Fasted rides: train "low", race "high"

Ask any athlete about training and nutrition, and most would repeat the same message – you need to make sure your energy stores are topped up in order to make the most of your training sessions. Right? Maybe not! Recent research, from a number of groups, suggests that training ‘on empty’ may actually be beneficial, flying in the face of conventional wisdom.

Muscle glycogen stores are essential for maintaining high intensity exercise. Typically, the body stores around 300 to 400g of carbohydrate in this form - that’s about 1600kcal of readily available energy (or 3 hours of good paced exercise).  Carbohydrate ‘loading’ can enhance exercise performance by allowing the athlete to race at their optimal pace for a longer period before fatiguing. Most of the research in this area quite simply stated suggests athletes should maintain high carbohydrates diets (10g per kilo of body weight per day), with particular care taken to post exercise nutrition. The market is swamped with industry claims about ‘recovery products’: companies waxing lyrical about ideal mixes of carbohydrates; additions of protein; protein types impacting on absorption rates – all of which necessitate the athlete to take in the product within the first 20 minutes after exercise. More detail on glycogen repletion strategies can be found in another of our PBscience factsheets.

So, what’s changed?

Low_glycogen_chartRecent research looking at how an athlete’s nutrition interacts with the body’s genes ‘switching’ on and off (which promotes muscle adaptations to training) has suggested that training with low or moderate glycogen levels may accelerate the transcription of several important genes. In one study involving 7 untrained volunteers1, greater increases in muscle enzyme content and exercise endurance were seen when training was performed with lower muscle glycogen stores. In a very elegant experimental design, the volunteers trained one leg twice a day, followed next day by rest. The other leg did the same training volume, but spread across two days. The research group reported that performance at 90% of maximum (~ 10 mile time trial intensity) improved significantly after 3 weeks of training in both legs – but, more remarkably, the ‘LOW’ glycogen leg (which had been training twice a day) improved by almost twice as much!

How does ‘training low’ affect the body’s systems?

Hansen and colleagues, the authors of this study, also took muscle biopsies from both legs, and were therefore able to measure the content of key enzymes involved in exercise metabolism. Their measurements in the LOW leg included:

  • Increased concentrations of enzymes key to oxidation and energy production;
  • Increased mitochondria, the power houses of the muscle tissue;
  • Increased concentrations of the stress hormones adrenalin and noradrenalin.

It would appear that depleting the muscle glycogen stores acts like a ‘cellular signal’ switching on genes that in turn, change the production of proteins in the cell. One affected protein is an enzyme called ‘AMP activated protein kinase’ or AMPK, which when activated, encourages the build-up of mitochondria within the muscles.  The greater the AMPK activation, the more mitochondria, and that results in a greater capacity for producing aerobic energy – possibly explaining the greater endurance ability. Having more mitochondria will help an athlete to be more efficient at burning fat: since this process relies predominantly on using oxygen. The hormone response was also interesting - training on low glycogen puts the body under extra stress, meaning it produces more adrenaline. Therefore the body adapts to dampen the body’s response to adrenaline. This hormone is strongly linked to lactic acid production – if the body is less sensitive to the adrenalin’s effect, we might see lowered lactic acid production which in turn helps to increase the lactate threshold to a higher intensity.

Does this mean I should throw away the sports recovery drinks?

It’s too early to be sure that this practice will help the performing athlete. A study just published2 wanted to make the research more applicable to the sporting scenario. It criticized the original work of Hansen and colleagues because:

  • The data was collected on untrained volunteers;
  • The mode of exercise (leg extension performance) was not relevant to athletes;
  • The 3 weeks of training was ‘clamped’ at a set intensity – in fact, it remained at 70% of max for an hour for every session;
  • Athletes are more likely to train using a variety of session types (steady work AND intervals);
  • The majority of athletes are reluctant to take a lot of rest days, making this type of training protocol unlikely;

In the study of Yeo and co-workers2, they designed a study which they view as having more ‘ecological validity’ (where what is done in the lab reflects true life practice in the field). They asked 18, endurance trained cyclists / triathletes to perform 3 weeks of training. Two groups were set up:

  1. HIGH group – trained 6 days per week, alternating sessions of one hour at 70% VO2peak with an 8 x 5 minute interval session;
  2. LOW group – trained twice per day, every second day using the same sessions as the HIGH group (with a 2 hour rest between).

The intervals consisted of the athletes going as hard as they could to maintain 5 minutes of high intensity training. This kind of ‘HIT’ has been shown to be a very effective training session3 and more importantly for this study, to deplete 50% of carbohydrate stores. Both groups were fed the same diet (8-9 per day of carbohydrate) to make sure that glycogen stores would only differ due to the training regimens. After the training, performance was monitored by asking the groups to ride for 60 min at 70% VO2max, and then to ride a 1 hour trial, sustaining the highest power they could. This second hour would be akin to a 25 mile time trial effort, or a 10 mile running race.Chaingang

The results, perhaps a bit more applicable to the athlete population than Hansen’s original paper, demonstrated some similar findings: Fat oxidation tended to be higher in the performance trial in the LOW group and the muscle enzymes involved in oxygen requiring energy processes increased more in the LOW group than in HIGH.

However, unlike in the Hansen paper, there was NO difference in the performance trial improvements between the two groups. What is worth noting here is that the HIGH group trained at a higher intensity during the HIT session: undoubtedly, this is a key finding of this paper – even with the metabolic effects of ‘training low’, without glycogen, training quality may be compromised.

A really interesting finding from this study was that already trained athletes had their glycogen levels FURTHER enhanced – the LOW group had significantly higher resting glycogen when measured after the training regimen. The authors linked the enzyme adaptation to this glycogen response: frequent fluctuations in low, high glycogen may have disturbed the body’s homeostasis (its internal thermostat) causing the gene switching to occur.

Practical implications

There is still a lot of interest in the sport science community how depletion of glycogen may change the way the body switches genes on and off after exercise, and the accumulative effect of this on fitness. However, it is too early for coaches to recommend this regimen to the training athlete. It is worth considering these points:

  • Training with low glycogen stores may lead to immune system suppression and the increased likelihood of falling ill.
  • Training intensity appears to be compromised when you have to train twice a day, on low muscle glycogen stores.
  • Although this novel research may prove to bear fruit during training, it is not a practice advised for athletes before competition: in other words, always “race high”!

However, since the training effects in the most recent study of Yeo and colleagues appear to be at least equal with and without sufficient glycogen stores, training twice a day MAY be a useful strategy for athletes with restricted time for training. This would make the nutrition on the intervening rest day critical. It might also be that the coach and athlete could specify times in the training year where small, intense blocks of this work are completed – the aim to be increased fat burning, NOT to increase power outputs around race intensity (e.g. early on in the periodised year).

So, for now, don’t throw away your protein recovery drinks!



1.   Hansen et al. J Appl Physiol 2005, 98, 93-99.

2.   Yeo et al. J Appl Physiol 2008, 105, 1462-1470.

3.   Lindsay et al. Med Sci Sports Exerc 1996, 28, 1427-1434.