For the past 4 months, I have been training for the Boston Marathon. My weekends off have consisted of running 16-20 miles through sub-zero wind chills, snowstorms, and unseasonable warmth. Even with the unpredictable weather, I have thoroughly enjoyed my preparation. After living in Boston for 4 years and spending much of my time in the hospital or on the esplanade running around the Charles, the marathon is a perfect finale.
In medical school and residency, you learn to critically evaluate scientific evidence As I was finishing my sweet potato waffles this morning to start my “carbo-load” for the Boston marathon, I wondered: Is this evidence-based?
This was a fleeting thought because, like most people, I don’t need much encouragement to increase carbohydrate intake. (Beer? Pizza? Muffins? I’m in!) A quick Google search provided an abundance of vague “evidence” referencing “studies” for and against the practice. I decided to put on my skeptical physician hat and check it out for myself.
Glucose is stored as glycogen in the liver and muscle. During moderately strenuous prolonged (>70 min) exercise, we use glycogen, then fat oxidation for energy. Since glycogen fuels our muscles, increasing our carbohydrate intake prior to the race, we will have more fuel for our muscles. We will run faster and HAHDAH (Read the last word with Boston accent).
The original study on carbohydrate loading (or “supercompensation”) was performed in 1971 in Sweden.1 The study took 10 students and had six of them eat a high carbohydrate diet for 3 days AFTER not eating ANY carbs for 3 days. The remaining 4 students ate a normal, mixed diet. All subjects ran a 30km race. Three weeks, later they reversed the diets and all participants ran a second 30km race. Each participant ran faster after following the high-carbohydrate diet, with a mean time improvement of 8 minutes.
In 1981, Sherman et. al took three different diets (15%, 50%, and 70% carbohydrates) and tested a 3 and 6 day carbohydrate-loading regimen.2 The runners then completed taper treadmill runs, followed by a day of rest, then the “performance run” of 20.9km on a treadmill. Muscle biopsies indicated increased glycogen levels in the runners eating a high carbohydrate diet. There was NO difference in training run performance.
A 1995 study involving a whopping 7 men and 8 women split participants into a high carbohydrate (75%) vs. low carbohydrate (55-60%) diet.3 Participants used a 3 day taper followed by a rest day. After the rest day, participants cycled to fatigue. The study showed that the men had a significant increase in time to fatigue when supplemented with a high carbohydrate diet. Women did not show a significant increase and when pooled together there was “no significant effect of diet on performance time.”
If these studies were performed to evaluate a medical therapy or intervention they would surely not convince me. The studies show little downside to carbohydrate loading and I feel it is a safe practice provided it is not taken overboard. The data does not show you HAVE to carbo-load for optimal performance. Additionally, there seems to be a trend towards an improved performance in male participants, suggesting there may be a difference in carbohydrate/fat utilization between the sexes.
The Bottom Line
If you like pasta and sandwiches go ahead and eat them before a race, however it will not be a panacea for poor training and nutrition in the training months before a race.
- Karlsson, J., and B. Saltin. Diet, muscle glycogen, and endurance performance. J. Appl. Physiol. 31:203-206, 1971.
- Sherman, W.M., D.L. Costill, W.J. Fink, and J.M. Miller. Effect of exercise-diet manipulation on muscle glycogen and its subsequent utilization during performance. Int. J. Sports Med. 2:114-118,1981.
- Tarnopolsky, M. A., S. A. Atkinson, S. M. Phillips, and J. D. MacDougall. Carbohydrate loading and metabolism during exercise in men and women. J. App. Physiol. 78(4):1360-1368, 1995.
- Sedlock DA. The Latest on Carbohydrate Loading: A Practical Approach. Curr. Sports Med. Rep., Vol. 7, No. 4, pp. 209-213, 2008.