Carbohydrate loading raises muscle glycogen above normal before an endurance event, and the modern version takes as little as one day of very high carbohydrate intake — around 10 grams per kilogram of body weight — combined with rest. A 2002 study by Bussau and colleagues found that one day at that intake supercompensated glycogen as effectively as the older multi-day protocols, without the depletion phase. The practice benefits events longer than about 90 minutes and does nothing useful for shorter ones.
24 News Click publishes information, not medical advice. Pre-race nutrition for athletes with diabetes or other conditions belongs with a clinician or sports dietitian.
Where did the classic depletion protocol come from?
The original protocol dates to Scandinavian muscle-biopsy experiments published by Bergström and Hultman's groups in 1967. They showed that a low-carbohydrate diet drained muscle glycogen, that hard exercise drained it further, and that carbohydrate eaten after depletion was stored at above-normal levels — the supercompensation effect.
The recipe that followed, popularized in the late 1960s and 1970s, was a brutal week: three to four days of low-carbohydrate eating plus exhausting training, then three days of very high carbohydrate and rest. It worked mechanically, but trials documented impaired training, irritability, and immune stress during the depletion phase, and its advantage over simpler schemes did not survive later research.
What replaced it?
A tapered version. By the 1980s, studies showed that the depletion phase was unnecessary: athletes could reach near-supercompensated glycogen with two to three days of rest or very light training plus about 7 to 10 grams of carbohydrate per kilogram per day. Sherman and colleagues' 1981 trial in runners is the classic demonstration, finding the moderate six-day protocol matched the depletion version for glycogen and was easier to tolerate.
Then Bussau's 2002 study compressed it further: one day at around 10 grams per kilogram with complete rest achieved full supercompensation in trained subjects. Later work confirmed that a single very-high-carbohydrate day, or a one-day protocol preceded by a brief hard session, restores and can overshoot glycogen in endurance-trained muscles, which store glycogen more readily than untrained ones.
How much benefit does loading actually produce?
The benefit is real but bounded. Trials and reviews place the typical endurance improvement in the range of roughly 2 to 3 percent for events beyond about 90 minutes — the difference between glycogen-fueled and glycogen-diminished racing in time trials. It will not transform a recreational runner's marathon, but at the pointy end of a field it is meaningful.
For events under about 60 to 90 minutes, loading shows no reliable benefit, because normal glycogen stores already cover that duration. Loading also adds roughly 1 to 2 kilograms of water mass, since each gram of stored glycogen holds water with it — a small but real cost in weight-sensitive sports, and a common reason athletes feel heavy on race morning.
What does a modern load look like in food terms?
For one to three days before the event: carbohydrate at roughly 8 to 10 grams per kilogram per day, with training rest or very short taper work, fat and fiber reduced to make room and protect the gut. Practical menus center on rice, pasta, bread, potatoes, juice, and low-fat sweets in ordinary portions repeated across the day.
For a 70-kilogram runner, 10 grams per kilogram means roughly 700 grams of carbohydrate daily — a demanding amount of food, which is why sports dietitians historically lean on juice, sports drinks, and simple foods during the load. The Mayo Clinic summarizes this modern carbohydrate-loading approach in its patient-facing guidance.
Does the anabolic window of load timing matter — one day or three?
One day works if intake is high enough and training stops, per the 2002 Bussau data. Three days at moderate-high intakes works as well and is gentler on appetite and digestion. Reviews present the two as interchangeable in outcome; the choice comes down to logistics and gut tolerance.
What does not work is loading without tapering: continued hard training during the load consumes glycogen as fast as it is stored. The rest component is as essential as the carbohydrate component, a point the original Scandinavian work made and later trials preserved.
Does loading interact with race-day fueling?
Loaded muscles still burn carbohydrate during the race, so event-day intake of 30 to 60 grams per hour remains the position-stand recommendation for events over an hour. Loading raises the starting store; mid-race fueling slows its drain. The two practices stack, they do not substitute.
Loaded athletes also pace differently in the research literature: with fuller stores, the early-race glycogen conservation that slow starts provide matters slightly less, though pacing strategy trials are separate from loading trials and the interaction is not deeply studied.
What does the evidence not show?
The evidence does not show benefits for events shorter than about 90 minutes, does not support the depletion phase under any modern reading, and does not show that fat-adaptation or low-carbohydrate preparation outperforms carbohydrate loading for endurance racing — trials of high-fat diets show preserved or impaired high-intensity capacity. The marketing of week-long loading products rests on a protocol the field abandoned decades ago.
Does loading work differently for women?
Most of the classic loading studies were done in men, and sports-nutrition reviews flag this gap honestly. What is known: women carry proportionally less muscle mass and, in some studies, showed smaller supercompensation responses at the carbohydrate intakes tested in early trials, which were set by body weight scaled to male physiology. Later work suggested higher total intakes can close the difference, but the trial base in women remains thin.
The practical implication is not that women should skip loading. It is that the published gram-per-kilogram targets were derived mostly in male subjects, and female athletes may land at the upper end of the ranges — or simply find the food volume demanding — which makes practicing the load in training doubly useful. Reviews also note that menstrual phase can shift metabolism and body water, adding noise that a one-off race experiment cannot separate.
This is a recurring theme in sports nutrition: where the evidence base is dominated by one sex, individual testing under realistic conditions replaces confident general rules.
What loading is not
Loading is not simply eating a large dinner the night before a race. A single oversized meal after a full training week cannot refill glycogen that was spent in the preceding days, and the water weight that appears on the scale the next morning is often misread by athletes as a successful load when the underlying stores are only partially restored. The documented protocols pair carbohydrate intake with reduced training over one to three days for exactly this reason.
FAQ
How long should a carbohydrate load last?
One to three days. A 2002 study found one day at about 10 grams per kilogram with rest fully supercompensated glycogen, while the classic 1981 Sherman trial showed two to three days at 7 to 10 grams per kilogram achieves the same result more comfortably.
Does carbohydrate loading help a 10-kilometer race?
Probably not. Benefits appear in events beyond roughly 90 minutes; normal glycogen stores cover shorter efforts. Loading also adds water weight, a net negative for short races.
Why do runners gain weight during a load?
Each gram of stored glycogen binds several grams of water, so a successful load typically adds 1 to 2 kilograms. The gain is water, not fat, and it is part of the mechanism by which glycogen is stored.
For more context, read Carbohydrate periodization: what endurance athletes need.
For more context, read marathon fueling plan.
For more context, read what to eat before morning training.