REF-111 · Tyler et al., 2015 · British Journal of Sports Medicine, Vol. · Paper

Cooling prior to and during exercise in the heat: meta-analysis (Tyler 2013)

A-fairly strong evidenceMeasured on a mixed groupWritten before October 2026, being re-checked

Tyler, Sunderland & Cheung pooled 28 articles on cooling in hot conditions (wet bulb globe temperature > 26 °C): 23 on precooling and 5 on cooling applied during exercise.

Precooling had a moderate overall effect on subsequent performance (d = 0.73), but the size depended on the test: sprint performance was impaired (d = −0.26), intermittent performance improved (d = 0.47) and prolonged exercise improved strongly (d = 1.91). Cooling during exercise also helped (d = 0.76).

Improvements appeared in studies with and without measurable cooling-induced physiological change, and the literature supports a dose-response relationship among cooling, thermal strain and performance.

Key Findings

ConditionEffect (Cohen's d)
Precooling, overall0.73 (moderate)
Precooling — sprint−0.26 (impairment)
Precooling — intermittent+0.47 (moderate)
Precooling — prolonged exercise+1.91 (large)
Cooling during exercise (per-cooling)+0.76
Studies were run in hot conditions (WBGT > 26 °C); 23 precooling and 5 per-cooling studies.
Dose-response: more cooling and more thermal strain relieved → larger improvement.

Practical Application

In a meta-analysis of 28 studies in hot conditions (> 26 °C WBGT), precooling improved prolonged exercise strongly (d = 1.91) and intermittent exercise moderately (d = 0.47) but impaired sprint performance (d = −0.26); cooling during exercise also helped (d = 0.76).
For a long hot event (multi-hour ride, marathon, long triathlon), precool and keep cooling during the effort where logistics allow; before a sprint or a very short effort, do not precool.
Not from this paper: which cooling METHOD works best (cold drinks, cooling packs, cooled room, vest, water) is in Wegmann 2012 REF-112; a practical race-day sequence is in REF-113.

Limitations

Pooled across many test types and small samples; effect sizes are unweighted-to-weighted summaries across heterogeneous protocols; only 5 per-cooling studies; no per-method breakdown in the abstract.
How this entry was checked and changed
Why this level

regraded 2026-09-27 from B+: meta-analysis pooling mixed test types and small trials

Content correction 2026-09-27
Title "Pre-cooling and heat acclimation strategies" -> the paper is a cooling meta-analysis; heat acclimation is not studied. Title corrected.
"Performance improvements: 1.5-13.1% distance, 0.9-6.9% power, 19-31.9% time to exhaustion" -> not in the abstract; deleted.
"Best method: cold water immersion + ice vest + cold drinks" -> the paper does not rank methods; pointed to REF-112 instead.
Added the overall precooling effect (d = 0.73), the per-cooling effect (d = 0.76), the 23/5 study split and the > 26 °C WBGT condition. Level B+ -> A- (meta-analysis).