Share this article on:

Effects of Glossopharyngeal Pistoning for Lung Insufflation in Elite Swimmers


Medicine & Science in Sports & Exercise: May 2007 - Volume 39 - Issue 5 - pp 836-841
doi: 10.1249/mss.0b013e31803191c1
BASIC SCIENCES: Original Investigations

Purpose: To investigate whether sessions of glossopharyngeal pistoning (GP) to lung volumes exceeding maximum inspiratory capacity in a group of extremely well-trained elite swimmers would affect maximum lung volumes and buoyancies.

Methods: Ten female and 16 male swimmers performed GP four times a week for 5 wk in addition to their regular swimming training program. Pulmonary function, chest expansion, hydrostatic weights (maximally inhaled and exhaled), and body composition (relative amounts of fat, bone, muscle, and fat-free tissue) were measured.

Results: Training compliance was 79% for the males and 82% for the females. Chest expansion increased significantly during the training period, by 1.0 cm and by 0.8 cm at the level of the xiphiod and the fourth costae for the males, and by 0.6 cm and 0.8 cm for the females. The buoyancy lifting force increased significantly by 0.17 and 0.37 kg for the males and the females, respectively. The females also increased their vital capacity significantly by 2%. No significant changes in body composition took place in either group.

Conclusion: The lung volumes and buoyancies of swimmers can be increased by sessions of GP.

1Division of Physiotherapy, Department of Neurobiology, Care Science and Society, Karolinska Institutet, Stockholm, SWEDEN; 2Department of Physical Therapy, Karolinska University Hospital, Huddinge, SWEDEN; 3Elite Sports Center, Swedish Sports Confederation, Lidingö, SWEDEN; 4Section of Exercise Physiology, Department of Physiology and Pharmacology, Karolinska Institutet, Stockholm, SWEDEN; 5Section of Environmental Physiology, Department of Physiology and Pharmacology, Karolinska Institutet, Stockholm, SWEDEN

Address for correspondence: Malin Nygren-Bonnier, PT, MSc, Department of Neurobiology, Care Science and Society, Karolinska Institutet 23 100, SE-141 83 Huddinge, Sweden; E-mail:

Submitted for publication April 2006.

Accepted for publication December 2006.

It is highly important in competitive swimming to reduce the total drag to reach peak swimming velocity. Swimmers streamline the body posture to reduce dynamic drag, which arises from frontal resistance and from eddy resistance (13). Body position (while not moving) in water is related to buoyancy; it depends partly on the density of the swimmer (which, in turn, depends on the relative amounts of fat, bone, and muscle) and partly on the volume of air in the lungs (20). Air in the intestines and other cavities also influences density (19). The passive drag of a swimmer is related to whether the swimmer has inflated or deflated lungs (20). Drag is lower when lung volume is higher; this partly explains why a large lung volume is beneficial for a competitive swimmer (10). Furthermore, swimmers competing for short distances (50 and 100 m) try to minimize the number of breaths, to reduce disturbances in stroking rhythm and body balance caused by breathing.

Today, a 50-m freestyle race at the international level is usually swum with one breath or without breathing. Swimmers with a larger lung volume have higher buoyancy and a greater store of oxygen available during such a sprint.

Opinions differ about whether it is possible to increase the vital capacity (VC) or the total lung capacity (TLC) by any form of training (2,4,7,17,18,29). Mahler et al. (23) report that static lung volumes and maximal inspiratory/expiratory flow rates are the same for elite athletes as they are for untrained people. In contrast, Armour et al. (2) found significantly larger lung volumes in elite swimmers than in elite runners and controls. The work of breathing is more pronounced during swimming than in other sports (2), but it is not clear whether elite swimmers have naturally larger lung volumes or whether they develop these larger volumes as a result of training the respiratory muscles during swimming (2,3). Results obtained from respiratory muscle training (RMT) and maximal inhalation maneuvers are confusing. Fanta et al. (17) have reported that VC can be increased to a small extent in healthy adults (nonathletes) by making inhalations to TLC with the glottis open, whereas other studies have found that RMT does not change lung volumes (11,27,28).

Glossopharyngeal breathing (GPB) is the use of glossopharyngeal pistoning (GP; the pistoning of small amounts of air into the lungs) by using glossopharyngeal muscles (12,22) to assist lung ventilation. The volume of each GP action or gulp has been reported to be up to 200 mL (4). It was described by Dail et al. (14) in the 1950s for patients with poliomyelitis. GPB is used for breathing by patients with neuromuscular disorders and can normalize tidal breathing. However, patients with weak inspiratory muscles as well as normal patients can practice GP. With greatly increased lung volumes that exceed inspiratory capacity, patients can cough more effectively and increase voice volume. Glossopharyngeal lung insufflation volume (GIV) is the lung volume added by GP over the inspiratory capacity. VCGP is the VC supplemented by GP. Thus, VCGP = VC + GIV. The added volume (GIV) can be used by breath-hold divers to increase both diving depth and duration (22). Some divers, such as those studied by Lindholm and Nyren (22), perform GP maximally on dry land as a stretching maneuver, claiming that GP improves the flexibility of the chest and diaphragm. Many competitive breath-hold divers have large lung volumes (22), but it is not known whether this is solely a result of the selection of individuals with a genetic advantage or whether performing GP plays a role.

It is well known that the VCGP volume of patients with reduced lung volumes can be higher than their VC (5,12), but it is not known whether a period of performing GP has any sustained effect on VC. It has recently been reported that 6 wk of GP sessions can increase the VC of healthy women (nonathletes) by 3% (24). The exact mechanism for this effect is not known, although a possible explanation could be that pulmonary compliance was increased by stretching, enabling the inspiratory muscles to inhale to a larger lung volume (21). A group of human subjects who can be considered to have highly trained respiratory muscles and who may be genetically predestined for large lung volumes are elite swimmers. Thus, elite swimmers could be used as a selected study group with highly trained respiratory muscles because of the nature of their sport, and any effect that RMT could have on VC (17) may already have been achieved by this group.

We hypothesized that 5 wk of GP sessions would increase the chest expansion of elite swimmers and that this would result in increases in VC, TLC, and buoyancy.

Back to Top | Article Outline


Back to Top | Article Outline


Thirty-one subjects were recruited from Swedish elite swimmers who had competed in the Swedish, European, or World Championships, or in the Olympic Games, in 2004. The subjects competed with different strokes and at different distances, and all trained by swimming for 8-10 sessions per week. In the beginning, the male group consisted of 20 subjects and the female group consisted of 11 subjects. Table 1 shows baseline characteristics for the subjects. The protocol conformed to the principles of the Declaration of Helsinki and was approved by the research ethics committee. All subjects gave their informed written consent to participate.

Back to Top | Article Outline


GP is the action of the glossopharyngeal muscles pistoning boluses of air into the lungs (12,14). Each subject learned GP from written information, an instructional video, and individual instruction from the same physiotherapist. All subjects gulped through the mouth; some subjects needed a nasal clip to avoid air leakage past the soft palate (5,12,14). Subjects first performed a short warm-up with stretching exercises for the chest (e.g., lateral flexion and rotation of the chest). Then, in a sitting or supine position, they performed 8-15 cycles of pistoning actions or gulping to VCGP. Each GP cycle comprised as many gulps of air as the subject deemed possible, followed by relaxation of the larynx, which resulted in passive expiration. Each session lasted 20 min, and the subjects trained every day during the first week and then four times a week for the remaining 4 wk. The numbers of pistoning actions of each cycle and the number of cycles were counted and recorded in a training diary, together with other training symptoms and details. The subjects graded their perceived tension in the chest on the Borg CR-10 scale (8) at a measurement session held every week. The subjects trained with supervision once a week and otherwise trained on their own. GP was used as a method only on dry land, for reasons of safety.

Back to Top | Article Outline

Outcome measurements.

All measurements were performed by the same researchers before and after 5 wk of GP sessions. Subjects were blinded to their own results throughout the full study. The experimenters for the respective tests (body plethysmograph, buoyancy, and chest expansion) did not recall baseline data when performing the posttests. The methods and measurements (pulmonary function with body plethysmograph and the weekly VC measurements) have previously been shown not to change VC in a control group of sedentary women (24).

Back to Top | Article Outline

Pulmonary function tests.

Static and dynamic spirometry were carried out using a body plethysmograph (Master Screen Body, Jaeger) according to the ATS standard (1). Reference values from Quanjer et al. (26) were used for comparison. The coefficient of variance (CV) of test-retest measurements of VC (within day) (N = 50) was 1.1%.

A portable infrared interruption flow sensor (SpirobankG, MIR, Italy) was used to measure VC before a training session. It was subsequently used to measure VCGP each week. Each subject's technique was evaluated each week.

Back to Top | Article Outline

Chest expansion.

Chest expansion was measured at the level of the xiphoid process and the fourth costae using a tape measure. The subjects were instructed to perform a maximal exhalation (to residual volume (RV)) and then an inhalation to maximum inspiratory capacity (MIC). Chest expansion was calculated as the difference between circumferences at RV and MIC (9). Chest expansion was also measured after gulping to VCGP. The CV of test-retest measurements (within day) at the level of the xiphoid process (N = 50) was 0.8%, and it was 0.4% for measurements at the fourth costae.

Back to Top | Article Outline

Hydrostatic weight and buoyancy.

The weight of each subject in water was measured at RV (HWRV) and after a full inhalation (HWTLC).

HWRV was measured in a procedure in which the lungs were nearly fully emptied with the head above the surface, followed by emptying of the expiratory reserve volume when the subject was fully submerged. The upper body was held in a position perpendicular to the water surface, and hip and knee joints were held at 90°. The top of the head was held at a distance of 30 cm below the water surface. HWTLC measurements were carried out using the same procedure, with the attachment of a 10-kg weight to a belt at the swimmer's waist. Subjects inhaled to full TLC with the head above the surface and then submerged. The subjects were weighed with a calibrated strain gauge (Model KRG-4, Bofors Electronic, Sweden), and the weighing was repeated three to five times to obtain reliable results. The CV of test-retest measurements (within day) for the two hydrostatic values were 1.5% at both the exhaled and inhaled conditions for the males, and 3.6 and 1.2% for the females at the exhaled and inhaled conditions, respectively. The CV were calculated between the second and third trials. The water temperature was 28 ± 0.5°C.

Back to Top | Article Outline

Body composition.

The percentage of body fat was calculated after measurements at four skinfold sites (16) using a Harpenden caliper (Harpenden Skinfold Calipers, H.E. Morse Co., British Indicators, Ltd). The radioulnar and femoral condylar widths were also measured, making it possible to calculate the relative amounts of skeleton and fat-free tissue (30).

Back to Top | Article Outline

Statistical analysis.

Descriptive statistics are presented as mean ± standard deviation. Student's paired t-test was used for dependent variables, and a value of P < 0.05 was accepted as statistically significant.

Pearson correlation coefficients were calculated to determine the relationship between buoyancy and TLC. A multiple regression analysis was used to predict the correlation between buoyancy (lifting force) and TLC, and between buoyancy and fat percentage (STATISTICA, 7.0, Stat Soft Inc, Tulsa, OK).

Back to Top | Article Outline


Four male swimmers dropped out, one because of airway infection and the other three because of lack of time. One female swimmer also dropped out because of lack of time. Therefore, these swimmers were excluded from the data analysis. Training compliance (self-reported) was 79% for the male group and 82% for the female group. Male swimmers performed an average of 9.4 (range 8-12) GP cycles per training session, with each cycle consisting of an average of 16 individual pistoning actions (range 6-35). Female swimmers performed an average of 10.6 (range 10-12) GP cycles per training session, with each cycle consisting of 10 individual pistoning actions (range 4-25). Borg CR-10 scale during GP averaged 4 (range 1-9) for the male group and 3 (range 0.5-6) for the female group. Some of the participants learned the technique immediately, whereas others required extra training sessions. One female swimmer could not master the technique adequately. This swimmer failed to insufflate her lungs to a volume greater than her VC during the sessions. Eight subjects used a nasal clip while training. Subjects occasionally reported that performing GP caused temporary symptoms such as sweating, tension in the chest, hacking cough, light-headedness, headache, and yawning. Four subjects reported isolated episodes of fainting during GP.

Back to Top | Article Outline

Pulmonary function tests.

VC increased significantly in the female group during the training period. The increase was 0.11 L (from 5.48 ± 0.68 to 5.59 ± 0.63, P < 0.05). There was no significant change in VC for the male group (Table 2).

Average maximal VCGP measured weekly with the portable spirometer was 9.05 ± 1.74 L in the male group and 6.07 ± 0.89 L in the female group. These results are 22.6% (1.64 ± 0.55 L; P < 0.001) and 14.8% (0.74 ± 0.26 L; P < 0.001) higher than the average VC in the two groups, respectively.

Back to Top | Article Outline

Chest expansion.

After the training period, chest expansion had increased significantly during a maximal inhalation to MIC from 7.9 (± 1.2) to 8.8 (± 1.4) cm (12%) at the level of the xiphoid process (P < 0.001) and from 6.5 (± 1.1) to 7.3 (± 1.0) cm (14%) at the level of the fourth costae (P < 0.001) in the male group, and from 6.9 (± 1.4) to 7.5 (± 1.2) cm (10%) at the level of the xiphoid process (P < 0.05) and from 6.4 (± 1.5) to 7.2 (± 0.9) cm (16%) at the level of the fourth costae (P < 0.05) in the female group. Chest expansion also increased by gulping from MIC to VCGP (Fig. 1).

Back to Top | Article Outline

Body composition and hydrostatic weight.

Anthropometrical results such as body mass, relative amounts of fat, muscle, bone, and fat-free tissue did not change during the training period. Baseline data are reported in Table 1.

Back to Top | Article Outline

Lung volumes and buoyancy.

The buoyancy (lifting force) increased significantly from 3.87 ± 0.88 to 4.04 ± 0.89 kg (P < 0.05) for the male group and from 4.08± 0.82 to 4.45 ± 0.92 kg (P < 0.01) for the female group (Fig. 2). Buoyancy at HWRV (sinking force) did not change. HWTLC expresses the influence of both lung capacity and body composition on the buoyancy. TLC was highly correlated with buoyancy at HWTLC (lifting force) for both the male group (r = 0.75) and the female group (r = 0.73) (Fig. 3).

Multiple regression analysis showed that the variables TLC and % fat explained about 70% of the variation in the buoyancy (lifting force) in both the male group and the female group (Table 3).

Back to Top | Article Outline


The chest circumference (Fig. 1) and the buoyancy (Fig. 2) of both groups when inhaling maximally increased after 5 wk of performing GP sessions. The VC of the female group also increased. The RV or buoyancy after exhalation did not change, nor did the percentage of body fat, and we believe that these findings are attributable to the subjects' increased ability to fill their lungs with air.

Figure 3 clearly shows the positive correlation between TLC and buoyancy, especially in the male group, where body fat percentage is lower. Increases in buoyancy decrease passive drag, as has been shown by manipulation of the volume of air in the lungs (20), and we speculate that increased buoyancy could increase swimming velocity in elite swimmers. Furthermore, we hypothesize that if a swimmer can increase TLC after a period of GP sessions, the swimmer would consequently achieve a higher water position. It is also possible that the increased chest volume would allow higher tidal volumes at reduced breathing frequency. This also would raise the average position of the swimmer relative to the water surface during a race.

Most studies dealing with passive and active drag relate to front-crawl swimming and have used a variety of low to submaximal velocities. Our speculations about the advantages of increased buoyancy and reduced drag pertain to competition conditions at high swimming velocities, as described by Yanai (31). Rotational effects (buoyant upper body and sinking legs) during floating (25) are, consequently, not applicable here. How different swim styles and distances would be affected by a change in lung volume is also beyond the scope of the present study.

Most subjects found that the method was easy to learn. The subjects in the present study performed GP for only 5 wk, and although their technique was initially suboptimal, their VCGP exceeded VC by 23 and 15% among males and females, respectively. This volume can be compared with results from elite breath-hold divers who insufflated their lungs up to 50% above their VC (22). These divers, however, had in some cases performed GP for many years. VC increased only in the female group in the present study, whereas buoyancy and chest circumference at TLC increased significantly in both groups. This suggests that the training period was too short for lung-volume changes large enough to be measured by means of spirometry to develop in the male group. Untrained females increased their VC significantly (3%) by performing GP sessions, in comparison with a control group, where no changes were seen (24).

Armour et al. (2) investigated the thoracic wall dimensions of elite swimmers, long-distance runners, and a control group. They also used spirometry and measured flow rates, lung volumes, mouth pressures, alveolar distensibility, and single-breath diffusing capacity. Swimmers had significantly larger lungs than both runners and controls, and this large size was a result of increased numbers of alveoli (2). The swimmers in the present study also had larger lung volumes than would have been predicted from their height, weight, and BMI (Table 2). It could, therefore, be argued that the elite swimmers in the present study comprise a preselected population with both a beneficial genetic heritage and the effects of intensive training on lung volume (2,3,15). We have measured standard clinical parameters (1) as well as functional parameters such as buoyancy. Both types of measurements show that the chest expansion and lung volume increased during 5 wk of GP. Therefore, although it may not be possible to increase lung volume by RMT (7,29), lung mobilization seems to accomplish this.

GPB and air stacking to lung volumes greater than inspiratory capacity have been practiced by patients for more than 50 yr, without any reports of pulmonary trauma (6). The overinflation strived for in normal subjects, however, could possibly cause barotrauma. Our subjects were informed of these risks. They carried out GP without experiencing any unacceptable discomfort or straining. We believe that when GP is performed in a moderate fashion, it does not significantly increase the risk of pulmonary trauma.

Because of the increase in intrathoracic pressure that occurs when a subject performs GP (12), there is a risk of orthostatic syncope (12,22). Four subjects fainted while performing GP; this emphasizes how important it is for susceptible subjects to take a supine training position, to avoid falling and subsequent injury in the event of becoming unconscious. The four swimmers who fainted or felt other discomfort while performing GP did not have higher CR-10 (8) ratings than the other swimmers; this suggests that the risk for syncope cannot be predicted by rating chest tension.

Back to Top | Article Outline


GP sessions increase buoyancy, lung volumes, and chest expansion of elite swimmers. We speculate that performance of GP may have a positive effect on a swimmer's maximal velocity when swimming with filled or partially filled lungs. This speculation requires further investigation. Because most of the studies on GPB have been done in patients with neuromuscular disorders, the full effects and possible complications from GP in healthy humans are not yet known.

The authors would like to express gratitude to all the elite swimmers who participated in this study, and their coaches.

This study was supported by the Swedish National Centre for Research in Sports, the Health Care Sciences Postgraduate School, and the Swedish Sports Confederation.

Back to Top | Article Outline


1. American Thoracic Society. Standardization of spirometry, 1994 update. Am. J. Respir. Crit. Care Med. 152:1107-1136, 1995.
2. Armour, J., P. M. Donnelly, and P. T. Bye. The large lungs of elite swimmers: an increased alveolar number? Eur. Respir. J. 6:237-247, 1993.
3. Astrand, P. O., B. O. Eriksson, I. Nylander, et al. Girl swimmers with special reference to respiratory and circulatory adaptation and gynaecological and psychiatric aspects. Acta Paediatr. 43(Suppl. 147):141-175, 1963.
4. Åstrand, P., and K. Rodahl. Textbook of Work Physiology. 3rd ed. New York, NY: McGraw Hill, 1986.
5. Bach, J. R., A. S. Alba, E. Bodofsky, F. J. Curran, and M. Schultheiss. Glossopharyngeal breathing and noninvasive aids in the management of post-polio respiratory insufficiency. Birth Defects Orig. Artic. Ser. 23:99-113, 1987.
6. Bach, J. R., A. S. Alba, and L. R. Saporito. Intermittent positive pressure ventilation via the mouth as an alternative to tracheostomy for 257 ventilator users. Chest 103:174-182, 1993.
7. Bancalari, E., and J. Clausen. Pathophysiology of changes in absolute lung volumes. Eur. Respir. J. 12:248-258, 1998.
8. Borg, G. A. Psychophysical bases of perceived exertion. Med. Sci. Sports. Exerc. 14:377-381, 1982.
9. Carlson, B. Normal chest excursion. Phys. Ther. 53:10-14, 1973.
10. Chatard, J. C., J. M. Lavoie, B. Bourgoin, and J. R. Lacour. The contribution of passive drag as a determinant of swimming performance. Int. J. Sports. Med. 11:367-372, 1990.
11. Clanton, T. L., G. F. Dixon, J. Drake, and J. E. Gadek. Effects of swim training on lung volumes and inspiratory muscle conditioning. J. Appl. Physiol. 62:39-46, 1987.
12. Collier, C. R., C. W. Dail, and J. E. Affeldt. Mechanics of glossopharyngeal breathing. J. Appl. Physiol. 8:580-584, 1956.
13. Counsilman, J. Competitive Swimming Manual. Bloomington, IN: Counsilman Co., Inc., pp. 142-143, 1977.
14. Dail, C. W., J. E. Affeldt, and C. R. Collier. Clinical aspects of glossopharyngeal breathing; report of use by one hundred postpoliomyelitic patients. J. Am. Med. Assoc. 158:445-449, 1955.
15. Donnelly, P. M., R. R. Grunstein, J. K. Peat, A. J. Woolcock, and P. T. Bye. Large lungs and growth hormone: an increased alveolar number? Eur. Respir. J. 8:938-947, 1995.
16. Durnin, J. V., and J. Womersley. Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 72 years. Br. J. Nutr. 32:77-97, 1974.
17. Fanta, C. H., D. E. Leith, and R. Brown. Maximal shortening of inspiratory muscles: effect of training. J. Appl. Physiol. 54:1618-1623, 1983.
18. Gaultier, C., and R. Crapo. Effects of nutrition, growth hormone disturbances, training, altitude and sleep on lung volumes. Eur. Respir. J. 10:2913-2919, 1997.
19. Holmer, I. Physiology of swimming man. Acta Physiol. Scand. Suppl. 407:1-55, 1974.
20. Holmér, I., and L. Gullstrand. Physiological responses to swimming with a controlled frequency of breathing. Scand. J. Sports. Sci. 2:1-6, 1980.
21. Kang, S. W., and J. R. Bach. Maximum insufflation capacity. Chest 118:61-65, 2000.
22. Lindholm, P., and S. Nyren. Studies on inspiratory and expiratory glossopharyngeal breathing in breath-hold divers employing magnetic resonance imaging and spirometry. Eur. J. Appl. Physiol. 94:646-651, 2005.
23. Mahler, D. A., E. D. Moritz, and J. Loke. Ventilatory responses at rest and during exercise in marathon runners. J. Appl. Physiol. 52:388-392, 1982.
24. Nygren-BonnierM., P. Lindholm, A. Markström, M. Skedinger, E. Mattsson, and B. Klefbeck. Effects of glossopharyngeal pistoning for lung insufflation on vital capacity in healthy women. Am. J. Phys. Med. Rehabil. 86:290-294, 2007.
25. Pendergast, D. R., P. E. Di Prampero, A. B. Craig, Jr., D. R. Wilson, and D. W. Rennie. Quantitative analysis of the front crawl in men and women. J. Appl. Physiol. 43:475-479, 1977.
26. Quanjer, P. H., G. J. Tammeling, J. E. Cotes, O. F. Pedersen, R. Peslin, and J. C. Yernault. Lung volumes and forced ventilatory flows. Report Working Party Standardization of Lung Function Tests, European Community for Steel and Coal. Official Statement of the European Respiratory Society. Eur. Respir. J. Suppl. 16:5-40, 1993.
27. Romer, L. M., and A. K. McConnell. Specificity and reversibility of inspiratory muscle training. Med. Sci. Sports. Exerc. 35:237-244, 2003.
28. Tzelepis, G. E., D. L. Vega, M. E. Cohen, and F. D. McCool. Lung volume specificity of inspiratory muscle training. J. Appl. Physiol. 77:789-794, 1994.
29. Wetter, T. J., and J. A. Dempsey. Pulmonary system and endurance training. In: Endurance in Sport, R. J. Shephard and P. O. Åstrand (Eds.). Malden, MA: Blackwell Science, pp. 59-61, 2000.
30. von Doebelin, W. Anthropometric determination of fat-free body weight. Acta Med. Scand. 165:37-40, 1959.
31. Yanai, T. Rotational effect of buoyancy in frontcrawl: does it really cause the legs to sink? J. Biomech. 34:235-243, 2001.


©2007The American College of Sports Medicine