The Scandinavian Experiments That Invented Carb Loading

Modern endurance nutrition rests on a handful of Scandinavian experiments from the 1960s, and the reason they were possible at all is a needle.

The technique that made it possible

For most of the twentieth century, studying what happened inside working muscle meant inferring it from blood samples and breath. You could measure what went in and what came out. What was actually happening in the muscle itself was largely guesswork.

In the early 1960s the Swedish researcher Jonas Bergström revived and refined a percutaneous needle biopsy technique — a way of taking a small sample of muscle tissue with a needle, quickly, repeatedly, and safely enough to do it to the same person several times in a day.

That is the whole story, in a sense. It turned muscle metabolism from something you modelled into something you measured.

Bergström and his colleague Eric Hultman promptly used it on themselves and on each other, which is a recurring feature of the era and one of the reasons the field moved so quickly.

What they found

The core observation: muscle glycogen — the stored carbohydrate in the muscle itself — is a limiting fuel for prolonged exercise, and the amount you start with strongly predicts how long you can keep going.

Then came the finding that produced the practice. When muscle was depleted of glycogen by hard exercise and then supplied with a high-carbohydrate diet, it did not simply refill. It overshot, storing considerably more glycogen than it had held originally.

The most elegant demonstration of the principle used a one-legged exercise design: exercise one leg to depletion, leave the other alone, then feed the subject a high-carbohydrate diet and biopsy both. The exercised leg had supercompensated. The rested leg had not.

Since both legs shared the same bloodstream, the same hormones and the same diet, the effect had to be local to the muscle that had been depleted. That is about as clean as physiology experiments get.

Related work by Bengt Saltin and colleagues extended it to running and to race performance, and the practical protocol followed almost immediately.

The original protocol, and why nobody uses it

The classic regimen that entered marathon culture in the 1960s and 70s went roughly:

1. Days 7 to 4 before the race: an exhaustive session to deplete, followed by three days of very low carbohydrate intake with continued training
2. Days 3 to 1: very high carbohydrate intake with minimal training

The depletion phase was genuinely awful. Athletes describe those three days as among the worst of their training year — exhausted, irritable, unable to concentrate, and running badly. Some got ill. Some got injured, which is a spectacular way to lose a race you have trained six months for.

It also worked. The supercompensation was real and the performance benefit was measurable. So people did it anyway for years.

How it got softened

Later research established the crucial practical point: trained athletes do not need the depletion phase.

Trained muscle is already good at storing glycogen. Simply reducing training during a taper and eating a high-carbohydrate diet for one to three days achieves most of the same increase in muscle glycogen, without spending race week hollowed out.

That is where the practice sits now:

• 1 to 3 days of high carbohydrate intake
• 8 to 12 g/kg body weight per day is the commonly cited range
• Alongside a taper, with training volume already reduced
• No depletion phase

The unglamorous detail that decides whether it works: those numbers are a lot of food. For a 70 kg runner, 10 g/kg is 700 g of carbohydrate a day. Most people who think they are carb loading eat a bowl of pasta on Saturday night and arrive nowhere near it.

The other practical point: high carbohydrate means low fibre for those days, not extra vegetables. And expect to gain a couple of kilograms, since glycogen is stored with water. That weight is fuel, not fat, and it is supposed to be there.

What remains true

Glycogen availability limits endurance performance. Comprehensively supported, and it is the mechanism behind the wall.

Loading helps for long events. Beyond roughly 90 minutes, starting with fuller stores delays depletion and improves performance. For a 5 km, it does nothing, because you were never going to run out.

Carbohydrate during the event matters too. This is the larger modern development. Race-day intake targets have risen substantially, and the ability to absorb high hourly carbohydrate rates is now understood to be trainable — the practice is in gut training.

Do not do it for the first time on race week. Rehearse it before a long training run. The failure mode is gastrointestinal and it is deeply unpleasant to discover at kilometre 25.

What did not survive

The depletion phase. Marginal benefit, large cost.

Carb loading for everything. It became a general-purpose ritual applied to 10 km races and gym sessions, where it accomplishes nothing except a large dinner.

The idea that fat adaptation replaces it. Low-carbohydrate approaches genuinely increase fat oxidation, and the evidence that this improves performance at race intensities is weak — with some indication that it impairs high-intensity capacity by downregulating carbohydrate use. For events where the pace is high enough to require carbohydrate, glycogen still matters.

What it says about the field

The Scandinavian experiments are one of the better examples of sports science getting something right and getting it right early — a clear mechanism, a clean experimental design, an immediate practical protocol, and a finding that has survived sixty years of scrutiny largely intact.

They are also an example of the usual refinement process. The first protocol was correct and unnecessarily brutal. Two decades of follow-up work established that most of the benefit was available at a fraction of the cost, and the brutal part was quietly dropped.

That is the same trajectory as Zátopek's repetition volumes and Lydiard's 100-mile weeks: a genuine discovery, an initial version that demanded far more suffering than necessary, and a modern version that keeps the physiology and discards the ordeal.

Eat a lot of carbohydrate for two days before a long race, reduce your training, and skip the part where you make yourself miserable first. Two Swedish physiologists took biopsies from their own legs so you would not have to.