Smith · Medicine and science in sports and exercise 2013 · Multicenter double-blind counterbalanced crossover trial · n=51

Curvilinear dose-response relationship of carbohydrate (0-120 g·h(-1)) and performance.

Cited 87 times in the scientific literature.

Level 2 - randomized trial

Individual randomized double-blind counterbalanced crossover trial

PubMed 22968309 · doi:10.1249/MSS.0b013e31827205d1 · record verified 2026-08-29

What was done

Fifty-one trained cyclists and triathletes (mean age 28 ± 7 years) across four research sites completed four trials. Each trial consisted of a 2-h constant-load ride at 95% of the workload eliciting a 4-mmol·L⁻¹ blood lactate concentration, immediately followed by a computer-simulated 20-km time trial. In a double-blind, counterbalanced design, twelve carbohydrate-electrolyte beverages (providing 10 to 120 g·h⁻¹ of 1:1:1 glucose:fructose:maltodextrin at a 1 L·h⁻¹ fluid intake rate; three doses per site) were tested against a noncaloric placebo. Data were natural log transformed and evaluated using mixed-model analysis.

What was found

Carbohydrate beverage ingestion showed a curvilinear dose-response relationship with endurance performance. Estimated incremental performance improvements were 1.0% at 9 g·h⁻¹, 2.0% at 19 g·h⁻¹, 3.0% at 31 g·h⁻¹, 4.0% at 48 g·h⁻¹, and peaked at 4.7% at 78 g·h⁻¹, with diminishing enhancement observed above 78 g·h⁻¹.

Why it matters

This study establishes an empirical dose-response curve for endurance athletes, demonstrating that ~78 g·h⁻¹ of a multi-transportable carbohydrate blend maximizes cycling time-trial performance during events lasting approximately 160 minutes.

Limits

Each participant received only three carbohydrate doses plus placebo rather than all twelve levels, requiring cross-site pooled modeling. The fixed 1 L·h⁻¹ fluid rate and 1:1:1 multi-source carbohydrate blend may not generalize to single-source carbohydrates, lower fluid volumes, running disciplines, or female-specific cohorts.

Cited by