REF-734 · Maunder, Podlogar & Wallis, 2018 · Medicine & Science in Sports & Exercise, · Paper · PubMed
Adding fructose to maltodextrin during a 4-hour recovery gave a 33% longer second run to exhaustion the same day
The point at which faster liver glycogen repletion stops being a mechanism and becomes a performance result
Protocol
Eight trained endurance runners and triathletes completed two identical trials in a single-blind, randomised, counterbalanced crossover. Each trial: treadmill run to exhaustion at 70% V̇O₂max (bout 1) → 4-hour recovery → the same run to exhaustion again (bout 2). The only difference between trials was the carbohydrate drunk during recovery, at 90 g/h in both:
GLU + MAL — glucose + maltodextrin (maltodextrin is a glucose polymer, so this arm is glucose-only in substrate terms)
FRU + MAL — fructose + maltodextrin at a 1:1.5 ratio, i.e. fructose is ~40% of total carbohydrate
Key Findings
| Outcome in bout 2 | FRU + MAL | GLU + MAL | |
|---|---|---|---|
| Time to exhaustion | 81.4 ± 22.3 min | 61.4 ± 9.6 min | P = 0.02 · ES 1.84 ± 1.12 · +32.4% ± 19.9% |
| Total carbohydrate oxidised | 223 ± 66 g | 157 ± 26 g | P = 0.02 |
| Oxidation rate of the carbohydrate ingested during recovery | greater | — | P = 0.001 |
| Total carbohydrate oxidation rate | no difference | no difference | ns in bout 1 and bout 2 |
| Plasma glucose, plasma NEFA | no difference | no difference | ns |
| Plasma lactate during recovery | higher | — | P = 0.001 |
| Self-reported nausea and stomach fullness | marginally better | — | not a formal result |
The primary result is the one to carry: a third more endurance capacity in a second bout four hours later, from changing the *type* of carbohydrate and nothing else. Total intake was identical.
The mechanism was inferred, not measured. Liver and muscle glycogen were not sampled in this study. The prior work it rests on — sucrose or glucose+fructose accelerating hepatic glycogen repletion without costing muscle glycogen resynthesis — is a separate literature (see REF-064).
What the ingested-carbohydrate oxidation data show: more of what was drunk *during recovery* was still available as fuel in the next bout. That is a carbohydrate-availability story, and it is the cleanest supporting result in the paper.
Higher recovery lactate is expected, not a warning. Fructose is metabolised in the liver, and lactate is a normal intermediate of that route.
Practical Application
When the next hard session is under ~6 h away, make ~40% of recovery carbohydrate fructose-containing; otherwise ignore the type and just hit the total. The trigger is the gap, not the sport — same-day double, back-to-back long weekend rides, a stage race, a camp with consecutive demanding days.
Total carbohydrate first, sources second. This trial gave 90 g/h in BOTH arms. It is evidence about composition at an already-adequate intake — it says nothing about whether a lower total plus fructose beats a higher total without it.
Aim for roughly 40% of recovery carbohydrate from fructose-containing sources, which is what the 1:1.5 fructose:maltodextrin ratio delivers. This sits close to the 1:0.8 glucose:fructose ratio the during-exercise literature converged on (REF-333) — the same ballpark, arrived at for a different reason.
Real food counts. Fructose-containing sources include fruit and juice, honey, agave, sucrose (which is half fructose) and regular soft drinks. Four hours of recovery shakes at 90 g/h is a laboratory dose, not a prescription.
90 g/h is the study's dose, not a target. Scale it to the session completed, body mass, and how soon the next session is.
Limitations
n=8, single-blind, one laboratory bout. Time to exhaustion at a fixed intensity is a sensitive but noisy measure, and the effect size confidence interval (1.84 ± 1.12) is wide enough to include a much smaller true effect. The direction is well supported; the magnitude is not a number to quote as "expect 33%".
The greater total carbohydrate oxidised in bout 2 (223 vs 157 g) is partly an artefact of duration — the FRU + MAL bout lasted 20 minutes longer, so more of everything was oxidised. Total oxidation *rates* did not differ. Do not cite the gram totals as an independent finding.
Trained runners and triathletes, and all data are from treadmill running. There is no cycling arm, no elite cohort, and the abstract does not report the sex composition of the group — so nothing here is scoped to one sex either way.
Glycogen was never measured, so the mechanism stated by the authors is an inference from earlier work, not a result of this study.
A four-hour recovery is the tested window. Nothing here speaks to overnight recovery, where the time available makes the composition question much less pressing.
Single-blind: participants could plausibly detect a difference in drink taste, and time to exhaustion is effort-dependent and open to expectancy.