Iwata · Journal of thermal biology 2024 · crossover trial · n=10

Effects of palm cooling on thermoregulatory-related and subjective indicators during exercise in a hot environment.

Level 2 - randomized trial

Crossover laboratory trial in humans evaluating physiological and perceptual responses

PubMed 38382413 · doi:10.1016/j.jtherbio.2024.103803 · record verified 2026-08-26

What was done

Ten active men (mean age 21 ± 1 years, height 172.2 ± 5.7 cm, weight 67.4 ± 7.2 kg) completed three experimental trials: bilateral palm cooling at 12°C (ICE12°C), palm cooling at 0°C (ICE0°C), and a control condition (CON). Following 30 minutes of rest at ambient temperature, participants performed 20 minutes of exercise in a hot environment (33°C, 60% relative humidity) with cooling applied before and during exercise. Researchers assessed rectal temperature, skin temperature, local sweat rate, heart rate, rate of perceived exertion, oxygen uptake, carbon dioxide production, respiratory exchange ratio, thermal sensation, and thermal comfort.

What was found

No differences were detected between the three conditions for rectal temperature, skin temperature, local sweat rate, heart rate, rate of perceived exertion, oxygen uptake, carbon dioxide production, or respiratory exchange ratio (exact numerical values were not provided in the abstract). Thermal sensation and thermal comfort improved in both the ICE12°C and ICE0°C trials compared to control, and the ICE0°C condition showed a longer duration of cold sensation than the ICE12°C condition.

Why it matters

Bilateral palm cooling before and during short exercise bouts in the heat provides subjective perceptual relief from heat discomfort, but it does not lower core body temperature or reduce physiological heat strain.

Limits

The sample size was very small (n = 10) and restricted entirely to young, active males. The exercise bout was brief (20 minutes), limiting generalizability to prolonged athletic or occupational heat exposure. The abstract omitted numerical values, effect sizes, and confidence intervals for all measured endpoints, and did not specify whether condition order was randomized.

Cited by