Somewhere on the internet right now, a heat suit is being sold to an amateur triathlete on the promise that it works like altitude camp without the plane ticket. The physiology behind that pitch is real: five weeks of heat sessions genuinely thicken your blood volume and can nudge your haemoglobin mass upward. What the pitch leaves out is that the single most rigorous trial ever run on the question found heat training no better than the same hours ridden in a cold room -- and that most of what you build disappears within two weeks of stopping.
The Simple Version
Heat training means deliberately raising your core temperature, either by exercising in a hot room or a heat suit, or by sitting in a sauna after training. The adaptations are well documented: plasma volume expands, heart rate at a given power drops, you sweat earlier and more, and your sweat carries less sodium. For racing in the heat, the evidence is strong and the payoff is large. For racing in the cool, the evidence is genuinely split -- one meta-analysis says a small benefit exists, its own statistical method has been publicly challenged, the strictest work-matched trial found nothing, and a five-week elite study found a real gain. This article separates what is settled from what is being argued about, and covers the safety side that marketing tends to skip.
How It Works
What Heat Actually Changes
Plasma Volume: The Fast, Reliable Adaptation
The first thing heat does is make you hold more fluid in your bloodstream. Losing sweat temporarily drops circulating volume, which triggers the renin-angiotensin-aldosterone system and antidiuretic hormone. Aldosterone tells your kidneys to hold sodium, and water follows it. In parallel, albumin accumulates inside the vessels, pulling water in behind it by osmotic pressure.
The effect is quick and large. Lorenzo and colleagues measured a 6.5% plasma volume increase after ten days. Scoon's sauna study found 7.1%. For a 75 kg athlete carrying roughly 3.5 litres of plasma, that is somewhere around 230 mL of extra fluid sitting in the circulation.
More plasma means more venous return, which by the Frank-Starling mechanism means a bigger stroke volume, which means your heart needs fewer beats to move the same blood. This is why heart rate at a fixed power drops after a heat block, and it is the mechanism that every optimistic claim about heat training ultimately rests on.
Sweating: Earlier, More, and Less Salty
Your hypothalamus lowers the core temperature at which sweating switches on, so you start cooling before heat has accumulated. Sweat glands also become more productive. A Bayesian meta-regression across roughly 75 studies put the average increase at 163 mL/h, with each additional heat exposure adding about another 9 mL/h.
At the same time the sweat glands reabsorb more sodium, so what reaches your skin is more dilute. In elite cyclists, sweat sodium fell by 14.1 mmol/L after a five-week block. This is the adaptation with the clearest practical consequence, and we will put numbers on it below.
Haemoglobin Mass: Real, Modest, and Short-Lived
Here the story gets more interesting. Short heat blocks of ten to fourteen days expand plasma but do not build red cells -- Scoon's runners gained 7.1% plasma volume while red-cell volume moved only 3.5%, which was not statistically significant.
The proposed mechanism for longer blocks is that the kidneys read the diluted blood as an oxygen-carrying shortfall and release erythropoietin. Five-week protocols do find something: Ronnestad's elite cyclists gained 2.6% and 2.4% haemoglobin mass in suit and chamber groups while controls lost 0.7%.
But two findings should temper the enthusiasm. In a 2025 study that ran altitude and heat side by side in elite cyclists, the heat subgroup gained 5.4% haemoglobin mass -- and it did not reach statistical significance (P = 0.105, n = 7). More importantly, both the altitude and heat groups were back to their starting haemoglobin mass within 10 to 14 days of stopping. Whatever you build here, you rent rather than own.
The Argument About Cool-Weather Racing
Most of us are not racing in Doha. The question that decides whether an amateur should spend five weeks on this is whether heat adaptation makes you faster in a normal, temperate race. The honest answer is that the field has not settled it.
Lorenzo's 2010 study is the origin of the optimistic view: tested at 13C, the heat group improved VO2max 5%, lactate threshold power 5%, and time-trial work 6%. The problem is the control group did not do matched work in the cold, so the gains could simply reflect a hard training block.
Mikkelsen's group built the design to close that hole. Their control group rode the same 28 hour-long sessions at the same 60% of VO2max, just at 15C instead of 40C. Tested in the cool, the heat group's time-trial power rose 6.0% -- and the control group's rose 5.5%. VO2max did not change in either. The heat added nothing. The authors also noted that chronic heat stress left some athletes too fatigued to execute their key interval sessions properly.
A 2021 meta-analysis of 28 studies came down on the other side, reporting a small but significant VO2max gain in thermoneutral conditions (g = 0.30). That conclusion was then publicly challenged in the same journal: the meta-analysis used fixed-effect models, which assume all studies estimate one true effect, and critics argued that the heterogeneity across protocols demanded random-effects models instead, under which the effect loses significance.
And then Ronnestad's five-week elite study measured performance in temperate conditions and found a pooled improvement of 4.9% against 1.7% in controls. Note that its heat sessions were added to normal training rather than substituted for matched work, so it answers "does adding heat help" rather than "is heat better than the same hours spent otherwise."
Four serious studies, four defensible positions. Anyone who tells you this is settled is selling something.
Example
Example: What a Five-Week Block Actually Buys You
Take a 75 kg age-group triathlete preparing for a 70.3, currently sweating about 1.0 L/h with fairly typical sweat sodium around 50 mmol/L. They run the protocol used in the elite studies: five 50-minute heat sessions per week for five weeks, at easy intensity.
What the research says they can expect:
| Adaptation | Expected change | Evidence |
|---|---|---|
| Plasma volume | +6.5%, roughly +230 mL | Lorenzo 2010, ten-day block |
| Sweat rate | +163 mL/h, so ~1.16 L/h | McDonald 2025, ~75 studies |
| Sweat sodium | -14.1 mmol/L, so ~36 mmol/L | Cubel 2024, five-week block, elite |
| Haemoglobin mass | +2.4-2.6% | Ronnestad 2022, elite, five weeks |
| Performance in the heat | Large, well established | Waldron 2021, g = 0.75 |
| Performance in the cool | Disputed, 0 to small | See evidence section |
The fuelling consequence, which is the least disputed practical gain. Over a five-hour race:
| Before the block | After the block | |
|---|---|---|
| Sweat rate | 1.0 L/h | 1.16 L/h |
| Total fluid lost | 5.0 L | 5.8 L |
| Sweat sodium | 50 mmol/L | 36 mmol/L |
| Total sodium lost | ~250 mmol (~5750 mg) | ~209 mmol (~4800 mg) |
So this athlete now needs roughly 800 mL more fluid across the race but roughly 950 mg less sodium. If they carried their old plan forward unchanged, they would arrive at the run underhydrated and over-salted. Rebuilding the hydration plan around the new numbers is a concrete, defensible reason to do the block -- and it does not depend on winning the cool-weather argument.
Note the assumed 50 mmol/L starting point is a typical value, not a measured one. The 14.1 mmol/L reduction is what the study measured.
What it costs. Twenty-five sessions across five weeks, roughly 21 hours. In the Mikkelsen trial, that load degraded the quality of athletes' key interval sessions. If those five weeks fall inside your sharpening phase, that trade is probably a bad one.
Practical Rules
Practical Rules
-
Decide what race you are actually preparing for. If it is hot, the case is strong and you should do this. If it is cool, understand that you are betting on a contested effect. The fuelling and hydration recalibration is a real gain either way, but the power gain may be zero.
-
Use the protocol that the studies used: 50 minutes, five times per week, for five weeks, at easy intensity. The intensity should be low -- this is a thermal stimulus, not a workout. Anything that turns it into a hard session buys you fatigue you did not want.
-
The post-exercise sauna is the cheapest way in. Scoon's protocol was about 31 minutes at roughly 90C immediately after training, around 13 times over three weeks. It has the practical advantage of letting you do your intervals cold and at full power, then take the thermal stimulus afterwards. Weigh that against the sample size: six runners.
-
Plan for maintenance or accept that it fades. Haemoglobin mass returned to baseline within 10 to 14 days in the Cubel study. Ronnestad kept adaptations going with three sessions per week during a three-week maintenance phase, though that group did not separate clearly from the untrained control -- so treat maintenance as sensible rather than proven.
-
Do not stack heat blocks against your most important sessions. Heat training is systemic stress. Put it alongside easy weeks and base work, not on top of a VO2max block, and keep it away from the final taper.
-
Do not chase a core temperature you cannot measure. The 38.5C target that circulates online comes from elite practice using ingestible or patch sensors, much of it published by a sensor manufacturer. Skin and ear thermometers are not accurate enough to guide this. Without a real measurement, use conservative proxies: the session should feel easy in the legs, you should be able to speak in sentences, and you should stop on time rather than on feel.
-
Know when to stop, and who should not start. Exertional heat stroke is diagnosed by a core temperature above 40.5C together with central nervous system dysfunction, and it is a leading cause of sudden death in sport. The early warning signs are neurological rather than thermal: confusion, irritability, disorientation, irrational behaviour, chills or goosebumps in the heat, sudden disproportionate fatigue. If any of these appear, stop and cool immediately -- cooling within ten minutes is what makes the difference in outcomes. Anyone with cardiovascular disease, uncontrolled hypertension, arrhythmia, or who is pregnant, unwell, running a fever, or taking medication affecting thermoregulation should clear heat training with a doctor before starting. If you are new to endurance training, build fitness first: the studies were run on athletes with VO2max between 55 and 75 mL/kg/min, and their conclusions do not transfer to a beginner.
Evidence Base
Evidence Base
The adaptations themselves are not in question. Plasma volume expansion, reduced heart rate at fixed workload, earlier and heavier sweating, and lower sweat sodium have been reproduced repeatedly. Performance in hot conditions improves substantially -- Waldron's meta-analysis puts the VO2max effect in the heat at g = 0.75 (95% CI 0.22-1.27, P = 0.005) against controls.
The cool-weather question is where the field divides, and it divides along methodology.
Lorenzo et al. (2010) reported 5-6% gains at 13C in trained cyclists after ten days. The control group in that design did not perform work-matched training in the cold, leaving open that the gains reflected the training block itself.
Mikkelsen et al. (2019) closed that gap with an active, work-matched control: 28 identical hour-long sessions at 60% of VO2max, differing only in ambient temperature (40C versus 15C). Both groups improved their time trial almost identically (6.0% versus 5.5%) and neither changed VO2max. The authors reported that cumulative heat stress compromised the quality of athletes' primary interval training.
Waldron et al. (2021) pooled 28 studies and found a small significant thermoneutral effect (g = 0.30, 95% CI 0.06-0.54, P = 0.014). Borg and O'Brien published a Comment in the same journal arguing the fixed-effect model was inappropriate given the heterogeneity of protocols, and that a random-effects re-analysis removes significance. The original authors published a Response defending their approach. We verified that the meta-analysis does state it used fixed-effect models; the re-analysis figures themselves sit behind a paywall and we could not confirm them at source.
Ronnestad et al. (2022) found a 4.9% pooled performance improvement in temperate conditions against 1.7% in controls, alongside haemoglobin mass gains. Because heat sessions were additional rather than substituted, this shows that adding heat helped, not that heat beats an equivalent alternative use of those hours.
Three limitations apply across the whole literature and matter for anyone reading it as an amateur.
Sex. The evidence base is overwhelmingly male. One meta-analysis of the heat shock protein response drew on 118 participants of whom 98% were men. Women appear to need longer induction -- nine days rather than four in one trial -- and menstrual cycle phase shifts baseline core temperature by roughly 0.3-0.5C, enough to distort before-and-after testing entirely.
Training status. Participants were typically trained athletes at 55-75 mL/kg/min. Extrapolating to less-trained athletes is not supported, and their heat illness risk is higher, not lower.
Commercial pressure. Heat suits and core temperature sensors are products. Some of the small case studies and practitioner content in this area originates with the companies selling the equipment. That does not make the findings wrong, but small-sample results from interested parties deserve more scepticism than a peer-reviewed trial with an active control.
One methodological note for anyone reading primary sources: haemoglobin concentration in g/dL is useless for tracking this. Heat expands plasma, so concentration can fall while total haemoglobin mass in grams genuinely rises. Only total mass answers the question.