Glycogen efficiently stores glucose in a branched, dense polymer. A study released in 2004 suggests that glycogen starvation during training increases endurance performance.
Glycogen is polymerized glucose. For athletes, glycogen has many advantages. When glycogen is broken down into pyruvate during glycolysis it results in a net gain of 3 ATPs. ATP is our body's and muscle's high-energy delivering molecule. When glucose is broken down during glycolysis it produces just 2 ATP. Thus, glycogen produces 50% more ATP during glycolysis than glucose. When the cell is in hypoxia (oxygen depletion) it is unable to break down the glycolysis product, pyruvate, further. Thus, for anaerobic exercise, glycogen breakdown results in a considerable energy benefit. In addition, as the image above shows, glycogen is branched every ten glucose monomers. This allows more glucose molecules to be available for glycolysis (only glucose molecules at the end of a chain can be cleaved). Over the course of animal evolution, glycogen has developed into the primary energy-storage molecule because of its denseness and efficient energy-release.
Glycogen has long been known to be beneficial to athletes; hence carbohydrate loading (carbo loading's affect on performance is debatable, but that is another story). Is glycogen beneficial during training? That is the question an elegant Danish study sought to answer. The study found that carbohydrate depletion during training was beneficial to performance (Skeletal Muscle Adaptation: training twice every second day vs training once daily; 2004; Anne K. Hansen, et al.)
The study took 7 healthy, untrained men and put them on a strict training schedule for ten weeks. The training consisted of leg extensions. Each participant worked out both legs for an hour, wait two hours under a fast, and workout one of the legs (low-glycogen protocol) for an hour. The next day, the participant would workout the other leg (high-glycogen protocol) for an hour. On this two-day sequence both legs got two hours of work, but one at lower levels of glycogen than the other. This is because following the first hour bout of exercise, glycogen levels in the muscle drop. The two-day sequence was repeated for ten weeks. The experimental sequence can be seen in the diagram below.
In response to 10 weeks of training, maximum power output increased significantly, being the same in the two legs. The endurance at 90% of this new maximum power output was markedly increased for both legs, but time until exhaustion was twice as long for low-glucose protocol trained legs compared with high-glucose protocol trained legs. In accordance, the actual work performed by low-glucose protocol trained legs was also markedly larger compared with high-glucose protocol trained legs. The results can be seen in the table below.
Muscle glycogen levels were raised significantly in low-glycogen protocol trained legs at the conclusion of 10 weeks of training. The authors of the study speculated that this may have been the result of a change in GS enzyme, a regulator of glycogen production. Mitochondria enzymes CS and HAD were both significantly higher in low-glycogen protocol trained legs.
The results of this study seem to support this hypothesis regarding glycogen: train low, perform high. However, these results must be treated with caution when applied to athletes. Other factors could be at play; training twice every other day compared to once daily may affect other pathways besides glycogen. In addition, glycogen starvation would potentially limit the amount of training one could conceivably do.
In summary, this study suggests, but does not prove, that glycogen starvation during training may increase endurance.