A Warm Environment Reduces Exogenous Glucose Oxidation and Endurance Performance during Cycling with Facing Airflow
32 °C vs 19 °C with realistic airflow: exogenous glucose oxidation −14-15%, ~15-min TT 15% slower
Participants
n=10 trained male cyclists/triathletes, 36 ± 6 y, VO2peak 55 ± 6 ml/kg/min
Protocol
VO2peak and familiarization trials, then two experimental trials: TEMP 19 °C and WARM 32 °C
Each trial: 2 h cycling at ~50% Wpeak (preload), then an ~15-min time trial
Fan-provided airflow covering the cyclist: ~29 km/h during the preload, ~35 km/h during the TT (the methodological point of the study — most lab heat studies give little or no airflow)
Glucose drink with [U-13C]-glucose every 20 min during the preload, 72 g/h
Measured: exogenous glucose oxidation (tracer), TT time, GI temperature, heart rate, RPE, GI comfort, intestinal fatty acid binding protein (I-FABP, GI integrity marker)
Key Findings
| Metric | TEMP 19 °C | WARM 32 °C | Δ |
|---|---|---|---|
| Exogenous glucose oxidation, mean 40-120 min | 0.56 ± 0.13 g/min | 0.48 ± 0.12 g/min | −15% (P = 0.015) |
| Exogenous glucose oxidation, peak | 0.79 ± 0.18 g/min | 0.68 ± 0.14 g/min | −14% (P = 0.008) |
| ~15-min TT time | 819 ± 47 s | 961 ± 130 s | 15% slower (P = 0.002) |
| GI temperature | lower | higher | P = 0.007 |
| Heart rate | lower | higher | P < 0.001 |
| RPE | lower | higher | P = 0.046 |
| GI comfort | — | — | not different (P = 0.659) |
| I-FABP (GI integrity) | — | — | not different (P = 0.094) |
Interpretive notes:
1
The performance loss and the drop in exogenous glucose use appeared despite airflow approximating outdoor speeds — so they are not an artefact of still-air laboratory heat.
2
GI comfort and I-FABP were not different between trials, so the lower glucose use in WARM was not accompanied by measurable GI damage in this protocol.
3
The authors attribute the impairment to thermoregulatory, cardiovascular and metabolic strain. Inference, not tested here: redistribution of splanchnic blood flow to the skin is the usual explanation for slower intestinal glucose delivery in heat.
Mechanism
Measured: higher GI temperature, HR and RPE in WARM at the same relative workload; lower exogenous glucose oxidation; slower TT.
Inference, not tested: warm ambient conditions raise skin blood flow and cardiovascular load; less splanchnic perfusion can slow glucose absorption; the combination lowers sustainable power. Whether more or different carbohydrate (glucose+fructose) restores oxidation in heat was not tested in this study.
Practical Application
In 10 trained male cyclists/triathletes (VO2peak 55), 2 h at ~50% Wpeak followed by a ~15-min time trial at 32 °C versus 19 °C — with fan airflow of ~29-35 km/h and 72 g/h glucose — cut exogenous glucose oxidation by 14-15% (peak 0.68 vs 0.79 g/min) and made the TT 15% slower (961 vs 819 s), with higher core temperature, HR and RPE but no difference in GI comfort or I-FABP — so in a hot race, count on burning less of what you drink and on a large performance drop even at road speeds; pace by RPE/HR rather than by cool-weather power.
The 72 g/h glucose-only drink was still oxidised at only ~0.5-0.7 g/min in the heat: do not assume that a hot-day fuel plan delivers what it did in cool conditions.
Inference, not tested here: because the loss was not accompanied by GI-integrity or comfort differences, simply cutting intake for "gut protection" is not supported by these data; keep fueling and manage heat.
Limitations
Male-only cohort, n=10 — female response untested
Lab-based, not a field race, although airflow approximated outdoor speeds
2-h preload + ~15-min TT — shorter than ultra-endurance events
Two ambient temperatures only (19 °C and 32 °C) — no dose-response across intermediate temperatures
Single drink: glucose only at 72 g/h — glucose+fructose blends and higher intakes not tested
Acute trial, not heat-acclimated subjects — the effect may be smaller after acclimation
Trained (Tier 2) cohort — elite response may differ
Power outputs, body-mass change, cortisol and funding are not reported in the abstract
How this entry was checked and changed
Why this level
regraded 2026-09-27 from A: crossover, n=10
Content correction 2026-09-27
Authors "Roberts KM ... James LJ", Decathlon funding -> Reynolds KM, Funnell MP, Collins AJ, Mears SA, Pugh JN, James LJ (Crossref); funding not in the abstract, removed.
"18 vs 30 °C, ~60% VO2max, 32 km/h, ~60 g/h 8% glucose+fructose" -> 19 vs 32 °C, ~50% Wpeak, airflow ~29 km/h preload / ~35 km/h TT, 72 g/h glucose only.
"TT 288 vs 250 W, exogenous −9.5%, cortisol elevated, body mass −0.7%" -> TT 819 vs 961 s (15% slower), exogenous oxidation −15% mean / −14% peak; cortisol, body mass and power values are not in the abstract, removed.
Participants n=10 trained male cyclists/triathletes (VO2peak 55) added; splanchnic mechanism now labelled inference; house advice list (power −10-15%, portions every 15 min, 100-110 g/h cap, heat-camp fueling) removed.
2026-10-01: citation of the superseded pointer SYNTH-14 repointed to SYNTH-02 (its "In the heat" section), which carries the cited content.