The more you eat, the faster you go (in ultraendurance)


As of September 2017, new Sweat Science columns are being published at Check out my bestselling new book on the science of endurance, ENDURE: Mind, Body, and the Curiously Elastic Limits of Human Performance, published in February 2018 with a foreword by Malcolm Gladwell.

- Alex Hutchinson (@sweatscience)


A few months ago, I blogged about a study that observed correlation between in-race carb intake and race time in Ironman triathletes. What was significant about that paper is that it looked at a topic that has been studied to death in the lab, and took it out into the real world. There are a lot of “problems” with the real world that make it hard to nail down causes and effects — but ultimately, the whole point of this type of research is to understand what’s happening in the real world. So these observational studies, despite their challenges, are very important.

That’s by way of intro for another small study, just published in the International Journal of Sport Nutrition and Exercise Metabolism, from researchers in New Zealand. They looked at the nutritional intake of participants in a brutal cycling race, the K4, which covers 384K and includes 4,600 metres of climbing. The average finishing time of the 18 study participants was 16 hours and 21 minutes! The key points:

  • The estimated calorie burn for the race was about 6,000 calories; the average intake was just 4,500 calories, so there was a big caloric deficit.
  • There was a significant inverse relationship (p=0.023) between number of calories consumed and finishing time. The more calories you managed to cram down your gullet, the faster you finished!

Is this a surprise? Given that the race was so long, it makes sense that taking in enough energy was a significant challenge. Obviously the same thing doesn’t apply during, say, a 100-metre sprint. The question is: where’s the breakpoint, beyond which energy intake becomes a significant independent predictor of performance? I think the general assumption is that it’s probably a bit below marathon distance — so it would be really interesting to see a study like this, with a very large number of participants, at a marathon.

15 Replies to “The more you eat, the faster you go (in ultraendurance)”

  1. Totally agree and see it with ultras…especially 100 milers. If you can manage to get down 300 cal per hour you’re usually doing pretty well. Breaking it down to a gel (or chew/blok) every 20mins or so, with top ups at aid stations (sometimes real food) seems to work well. Of course heat, hydration and electrolytes play a large roll in how well you can keep everything down too (I’m a puker). Cooler weather makes it that much easier to consume more calories per hour.

  2. Again like the ironman study – there’s no indication either way as to whether higher caloric intake results in faster performances or whether faster performances result in higher caloric intake (due to rate-limiting fat metabolism factors).

  3. @Mike W., it is indeed a bit thin; most of it would be accounted for by the climbing alone.

    Suppose that a representative rider weighs 70kg and his bike, clothes, food, water & sundries weight 10kg. Taking g to be 9.81 N/kg, it costs 4600 x 80 x 9.81 / 1000 = 3610 kJ to lift the rider and his stuff 4600m. That is about 3610/4.2 = 860 kCal. We may approximate the bicycle as 100% efficient, but the rider is nowhere near that; human cycling efficiencies vary over a range something like 18%-24%. It is normal to take 21% as the point estimate when the true value is unknown. So the energy cost of the climb to the cyclists, ignoring air resistance, is about 4090 kCal.

    That leaves less than 2000 calories to pedal the bike over 384km. It is true that the headline average speed, 384/16.5 = 23.3 km/h, is rather slow (slower is more efficient, within reason), but presumably the total time is elapsed and includes stops. Also, the climbs lower the average speed, so that the representative speed pedaling on the flat would be higher. All in all, it seems like a very tight budget.

  4. Re. calorie calculation: I admit was I surprised by the low number too. Here’s the passage from the paper:

    “The mean (SD) total energy intake of the cyclists was 18.7 (8.6) MJ, with the distribution from carbohydrate, protein, and fat 75.6% (12.0%), 14.5% (9.2%), and 6.1% (6.3%), respectively. The mean energy requirement for the race was estimated at 25.5 (7.4) MJ.”

    Did I do the conversion to calories correctly?

  5. @Tom: Yes, observational studies can indicate only correlation, not causation. But you’ll note that no one is actually replacing their energy loss. My understanding is that, in ultraendurance events like this, energy intake is almost universally limited by the ability of the athlete to stomach more calories, rather than being driven by hunger.

  6. Hi,
    I know the K2/K4 ride, but not yet done it. I do how ever do the Lake Taupo Enduro each year, 320km with around 3200m of climbing. My time is around 12 hours, so not fast. I ride with a Garmin 705 which shows calories burnt and it indicates around in the first 9 hours I used 9500cal,sadly my battery died before the finish. But based on this it indicates around 1000cal per hour. I’ve also been a study subject at Massey University Sports science department, on their erg bike and the figures seem to be consistent with theirs. I’m guessing the Author of the study are the same department?? I currently use Hammer Perp, at the rate of 2 scoops per hour, supported by gels, dried fruit and nuts to try to eat at a rate of around 600cal per hour, or more if I can. Im a 92kg rider , my av speed is around 25kmh.

    After reading the article Im thinking that either my Garmin is wrong or the information from the study has been mis-interpreted.??

  7. My 24hr time trial (I covered ~720k) energy use according to my Powertap was 16102 kJ (670kJ per hour).

    From my team’s food log I consumed approx 10000 kJ (416kJ per hour).

  8. Is there a similar study that takes into account ketosis? I’m running my first ultra in July and hope to not need as much gel since I’ll be a fat burning machine.

Comments are closed.