Tip · Oxygen

CO2 tolerance training

Q: Does my BOLT score say anything about my fitness?

According to two independent studies, no. In 49 speed skaters, the BOLT showed no significant correlation with any parameter from the Wingate test and cardiopulmonary exercise testing; the values lay between minus 0.172 and 0.013. In 28 active students, it correlated neither with VO2max nor with time to exhaustion nor with weekly activity. The value is not suitable as a fitness surrogate.

Q: Where does the target mark of 40 seconds come from?

From the Buteyko method itself, where a breath-hold pause of at least 40 seconds is considered normalized and is also evaluated that way in studies. There is no investigation that has calibrated this value against a health or athletic outcome. Anyone who uses the value as a measure of training progress is mainly measuring how good they are at the test.

Q: Does the training help with asthma?

Yes, to a small extent and as an add-on. A meta-analysis of seven randomized trials found a standardized mean difference of minus 0.33 for asthma control with a heterogeneity of zero, but no advantage for overall quality of life. In a three-month randomized trial, the symptom score and the use of β2 agonists and inhaled corticosteroids fell by about 20 % each. Always in addition to the prescribed therapy, never as a replacement.

Q: Does lung function improve as a result?

No. In the three-month study, neither FEV1 nor exhaled nitric oxide changed significantly, while symptoms and medication needs declined. The effect lies at the level of symptoms and perception, not in the measured airway obstruction or inflammation.

Q: Is breath-holding dangerous?

Sitting down and without prior hyperventilation, no. In the water it can be fatal, and precisely through the mechanism the training aims for: hyperventilation lowers the CO2 partial pressure and delays the urge to breathe while oxygen continues to be consumed. The urge to breathe may then no longer come in time before loss of consciousness. In freedivers, a blackout at a saturation of 51 % has been documented.

Q: Can the test at least be measured reliably?

There are no data on that. No paper was found on the reliability and repeatability of the BOLT in healthy people. The authors of the largest investigation themselves describe the test as subjective and submaximal and write that it has not yet been scientifically validated. At least the procedure can be standardized: 15 minutes of rest beforehand, after exhaling, until the first urge to breathe.

A measurement from the breathing and diving scene. That training on it improves oxygen utilization has not been shown in humans.

The BOLT is a measurement from the breathing and diving scene: the time from exhalation to the first urge to breathe. It is traded as a fitness marker. This page sets out what it tells you — and why training on it shifts precisely the warning threshold that saves lives in the water.

In short

The BOLT is not related to performance. In 49 speed skaters, there was no significant correlation with any parameter from the Wingate test and cardiopulmonary exercise testing, and in 28 active students none with VO2max, time to exhaustion or weekly activity. The authors themselves write that the test has “not yet been scientifically validated”. The mark of more than 40 seconds comes from the Buteyko method. The method behind it is supported in asthma, to a small extent. And the safety point belongs here too: the training lengthens the time until the urge to breathe — that is, exactly the warning threshold that saves lives during a blackout under water.

How the test works

The urge to breathe arises mainly via the partial pressure of carbon dioxide, not via oxygen deficiency. The urge to breathe occurs when the arterial CO2 partial pressure reaches about 45 to 60 mmHg; loss of consciousness only threatens when the oxygen partial pressure drops to 25 to 30 mmHg. The training targets the gap between these thresholds: anyone who lengthens the time to the first urge to breathe shifts the subjective CO2 threshold — not the oxygen reserve.

In the study version, it is measured like this: sit quietly for 15 minutes, then take a few normal breaths through the nose, hold after exhaling until the first distinct urge to breathe, nose clip, time to one decimal place. Explicitly not up to the maximum breath-hold time. The authors themselves call the test “a subjective and submaximal assessment”.

What the value has to do with performance

Nothing measurable. In the most direct validation study, 49 female and male speed skaters completed the BOLT, the Wingate anaerobic test and cardiopulmonary exercise testing. No correlation was significant, r(47) ranged between −0.172 and 0.013; the model explained R² = 0.08 with a mean prediction error of 9.78 seconds.

The cross-check in a less selected group turned out the same way: in 28 active students, the BOLT correlated neither with VO2max (ρ = 0.31; p = 0.10) nor with time to exhaustion (ρ = 0.27; p = 0.16) nor with reported activity (ρ = 0.05). Two independent studies point in the same direction.

Where the method behind it can do something

The test comes from the Buteyko method, and that has a robust, if small, evidence base in asthma. A meta-analysis of seven randomized trials found a standardized mean difference of −0.33 for asthma control (95 % CI −0.54 to −0.13; I² = 0 %). The heterogeneity of zero is remarkable: the studies all point in the same direction. There was no advantage for quality of life (SMD 0.13). This applies to the method as an add-on to therapy, not to the measurement as a training goal in healthy people.

What is well supported

Two things hold up here. First, the negative findings on the measurement itself, and they are clean: 49 highly trained female and male athletes with three standardized tests in the same group, plus 28 active students with a graded treadmill test and respiratory gas analysis — both times no relevant correlation with performance. Second, the clinical effect of the method in asthma: in a randomized trial with two groups of 30 adults each, the voluntary breath-hold pause as the primary end point increased clearly (F(1;58.09) = 28.70; p < 0.001). Secondarily, the Asthma Control Questionnaire (d = −0.63) and the Nijmegen Questionnaire (d = −0.52) improved, and the use of β2 agonists and inhaled corticosteroids fell by about 20 % each.

What the studies show

BOLT versus performance in highly trained athletes (Kowalski et al. 2024)

49 female and male speed skaters completed the BOLT, the Wingate test and cardiopulmonary exercise testing. Peak power, total work, power drop, time to exhaustion and maximal oxygen uptake were measured. Result: no significant correlations, r(47) = −0.172 to 0.013 at p = 0.248 to 0.984; age, body measurements and training experience were not significant either.

The cross-check in active people (Marko et al. 2026)

28 healthy, physically active students (15 men, 13 women, 22.1 ± 2.5 years) completed a graded treadmill test according to the Bruce protocol with respiratory gas analysis. The BOLT correlated neither with VO2max (ρ = 0.31; p = 0.10) nor with time to exhaustion (ρ = 0.27; p = 0.16) nor with physical activity (ρ = 0.05; p = 0.59). No difference between the sexes.

Three months of training on the breath-hold pause (Vagedes et al. 2024)

Randomized controlled trial with two groups of 30 adults with asthma each. The primary end point was the voluntary breath-hold pause after three months, and it rose clearly (F(1;58.09) = 28.70; p < 0.001). Symptom score and medication use declined, the objective values did not: neither FEV1 nor exhaled nitric oxide changed significantly.

What the value does not show

The test is not validated — that is not an interpretation but the literal finding: “it has not yet been scientifically validated or linked with exercise performance in highly-trained individuals”. There is no paper that has established reliability or a reference range in healthy people. The target mark of 40 seconds is also a convention from the method itself.

That training on this value improves oxygen utilization has been shown nowhere. No paper measured a change in VO2max or oxygen extraction; what was measured were symptom scores, medication use and the test value itself. Outside the asthma context, the group difference is missing: in 90 adults with acute coronary syndrome, sleep quality, fatigue and oxygen saturation improved within the groups under both Buteyko and diaphragmatic breathing, but not between them.

How to do it

The test as it was measured in the study: sit quietly for 15 minutes, then take a few normal breaths through the nose, hold your breath after exhaling — until the first distinct urge to breathe, not to the limit. Nose clip, time to one decimal place. Without these specifications, different people measure different things.

In the studies, the training was guided and ran over weeks to months. In a clinical study of 90 children with asthma, there were two sessions of 30 to 45 minutes each while medication continued; afterwards, 60.0 % of the Buteyko group reached a breath-hold pause of at least 40 seconds, compared with 0 % in the comparison groups. A step-by-step protocol with breath-holding is not found in the studies.

Safety

The calm test itself is unproblematic: by definition it ends at the first urge to breathe, that is, far before any limit; no adverse events were reported in the Buteyko literature. It becomes dangerous as soon as water or hyperventilation come into play — through exactly the mechanism the training aims for. Hyperventilation lowers the CO2 partial pressure and delays the urge to breathe while oxygen continues to be consumed; the breathing impulse may then no longer come before loss of consciousness occurs. Anyone who trains CO2 tolerance therefore deliberately shifts the warning threshold that saves lives in the water. This is not theory: a trained freediver suffered a blackout at the end of a 100-meter dynamic apnea dive at an oxygen saturation of 51 %. Therefore: never in or near water, never in the car or while cycling, never after hyperventilation. Caution in pregnancy, epilepsy, coronary heart disease, cardiac arrhythmias, panic disorder and poorly controlled asthma — in the studies, the method was used without exception in addition to standard therapy.

BK-Score Hype far ahead of evidence

Human evidence		2
Mechanism		4
Safety data		6
Hype gap		3
Track record of use		5

The CO2 tolerance test is a measurement from the diving and breathing scene, not a validated parameter. That training on it improves oxygen utilization has not been shown in humans.

The score rates the state of knowledge, not the effect. “Safety data 9” means well studied – not harmless.
Subjective assessment by Biohacking Kompakt based on published scoring rules – not a scientific rating and not a medical recommendation. Rules and all ratings (German)

Frequently asked questions about CO2 tolerance training

Does my BOLT score say anything about my fitness?