REF-056 · Ivy et al., 1988 · Journal of Applied Physiology, Vol. · Paper

Carbohydrate Timing After Exercise: Immediate vs Delayed Consumption

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

Participants

n=12 male cyclists. On two occasions they cycled 70 min at 68% VO2max with six 2-min intervals at 88% VO2max, then ingested a 25% carbohydrate solution (2 g/kg body weight) either immediately after exercise (P-EX) or 2 h after exercise (2P-EX). Vastus lateralis biopsies at 0, 2 and 4 h post-exercise.

Overview

Ivy et al. established the timing principle: with the same carbohydrate dose, muscle glycogen storage in the first 2 h after exercise ran at 7.7 μmol·g wet wt⁻¹·h⁻¹ when the carbohydrate was taken immediately, but only 2.5 μmol·g⁻¹·h⁻¹ when it was delayed 2 h (i.e. that group had not yet eaten).

During the second 2 h the immediate group's rate slowed to 4.3 while the delayed group — now fed — reached 4.1 μmol·g⁻¹·h⁻¹, still 45% slower (P < 0.05) than the immediate group's first-2-h rate, despite significantly elevated plasma glucose and insulin. Glycogen immediately post-exercise did not differ between trials.

Conclusion: delaying carbohydrate after exercise reduces the rate of muscle glycogen storage.

Protocol

Within-subject design on two separate occasions: 12 male cyclists cycled 70 min on a cycle ergometer at 68% VO2max, interrupted by six 2-min intervals at 88% VO2max (from the abstract)
A 25% carbohydrate solution (2 g/kg body weight) was ingested either immediately post-exercise (P-EX) or 2 h post-exercise (2P-EX)
Vastus lateralis biopsies at 0, 2 and 4 h post-exercise; venous blood before, during and after exercise
Measures: rate of muscle glycogen storage; plasma glucose and insulin

Key Findings

First 2 h: 7.7 μmol·g⁻¹·h⁻¹ (immediate) vs 2.5 (delayed, unfed) — the delayed rate was roughly a third of the immediate one (arithmetic)
Second 2 h: 4.3 (immediate) vs 4.1 (delayed, now fed) — the fed-late rate was 45% slower than the immediate group's early rate
Post-exercise glycogen was the same in both trials; the delayed trial's slower storage occurred despite higher plasma glucose and insulin
Dose: 2 g/kg carbohydrate as a 25% solution

Practical Application

In 12 male cyclists, 2 g/kg of carbohydrate taken immediately after a 70-min ride stored muscle glycogen at 7.7 μmol·g⁻¹·h⁻¹ over the next 2 h versus 2.5 when the same drink was delayed 2 h; even once the late group ate, its rate (4.1) stayed 45% below the immediate group's early rate — eat carbohydrate as soon as the session ends when the next hard session is close.
Not from this paper; the base's own practice: after a hard or long session, carbohydrate in the first 30 min, on the order of 1.0-1.2 g/kg in the first hour with some protein; plan the post-ride food before the ride so "immediately" is possible.
Inference, not tested here: the advantage is a rate advantage that matters when recovery time is short (same-day or next-morning session); with 24 h and adequate total intake, glycogen catches up.

Limitations

n=12 men, 1988, one dose (2 g/kg), one delay (2 h), 4-h window; no measurement of how the difference resolves by 24 h.
The "delayed" group's first-2-h rate is the fasted rate — the comparison mixes "fed vs unfed" with "early vs late" in that window; the second-2-h comparison (4.1 vs 7.7) is the cleaner timing effect.
How this entry was checked and changed
Why this level

regraded 2026-09-27 from A: crossover, n=12

Content correction 2026-09-27
"68% faster" -> the abstract's numbers: 7.7 vs 2.5 μmol·g⁻¹·h⁻¹ in the first 2 h, and 4.1 vs 7.7 (45% slower) once the delayed group was fed; added the second-2-h rates and the dose.
Removed the invented "effect independent of carbohydrate type (glucose, fructose, maltose equivalent)" — carbohydrate type was not varied.
Removed "miss the 30-min window = lose 60-70%", "extends to 6+ hours", "Z3+ or >90 min", "<15 min"; dose advice labelled as the base's own practice. Level A -> B (crossover, n=12).
2026-09-28: added Protocol from the abstract (Europe PMC/PubMed PMID 3132449).