Archive for February, 2011

How should a runner breathe?

February 22, 2011

Usually a beginner’s experience of running is dominated by shortness of breath and not surprisingly,  one of the first questions the novice asks is: how can I improve my breathing?  The shortness of breath is due to inadequate supply of oxygen to the muscles and inadequate removal of carbon dioxide, which results in stimulation of carbon dioxide sensitive nerve endings in the walls of the arteries and drives the expiratory centres in the brain stem, creating a sensation of shortness of breath.  However for the novice, it is rare that breathing is the bottleneck in the pathway that transports oxygen to the muscles where fuel is oxidized releasing carbon dioxide which is then transported back to the lungs.  The bottleneck is usually in the cardiovascular system: the heart is unable to pump blood at an adequate rate and the capillaries within the muscle are inadequate to deliver the blood effectively to the muscles. The tyro distance runner soon learns to focus on improving cardiovascular function, and monitors his/her progress by recording heart rate.  Breathing itself ceases to be a focus of attention.

After a period of several months the progress in increasing fitness slows down.  Although capillary development continues slowly over a period of years, the increases in aerobic capacity are only perceptible on a timescale of months.  The athlete scarcely remembers those early days when the most pressing question appeared to be how to improve breathing.  By this stage, many are preoccupied with an attempt to identify the training program that will eke out further improvements in the capacity of the heart to pump blood and of the muscles to extract and use the oxygen delivered by that blood.   A few athletes get diverted into experimenting with changes in style – such as changing from heel strike to landing on the mid-foot or fore-foot.  Despite the intense discussion about running style on blogs and forums, there is remarkably little evidence that adjustments of style lead to major improvement in performance.  Very few give a second thought to the possibility of improving the efficiency of oxygenation the blood in the capillaries in the lungs or the removal of carbon dioxide from the lungs.  Yet the few published scientific studies of the effects of training directed at improving breathing technique indicate that this can produce worthwhile additional improvements in performance in well-trained athletes.

Training to increase ventilatory capacity

For example, Leddy and colleagues (2007)  used a technically complex device that  allowed hyperventilation while carbon dioxide level was maintained constant, to train a group of experienced runners to breathe more deeply.  After  10 hours of training spread over 4 weeks, the runners had increased their ventilatory capacity and achieved a 50% increase time to exhaustion during  treadmill running test at 80% of VO2 max on the day after training, and also on further testing 1 week later.  Furthermore 4 mile run time decreased by 4% and this improvement was maintained for a period of at least 3 months.  The findings demonstrate that experienced runners can be trained to increase their ventilatory capacity and that this increase is associated with improved running performance, sustained for several months after the completion of the training.

Although the procedure used by Leddy was too cumbersome for everyday use by amateur runners, the findings suggest the possibility that simply focussing on breathing more deeply while running might result in better performance.  In fact I had discovered many years ago that the most effective strategy to maximize my speed in the final few hundred metres of a long distance race, such as a marathon, is to devote my conscious attention to breathing a deeply as possible in the final Km.  Typically during a race such as a HM or marathon I breathe at a rate 3 steps per inspiration and 3 steps per expiration for most of the race; during the final Km I increase the breathing rate to a 2:2 pattern.   While the capacity to increase pace at the end of a marathon is important if the main goal is to beat an evenly matched opponent, when the primary goal is to record a fast time, the average pace during the first 41 Km is more important than the pace in the final Km.  So it is more pertinent to ask whether or not efficiency when running below lactate threshold can be improved by more effective breathing.

Fractional breathing

I was therefore intrigued when Rick raised the issue of breathing patterns on his blog a few months ago.   He provided a link to an article by a Russian chemist, Alexander Streltsov (1992), reporting that breathing efficiency could be improved by a technique called fractional breathing: a breathing pattern in which the duration of inspiration is longer than expiration.  The runner makes 4 consecutive inspiratory efforts followed by an expiration, thereby prolonging and deepening the inspiration.   The rationale for the method does not make much sense to me.  Streltsov states that the reason from prolonging inspiration is to ensure that the inspiratory period matches the time it takes for oxygen to attach to haemoglobin (approximately 0.8 seconds).  His argument appears to imply that oxygen can only be transferred to blood during inspiration.  While it is true that the alveoli expand during inspiration and deflate during expiration, they do not deflate entirely due to a natural lubricant called surfactant, which reduces surface tension in the alveolar walls.  The volume of air in the alveoli rises to a peak at the end of inspiration and falls to a minimum at the end of expiration, but the average volume during expiration is not greatly different from that during inspiration.  Furthermore the concentration of oxygen rises and falls during the breathing cycle but the average value during inspiration tends to be similar to that during expiration.  Therefore, oxygen exchange can occur during both the inspiratory and expiratory phase.  As I understand it, the efficiency of oxygen transfer will be maximised by ensuring the concentration of oxygen averaged over the entire cycle is a high as possible.  One of the most important issues is the fact that during the first part of inspiration, the air reaching the alveoli is the stale dead-space air that had remained within the bronchial tree following the previous expiration.  So breathing will only be effective if the volume of air inspired in each breath is much greater than the dead-space volume.  This certainly suggests that deep breathing is likely to be more effective, but does not provide a clear justification for unequal breathing. However it is possible that extending the duration of inspiration will result not only in maximum filling of the lungs but also in a rapid expulsion of air during expiration, so perhaps fractional breathing is an effective way of maximising the depth of breathing.

Steltsov provides some quite impressive evidence for improvement of respiratory efficacy achieved by several international Russian athletes when using fractional breathing.   For example, middle-distance runner, Olga Nelubova, decreased her HR from 118 to 111 when running at an easy  pace (3 m/sec) in the low-aerobic zone, after only 4 days of training in the new breathing technique – a far more rapid improvement than would be expected simply to the cardiovascular improvement produced by 4 days of aerobic training.

My experiences with  fractional breathing

Despite being unconvinced by the explanation proposed by Streltsov to justify his proposed technique, the data suggest that his technique is worth further examination.  I decided that it would be worthwhile to try fractional breathing myself to see if it improved my efficiency.  Because I anticipated it would be easier to regulate my breathing while exercising in the low aerobic zone, it seemed sensible to start with a training program in this zone.

Steltsov maintains his technique can be applied to various forms of aerobic exercise.  Because I can assess heart rate at a fixed work-rate most precisely on the elliptical cross trainer I decided to start with a brief training program on the elliptical cross trainer.  I employed a testing procedure consisting of 26 minutes at a power output of 100 watts, after a standard 6 minute graded warm-up and a 4 minute ‘HR stabilization’ period at an output of 100watts.  The demands of this test closely resembled the demands of the low-aerobic recovery sessions that I have performed intermittently over the past two years, and during which my HR had achieved a reasonably stable plateau in recent months, so I could reasonably attribute any improvement during a brief respiratory training program to the respiratory training itself, rather than to a general increase in fitness.

The primary measure of efficiency was heart rate averaged over the 26 minute period.  Before I began the breathing training program, I recorded HR during three sessions using my usual pattern of breathing in the low aerobic zone: namely three steps per inspiration and three steps per expiration.  Then I did four 40 minutes training sessions on consecutive days,  in which I practised fractional breathing, taking 4 steps during  inspiration and 2 steps during expiration. After four training sessions I found it fairly easy to maintain the 4:2 breathing pattern.  It should be noted that unlike the Russian athletes described by Streltsov, I did not produce an overall slowing of breathing rate.  In the week following training I performed three test sessions: two using my former 3:3 breathing pattern and one using the fractional 4:2 breathing pattern.  I included tests using both the unequal and equal patterns to test whether or not the inequality was essential for achieving the benefits. Then six weeks later, I did a further test session, using my usual 3:3 breathing pattern.

In the two baseline tests prior to training, average HR during the 26 min at 100watts was 116 and 117 BPM.  In the two tests in which breathing pattern was 3:3, in the week following the four training sessions, average HR during the test was 111 and 110, while in the test in which the breathing pattern was 4:2, the average HR was 108.  Six weeks after the training, average heart rate during the test when breathing with a 3:3 pattern was 111 bpm.  These results indicate an improvement of approximately 6 percent after training.  This improvement was maintained for six weeks. Furthermore after training, my efficiency was only marginally better when using the 4:2 fractional breathing pattern, than using my usual 3:3 breathing pattern.  It appears I had increased the effectiveness of breathing, probably by increasing the depth of each inspiration and the forcefulness of expiration.

Interpreting the evidence

This evidence is little more than anecdotal, but it is of noteworthy that the magnitude of the improvement in my HR was similar to that achieved by the Olga Nelubova. It is not clear whether or not I achieved improvement by exactly the same mechanism as the Russian athletes. Unlike Olga Nelubova, I exhibited no overall slowing of my respiratory rate.  My breathing rate was 27 breaths/min before and after training, whereas Nelubova’s rate was 33 before training and 21 after training.   The main reason that I did not adjust my overall breathing rate is that on the elliptical machine, it is easiest to maintain a constant work-rate by employing a constant cadence, and I naturally adjust breathing rate to be an integral multiple of step rate.  Therefore producing a gradual change in breathing rate as training proceeded would have required an unnatural decoupling of step rate and breathing rate

The fact that I exhibited a similarly improved performance when tested using a 3:3 pattern indicates that unequal duration of inspiration and expiration is not essential, though it is possible that the use of an unequal pattern during training promoted both deeper inspiration and more forceful expiration and hence increased the efficiency of the training.   I was definitely more conscious of my abdominal wall moving out during inspiration and inwards during expiration both during training and in the testing sessions.   So, I suspect that in my case the major factor contributing to the improvement was simply due to moving a larger volume of air during each breath.

Exercise in the low aerobic zone is not directly relevant to racing.  However if conscious focus on breathing improves efficiency of exercise even in the low aerobic zone, it is likely that even greater benefits would be obtained in the mid and upper aerobic zones that are characteristic of marathon  racing.   Streltsov reported benefits throughout the aerobic range.

It would be premature to draw any definitive conclusions, especially as it is unclear how similar the mechanism of change in my breathing were to that in the Russian athletes.  Nonetheless at this stage I am inclined to think that it is worthwhile paying greater attention to breathing technique


Leddy JJ et al (2007) Isocapnic hyperpnea training improves performance in competitive male runners Eur J Appl Physiol 99:665–676

Streltsov , A (1992)   The endurance reserves,