The "train low, race high" paradigm has been widely discussed over the past few years. The idea is that you do some training sessions with depleted carbohydrate stores, and this helps teach your body to burn fat more efficiently. Then when it comes time to race, you make sure to have fully stocked carbohydrate stores, which allows you to take advantage of "metabolic flexibility" to burn both fuel sources efficiently.
(One point that always needs clarification: this is not about eating a chronically low-carb diet. It's timing your workouts and food intake to be in a low-carb state a few times a week for specific sessions.)
The research into "train low" has produced mixed results. An initial study found enhanced performance, but it was a pretty odd design: one-legged kicking with different training protocols for each leg. Subsequent studies have found that training low can produce some apparent metabolic benefits, but in general these benefits haven't translated into actual increased performance. Meanwhile, others have raised concerns that training low can deplete your immune function and break down muscle protein. So the jury is definitely still out.
A new study from Martin Gibala's group at McMaster University, published in the International Journal of Sport Nutrition and Exercise Metabolism, offers some interesting new twists. The protocol involved two weeks of twice-a-day, three-times-a-week interval workouts; each workout was 5 x 4:00 cycling at 60% of peak power with 2:00 rest. The 18 subjects were divided into two groups. During the three hours between the two workouts, one group fully replenished its carb stores (they took in 195 grams of carb), while the other didn't (they took in only 17 grams of carb). That was the only difference between the two groups: the workouts remained identical.
After two weeks, the train-low group had improved its performance on a ~20-minute time trial by 16 percent, while the train-high group had improved by only 8 percent – the first demonstration of enhanced performance in full-body exercise from a train-low protocol. Interestingly, the train-low group didn't fare significantly better in a repeated-sprint test, which may suggest that train-low is most effective for endurance tasks (not surprising if it boosts fat-burning).
Contrary to expectations, muscle biopsies didn't show any advantage in mitochondrial content, which is usually the key driver of endurance adaptations. So why did performance improve? The short answer is that it's impossible to know for sure. One feature of this particular study is that the workout intensities were fixed: everyone did 5 x 4:00 at 60 percent of peak power, regardless of how they were feeling. The workouts were presumably harder for the carb-starved group, which may have produced a greater training stimulus. Or at least, the workouts would have felt harder, which may have produced the same effect: "subjects in the [train low] group may have been prepared to 'push harder' in the time trial," the authors note.
From a practical perspective, the short duration of the protocol (two weeks) is one of the key features of the study. Given the concerns about lowered immune function and so on, training low for long periods of time may not have a great risk-benefit ratio. The biggest caveat is that the subjects were recreationally active, not trained athletes. As the authors point out, it's far easier to produce improvements in untrained people than in well-trained athletes.
So where does that leave us? There are still plenty of unanswered question about how, and indeed if, training low helps. But this seems like an encouraging sign, and makes an interesting case for the benefits of very short train-low cycles. Next question: when should you insert those two weeks relative to a race?
Find runs near you