The young distance runner Jake Smith was affiliated with a small company his agent connected him with ahead of the World Half Marathon Championships in Gdynia, Poland, last year. “They literally told me that I needed to eat more,” he recalls, referring to his performance in the London Marathon two weeks earlier, where he struggled as a pacer.
An adhesive patch with a thin filament embedded in it adorned the 22-year-old’s upper right arm. An Atlanta-based start-up called Supersapiens in collaboration with medical device giant Abbott Laboratories created a continuous glucose monitor (CGM), designed to track blood sugar levels in diabetics, and repurposed it for athletes. Smith’s glucose levels were middling when he started the race and then steadily declined during it, according to the data he uploaded after the race. Toward the end of the run, he says, he felt like, “This shouldn’t feel so bad.”. He ate pasta, rice, chicken, vegetables, and fruit the day before the race, and he kept a watchful eye on his phone’s Supersapiens app. He ate more whenever his levels dipped.
His challenges began the next morning after he ate a breakfast of two bagels with Nutella, Biscoff spread, and peanut butter. He achieved a personal best of 1:00:31, smashing his own British under-23 record and coming in 18th overall. It’s impossible to know how high his glucose levels were because they maxed out the sensor through the entire race. The app wouldn’t allow Smith to go higher than that, he says. “They’d love to find out.”
An entrepreneur named Brian Davis sent me a LinkedIn message about a new company he and his partners were starting and asked me to meet for coffee. A CGM, he told me after I’d agreed to sign the NDA, would give athletes real-time information about how well-fueled they were, as well as when and what to eat.
I live in Toronto, so Davis was in town to talk with Michael Riddell, a researcher at York University who is one of the world’s leading experts on how people with diabetes respond to exercise. Diabetes results from an inability of the body to control glucose levels due to insulin being inefficient or absent. Insulin is the body’s main tool for channeling glucose out of the bloodstream and into muscle cells or fat cells. During the past decade, CGMs have been developed and improved to make it easier for people with diabetes to maintain safe blood glucose levels. They have been critical for Team Novo Nordisk, a pro cycling team whose members all have Type 1 diabetes, not only for the health and safety of the riders, but also for their performance. Supersapiens, the 2019 cycling team founded by Phil Southerland, was conceived from that insight. It’s not just diabetic athletes who worry about bonking, he reasoned.
It’s not completely unheard of to fit CGMs to healthy individuals. Peter Attia, the podcaster and physician who wrote about blood sugar levels in endurance athletes back in 2017, called his CGM “one of the most informative inputs I’ve ever had.” But Supersapiens had some significant challenges to overcome in order to sell to athletes. CGMs are typically prescribed by doctors around the world because they are regulated. I talked to Davis in 2019 about non-prescription sales they hoped to get approved by mid-2020. As of last fall, Supersapiens was only available in Europe, not the United States. As a result of delays caused by COVID at the Food and Drug Administration, it is not expected to be approved until next year.
Another obstacle-which is even more challenging-is that there is no definitive link between blood sugar and performance. Our lives aren’t like that of an automobile, where the fuel runs until the tank runs out. Muscles instead are fuelled by a complex mixture of fats and carbohydrates (of which glucose is just one type) stored in various places (the bloodstream is also one), blending according to the intensity and duration of the task and the relative level of the fuel tanks within. When stress, fatigue, dehydration, and a slew of other factors impact glucose levels in people with diabetes, they’re even more complicated in non-diabetics. The mere fact that you have low blood sugar does not mean you’re about to collapse. In reverse, just because your blood sugar is high does not mean that you will not crash.
Still, Supersapiens’ pitch is that it’s better to have some information than no information. Union Cycliste Internationale banned glucose monitors from competition in June, signaling the support of this pitch – a ban partially applied to Supersapiens and implicitly based on the assumption that knowing your glucose levels gives you an edge. Mick Rogers, the UCI’s innovation manager, told Cycling Weekly that bike racing fans don’t want Formula One-style racing. It is their wish to be surprised.e surprised. “They seek unpredictability.”
While Supersapiens has plans to sponsor the Ironman World Championships in Hawaii where CGMs are still allowed, it has signed partnerships with cycling teams including Canyon-SRAM and Ineos (whose riders can still use them in training). Athletes like marathoner Eliud Kipchoge have also signed up, and the company is analyzing the data from their uploads. Do non-diabetics have high glucose levels? It’s not something we’re familiar with,” admits Riddell, who is now a scientific advisor to the company. Elite-level training and racing adds some unique variables that were never considered before, he explains: “Sometimes the data is quite high, while other times it is quite low.”. There isn’t anything abnormal about it, but it’s out of the ordinary.”
Trent Stellingwerff, a highly respected sports scientist at the Canadian Sport Institute Pacific who specializes in sports nutrition and metabolism, says extracting actionable advice from this data flow will be the biggest challenge for Supersapiens. A total of 12 full-time scientists have been hired by the company to meet this need, bringing the total to over 70 employees. Researchers are getting tantalizing glimpses of, say, Kipchoge’s bloodstream’s glucose levels during the Hamburg Marathon this spring. However, can Kipchoge learn anything from those data about how to improve his performance next time? Stellingwerff is confident that the unit measures accurately. I would like to know: Why?”
In your bloodstream, you have approximately a teaspoon of sugars, and your body is designed to maintain that level. You will receive insulin after eating three scoops of ice cream, enabling your body to store the excess sugar in fat and muscle cells. You will receive quick fuel from your muscles when you are chased by a lion by stress hormones released by the liver into your bloodstream. You burn glucose 100 times faster during exercise than you do at rest, but a delicate balance keep blood glucose levels within a narrow range between 70 and 140 milligrams per deciliter. Low glucose levels don’t necessarily indicate low fuel levels.
A person’s muscles have a limited ability to store carbohydrates and fat, so both must be consumed to sustain them. Below is an illustration of the fuel mixture for different intensities of exercise, from a recent review in Nature Metabolism. Carbohydrates and fats are stored in the muscles in the form of muscle glycogen and muscle triglycerides. Fats and sugars are transported in the blood as plasma FFA (free fatty acids) and plasma glucose.
As an easy walker, fat provides most of the fuel at low intensity. At the highest intensity, equivalent to a brisk run, you’re burning mostly carbohydrate, but predominantly in the form of muscle glycogen rather than glucose. Looking at a graph like this, you might wonder why anyone would care about glucose levels.
On the trails or while riding your bike, the picture gradually changes as you stay out there longer. Approximately 90 to 120 minutes of intense exercise is all you can store in your muscles as glycogen. Glucose is used more often when those supplies run out. The following graph shows how the fuel mix shifts during prolonged exercise based on studies conducted by University of Texas researcher Edward Coyle and others during the 1970s and 1980s:
Usually after eating something every three hours or so, you’ll start burning 40 percent glucose-that is, if you make sure your glucose levels stay high. You will suffer performance and glucose levels if you simply drink water. Based on Coyle’s 1983 paper, the entire sports drink industry is built on this observation.
The message from Gatorade is straightforward: drink as much sports drink as possible so as not to experience any drop in glucose levels. A more sensible recommendation is given in Supersapiens: eat and drink only enough to meet your needs. It takes time and can lead to stomach upset or worse if we down gels or drinks on the move. Choosing your optimal performance zone is a process of trial and error that can be used to discover your optimal performance zone. Southerland says that below 110 he struggles with longer rides. I feel the best at 140-180 lbs. “However, these levels are very specific to an individual.” Since Smith ran his breakthrough half-marathon almost entirely above 200 mg/dL, the app’s upper threshold, his zone looks more linear. While Kipchoge’s data remains confidential, company president Todd Furneaux is prepared to speak in general terms: “There are 180-200 pound athletes in every Ironman,” he says. Their performance is flat.”
Supersapiens produces an Abbott sensor called the Libre Sense, which is described as a “glucose sport biosensor.” While it most closely resembles the FreeStyle Libre 2 CGM marketed for people with diabetes, there are some notable differences. Unlike the regular version, the sport model sends minute-by-minute updates via Bluetooth to the app (or the upcoming wrist display). This is likely to reassure regulators that it will not be used as a medical device. The measurement range is capped at 200 mg/dL, which is much lower than what you would need to monitor your levels with diabetes safely. Having applied the device to your arm, each unit lasts for 14 days once it has been purchased for 65 euros (roughly $77).
As a result of athletes such as Smith exceeding the upper limit, the real-life data from athletes appears to be different from what the company would have expected. Initially, Furneaux and his team thought we were working on ways not to bonk. The study collected continuous glucose monitoring data during exercise from people without diabetes, and Riddell notes that some went well below 70 mg/dL. He suggests that this range is associated with impaired cognitive and physical function. CGMs could have told these people they needed more fuel, which would have improved performance, although this claim hasn’t been proven.
Whether or not elite athletes are subject to the same observation is ambiguous. Louise Burke, an exercise nutrition researcher at Australian Catholic University working closely with Olympic athletes for 40 years, has observed athletes drop below 50 mg/dL without showing any symptoms, while other athletes display clear symptoms around 75 mg/dL. The level of the athlete may determine this, she adds. Athletes who are really elite sometimes manage to push themselves lower than they would otherwise be able to.es lower than they would otherwise be able to. We don’t know what to do, but mostly we don’t.”
There’s more to it than that, though. Kurt Burke investigated whether blood glucose monitors could detect chronic low energy availability, which was associated with health problems and overtraining, earlier this year with 14 elite Australian racewalkers. While it is difficult to draw meaningful conclusions from fluctuating glucose levels during the day, Burke believes overnight levels may be a better indicator of whether you are getting more calories to fuel your training. Therefore, Burke is interested but not convinced, until the results are analyzed. I don’t mean to imply the information won’t be valuable, but it does need to be validated.
Jake Smith, the British half-marathoner, used the CGM to fine-tune his carbohydrate loading before a major race. To properly prepare your muscles for competing, modern protocol recommends intense carbohydrate intake for a couple of days before starting. An athlete weighing 150 pounds wants eight to 12 grams of carbohydrate a day, which works out to about 16 cups of pasta a day. Although you cannot measure glycogen directly with a CGM, the Supersapiens app provides a trailing 24-hour glucose average. Furneaux believes that number could serve as a proxy for muscle glycogen stores, since if it’s higher than normal, it means the surplus glucose has nowhere else to go.
It can also be tricky to get in a workout or race in the last few hours before it. A phenomenon called rebound hypoglycemia causes temporary feelings of lightheadedness and weakness after a few minutes of exercise in 30 percent of endurance athletes. Simple carbohydrates trigger a rise in insulin levels that lasts for at least an hour after eating them 30 to 60 minutes before exercise. You have two levers working together, insulin and exercise, to lower your blood glucose levels simultaneously, causing them to drop too rapidly as you exercise. Supersapiens data shows this a lot, says Riddell. “People don’t fuel themselves properly.” One way to combat this is to eat quickly before exercising so your insulin levels do not rise. But a CGM helps you also discover precisely how your glucose levels respond to specific foods and pre-workout timings.
Can you imagine all the interesting uses this could have? Verify. But what about actual evidence that sticking one of these things on your arm will make you faster? The Libre Sense’s website states it will “inform athletes how to fuel appropriately, to refill their glycogen stores before a race, and to know when to replenish during a race in order to maintain athletic performance.” Follow the relevant footnotes, and you will find two reviews from 2015 about refueling after exercise, and a Swedish undergraduate thesis from 2016 about four national-class swimmers who wore a CGM for a week without intervention or performance monitoring.
Literature on the topic lags sometimes behind elite practices, of course. A Dutch sports nutritionist who revitalized Kenenisa Bekele’s marathon career, Armand Bettonviel, responded to my email to share his views. The first thing Bettonviel stressed when talking about Supersapiens was that interpreting data from the CGM was still “not yet hard science.” He is using it to develop a more comprehensive picture of how Kipchoge’s body produces and uses glucose under different challenges. The general insights provide him with the foundation for drilling down into the specifics of Kipchoge’s in-race drinking protocol, which he carefully optimized while trying to run the marathon in under two hours.
The caveats are, however, there. In order to get Kipchoge to his “optimum blood glucose range,” Bettonviel is looking for the best pre- and in-race nutrition protocol that keeps him there. Bettonviel says metabolic flexibility could also be a key performance indicator for endurance athletes. In addition, he has found that Kipchoge’s glucose responses don’t necessarily translate to other athletes, making it difficult to formulate generalized rules. Our team is still analyzing and learning, he says. We haven’t made any big changes yet and will make small ones as soon as possible.
There are very few studies on athletes wearing CGMs in the published literature, and nearly all of them focus on health rather than performance. In a recent study published in Cell Metabolism, 15 athletes recruited from the national-team endurance program were given CGMs for a period of up to two weeks. Their average blood glucose level was lower at night, mostly during the middle of the night, than non-athlete controls, but higher during the afternoon. They also spent more time above the upper glucose threshold, usually about 140 mg/dL. They generally remained within normal limits during their training sessions, however.
There has been a resurgence of interest in CGMs for people who are healthy and do not have diabetes. There are some controversies surrounding it, however. Tom Hughes, a medical doctor and sports science lecturer at Leeds Beckett University in Britain, was cautious when Supersapiens announced its title sponsorship of the Ironman World Championships this spring. During Ironman races, he says, he hasn’t observed significant drops in blood glucose, as he has taken old-school finger-prick readings of blood glucose when feeling bonked and observing levels well over 100 mg/dL. As for obsessively tracking your glucose readings throughout the day, he doesn’t believe you’ll find anything healthful about it. In reality, he says, it’s just another opportunity to “stress about another number.”
I was surprised to learn that Riddell, the diabetes researcher and adviser to Supersapiens, shares my view. “It’s worth writing about the obsession with numbers,” he says. Even among people with diabetes, the patient is often the one who dislikes the continuous glucose monitor.” After all, there’s a stream of continuous data that appears to be judging you, often negatively, at each meal and snack. And when you try to “fix” your behavior, your glucose levels don’t always respond in the way you expect. At least 40, and possibly 200, factors contribute to glucose levels, making it difficult to decide which signals matter most. If Supersapiens is to succeed as an athletic aid, it has to hone its “so what?” skills.”
In addition to their 12 scientists pouring over the data samples from the company’s athlete ambassadors, they look for patterns, trends, and telltale signs-and perhaps even new discoveries. Data are already changing our understanding of what glucose looks like in serious athletes. In the Scandinavian study, glucose values remained in the normal range even during heavy training. During competitions, Jake Smith and others produce values that no one expected. Riddell explains that “in non-diabetics, glucose homeostasis is not disturbed by exercise.” It’s wrong to say that! This is true! That’s 50 years out the window. Our textbooks will be rewritten.”