To breathe or not to breathe: Can dynamic apnea training be beneficial for endurance?

We are living in the ocean of air like aquatic creatures live in the water. And breathing is the way of living. If we stop to breathe, we will die.  Need of air is especially noticeable during physical activity.
When you  exercise your body is crying for oxygen. Even moderate run results in tenfold increase in ventilation compare to resting level and with exhaustive run can it can rise to thirty.
So, the idea to deliberately deny the body of oxygen while exercising looks a little bit weird. Sportspeople nevertheless are ready to explore everything that may give them the edge in their fight for victory.
Maybe, to your surprise, there is a method, when athletes intentionally limit the supply of oxygen, holding their breath during training.
Despite this a slightly strange approach, people are evolutionarily not entirely new for it, because for thousands of years we have adapted to actions with limited or even completely impossible oxygen availability.
In this article, to understand the rationale for breath-holding training, we will look at two more traditional hypoxic activities. One of them leads us to the mountains; this is a high-altitude training, and another, free-diving, down under the water.

Photo: Vasil Lomachenko “sleeps” at the bottom of the pool.

High altitude training.

Air becomes thinner with the altitude; thus it contains less oxygen.
Decreasing partial atmospheric oxygen pressure makes it more challenging to saturate the blood with oxygen. That becomes especially noticeable in exercise because blood flows faster and time for the gas exchange in the lungs becomes limited. Consequently, oxygen supply for energy production impairs.

As compensation for this problem, the body begins to develop some adaptations. These include adjustments in oxygen delivery (central adaptation), more efficient extraction and use of oxygen (peripheral adaptation), and improved buffer capacity.
The most immediate adaptation is increased ventilation. Pumping more air through the lungs helps, to some extent, compensate for the drop in oxygen in the atmosphere.

Another important central adaptation is the increase in red blood cell production, which allows a given amount of blood to be saturated with more oxygen. However, this change in blood content is not immediate. It takes considerable time at altitude; most of the recommendations are that it should be at least two weeks (the more, the better) and about 2500 m high.

Peripheral adaptations include an increase in muscles capillary density, mitochondria volume and efficiency, enhanced oxidative and glycolytic enzymes activity, and increased myoglobin content
(Millet, Roels, Schmitt, Woorons, & Richalet, 2010).

An upturn in the buffer capacity inside the muscle cells and blood help to tolerate the accumulation of by-products and acidosis, which develop faster with hypoxic stress.

However, it is necessary to note that these adaptations:
1. Are very individual and are not universally present.
2. Depend on the level of altitude and time spent at it.
3. Depend on the level of activity at altitude.

The idea of altitude training is to use hypoxic adaptations for improving sports performance not only at altitude but at sea-level as well. There are a few problems, however.

First is intensity.
It is difficult to maintain the same level of training intensity in hypoxic conditions; thus, some undesirable detraining effect may be possible.

Secondly, there are altitude sickness and logistical problems.
If altitude is high (more than 3000 m) some unpleasant health complications may occur. It is called altitude sickness. Usually, it disappears with acclimatisation in a few days but, anyway, this can spoil altitude training camp.
It is not always easy to organise altitude training properly because it demands changes in athlete training routine and lifestyle.

And finally, altitude-induced adaptations are short-lived, and their effect may disappear in just two weeks after return to a sea-level.

Different strategies.

To overcome these shortcomings, several approaches to training at altitude have been developed.

The traditional approach is to live high and to train high. It is an altitude training camp. This is probably the best way for preparation to competition at altitude.
However, its effect on sea-level performance is doubtful precisely because all mentioned above problems remain unresolved.

To live high and to train low.

This method was introduced as a revolutionary approach some time ago.
The idea is that living high induces an increase in blood red cells content while training low allows preserving intensity.
However, it becomes clear that this method has limited benefits. It looks like just living high does not evoke great positive changes. Better adaptation is evident if the athlete is active at altitude.

To live low and to train high.

Advantages of this approach are avoiding mountain sickness and preserving traditional training regime. At the same time, adding high-altitude training several times a week may induce useful hypoxic adaptation.

This method basically is Intermittent Hypoxic Training (IHT). In case if there are no mountains near, athletes may use artificially created hypoxic conditions (e.g. hypoxic chambers). With IHT perhaps it is difficult to expect significant changes in blood red cells content because hypoxic exposure is too short. However, positive adaptations at muscles level may be possible.

Some authors argue that an inclusion of high-intensity exercises in IHT is a necessity.
This accentuate hypoxic stimulus and promotes positive changes in fast-twitch fibres.
Especially high-intensity IHT is relevant for sports games and martial arts athletes because maximal efforts in under-recovery (thus hypoxic) conditions are quite common in these sports (Brocherie, Girard, & Millet, 2018).

However, it is still debatable whether IHT has noticeable benefits for performance compare to traditional sea-level exercises. Especially if we are talking about already highly trained athletes

Chronic altitude exposure.

Success of Kenyans and Ethiopian/Eritreans runners who are natives of highlands 2400-2600 m promotes a belief of magic influence of altitude training. Yes, it is quite possible that chronic altitude exposure combined with high-intensity and high-volume workouts contributes significantly to the success of these athletes. However, it is unlikely that altitude is the most important factor. Otherwise, Himalayans and Andeans would be the best runners. Genetic, climatic, social, and cultural factors are more likely to play an important role (Wilber & Pitsiladis, 2012).

Conclusion for altitude training:

Altitude training is certainly useful for preparation to competitions at altitude.
However, sea-level performance improvement after short altitude exposure is questionable. There is always a trade-off between possible benefits of hypoxia and loss of training intensity. And, if there are any benefits from short-term altitude training, they are short-lived.

Addition of intermittent hypoxic training to usual training schedule may be beneficial for sea-level performance. High-intensity exercises should be included to emphasise the effect of hypoxic adaptations. However, the significance of these improvements is dubious, and it should be weighed against costs and inconveniences associated with the implementation of IHT.

Altitude training should be individual, especially for team-sport athletes. A team consists of athletes of different physiological characteristics and abilities. It is inefficient to use the same approach to all (Girard et al., 2013).

Chronic altitude exposure combined with high-intensity and high-volume training is the best way to use altitude.

Regarding the topic of this article, we can learn from studies and experience of high-altitude training that the main adaptations to hypoxia may be:
Increase in blood oxygen-carrying capacity, more efficient oxygen extraction and expenditure (more ATPs are produced per molecule of O2), and, possibly, improvements of buffer capacity.

The problem is that most of these adaptations are individual, temporary, and often small. Thus, the question is, can even shorter hypoxia that occurs during breath-holding exercises  be beneficial?

Free diving.

There is another form of human activity when the availability of oxygen is not only limited but also wholly impossible – this happens when we are underwater.

Humans dived long before scuba gear was invented. They do this using a breath-holding technique; just like many aquatic mammals and birds do.

Even now, there are some groups of people who continue to use diving for a living.
Ama divers in Korea and Japan spend hours in sea and half of that time they are underwater. Despite such long overall breathless time, they probably do not push themselves out of the aerobic zone. Each dive is relatively short and does not lead to significant deoxygenation. This notion is supported by the normal blood PH levels in Ama divers (Erika Schagatay, 2010). However, there are people who take breath-holding to the extreme and achieved remarkable results.

Sports free-diving.

An ordinary person can hold their breath for approximately 1 min and may swim underwater around one swimming pool length— 25 m.
Now compare this with World records in some of the free-diving disciplines:
Static apnea (holding breath without physical activity)—  than 11 min 54 sec.
Depth dive with fins — 130 m.
Dynamic apnea (swimming underwater )—316.53m with fins.

It is obvious that these athletes perform in extreme hypoxic conditions, likewise mountaineers at high altitude. However, there are two principal differences from the altitude exposure:
Firstly, it is impossible to compensate for hypoxia by increased ventilation. Secondly, it is the presence of significant hypercapnia (accumulation of CO2 in the blood). Consequently, in this sport, it is necessary to spend oxygen economically and to increase its stores. Another necessity is to develop the ability to tolerate hypercapnia.

For decreasing oxygen consumption body lowers heart rate (bradycardia) and redirects blood flow to vital organs (mainly to brain) while shut down periphery (skin and muscles). Interestingly this “diving reflex” is universal for most diving species.
For improving oxygen storing capacity, free-divers usually have bigger lung volume. They may have a bigger spleen and enhanced ability to contract it. Spleen contains blood red cells reserve, therefore this adaptation increases blood oxygenation (E. Schagatay, Richardson, & Lodin-Sundstrom, 2012).

Elite divers can store much more CO2 in their blood (improved buffering) than ordinary people. Finally, they learned to suppress the urge to breathe caused by the accumulation of CO2. Thus experience divers show less blood desaturation compare with ordinary people during the same duration apneas. At the same time, they can achieve extremely low blood saturation during their maximal apneas (up to 30%) and still remain conscious.

However, it is necessary to note that:
Free-divers are phenomenal when there is a need to perform “longer” on limited oxygen.
Nevertheless their adaptations are specific. They are not the best athletes when the need is to perform “faster”.
Thus it is questionable whether  free-diving benefits can be transferred  to the “fast” sports (Erika Schagatay, 2010).

Holding breath during exercise.

It became clear long ago that performance may be limited by the availability of oxygen and the ability to tolerate its shortage. Thus athletes and coaches started to experiment with the training methods which can improve hypoxic resistance even before high-altitude training became a fashion. For example, great runner Emil Zlatopec allegedly used breath-hold during his training sessions back in 50-th. Some swimming coaches were among pioneers of this method perhaps because swimming itself predisposed for developing breath-hold technique.

Relatively recently, breath-hold got “the second wind” thanks to efforts of a group of scientists, among whom Xavier Woorons probably is the most prominent.
So what is the idea?

It is simple. Breath-hold is probably the easiest way to create hypoxic conditions during the workout to invoke adaptations which will be beneficial for performance.
To some extent, it is the same as IHT.

Although hypoxia exposure is much shorter during breath-holding exercises than in altitude training, it is possible to achieve significant level of desaturation. For example, once I got 30% of pulse oxygen saturation (SpO2) during breath-hold running if my finger oximeter can be trusted (they may be not reliable at low saturation). At normal values of 95-100%, I was really scared, honestly. Usual in my apnea training saturation of 50-60%  corresponds to 7000 m altitude.

The advantage of this method is that you don’t need to travel to the mountains or pay for the expensive hypoxic chamber.
There are also some similarities to free-diving due to the use of breath-holding techniques and the presence of significant hypercapnia.

Thus, we may expect the same adaptations as during IHT. These mostly should be in muscles because hypoxic exposure in breath-holds is even shorter than in IHT, thus perhaps not long enough for changes in the blood.
Presence of hypercapnia may provide additional stimulus for developing buffer abilities.
(Woorons et al., 2010)

Recommendations for breath-hold training.

If a coach decides to proceed with breath-hold training, what can we advise on its technique, mode, and dose? Well, we have limited scientifically based knowledge about the topic. Thus, the following recommendations are mostly empirical and based on Woorons and his colleague’s studies and my own personal experience.

Inhale-hold or exhale-hold?

Some studies which explore breath-holding exercises found that when athletes held their breath with the full lungs, they actually did not achieve significant hypoxia. Rather CO2 accumulation is the most significant stress-factor  (Dicker, Lofthus, Thornton, & Brooks, 1980).
These findings made it plausible to suggest exhale-hold technique when subjects hold their breath after “natural” exhale to functional residual capacity. In this case deoxygenation should occur earlier and to a greater extent.

I think this is a question of the length of breath-hold. Free-divers can achieve extreme deoxygenation, holding breath with the full lungs. When I experimented with breath-hold myself, I could achieve a similar level of oxygen desaturation (around 50%) with both techniques. I just increased the length of breath-hold with full lungs 1.5 times. For example, during apneic jogging, I held the breath for 20 steps in exhale mode and for 30 steps if I did it after inhale.
Maybe during breath-holding sprints, when apnea is short, exhale-hold technic is better.

Static apneas.

In my opinion, it is unlikely that static apnea alone gives some noticeable benefits for endurance. To achieve significant hypoxia, the duration of this exercise must be really extreme (around 5 min). I had reached just 90% SpO2 after 3.5 min. And because there is no physical activity, I don’t think adaptations in muscles are possible. However regular repeated static apneas may be the way for preparation to breath-hold exercises. Additionally, this may help to increase spleen volume and enhance its contraction abilities (Bakovic et al., 1985; Lodin-Sundstrom & Schagatay, 2010).

Static apnea may be a kind of psychological training —meditation if you like. For instance, one of the best boxers in the World —Vasil Lomachenko includes apnea in his training and can hold it around 4 min (see photo). Although he admits that he does not like it.

Forced hyperventilation.

Hyperventilation enhances CO2 removal. Level of CO2 in the blood plays a major role in breath-holding because CO2 accumulation makes you breathe. Sometimes we can see that people before diving make a few forceful exhales for reducing CO2 level.
Although that can to some extent make apnea easier, generally it is not recommended to do this.
CO2 is a guarding dog which prevents us from “forgetting to breathe.” I am not joking. As I will discuss below, blackouts during apnea is quite common, and having something that makes you to breathe can be really useful. In addition, CO2 stimulates the dilation of cerebral vessels, which can also help to prevent blackouts (Fielding, Pia, Wernicke, & Markenson, 2009).

Control.

Of course, you cannot run for a long on one breath. Thus technique with reducing breath frequency is used.
The best way to control the time of repeated apneas in the running is to count steps during breath-holds intervals and to count breathes between apneas.
For example, I use the following mode: 20 steps (around 7.5 sec) exhale-hold— three breathes— 20 steps exhale-hold—three breathes and so on.
I use 30 steps exhale-hold apneas during sprints otherwise holding period will be too short. However, in this case, I make five breathes pause instead of three.
The same principal approach may be used in cycling ( counting revolutions) and swimming (strokes).

Intensity.

Perhaps it is plausible to combine different intensity for breath-hold training. With light jogging, it is possible to spend more time in hypoxia. However, adding apneic sprints may help induce another/additional positive adaptation. Basically, this has the same rationale as adding high-intensity to IHT at altitude.

Duration and desaturation.

Keeping apnea during exercise is quite difficult, unpleasant, and possibly even risky method. All recommendations about its duration are more speculative than based on scientific rationale. So take it with the caution and listen to your body.

In (Fornasier-Santos, Millet, & Woorons, 2018) study, athletes made 2-3 series of 8 x 40 meters sprints. Sprints departed every 30 sec. Thus athletes exhale-hold breath for 6-7 sec during sprints and breathe normally for 23-24 sec between them. Desaturation was around 90%.

In sub-maximal running ( 70% of VO2max) subjects made four 5 min bouts with reduced breath frequency (one breath in 4 sec). At the end of each minute inside bout they breathe normally for 15 sec. So basically reduced breath was kept for 45 sec then 15 sec normal breathing then again 45 sec reduced breath and so on. The break between bouts was 1min
They reported desaturation around 85% SpO2 (Woorons et al., 2008)

In swimming study (Xavier Woorons, Patrick Mucci, Jean Paul Richalet, & Aurelien Pichon, 2016) athletes performed 10-20 x 25 meters exhale-hold swimming sprints. During the sprint, athletes hold breath until the urge to breathe then inhale and finish remaining part. After 10-15 sec break, they repeated sprint.
On average, athletes were in moderate hypoxaemia (88-91%) around 90 sec and in severe ( <88%) around 55 sec. However, there was significant inter-individual variability.

In cycling: two sets of 8x 6 sec sprints, set-off every 30 sec. Break between sets — 3 min. SpO2 reached 86% (Woorons, X., Mucci, P., Aucouturier, J., Anthierens, A., & Millet, G. P., 2017).

I could achieve very severe hypoxia (up to 40 %) in sub-maximal (70% HRmax) run 3-5 min with repeated cycles of exhale breath-holding. Mode: Exhale-hold 20 steps – 3 breathes -exhale- 20 steps holding, and so on. I repeated this runs four times with around 2 min passive rests between bouts. Usually, most severe hypoxia was after the first run. After that, possibly due to increased red cells influx from spleen, exercise became easier. However, saturation was never higher than 60-70%. After that, I usually supplemented breath-hold training with two series of ten 50 meters (30 steps) sprints. Mode: 30 steps breath-hold with 5 breathes between. Desaturation usually was around 70-80 %.

 

60% SpO2 after my 3 min apneic run. 

 

As for my training, however, I want to emphasise that such a severe level of hypoxia is probably dangerous. Also, it is unclear whether greater deoxygenation leads to greater effect.

Different studies reported different levels of blood lactate (BL) during breath-hold exercise. I think this is because BL is secondary not only to hypoxia but to intensity as well. Usually BL level is moderate because it is difficult to exercise for a long and with significant intensity in apneic conditions. I had BL of 3.5 mmol/L after 4×3 min 70% HR breath-hold run with 1 min rest between.

So, basically, breath-hold training incorporated into the normal session unlikely can be longer than 30 min. It is better to split it into 2-4 sets with periods of normal breathing between.
This is physically and, even more mentally, difficult exercise. Hence it should be considered how it may influence the main session.

Dangers.

Despite luring benefits of training in hypoxia, some scientists warn that negative consequences may override positive adaptations.
Due to fact that dry-land breath-hold training remains quite exotic and not widely understood the possible dangers of it implementation can be obtained from the studies on pathological hypoxia, altitude exposure and free-diving.

However, it is necessary to note that these are different conditions; therefore, all comparison should not be taken for granted.

For example, cognitive impairments at high-altitude are well-known. Especially if subjects are not acclimatised.  However it is clear that breath-holding is different. It has  rapid onset of hypoxia (which may be considered as bad) and shorter overall duration (may be good).

Dempsy and Morgan based on relatively well studied health impairments from oscillatory sleep apnea argued that IHT is basically the same. Additionally, they pointed out that response to hypoxic training has high individual variability even among high-level athletes (Dempsey & Morgan, 2015).

On contrary Bain et al. pointed out that unlike to pathophysiological sleep patients trained apneists do not lose cerebrovascular CO2 reactivity and their sympathetic baroreflex gain and respiratory muscle sympathetic nervous activity modulation are also unchanged (Bain, Drvis, Dujic, MacLeod, & Ainslie, 2018).

Repeated exposures brain to hypoxia may have negative consequences. For example, some studies found short-term memory impairments in elite free-divers (Bain et al., 2018).
Even such seemingly positive response to hypoxia and hypercapnia as dilated brain vessels may cause elevated intracranial pressure and consequently brain micro-damages (Dempsey & Morgan, 2015).

However, in their  study (Ridgway, L. and McFarland, K., 2006) found no neurocognitive impairments in twenty one elite free-divers. And this despite the fact that these athletes competed for 1-20 years and had multiple adverse neurological events. In addition, authors pointed out that there is a big difference between pathological hypoxia which is often accompanied by ischaemia  and hypoxia with the normal blood flow.

Heart arrhythmias are quite common in dry apneas and free-diving. Reasons for this are still unclear.
Cardiac arrest may be one of the reasons for fatal incidents among amateur divers, especially in cold water. Although, to my knowledge, there are no such incidents in elite free-diving, sometimes even trained divers show unusual HR patterns during  apneas (Lemaitre et al., 2005; Lemaitre, Lafay, Taylor, Costalat, & Gardette, 2013).

Perhaps it is worth noting that when holding your breath with significant physical exertion, the load on the heart can differ both from free diving and from training at altitude.

At altitude, HR increases due to hypoxia. On the contrary, in free-diving, it slows-down aiming to decrease metabolic rate.
Physical work in free-diving is relatively small. During breath-hold running, cycling, etc. intensity is much higher. Hence there may be a conflict between a tendency to apneic bradycardia, and the tendency to increasing HR due to physical activity. For example, in my sub-maximal breath-holding runs HR oscillated in a sinusoidal manner; going down during breath-hold periods and up during normal breathing. Honestly, I don’t know, is it good or bad for the heart.

 

HR oscillation during my sub-maximal apneic run. 

 

Black-out due to severe brain hypoxia is a frequent event in free-diving.
This may be not so in dry-land apneas but is still possible. Sometimes I felt a bit dizzy in apneic runs. Therefore,  some precautions are necessary.

As I mentioned before it is better, do not forcibly exhale and to keep CO2 level high. Firstly this makes brain vessels dilated and secondly it is a safety mechanism which urges you to breathe before you reach dangerously low levels of hypoxia.
It is good to have somebody near when you are exercising with breath-hold just in case you have a syncope. This person has to know what to do for helping you regain consciousness.

In general, how breath-holding exercises affect health is, in my opinion, still unclear. Coaches should carefully monitor athlete (especially heart) and weigh potential risk against possible benefits.

Any proven results so far?

In my opinion, there are not, honestly.

There are just a few studies mostly conducted by Woorons and colleagues.
Usually, studies were short, just 6-15 sessions and had small samples. Often  limited physiological data were collected thus we don’t know where exactly adaptations occurred and to what extent. It is impossible to make such study blind because participants certainly know whether they get treatment or not. They breathe or not. Thus there may be a strong placebo effect. This is particularly important because tests used for evaluation of treatment effect (repeated or long sprints) strongly depend on motivation. Subjects who have got some training have stronger motivation to perform better in a final test.

Chasing a result.

(Fornasier-Santos, C., Millet, G. P., & Woorons, X.,2018) trained elite rugby players with apneic sprints.
Authors reported remarkable improvement of 64% in “open-loop” repeated sprint test (a test where sprints are repeated till speed falls to the preset percent of reference speed) after four weeks training.

Honestly, in my opinion, such test design is open to statistical manipulations. It hugely depends on the choice of reference speed and failing criteria. Not surprisingly, studies which used this test often reported remarkable (and unreal) results.

I have found support for my point of view in (Montero, D. and Lundby, C., 2015) comment on cross-country skies study. There was 55% improvement in open-loop test after just six sprint sessions in hypoxia for experimental group and no improvement in controls (Faiss, R., Willis, S., Born, D.P., Sperlich, B., Vesin, J.M., Holmberg, H.C. and Millet, G.P., 2015). However Montero and Lundby showed that this is result of flawed criterion for task failure and there were no difference between groups in reality.

As for rugby study, listen, to improve some “real” physiological quality by 64% in already well-trained athletes after just seven 20 min sessions?! I think it is doubtful and was more likely due to some mentioned statistical limitations, the possibility of learning the test, presence/absence of motivation and possible use of pacing which is difficult to avoid in repeated sprint tests (F. Billaut et al., 2013).

Interestingly, the authors admitted that many breath-hold studies where the final test had different protocol compare to training exercise failed to achieve improvements.
They consider the difference in protocols as a weakness of these studies. In their experiment they used the same 40 m sprints departed every 30 sec as a training exercise and as a final test.

In my opinion, using test non-similar to training exercise is not a weakness. If some real physiological improvement has been achieved, it can be checked by various tests from the same “physiological group”, since it has universal significance. Otherwise, the participants most likely only learned a specific test. However, in the end, coaches are interested in improving the qualities necessary for the performance, and not in improving the result of one specific test.

I really respect scientists who devoted significant time and efforts to the exploration of some unusual methods, but the unbiased approach must be of paramount importance. Therefore, when I see figures such as a 64% improvement after seven sessions or 55% after six, I, as a practitioner, am very skeptical. I already wrote about “result-bias”  problem in sports science.

Other studies.

(Trincat, L., Woorons, X., & Millet, G. P. , 2017) trained swimmers. Eight subjects in the experimental group and eight in control. The study had uneven gender distribution (4 males in exp. group and 5 in contr.). Researchers gave slightly different protocols for training and testing. For training it was two sets of 16×15 m sprints with 30 s send-off) whereas for testing they used open-loop 25 m repeated sprint test with departure every 35 sec.

Now after six sessions, authors reported around 35% improvement, which is 2.5 sprints ( from 7.1 to 9.6) in training group and no improvements in controls. However, the actual difference between the training group and controls in the final test was insignificant, just 0.9 sprints (9.6 vs 8.7). Considering that both groups were the same before the experiment (7.1 vs 8.0) and remained the same after, we may conclude that training did not have an effect. Indeed, in fact, we have two unequal gender groups; one swam insignificantly better (0.9 sprints) before training and another the same 0.9 after. It may be just natural variations and/or stronger motivation. There is no proven effect of training, in my opinion.

(Xavier Woorons, et al., 2016) trained triathletes in swimming pool. Eight athletes, ten sessions 12-20 x 25 m breath-hold sprints per session.
Reported improvements were 3.5-4.4% on 100 m 200 m and 400 m.
Again authors did not present individual statistics. Standard deviation was big, especially on 100 and 200 m. This means that results were variable, and some athletes did not improve. More importantly, reported improvements were achieved compared to pre-training test. However, it would be interesting to know if athletes actually improved their personal bests? Otherwise, they might swim a little bit slower before the experiment and add a little bit more motivation in final testing after.

In already mentioned study with runners (Woorons et al., 2008), sub-maximal (70% VO2 max) breath-hold running was used. Four 5 min bouts in session, 12 sessions in four weeks. Seven experimental subjects and 8 controls. No improvements in the final test (time to exhaustion at VO2 max) were found.

The only long-term training study that I have found,  was ( Joulia, Fabrice & Steinberg, Jean & Faucher, Marion & Jamin, Thibault & Ulmer, Christophe & Kipson, Nathalie & Jammes, Yves, 2003). They used dynamic repeated apneas to train triathletes. Exercise was really long ( 1 hour) but with low intensity ( 30 % VO2 max cycling). Apnea consisted of 20 sec breath-hold (I suppose inhale-hold) separated by 40 sec of normal breathing. The training was carried out 3 times a week and lasted 3 months. The authors reported enhanced tolerance to hypoxaemia, however no improvements in incremental cycling test.

Overall I am not convinced that short-term breath-hold training with very limited hypoxic exposure can result in significant achievement in real performance. Indeed, even much higher hypoxic dose during IHT or high-altitude camp usually has minor, if any, benefits for performance compared to traditional training methods (François Billaut, Gore, & Aughey, 2012).

Perhaps for achieving something you need to use apneic training for months if not for years. Whether it is worth it or not, and how it can affect health, is another matter.

Summary.

The main question when you are going to suggest an athlete some “revolutionary” method is: does it give some real advantage compared to traditional methods, which are more familiar for the athlete and safer?

Yes, breath-hold training has a scientific rationale behind it; however, to expect significant gains in performance in a short time would be naive. For example, whereas chronic moderate-altitude training, which allows high-intensity, may provide some small additional gains, the usefulness of intermittent and short-term hypoxic training, is unclear (Faiss, Girard, & Millet, 2013).

Success in sport depends on many factors. Pushing athletes do not breathe on training may have rather negative psychological consequences, whereas small possible physical gains may be really insignificant for overall performance.

Therefore, my advice is never to force athletes if they do not want to do this workout. For those who want to explore – well, you can try. Perhaps some interesting method could be a combination of apnea exercises with various forms of training with normal breathing (for example, high-intensity bouts). Very interesting is the psychological aspect of apnea. Breath-hold exercises, static and dynamic (especially in the water) may be a good form of meditation and relaxation.

In general, although breath-hold training is an interesting method, at the same time, it is still an unexplored and probably dangerous area. Further research requires an objective and cautious scientific approach.

References.

Bain, A. R., Drvis, I., Dujic, Z., MacLeod, D. B., & Ainslie, P. N. (2018). Physiology of static breath holding in elite apneists. Exp Physiol, 103(5), 635-651.
Bakovic, D., Valic, Z., Eterovic, D., Vukovic, I., Obad, A., Marinovic-Terzic, I., & Dujic, Z. (1985). Spleen volume and blood flow response to repeated breath-hold apneas. J Appl Physiol, 95(4), 1460-1466.
Billaut, F., Gore, C. J., & Aughey, R. J. (2012). Enhancing team-sport athlete performance. Sports Medicine, 42(9), 751-767.
Billaut, F., Kerris, J. P., Rodriguez, R. F., Martin, D. T., Gore, C. J., & Bishop, D. J. (2013). Interaction of central and peripheral factors during repeated sprints at different levels of arterial O2 saturation. PLoS One, 8(10).
Brocherie, F., Girard, O., & Millet, G. (2018). Altitude and football: What are new methods and opportunities to maximize players’ fitness?
Dempsey, J. A., & Morgan, B. J. (2015). Humans In Hypoxia: A Conspiracy Of Maladaptation?! Physiology, 30(4), 304-316. doi: 10.1152/physiol.00007.2015
Dicker, S. G., Lofthus, G. K., Thornton, N. W., & Brooks, G. A. (1980). Respiratory and heart rate responses to tethered controlled frequency breathing swimming. Med Sci Sports Exerc, 12(1), 20-23.
Faiss, R., Willis, S., Born, D.P., Sperlich, B., Vesin, J.M., Holmberg, H.C. and Millet, G.P., 2015. Repeated double-poling sprint training in hypoxia by competitive cross-country skiers. Medicine and science in sports and exercise, 47(4), pp.809-817.
Faiss, R., Girard, O., & Millet, G. P. (2013). Advancing hypoxic training in team sports: from intermittent hypoxic training to repeated sprint training in hypoxia. British Journal of Sports Medicine, 47(Suppl 1), i45-i50. doi: 10.1136/bjsports-2013-092741
Fielding, R., Pia, F. A., Wernicke, P., & Markenson, D. (2009). Scientific review: Avoiding hyperventilation. International Journal of Aquatic Research and Education, 3(4), 10.
Fornasier-Santos, C., Millet, G. P., & Woorons, X. (2018). Repeated-sprint training in hypoxia induced by voluntary hypoventilation improves running repeated-sprint ability in rugby players. European journal of sport science, 18(4), 504-512.

Joulia, Fabrice & Steinberg, Jean & Faucher, Marion & Jamin, Thibault & Ulmer, Christophe & Kipson, Nathalie & Jammes, Yves. (2003). Breath-hold training of humans reduces oxidative stress and blood acidosis after static and dynamic apnea. Respiratory physiology & neurobiology. 137. 19-27. 10.1016/S1569-9048(03)00110-1.

Girard, O., Amann, M., Aughey, R., Billaut, F., Bishop, D. J., Bourdon, P., . . . Schumacher, Y. O. (2013). Position statement—altitude training for improving team-sport players’ performance: current knowledge and unresolved issues. British Journal of Sports Medicine, 47(Suppl 1), i8-i16. doi: 10.1136/bjsports-2013-093109
Lemaitre, F., Bernier, F., Petit, I., Renard, N., Gardette, B., & Joulia, F. (2005). Heart rate responses during a breath-holding competition in well-trained divers. Int J Sports Med, 26(6), 409-413.
Lemaitre, F., Lafay, V., Taylor, M., Costalat, G., & Gardette, B. (2013). Electrocardiographic aspects of deep dives in elite breath-hold divers. Undersea Hyperb Med, 40(2), 145-154.
Lodin-Sundstrom, A., & Schagatay, E. (2010). Spleen contraction during 20 min normobaric hypoxia and 2 min apnea in humans. Aviat Space Environ Med, 81(6), 545-549.
Millet, G. P., Roels, B., Schmitt, L., Woorons, X., & Richalet, J. P. (2010). Combining hypoxic methods for peak performance. Sports Medicine, 40(1), 1-25.
Montero, D. and Lundby, C., 2015. Enhanced performance after repeated sprint training in hypoxia: false or reality?. Medicine and science in sports and exercise, 47(11), p.2483.

Ridgway, L. and McFarland, K., 2006. Apnea diving: long-term neurocognitive sequelae of repeated hypoxemia. The Clinical neuropsychologist20(1), pp.160-176.

Schagatay, E. (2010). Predicting performance in competitive apnea diving. Part II: dynamic apnea.
Schagatay, E., Richardson, M. X., & Lodin-Sundstrom, A. (2012). Size matters: spleen and lung volumes predict performance in human apneic divers. Front Physiol, 3(173).
Smith, K. J., & Billaut, F. (2010). Influence of cerebral and muscle oxygenation on repeated-sprint ability. European Journal of Applied Physiology, 109(5), 989-999.
Trincat, L., Woorons, X., & Millet, G. P. (2017). Repeated-Sprint Training in Hypoxia Induced by Voluntary Hypoventilation in Swimming, International Journal of Sports Physiology and Performance, 12(3), 329-335.

Wilber, R. L., & Pitsiladis, Y. P. (2012). Kenyan and Ethiopian distance runners: what makes them so good? Int J Sports Physiol Perform, 7(2), 92-102.
Woorons, X., Bourdillon, N., Vandewalle, H., Lamberto, C., Mollard, P., Richalet, J. P., & Pichon, A. (2010). Exercise with hypoventilation induces lower muscle oxygenation and higher blood lactate concentration: role of hypoxia and hypercapnia. Eur J Appl Physiol, 110(2), 367-377.
Woorons, X., Mollard, P., Pichon, A., Duvallet, A., Richalet, J. P., & Lamberto, C. (2008). Effects of a 4-week training with voluntary hypoventilation carried out at low pulmonary volumes. Respir Physiol Neurobiol, 160(2), 123-130.
Woorons, X., Mucci, P., Aucouturier, J., Anthierens, A., & Millet, G. P. (2017). Acute effects of repeated cycling sprints in hypoxia induced by voluntary hypoventilation. Eur J Appl Physiol, 117(12), 2433-2443.
Woorons, X., Mucci, P., Richalet, J. P., & Pichon, A. (2016). Hypoventilation training at supramaximal intensity improves swimming performance. Medicine and science in sports and exercise, 48(6), 1119-1128.