Cold and heat exposure 01: The daily cold plunge

Short answer. The metabolic and hormonal claims for daily plunging come from long, mild exposures, one to two hours at a time, that a three-minute ice plunge does not resemble. What survives in general-population trials is less stress at twelve hours, better sleep quality and better quality of life, not better mood. If you plunge, keep your head above water and do not hold your breath.

What it is

Deliberate cold exposure taken as a standing habit rather than as post-training recovery. Morning plunges, cold showers, winter swimming, cold rooms. Typically shoulder-deep, typically brief, typically nowhere near a training session.

This is a different practice from the post-exercise ice bath, with a different evidence base and different findings. If the question is what cold does to muscle growth, that is covered elsewhere. (→ Inflammation 01: Cold water immersion)

The exposure question matters more here than almost anywhere else in this category, because the gap between what was studied and what people do is unusually wide. We will keep returning to it.

The proposed mechanism

Three chains get proposed, and they are largely independent of each other.

Catecholamines. Cold triggers a sympathetic response. Noradrenaline rises sharply. Since noradrenaline is involved in alertness, mood and focus, the argument runs that repeated cold exposure trains a beneficial stress response.

Brown adipose tissue. Cold recruits brown fat, which burns glucose and fatty acids to produce heat. Repeated recruitment is proposed to improve glucose handling and insulin sensitivity.

Hormesis. A controlled, survivable stressor produces adaptations that generalize beyond the stressor itself.

All three are real physiology. The question is what dose produces what outcome, and that is where the picture thins out.

What has actually been tested

The catecholamine data. The frequently cited figures come from a study of one-hour, head-out immersions at three water temperatures in healthy men. At 14°C, noradrenaline rose by roughly 530% and dopamine by roughly 250% against pre-immersion values, alongside a large increase in metabolic rate and a marked diuresis. Cortisol did not rise at any temperature (Šrámek et al., 2000).

That is one hour. Most people plunge for two to five minutes. Nothing in that study establishes what a three-minute exposure produces, and nothing in it measured any downstream outcome.

The metabolic data. Cold acclimation studies do show brown fat recruitment. Four weeks of two-hour exposures at 10°C, five days a week in a liquid-conditioned suit, increased both the volume of metabolically active brown fat and its oxidative capacity in six non-acclimated men (Blondin et al., 2014). Acute moderate cold at around 18°C, delivered by water-perfused suit for two hours with shivering excluded, improved peripheral glucose uptake and insulin sensitivity by around 20% in fifteen healthy men (Iwen et al., 2017).

Then the pooled analysis. Across seven studies and 85 participants comparing cold exposure against thermoneutral conditions, fasting glucose, insulin and triglyceride concentrations showed no significant differences. Free fatty acid concentrations did increase (Tabei et al., 2024).

Note the protocols in all of the above: water-perfused suits and controlled chambers, two hours at a time, temperatures chosen specifically to sit below thermoneutral but above the shivering threshold. That is a laboratory exposure designed to recruit brown fat without triggering the acute stress response. It is close to the opposite of an ice plunge, which is short, extreme, and shivering-inducing by design.

The general-population outcome data. A meta-analysis of randomized trials in healthy adults, specifically excluding athletes and excluding immersion taken before or after exercise, pooled eleven studies and 3,177 participants. Exposures ran from 7°C to 15°C and from 30 seconds to two hours, at or above chest level.

The findings: inflammation increased in the short term, immediately after immersion and one hour later. Stress was reduced at twelve hours. Immune function showed no significant effect immediately (standardized mean difference, SMD, -0.16; 95% confidence interval, CI, -0.82 to 0.51) or at one hour (SMD -0.18, 95% CI -1.09 to 0.74). Sleep quality and quality of life improved. Mood did not. Narrative synthesis noted a 29% reduction in sickness absence among people taking cold showers (Cain et al., 2025).

The inflammation result is worth sitting with. In the training literature, cold is claimed to reduce inflammation and does not measurably do so. Here, in a completely separate literature with different participants and different researchers, cold exposure raised inflammatory markers acutely. Two independent lines, neither supporting the popular claim.

Our reading

The gap between what was studied and what people do is wider here than anywhere else in this category, and it runs in a specific direction.

The metabolic case does not transfer. Brown fat recruitment and improved glucose handling came from two-hour exposures in the mild range, deliberately set below the shivering threshold. A three-minute plunge at 4°C is a different physiological event, and the pooled fasting metabolic outcomes did not separate from thermoneutral anyway.

The hormonal case is one hour of data being applied to three minutes. Nothing establishes that the catecholamine response scales down proportionally, or that it produces any downstream outcome. The general-population trials measured mood directly and did not find an improvement.

What does survive is worth having. Stress reduction at twelve hours, better sleep quality, better quality of life. Those are outcome-level findings in the population that does this.

So: if you plunge and enjoy it, the evidence supports continuing for reasons of sleep and stress, not for metabolism or inflammation. The one finding worth changing behavior over is the risk profile, and it is specific rather than general: keep your head above water and do not breath-hold. That distinction separates the roughly 5% arrhythmia rate seen without breath-holding from the majority reported with it (Lundström et al., 2025).

Where the goals diverge

Wanting a metabolic effect. The inflammation claim attached to the same practice is covered in the training literature. (→ Inflammation sub-hub) The exposure that produced the metabolic findings is two hours of mild cold, not three minutes of extreme cold. If the metabolic literature is your reason for plunging, the plunge is not the intervention that was studied. And the pooled fasting outcomes did not separate from thermoneutral.

Wanting to feel better. This has the most support in the general-population data, and the support is for stress reduction at twelve hours, sleep quality and quality of life. Not mood. The distinction is real: the trials measured both, and they came apart.

Wanting fewer sick days. The sickness absence figure is the most interesting result in this literature and it comes from narrative synthesis, not from the pooled analysis, and it sits alongside a null result for measured immune function. Something happened. Nobody has established what.

Training for muscle. Timing is the whole question, and it is unresolved. Covered separately. (→ Inflammation 01: Cold water immersion)

Wanting the opposite. Heat after training has its own literature, its own small effect, and a point estimate on the other side of zero. (→ Cold and heat exposure 02: Heat after training; Cold and heat exposure sub-hub)

The case against

The exposure mismatch runs through everything above. Almost every mechanistic finding cited in support of daily plunging came from a protocol that nobody practices. This is not a minor caveat about generalizability. It is the central problem with the entire popular case.

The general-population evidence base is thin. Eleven randomized trials. The authors flag small sample sizes and a lack of population diversity, particularly the under-representation of women, and call for higher-quality trials on long-term effects.

The acute risk is better characterized than most of the benefits. In healthy adolescents, face immersion at 10°C produced supraventricular extrasystoles in six of 54 and ventricular bigeminy in two; whole-body immersion at 2°C produced supraventricular extrasystoles in four of 20 (Lundström et al., 2025). The proposed mechanism is simultaneous activation of the sympathetic cold shock response and the parasympathetic diving reflex, an autonomic conflict associated with arrhythmia (Shattock and Tipton, 2012).

The conditions matter and they cut in a useful direction. In a field study of a mass-participation cold swim, where participants kept their heads above water and did not breath-hold, one of twenty showed a documented arrhythmia, against 69% to 92% reported in studies involving breath-hold submersion. The risk concentrates around apnea and face immersion, not around cold water as such.

What is not known

Whether short extreme exposure produces any of the effects that long mild exposure produced is untested. Nobody has run the plunge protocol against the laboratory protocol on the same outcomes.

Whether the catecholamine response scales down proportionally at three minutes, or has a threshold, or plateaus, is unknown. The dose-response curve for the most-cited finding in this area has not been drawn.

What produced the sickness absence result is unknown, given that measured immune function did not move.

And essentially nothing here has been tested past a few months. Every trial is short.

Safety

Cold shock on entry can trigger involuntary gasping and a sharp rise in cardiovascular load within the first minute (Tipton et al., 2017). Do not breath-hold, do not submerge your face, and do not do this alone. Anyone with a cardiovascular condition, uncontrolled blood pressure, or a cold-triggered disorder should speak to a doctor before starting.