Skip to content

Carbohydrate mouth rinsing showed a small exercise effect after eating

A review of randomized trials found a small performance advantage over placebo in fed adults, with low overall certainty.

By 100HP editorialAbstract-based explanation checked

Based on the published abstract. The full paper may contain additional methods, results and limitations.

The 30-second takeaway

Across the exercise trials, carbohydrate mouth rinsing produced a small average performance advantage over placebo when adults had eaten beforehand. Evidence certainty was low, and the standardized result does not reveal a specific gain in speed, strength, or endurance. Cognitive findings were more uncertain and came from a much smaller evidence base.

Keep in mind

Overall evidence certainty was low, and cognitive evidence certainty was very low. This abstract-based summary lacks task-specific results and does not establish a reliable benefit for every exercise setting.

THE NUMBERS, WITH CONTEXT

What researchers found

Hedges' g = 0.18 (95% CI 0.09 to 0.28)

Standardized effect on exercise performance

Carbohydrate rinse versus placebo during exercise after eating; task durations were not reported.

J Int Soc Sports Nutr, 2025 · Original source ↓

What the pooled effect means

The analysis combined different exercise performance measures on a standardized scale. Its small positive result describes an average across studies, rather than a percentage improvement or a promise of better performance for an individual.

Cognition needs separate caution

The cognitive findings were summarized narratively instead of being pooled with the exercise outcomes. Although the review described improvement compared with placebo, it rated this evidence very uncertain, limiting conclusions about cognitive benefit.

CHECK THE ORIGINAL

The original publication

Fed, not fasted: is carbohydrate mouth rinsing still ergogenic? A three-level meta-analysis.

Deng H, Fan X, Liu P et al.
J Int Soc Sports Nutr · 2025

PubMed ID
41199504
Record checked

AI-assisted research and writing. This explains one selected publication; it is not a complete review of everything known. Our approach.