REF-957 · Mendonça et al., 2026 · Research in Sports Medicine, pp. 1-18 · Paper · PubMed

Carbohydrate mouth rinse and performance in cyclists: effects of different types of carbohydrates and the placebo effect induced by mouth rinse

Blimited evidenceMeasured on trained athletesWritten before October 2026, being re-checked

Does rinsing the mouth with a carbohydrate (CHO) solution — maltodextrin or fructose — help male amateur cyclists in a time trial, and does the rinse itself work as a placebo? Twelve male amateur riders each did four conditions in a double-blind, placebo-controlled, randomized crossover: a maltodextrin rinse, a fructose rinse, a placebo rinse and no rinse at all (control).

Each session was 15 minutes at 65% of Wmax and 90 rpm, then a 15-minute time trial. Time-trial power output showed no difference between conditions (p = 0.989).

Participants

Twelve male amateur cyclists (n = 12). Age and fitness values are not given in the abstract.

Protocol

Randomized, double-blind, placebo-controlled crossover; four conditions: maltodextrin rinse, fructose rinse, placebo rinse, no-rinse control.
15 minutes of riding at 65% of Wmax, cadence 90 rpm, then a 15-minute time trial with the instruction to do as much work as they could.
Performance assessed by power output. Rinse concentration, volume, duration and frequency are not given in the abstract.

Key Results

Time-trial power showed no difference between conditions: 247.0 ± 42.0 W with maltodextrin, 242.8 ± 38.1 W with fructose, 246.8 ± 38.8 W with placebo and 243.2 ± 38.9 W with no rinse (control) (p = 0.989).
Authors' conclusion: neither a maltodextrin nor a fructose CHO rinse made the 15-min cycling time trial any better, and rinsing as such produced no placebo effect.

Practical Application

"Will rinsing my mouth with a carbohydrate drink make me faster in a short time trial?" In twelve male amateur cyclists (randomized, double-blind crossover), no: after 15 minutes at 65% of Wmax, power in a 15-minute time trial was the same with a maltodextrin rinse (247.0 ± 42.0 W), a fructose rinse (242.8 ± 38.1 W), a placebo rinse (246.8 ± 38.8 W) and no rinse at all (243.2 ± 38.9 W; p = 0.989).
The type of carbohydrate made no difference, and rinsing with anything was no better than not rinsing — no placebo effect of the act itself.
Not from this paper: the pooled evidence in REF-483 found a small gain in mean power and none in completion time, and REF-066 places the rinse at sustained efforts of about an hour; a twelve-rider null does not overturn a pooled estimate, but it agrees that the effect, if any, is small.
Not from this paper; the base's own practice: treat the rinse as optional — it adds no fuel, and normal fuelling by duration is in SYNTH-02.

Limitations

Twelve riders: underpowered to detect a small effect; "no difference" is not proof of equivalence.
A 15-minute time trial after 15 minutes of steady riding — shorter than the efforts for which a rinse is usually recommended.
Men only; amateur level not quantified; fed or fasted state and the rinse protocol are not given in the abstract.
Online ahead of print. Written from the abstract and record metadata only (full text not read).
How this entry was checked and changed
Why this level

crossover with twelve cyclists; a null result at this size cannot rule out a small effect