Why You Sleep Hot: The Mattress Isn't Cooling You

A cooling mattress cannot cool you. Nothing you sleep on generates cold. What actually happens at bedtime is that your body opens the blood vessels in your hands and feet and dumps heat out through them, which drops your core temperature and lets sleep begin. Everything a mattress does is downstream of that: it either lets the heat go or it holds it against you.

What your body actually does

Before you fall asleep, something visible starts happening at your extremities. The blood vessels in your hands and feet dilate, warm blood moves to the surface, and heat leaves your body through the skin. Researchers call this vegetative sleep preparedness. Peer-reviewed — Kräuchi et al., 2000.

How far in advance is less settled than it looks. Laboratory work has described the process beginning well over an hour before sleep onset; a 2026 study tracking 42 adolescents and 45 older adults with wearable rings in their own homes found the distal temperature rise starting about nine minutes before they tried to sleep. Peer-reviewed — Jouvencel et al., 2026. The mechanism is consistent. The timing depends on how and where you measure it.

The consequence is counterintuitive. Your hands and feet get warmer, and that is precisely how your core gets cooler. Warm extremities are not a sign that you are overheating; they are the radiator working.

Sleep onset follows the drop in core temperature that this heat loss produces. So the useful question is not "how cold is my bed" but "can my body get rid of heat right now."

The measurement that predicts sleep onset

The research measures this as the distal-proximal skin temperature gradient: the difference between the temperature of your hands and feet and the temperature of your trunk.

That gradient predicted how quickly people fell asleep better than core body temperature, melatonin onset, heart rate, or how sleepy people said they felt. Peer-reviewed — Kräuchi et al., American Journal of Physiology, 2000.

Sit with that for a moment. Of everything that could plausibly forecast sleep onset, including the hormone we sell as a sleep aid, the strongest single predictor was how effectively the body was shedding heat through its extremities.

Why the room matters more than the mattress

Heat only moves from warm to cool. If the room is warm, there is nowhere for your body heat to go, and no surface you lie on changes that. A mattress marketed as cooling, in a hot room, is being asked to solve a problem it has no access to.

The measured effect is consistent. In a longitudinal study of community-dwelling older adults, sleep was most efficient and restful when the bedroom sat between 20 and 25°C, with a 5 to 10% drop in sleep efficiency as the room went from 25°C to 30°C. Peer-reviewed — Baniassadi et al., 2023. A systematic review of ambient heat and sleep reached the same conclusion globally, and found little evidence that people adapt to it. Peer-reviewed — Chevance et al., 2023.

This is why the room is the primary variable and the mattress is the secondary one. Most coverage of sleeping hot has that backwards, because there is a product to sell in one of them and not the other.

What foam has to do with it

Given the mechanism, what a sleep surface can legitimately do is get out of the way. Two properties matter: how much of your body it encloses, and whether air can move through it.

Memory foam is the difficult case on both counts. It is designed to conform, which means more of your skin is in contact with material and less of it is exposed to air, and it is a dense closed structure that air does not readily pass through. That is a description of the material, not an accusation: the same conforming behaviour is exactly what people buy it for.

We have written separately about what memory foam is made of and what the evidence does and does not support there. Heat is a distinct question from chemistry, and it is the one more people actually notice.

Claims about phase-change materials, gel infusions and cooling covers are harder to assess, because the marketing rarely states the condition under which the effect was measured. A material that absorbs heat has a finite capacity and a duration; "cools all night" is a claim about duration that we have not seen substantiated. Inferred

What this does not establish

It does not establish that mattress materials are the main reason anyone sleeps hot. Room temperature, bedding, sleepwear, alcohol, illness, hormones and medication all affect thermoregulation, and several of those have larger effects than what you lie on.

It does not establish that any specific product fails. We have not tested mattresses for heat retention, and we are not going to imply a measurement we have not made.

And the research described here is about the physiology of healthy sleep onset. It is not a diagnosis. Night sweats that are new, drenching, or come with weight loss or fever are a reason to see a doctor, not to buy a mattress.

What actually helps

Cool the room for the night. The 20–25°C band is where measured sleep efficiency is highest, and it is usually the cheapest variable to change. Peer-reviewed — Baniassadi et al., 2023.

Warm yourself before bed, which is not the contradiction it sounds like. This is where the intuitive version of the advice goes wrong. Because heat loss drives sleep onset, it is tempting to conclude you should keep your hands and feet cool and uncovered. The experimental evidence does not support that. When researchers directly manipulated skin temperature, warming the trunk shortened sleep-onset latency by 26%, while manipulating distal skin temperature and core temperature had no effect. Peer-reviewed — Raymann et al., 2005.

Warming works because it causes the vasodilation that dumps heat. A bath is the best-evidenced version: across 1,094 older adults, bathing one to three hours before bed was associated with shorter sleep-onset latency and a measurably higher distal-proximal gradient. Peer-reviewed — Tai et al., 2021.

So the sequence that matches the evidence is: warm before bed, cool room to sleep in. Not cold feet.

Prefer a surface that encloses less of you and lets air move. That is a property of construction rather than a brand.

What we would not spend money on is a cooling claim with no stated test condition. If a product does not say how long the effect lasts or what room temperature it was measured at, there is nothing there to verify.

Frequently asked questions