THE RESEARCH BEHIND KYLA.

Foundational review of arteriovenous anastomoses (AVAs), the dense vessel network in the palms, soles, and face that moves blood between core and surface to regulate temperature. AVA density reaches roughly 600 per cm² in the nail beds and around 100 per cm² in the palm. The thermoneutral zone, where AVA flow alone controls core temperature, sits between 26 and 36°C. Outside that window, the body shifts to sweating or shivering.

Original Stanford research on palm cooling between bouts of aerobic exercise in hot conditions. Time to exhaustion increased by 20 to 40% when palm cooling was applied between efforts, with a slower rise in core temperature and a lower rate of perceived exertion at equal workloads.

Randomized controlled trial with 10 men walking in 40°C heat in insulated military overgarments until oesophageal temperature reached 39.2°C. Each man did all three conditions on separate days. Both cooling arms used three sites at once: the traditional arm covered neck, groin and axillae together, the glabrous arm covered cheeks, palms and soles together. Cooling applied to glabrous skin lowered core temperature by 0.30°C every 10 minutes, against 0.17°C for cold packs on the traditional neck, groin, and armpit sites, and 0.12°C with no cooling. Direct evidence that the palms and other glabrous surfaces are the body's most effective route for moving heat out of the body.

Stanford follow-up on palm cooling between sets of resistance training. Subjects gained 40% more bench press volume over 3 weeks (vs 13% in the control group), 144% more pull-up volume in trained subjects (vs 5% control), and 22% in 1-rep max bench press over 10 weeks of pyramid training.

Follow-up paper from the Stanford lab reviewing the broader applications of glabrous skin cooling during strength and endurance training. Reports approximately 0.5°C reduction in core temperature with palm cooling between efforts, leading to extended exercise capacity and improved recovery between work bouts.

Coverage of the Stanford palm cooling research and its adoption by Stanford athletics, the San Francisco 49ers, the Oakland Raiders, and Manchester United. Reports approximately 20% performance improvement during activity using palm cooling, alongside significantly faster recovery between efforts.

Randomized study on leg press, comparing no palm cooling, palm cooling between sets, and palm cooling both between sets and after exercise. Cooling between sets cleared blood lactate faster and slowed the drop in average power, with the cooled condition holding higher power through the fourth set. Independent support, from the University of Tulsa, for cooling between efforts to limit fatigue.

Study of 34 subjects completing rowing ergometer workouts with and without intermittent palm cooling, applied through gloves. The cooled workouts produced greater distance rowed, alongside lower heart rate and lower blood lactate, which the authors attribute to reduced fatigue. Extends the evidence into an endurance, repeated-effort sport.

Sixteen resistance-trained men performed four sets of bench press to fatigue, with palm cooling, palm heating, or no treatment during the rest intervals. Total training volume reached 2,480 kg with palm cooling, against 1,972 kg with no cooling, a gain of about 26%. Muscle activation was higher, and core temperature and perceived exertion were lower. An independent replication of the Stanford resistance findings, from a separate lab at the University of New Mexico.

Randomized controlled trial with 15 subjects on repeated 60-meter sprints with directional changes. Subjects using palm cooling completed the final sprint 2.76% faster, with heart rate roughly 14 bpm lower at the end of recovery, and reported significantly less delayed-onset muscle soreness 48 hours after testing.

Randomized crossover study with 11 active men cycling in 33°C heat, with hands and forearms immersed in 15 to 17°C water during a 15-minute break. Mean power output in the second half was significantly greater than control. Rectal temperature dropped by 0.54°C, heart rate dropped by 16 bpm, and skin blood flow decreased by approximately 40%.

Fourteen recreationally trained men did four sets of biceps curls to failure at 70% of one-rep max under three randomised conditions: hands immersed in cold water (10 degrees Celsius) for 60 seconds of each 90-second rest, a thermoneutral immersion presented to them as a treatment, and no cooling at all. Cooling beat the control condition (p = 0.002) and beat the sham (p = 0.013). The sham also beat the control condition (p = 0.003).

We include it because it is the first study in this field to ask that question directly, and because the honest reading is that both things are true: the cooling did more than the sham, and the sham did more than nothing. Worth noting that this was whole-hand immersion in water rather than a held object, on a single-joint exercise, with 14 men.

Eight resistance-trained women did four sets of bench press to failure at 85% of one-rep max, with three-minute rests, on three separate days: palms cooled to 10 °C, palms heated to 45 °C, or nothing. Both cooling and heating raised the total volume lifted, cooling by about 17% over the thermoneutral day. The authors read it as a central effect: a strong signal from the palms changes how the brain paces fatigue. Small sample, one exercise, same group that published the 2010 study in men.

Twenty female collegiate athletes ran a modified Loughborough Intermittent Shuttle Test and then a 20-metre repeat sprint test, in a randomised crossover. During the rest intervals they held a portable palm-cooling device at 7 to 15 °C, or nothing. With cooling they completed significantly more sprints at 90% or more of their maximum velocity. Heart rate, blood lactate and perceived exertion did not differ between the conditions. A small handheld device, no vacuum, no water bath.

Meta-analysis of 28 studies of men exercising in the heat (above 30°C), 20 cooling before exercise and 8 cooling during it. Both approaches improved performance by a similar amount, and among the cooling-during-exercise methods an ice vest was the most effective. These are whole-body or torso cooling methods applied before or throughout exercise, not palm cooling taken between efforts, so the results support the general case for cooling rather than KYLA's specific method. The review also found no correlation between finishing core temperature and performance.

Six male university soccer players completed a 90-minute intermittent treadmill protocol with and without a pre-cooling cold shower. The pre-cooling lowered rectal temperature before exercise, but the authors found no evidence of benefit to oxygen consumption, heart rate, perceived exertion, or blood markers under normal laboratory conditions. This is whole-body pre-cooling before intermittent exercise, not palm cooling between efforts. It is a useful counterpoint, showing that cooling is not automatically ergogenic and that the effect depends on the method and the heat load.

This review pooled 59 studies (563 athletes) of men exercising in heat above 30°C, comparing cooling before exercise and during it across self-paced and fixed-intensity protocols. Cooling consistently reduced the performance decline, but the benefit was larger for constant-workload exercise and for pre-cooling specifically. As with the other entries here, these are whole-body and pre-exercise cooling methods rather than palm cooling between efforts. It is a current summary of how much the effect depends on context.

Seventeen competitive runners completed two simulated 5-km treadmill runs in hot, humid conditions (32°C, 50% humidity), warming up either in a T-shirt or in a vest filled with ice. The vest lowered body temperature, heart rate, and thermal discomfort during the warm-up, and the runners were about 13 seconds faster over 5 km, with most of the advantage appearing later in the run. This is torso cooling applied before the run, not palm cooling between efforts, and the sample is small, but it is a clean example of pre-cooling producing a measurable performance gain.

Nine male rugby players performed a 2 x 30-minute intermittent-sprint protocol in warm conditions (32°C) under three conditions: control, ice vest, and ice bath plus ice vest, with cooling applied before exercise and at half-time. Maximal sprint times did not differ between conditions, but the players covered more ground during the harder sub-maximal running after ice-bath cooling. This is whole-body pre-cooling for team-sport exercise, not palm cooling between efforts, and it shows the benefit of cooling can be real but modest and limited to certain parts of performance.

Twelve subjects, continuous treadmill running to exhaustion at 30°C. A different device, the BEX Runner, worn during the run rather than held in the rest between efforts. Time to exhaustion was shorter with cooling, 41.3 minutes against 46.7, and core temperature rose at the same rate either way.

The largest palm cooling study in resistance training. 41 physically active adults, three randomised crossover protocols, no external funding. Cooling ran the full three-minute rest between sets at 13 to 15°C, the same window KYLA holds. Pull-ups fell from 38.2 repetitions to 32.2. Push-ups and leg extensions showed nothing either way. Core temperature never moved across six sets, and the authors put the pull-up drop down to grip rather than temperature.

Several episodes discuss glabrous skin cooling, the AVA mechanism, and applications for performance and recovery. Notable discussions on how palm cooling can be used to support recovery and training capacity, with references to the Stanford research and practical protocols for athletes.

In addition to the published literature, we run our own athlete testing across several sports, including HYROX, CrossFit, hockey and endurance running. So far the direction has been consistent: deeper heart-rate recovery between efforts, lower blood lactate at matched pace, and output held across repeated efforts. The samples are still small. The figures, how they were measured, and what we have not measured yet are on the methodology page. This is our own testing, not independent research.