies (three to 10 weeks) have been too short to adequately assess muscle adaptation and performance.
3) Researchers have been unable to measure performance with enough sensitivity to detect changes that would be worthwhile in real-life competitive events.
It is also possible that potential adverse effects of "training low" counteract the presumably beneficial muscle and metabolic adaptations. "Training low" may interfere with the intensity and/or volume of endurance training by increasing the perception of effort and reducing power output. "Training low" may also decrease immune function and increase the risk of illness and/or injury (1,2).
Most elite endurance athletes follow a complex, year-long periodized program of training and nutrition. Some training sessions are undertaken with low carbohydrate availability (overnight fasting, several sessions a day, and/or little carbohydrate intake during the workout), while others are undertaken with high carbohydrate availability (more recovery time, meal post-exercise and/or carbohydrate intake during the session) (7).
Despite being physiologically and psychologically challenging, it is possible that "training low" may further enhance training adaptations and improve fat oxidation in elite endurance athletes (6,7). Incorporating "training low" strategies is most suitable during lower intensity sessions and/or at the beginning/middle of a training cycle (2,7).
The ability to generate high power outputs and work rates is a critical component of a periodized training program. Furthermore, the strategic moves that occur during competition depend on the athlete's ability to work at high intensities, which are fueled by carbohydrate. Thus, a diet with high carbohydrate availability is recommended during periods of heavy training and when an athlete is preparing to peak for competition (1,2; see also the companion paper on this site by M. Saunders, which discusses the role of dietary carbohydrates during heavy training).
1. Hawley JA, Burke LM. Carbohydrate availability and training adaptation: effects on cell metabolism. Exerc Sport Sci Rev. 2010;38:152-160.
2. Burke LM. Fueling strategies to optimize performance: training high or training low? Scand J Med Sci Sports. 2010;20(Suppl 2):48-58.
3. Hansen AK, Fischer CP, Plomgaard P, Andersen JL, Saltin B, Pedersen BK. Skeletal muscle adaptation: training twice every second day vs. training once daily. J Appl Physiol. 2005;98:93-99.
4. Yeo WK, Paton CD, Garnham AP, Burke LM, Carey AL, Hawley JA. Skeletal muscle adaptation and performance responses to once a day versus twice every second day endurance training regimens. J Appl Physiol. 2008;105:1462-1470.
5. Hulston CJ, Venables MC, Mann CH, Martin C, Philp A, Baar K, Jeukendrup AE. Training with low muscle glycogen enhances fat metabolism in well-trained cyclists. Med Sci Sports Exerc. 2010;42:2046-2055.
6. Stellingwerff T, Boit MK, Res PT. Nutritional strategies to optimize training and racing in middle-distance athletes. J. Sports Sci. 2007; 25(Suppl 1):S17-S28.
7. Stellingwerff T. Case study: nutrition and training periodization of three elite marathon runners. Int J Sport Nutr Exerc Metab. 2012; 22:392-400.