You’ve seen the headline: elite marathoners run 2.6% more efficiently when they eat 120 grams of carbs per hour. It’s real data, but it’s narrower than it sounds. It compared 120 g/h to 60 g/h in just 8 elite men, and nobody measured a finish time.
So should you copy it? For most marathoners, not yet. Here’s what the study showed, what it didn’t, and a safer way to move toward higher fueling.
What the 2.6% study measured
Researchers at Liverpool John Moores University ran 8 elite men through three 120-minute treadmill runs (Ravikanti et al., J Appl Physiol 2025). Their marathon bests averaged 2:22:54. Each man took 60, 90, or 120 grams of carbs per hour (g/h) in a labelled drink, so the team could track what burned.
Running cost about 2.6% less oxygen at 120 g/h than at 60 g/h (P = 0.021). Oxygen cost is how much air your body needs to hold a given pace. In short: at the same speed, the 120 g/h runners worked slightly less hard on the inside.
Notice the limits. The gap was against 60 g/h only. It wasn’t shown against 90 g/h. The 60 g/h drink also used a different carb type than the 120 g/h drink, so dose and type are tangled together. No one raced anything.
Critics have already pushed back. A 2026 paper in the same journal questions the interpretation (DOI 10.1152/japplphysiol.00303.2026). Eight runners is a small pool.
One small lab test isn’t a rule.
Why more carbs burned doesn’t mean a faster race
The same study tracked exogenous carbohydrate oxidation, meaning carbs from your drink that your muscles burn. It rose with every step up in intake.
At 1.68 g/min, those runners burned about 101 g of drink carbs per hour. Think of your gut as a highway with toll booths. Glucose uses one set of booths, and fructose uses another. Mix the two and more cars get through, but only until every booth is busy.
Faster absorption sounds like a win. Yet a 2022 cycling study (Podlogar et al.) found 120 g/h burned more drink carbs than 90 g/h without sparing any of the carbs stored in the muscle. In short: more fuel in doesn’t always mean more saved up.
The 120g carbs per hour marathon problem: your gut
Here’s what the headlines skip. Every runner in the Liverpool study reported at least one moderate or severe gut symptom in every trial. That includes 60 g/h. Nausea, fullness, and cramps peaked at 120 g/h.
These were elite men who had loaded carbs for days. The authors say gut training is needed before real-world use.
Survey data shows how fast the risk climbs with duration. In a study of 221 athletes (Pfeiffer et al., 2012), serious gut trouble hit about 31% of Ironman athletes, 14% of 70.3 athletes, and 4% of marathoners.
That last caption matters. Marathoners in that survey ate less. Low symptoms reflect low intake, not an iron gut.
Heart rate won’t warn you when your stomach is the limit. Your belly will.
What happened in the field
Race-style tests of 120 g/h in runners are rare. In a 42 km mountain marathon, 20 elite male trail runners finished after eating 60, 90, or 120 g/h (Viribay et al., 2020). Six more dropped out, three of them for gut problems.
The 120 g/h group showed less muscle damage a day later. Their creatine kinase (a muscle damage marker) rose 156%, versus roughly 970% in the other groups. Race time, effort rating, and heart rate didn’t differ. Translation: they recovered better, but they didn’t finish faster.
For a deeper look at that result and the gut-adaptation plan behind it, see gut training for 120 g/h.
Sub-2 and the elite template
Maurten reports that Sabastian Sawe ran 1:59:30 at London in April 2026 on roughly 115 g/h. That’s company-sourced, not peer-reviewed. Sawe also trained his gut under supervision with lab tests.
You aren’t Sawe. His gut, pace, and support crew are one of a kind.
How many carbs for 3:00, 3:30, and 4:30 finishers
No study has tested amateur finishers at these doses. What follows is reasoning, not study data.
The drink-carb ceiling comes from your gut’s transporters, not your pace. That ceiling is about 90 to 105 g/h with a glucose-fructose mix, and it’s likely similar for slower runners. Slower runners are on course longer, so they need more total carbs. Yet they burn glycogen at a lower rate. The extra grams may help less per gram.
| Finish time | Longer-run starting range | Stretch target after gut training | Total at 60 / 90 / 120 g/h |
|---|---|---|---|
| 3:00 | 60 to 75 g/h | 90 g/h | 180 / 270 / 360 g |
| 3:30 | 60 to 90 g/h | 90 to 100 g/h | 210 / 315 / 420 g |
| 4:30 | 60 to 90 g/h | 90 g/h | 270 / 405 / 540 g |
These tiers follow the classic advice of 30 to 60 g/h for shorter efforts and up to 90 g/h past 2.5 hours (Burke et al., 2011). Look at the right column. A 4:30 runner at 120 g/h must carry or grab 540 g, roughly 22 gels. That’s a lot of sticky packets.
For a full race-day schedule, see how many grams of carbs per hour for your marathon.
A careful 90 to 120 g/h test
Treat 120 g/h as an experiment, not a goal.
- Pick a long run at marathon effort, 2 hours or more.
- Start at your current dose. Add 10 g/h each week.
- Take small amounts every 15 to 20 minutes, not big gulps.
- Stop climbing if you get cramps or nausea. Hold or drop back 10 g/h.
- Keep what works. Never try a new product on race morning.
Picture Dana, 41 (an illustrative example, not a real athlete), who ran her first marathon in 3:52 on 60 g/h and felt flat after mile 20. Her coach added 10 g/h each week on Sunday long runs. At 90 g/h she felt strong. At 110 g/h she got stomach cramps, so she backed off. She raced at 90 g/h and finished in 3:41.
Her best number wasn’t the biggest one. It was the highest she could keep down. Train easy runs in Zone 2 and save the heavy fueling for the long, hard days.
Your best dose is the one your gut keeps.
Where AthleteOS fits
AthleteOS builds your training plan around your goal race. It also lets you log fuel for each workout. Log what you eat on every long run, and you’ll see which dose you tolerated before race day.