A study in the Journal of Applied Physiology has been doing the rounds in running circles, and the headline everyone has taken from it is that 120 grams of carbohydrate per hour beats 90, which beats 60. That is accurate. It is also the least useful sentence in the paper.
The design: eight elite male marathoners, personal bests averaging 2:22:54, each completing three 120-minute treadmill trials with 13C-labelled glucose-fructose drinks at 60 g/h (maltodextrin to fructose 1:0), 90 g/h (2:1) and 120 g/h (1:1). Every trial was preceded by 24 hours of high-carbohydrate eating at 8 g/kg and a pre-run meal at 2 g/kg. The runs were not easy: 15 minutes at 95% of lactate threshold, then 90 minutes at 94% of lactate turnpoint, then 15 minutes back down.
The results separated cleanly. Whole-body carbohydrate oxidation and second-hour exogenous oxidation differed between all three trials in the same order — 120 > 90 > 60 — with exogenous oxidation at the top dose reaching 1.68 ± 0.16 g per minute. Running economy improved too: at 120 g/h, the oxygen cost of running was 8.1 mL O₂/kg/km lower than at 60 g/h. Same pace, meaningfully less oxygen. That is a real result, and it is the one being quoted.
The finding underneath the finding
Here is the line that should change what you actually do. The incidence of moderate or severe gastrointestinal symptoms was high in all three trials. Peak nausea, stomach fullness and abdominal cramps were worst at 120 g/h — no surprise — but the low dose did not spare anyone either. Sixty grams an hour, the figure most recreational runners treat as the safe conservative option, still produced moderate-to-severe symptoms in this group.
And consider who this group was. Elite men averaging a 2:22 marathon, who had spent 24 hours deliberately carb-loading, running for two hours in a controlled lab. If that population reports common gut distress across the board, the variable being tested is not really the number on the gel packet. It is the gut’s trained capacity to process anything at all while blood is being diverted to working muscle.
Our take: The paper does not say “take 120 grams an hour.” It says the ceiling on carbohydrate delivery is metabolically higher than we assumed, and that almost nobody’s gut is currently equipped to reach it. The authors say so plainly: gut training is warranted before applying intake rates this high in practice. Treat 120 g/h as a destination you train toward over months, not a number you adopt on race morning because a study said it was faster.
What this does not tell you
Eight participants. All male. All elite. Two hours, not four. The trials measured metabolism and economy, not race performance — nobody in this study raced anything. Running economy improving by 8.1 mL O₂/kg/km is a strong physiological signal, but the leap from “cheaper oxygen cost in the lab at hour two” to “faster marathon at hour four” is an inference the data does not make for you.
The practical asymmetry also matters. The upside of getting fuelling right is a few percent. The downside of getting it wrong is vomiting at 30km and walking home. Those are not symmetrical outcomes, and a protocol you have never rehearsed at race intensity is the single most common way runners find the downside.
What to do with it
- Find your current ceiling before chasing a higher one. Log intake and symptoms across your next four long runs at the intensity you will actually race. Most people have never measured this honestly.
- Raise the dose in training, not in racing. Add roughly 10 g/h every couple of weeks in long sessions and hold there until symptoms settle.
- Mind the ratio, not just the total. The 120 g/h trial used a 1:1 maltodextrin-to-fructose blend, not the 2:1 used at 90. Fructose uses a separate intestinal transporter, which is what makes the higher rate absorbable at all.
- Match the dose to the duration. A two-hour effort at threshold is the context tested here. For a 90-minute race, the delivery ceiling is not your limiting factor.
None of this is medical or nutritional advice for any individual, and anyone with a history of gut issues should be talking to a professional rather than a study abstract.