Multi-carb fueling for runners: How to sustain high carbohydrate oxidation during marathons and ultras

By now, it’s more or less common knowledge that you should “fuel” during marathons and ultramarathons by taking carbohydrates. For marathoners, this means sports drinks, energy gels, or energy blocks; for ultramarathoners, “real” food like granola, pretzels, potato chips, and peanut butter sandwiches are popular as well—your stomach can only handle sugary syrup for so long.

Knowing that you should fuel leads, naturally, to an obvious follow-up question: how much should you be fueling? In other words, what should your fueling rate be? I’ve been digging into the research on this question lately as it applies to carbohydrate intake.

For a long time, the default recommendation from sports nutrition experts has been 60 grams of carbohydrates per hour (60 g/hr) for events lasting two to four hours [1].

This is still a reasonable starting point for first-time marathoners, but I want to cover some of the latest developments on fueling, and especially the concept of multiple carbohydrate fueling—consuming sources of both glucose and fructose to sustain a higher rate of fuel oxidation.

Multiple carbohydrate fueling is the key to understanding why very long races like ultramarathons (as well as long cycling races, like Tour de France stages) have seen major breakthroughs from the more systematic application of scientifically-based fueling strategies.

Why fueling improves running performance (in events lasting 90 minutes or more)

Without getting too bogged down in details, it’s worth reviewing why fueling actually helps in long races. There are two reasons why

Fueling with carbohydrates helps improve performance for two reasons: first, and more obviously, taking “exogenous” carbohydrates allows your internal carbohydrate stores to last longer: liver and muscle glycogen last longer, so you can run further and faster before your overall carbohydrate supplies run low.

Fueling with carbs also shifts your body’s fuel preferences towards burning more carbohydrates and less fat for a given running intensity [2].^[1] This might seem paradoxical, or counterproductive—wouldn't burning more carbs offset some of the extra carbohydrate availability?—but this shift in substrate utilization has an independent benefit on performance, because you get about 7% more energy per unit of oxygen if you burn carbohydrates versus if you use the same amount of oxygen to burn fats [3].^[2]

So, fueling with carbs not only helps you last longer at the same speed, but it also boosts the amount of energy your body can put to use at a given rate of oxygen consumption.

Why carbs only matter for events lasting ~90 minutes or longer

The break-even point where it starts making sense to fuel at all is in all-out events lasting about 90 minutes or longer [4]. From the two main benefits of fueling above, we can infer why.

First, in a shorter event, you won’t be running far enough to experience severe glycogen depletion, so whole-muscle glycogen depletion is less of a limiting factor for performance.

Second—and again, more subtly—in shorter events, your muscles are already burning close to 100% carbohydrates, so the usual shift to burning more carbs and less fat no longer occurs.

The plot below shows this effect in action: as running intensity approaches 90% VO2max (which, for a well-trained running, might be anywhere from 10k to half marathon pace), fat oxidation rapidly approaches zero.^[3]

Now, 90 minutes is a rough guideline—for most runners, that corresponds to about a half marathon, and there’s only one study that I’ve seen on fueling for the half (which found that it made no difference for performance) [5].

My advice for the athletes I coach is not not fuel for the half marathon, unless we’re doing the half as part of a marathon buildup and they need practice fueling while running at fast speeds. In contrast, aggressive fueling was a big part of Eliud Kipchoge’s 1:59 marathon world best, so it’s hard to believe that fueling doesn’t matter at all for a 90 minute event, but is critical for a 119 minute event. I’d like to see more research on fueling—especially aggressive fueling rates—for half marathon-type distances, but until then, the 90 minute cutoff is a decent guideline.

Single versus multiple carbohydrates: glucose versus fructose

There are many different kinds of carbohydrates, but when it comes to the kind you’d take as fuel during a race or workout, we really only need to concern ourselves with two: glucose and fructose.

Glucose inside the body

Glucose (dextrose on some nutrition labels) is a simple sugar that’s used for energy in every cell in your body. It’s found in copious amounts in sweet foods like honey, fruits, and is the building block for most complex carbohydrates.

Once glucose makes its way to your small intestine, it’s absorbed through a special transporter protein called SGLT1, which shuttles glucose across the intestinal wall and into the blood vessels that run from the intestines to the liver. The liver lets most glucose pass on through into general circulation, where it shows up as blood sugar and is directly available for muscles to oxidize.

Fructose inside the body

Fructose is another simple sugar that, as its name suggests, is often found in fruits (almost always alongside glucose). In terms of its metabolism during exercise, fructose differs from glucose in two ways: first, it is absorbed through a different transporter protein in the small intestine (GLUT5), and second, when fructose passes through the liver, most of it (~70–90%) is taken up by the liver [6].

Inside the liver, around 30% of this fructose ends up getting converted back to glucose and sent out into the bloodstream right way; 40% gets turned into liver glycogen and stored locally (and perhaps later turned back into glucose and released into the bloodstream) and 30% gets processed into lactate and released into the bloodstream [7]. So, while fructose is processed a little more slowly, it does end up becoming available to your muscle, just through a more circuitous route.

A quick note on sucrose, maltodextrin, and other rapidly-digestible starches

Sucrose—table sugar—is just a glucose molecule bonded with a fructose molecule. As soon as sucrose gets to the small intestine, an enzyme breaks this bond very quickly, so fueling with sucrose has the same effect as fueling with a 1:1 mix of glucose and fructose. High-fructose corn syrup is a glucose-fructose mix, consisting of 42–55% fructose and the remainder as glucose—basically just “pre-split” sucrose. 

A similarly quick splitting of molecular bonds occurs in the case of rapidly-digestible starches like maltodextrin, potato starch, and cornstarch, which are made up of long chains of glucose molecules. Once these hit the small intestine, an enzyme breaks down these binds, making fueling with these carbs essentially equivalent to fueling with pure glucose.

The situation is different for slowly-digestible starches, like those found in whole grains and products specifically engineered to be resistant to digestion (e.g. UCAN gels), but that’s a topic for another day.

So, it should be clear that if we restrict ourselves to the fuels most runners would take in a marathon or ultramarathon, there are really only two effective types of fuel: glucose and glycogen.

How multiple carbohydrate fueling supports high fuel oxidation rates

The idea behind multiple carbohydrate fueling is to leverage one of the key differences between glucose and fructose: the fact that they’re absorbed through different intestinal transporter proteins. By leveraging both transporters in parallel, you can absorb a given amount of carbohydrates more quickly, leading to a greater influx of carbs to your muscles, and—hopefully—a higher fuel oxidation rate: the amount of carbs you take that end up getting burned with oxygen for energy.

This idea isn’t just theoretical: it has robust experimental support. The plot below shows data from 69 different studies on glucose-only versus glucose + fructose fueling, compiled by Asker Jeukendrup.

These studies directly measured fuel oxidation rate by giving athletes specially-tagged carbohydrates—radioactive Gatorade, more or less—at different rates, then measuring how much of the carbon from these carbohydrates wound up as CO2 in the air exhaled by the athletes.

Notice how fueling with a glucose-fructose mix leads to clearly higher carbohydrate oxidation (“fuel oxidation rate”), even for the same rate of carbohydrate consumption (“fueling rate”).^[4] The dashed gray line denotes the highest possible oxidation rate for a given fueling rate—i.e. burning 100% of the carbs ingested.

A few things are clear from this plot:

1. If you want to achieve high fuel oxidation rates, you must fuel with glucose and fructose.
2. There are theoretical benefits to multiple carbohydrate fueling at rates of up to 120–150 g/hr, even though multi-carb fueling has diminishing returns as well.
3. There are rapidly diminishing returns beyond about 60 g/hr if you only fuel with glucose. No matter how high your fueling rate, it’s essentially impossible to crack a fuel oxidation rate higher than 60 g/hr with glucose alone.

Multi-carb fueling enables a higher fuel oxidation rate for the same fueling rate

It’s really important to drive home the fact that a glucose + fructose mixture yields more carbohydrate oxidation at the same fueling rate.

So, if you are taking one standard gel (25 g of carbs) every 20 minutes (a fueling rate of 75 g/hr), significantly more of those carbohydrates actually get burned if that gel is a glucose-fructose gel as opposed to a glucose-only gel (or maltodextrin-only gel).

Where do the rest go? They’re probably sitting in your small intestine, waiting to be absorbed.

The maximum possible fueling rate for runners

Although fueling rates of 120–150 g/hr look promising on paper, these extraordinary fueling rates are probably only achievable by cyclists, and maybe some ultramarathoners.^[5] Feel free to prove me wrong on this one, though—let me know in the comments if you can achieve fueling rates of ≥120 g/hr during running.

In practice, most runners end up limited by what their stomach can handle, not by any theoretical maximum fueling benefit. Even Eliud Kipchoge “only” fueled at ~100 g/hr—and he was able to take fuel and fluids as needed from support staff on a bike, which makes aggressive fueling a lot easier.

I personally have found that most of my marathon athletes top out around 70 g/hr under the more realistic constraints of sub-elite marathoning, like carrying your own gels and only having fluids every few miles.

Trent Stellingwerf, a top physiology researcher who works with some of Team Canada’s top athletes, has a short write-up on the fueling strategy for three Canadians in the 2:11–2:16 range, who ended up choosing fueling rates from ~50–80 g/hr after lengthy experimentation in training [8].

More realistic fueling rates: when does it make sense to use multi-carb fueling?

If we zoom in on the plot from above to look at fueling rates that are more realistic for road runners, we can see that the real benefits of multi-carb fueling don’t emerge until you are above about 45 g/hr.

Beyond that fueling rate, multi-carb fueling quickly becomes dramatically more effective.

The optimal glucose to fructose ratio for multiple carbohydrate fueling

Research is a bit more spotty on the exact ratio of glucose to fructose that’s optimal, but a few studies suggest that something in the range of 0.8:1 to 1:1 is about right [9,10]. Getting it perfectly dialed in is less important than fueling with some glucose and some fructose.

Why is a 1:1 ratio, or thereabouts, optimal? A rigorous scientific explanation is lacking, but something more akin to a “just-so story” comes from a paper by a tea of researchers in Switzerland and Singapore [11]—it just so happens that a glucose:fructose of 1:1 is about what you find naturally in the kinds of fruits and sweet vegetables often sought out by hunter-gatherers!

Plot from Rossi et al. 2017.

The authors point out that the ratio is closest to 1:1 in the kinds of fruits and veggies you find fresh and above ground: so, not the things you have to dig up, mill, or cook.

This theory is absolutely a bit on the wacky side, but it’s just too much fun not to mention: a handful of wild blueberries makes for the perfect mid-run snack when you’re closing in on a kudu after a multi-hour persistence hunt across the savanna.

(In reality, actual fruits tend to have a lot of fiber, so for marathons I recommend sticking with "fake" food like gels and sports drinks, though fresh fruits and vegetables are great for refueling after your workouts)

Conclusion

If you’re trying out a fueling rate above about 45 grams per hour, you need to be using both glucose and fructose, ideally at around a 1:1 ratio, for optimal benefits. The exact ratio is not super critical, though.

Strong experimental evidence shows that this multiple carbohydrate fueling strategy leads to a greater fuel oxidation rate, even for the same fueling rate—meaning your body gets more out of the same amount of carbohydrates.

In practice, you’ll probably be limited by what your stomach can handle, but if you run marathons or ultramarathons and still use a fairly conservative fueling strategy, it’s worth experimenting with a higher fueling rate, based on multiple carbohydrates, to see if your performance improves.

Learn more about fueling for long races

If you enjoyed this article, subscribe to my free email list! It’s the best way to find out when I’ve got a new article on the science of training and racing, or a new calculator or training tool.

I also have a book, Modern Training and Physiology for Middle and Long-Distance Runners, that focuses on the science of performance and training for events from 800m to the 10k. Check it out!

Footnotes

---

^[1] The mechanistic explanation for the shift in fuel preferences is likely the increase in blood glucose levels when you take carbs during a workout: higher blood glucose acts as a signal to your muscles to shift towards oxidizing more carbs and less fat. It’s useful to think of this and other carb-related adaptations as a distributed system responding to local conditions, as opposed to something that’s being centrally controlled by the brain.

^[2] This point about carbs being a more efficient source of energy may seem confusing, because of the well-known (and correct) fact that fats contain more caloric energy per unit mass—a gram of fat contains 9 kcal of energy, while a gram of carbs contains only 4. However, if you’re limited by oxygen supply (as is the case in running), carbs come out ahead: if you use one liter of oxygen to burn carbohydrates, you regenerate about 7% more ATP than if you’d used that same liter of oxygen to burn fats.

^[3] The drop in fat oxidation at higher percentages of VO2max is mediated in part by the presence of lactate molecules inside muscles fibers: here, lactate acts essentially like a hormone, signaling to muscle fibers that they should dial back on beta oxidation and shuttle more lactate and pyruvate into the mitochondria for aerobic respiration. This is another example of the “distributed intelligence” of the body during exercise.

^[4] The fitted lines come from a pretty sophisticated statistical model: the data are fit with a shape-constrained additive model (using R package ‘scam’) fitted to the data from Jeukendrup’s 2010 review on multiple carbohydrate fueling, using a factor-smooth interaction to model glucose-only fueling versus glucose + fructose fueling. This model is constrained to start at zero (since exogenous carbohydrate oxidation must naturally be zero when you aren’t fueling at all) and is constrained to increase monotonically (a reasonable assumption), but within these constraints, the model can accommodate arbitrary curves. The shaded regions are (approximately) valid 95% confidence intervals—they don’t account for the fact that the N=69 different studies have different sample sizes, and therefore should be weighted according to their uncertainty.

^[5] Ultra running is more amenable to aggressive fueling in part because the pace is slower, and in part because some segments of the race—like steep ascents and mandatory medical checkups—are good opportunities for ramping up fuel intake. You also stand more to gain from a performance benefit because of the length of the race.