Yoga Breathing Techniques: Implications for Stress Management, Health, and Psychophysiological Research
James E. Kennedy
Published on the internet in pdf and HTML at
http://jeksite.org/yoga/resp.htm Copyright 1990, 1994 James E. Kennedy
Understanding and application of various respiratory practices are impeded by the many interacting physiological and psychological variables. Yoga techniques may offer insights into useful breathing practices and control of important variables. This review integrates relevant data from (a) the psychophysiological/psychological literature, (b) the physiological/medical literature, and (c) studies of yoga. The available data indicate that yogic slow breathing practices promote dominance of the parasympathetic system, can help control stress, and can contribute to treatment programs for some chronic diseases. Basic research is needed on yogic rapid breathing and alternate nostril breathing techniques. Yogic claims about nasal airflow laterality and cognitive laterality have partial support. Psychological factors such as anxiety and distraction, as well as the physical details of breathing techniques, are important variables in psychophysiological research on respiratory practices.
NOTE OF MAY, 1994: Several lines of research have progressed since this unpublished manuscript was prepared four years ago. These include: (a) several new studies of yoga breathing techniques have been reported, but they do not significantly alter the conclusion in this manuscript, and (b) the effects of deep breathing on the autonomic system and the effects of hyperventilation appear to be more complicated and/or variable than was recognized when the manuscript was prepared.
Various respiratory patterns and maneuvers can provide striking influences on the autonomic nervous system and may exacerbate or reduce adverse responses to stressors. For example, increased breathing rate is a typical response to stressful situations (Grossman, 1983; Magarin, 1982). This tendency can lead to breathing in excess of metabolic needs (hyperventilation), which causes reduced blood carbon dioxide concentrations. The reduced carbon dioxide causes psychophysiological and psychological effects that include (a) enhanced arousal and anxiety, and (b) decreased cerebral and coronary blood flow, which can lead to a variety of clinical symptoms including dizziness, poor performance, headache, chest pain, cardiac abnormalities, and sleep disturbance (Brown, 1953; Fried, 1987; Grossman, 1983; Lum, 1976; Magarin, 1982). Certain other respiratory patterns that modestly elevate blood carbon dioxide concentration appear to promote the opposite effects, including reduced anxiety and increased or well-maintained cerebral and coronary blood flow (Grossman, 1983).
However, practical applications of breathing techniques are hindered by the lack of understanding and control of the many interacting variables. A recent review of physiological mechanisms for respiratory influences on the cardiovascular system described a maze of interwoven and dramatically interacting control mechanisms (Daly, 1986). As indicated above, psychological factors can strongly interact with these physiological mechanisms. With the present state of knowledge, the psychophysiological effects of novel respiratory practices cannot be reliably predicted and replications of basic experiments are often inconsistent due to uncontrolled variables (several examples are given in following sections).
Yoga breathing practices may provide insights into valuable respiratory techniques and control of important variables. These practices are intended to maintain optimum health—with particular emphasis on stress reduction—but have received little scientific attention. According to yoga tradition, the practices were developed by extensive personal experimentation and keen introspection of the results. The breathing practices, or pranayama, are one component of hatha yoga, which is intended to give one a healthy body and mind.
Reduction of hypertension (Irvine, Johnston, Jenner, & Marie, 1986; Patel, Marmot & Terry, 1981; Patel & North, 1975) and dramatic improvement of heart disease (Ornish et al., 1979; 1983; 1990) have resulted from integrated treatment programs that included yoga breathing practices. However, the roles of individual treatment components have not been delineated in these studies. A review of the scientific information related to yoga breathing practices may be useful for evaluating the role of breathing practices in these programs and for improving the practices or adapting them to special cases.
According to yoga tradition, certain breathing practices induce relaxation and calmness, whereas others are invigorating and arousing. In addition, certain practices are claimed to influence cognitive functioning of the brain hemispheres.
This article is intended to (a) describe basic yoga breathing practices, (b) summarize the available scientific information relevant to the effects of these practices, and (c) identify topics needing further research. Health threats from potential misuse of certain powerful respiratory techniques are also noted. The techniques are discussed individually in the sequence they are commonly practiced. Studies that combined several practices without isolating the effects of individual breathing techniques are included in the final discussion and conclusions section. The review focuses on the common basic breathing techniques, with emphasis on beginning to intermediate levels. Less common practices and extremely advanced practices are beyond the scope of this review.
The basic mode of respiration used in many yoga practices and recommended for normal daily activities is slow, smooth breathing using the diaphragm rather than the respiratory muscles of the chest (Christensen, 1987, p. 136; Samskrti & Veda, 1985, p. 10). This breathing pattern is sometimes referred to as abdominal breathing, although, as noted below, the abdominal muscles may play a minor role. Breathing is through the nose rather than the mouth.
The diaphragm is the dominant respiratory muscle for quiet breathing in awake healthy adults, but increased use of chest muscles and increased breathing rate are common results of stress and may become habitual. Slow diaphragmatic breathing appears to reduce adverse effects of stress and promote parasympathetic cardiovascular dominance. The opposite effects are induced by more rapid breathing using the chest muscles. Before discussing the available data, a brief review of the respiratory process may clarify the nature of these modes of breathing and the methodological issues in their investigation.
The Respiratory Process
Three muscle groups can be used in breathing: (a) the diaphragm, (b) the muscles of the rib cage, and (c) the abdominal muscles. This summary of the roles of these muscles is based on Collett, Roussos, and Macklem (1988); Grassio & Goldman (1986); Guyton (1986); and Troyer and Loring (1986).
The diaphragm is the most important muscle for inhalation. It is a thin sheet of muscle separating the chest and abdominal cavities. When relaxed the diaphragm forms an open-bottomed cylinder that extends up into the lower part of the rib cage along the sides of the rib cage, and forms a dome on top. Diaphragm contraction shortens the cylindrical portion and pulls the dome down. This movement expands the lungs by pulling them down, which creates a partial vacuum that causes inhalation if the airway is open. When the diaphragm relaxes, the elastic recoil of the lungs pulls it upward, causing exhalation. The downward pressure on the abdominal viscera from contraction of the diaphragm forces the abdominal wall to extend forward and/or the lower rib cage to expand to the sides. The term abdominal breathing derives from this easily observed movement of the abdominal wall, but can also refer to the use of abdominal muscles described below.
Although upper chest movement is relatively inconspicuous in quiet breathing for a relaxed person, some thoracic muscles play a role. The external and parasternal intercostals (joining adjacent ribs) and the scaleni (connecting the shoulder area and spine) are activated during inspiration to hold the ribs in an expanded position that compliments the force of the diaphragm. However, the exact roles of each of the muscles are not yet resolved. The minimal chest movement combined with the fact that some chest displacement could be a result of diaphragmatic action have contributed to the difficulty in resolving this question. The internal intercostals may sometimes play a role in exhalation.
The abdominal muscles are the most powerful and important muscles for forced exhalation, but are normally not used in quiet breathing. Contraction puts inward pressure on the abdominal viscera, which then push the diaphragm up and reduce lung volume. In addition, these muscles may assist expiration by pulling down and deflating the lower rib cage. The important abdominal muscles for respiration are the rectus abdominous, the transverse abdominous, and the external and internal obloquies.
Abdominal muscles can contribute significantly to inhalation by pushing the relaxed diaphragm farther into the rib cage. This action (a) places the diaphragmatic muscle fibers on a more favorable part of their length-tension curve, and (b) converts some of the respiratory system expiratory elasticity to inspiratory forces.
Only about 10 percent of total respiratory capacity is used on each breath in quiet breathing. The volume of each exhalation or tidal volume is about 500 ml for a quiet adult male. Most of this tidal volume goes to lung areas that exchange oxygen and carbon dioxide with blood, but about 150 ml is dead space from passages that cannot contribute to gas exchange. Dead space volume is relatively constant whereas tidal volume varies greatly with physical exercise, breathing pattern, and other factors. Thus, larger tidal volumes have a smaller proportion of dead space. Dead space can increase significantly with lung disorders.
During normal quiet breathing, exhalation is driven by the elastic forces of the lung. Muscles used for inhalation contract to slow and control the rate of exhalation. The position of the relaxed diaphragm and corresponding lung volume after exhalation depend on a balance between the elastic forces collapsing the lungs inward and the elastic forces expanding the chest outward. This lung volume at the end of relaxed expiration is called the functional residual capacity (FRC).
About one fourth of the respiratory capacity not used with quiet breathing can be accessed with additional exhalation and three fourths with additional inhalation. If abdominal muscles force maximum reduction of lung volume, the expiratory reserve volume of about 1100 ml of air below FRC for an average male is expired. This combines with the tidal volume (500 ml) and the inspiratory reserve volume of about 3000 ml to give a vital capacity of 4600 ml. In addition, a residual capacity of about 1200 ml of air remains in the lung after maximum exhalation. These values are typical for a young adult male. The volumes are about 25 percent less for an average female, and vary with body size, posture, and physical condition.
Adequate air flow or ventilation of the lungs can be achieved with slow breathing rate and large tidal volume or fast rate and small tidal volume. The ventilation rate is normally set to provide oxygen and remove carbon dioxide in accordance with metabolic needs.
The abdominal and chest muscles also have important functions for posture, locomotion, and verbalization that must be integrated with and may modify respiratory functions.
In a wide-ranging, extensive review of the literature related to respiration and stress, Grossman (1983) concluded:
A breathing pattern characterized as rapid, low-tidal volume, predominantly thoracic ventilation with relatively low alveolar and blood concentrations of carbon dioxide . . . is associated with psychological characteristics of anxiety, neurosis, depression, phobic behavior, and high levels of perceived and objective stressors. Voluntary performance of this breathing pattern seems to intensify subjective and physiological indicators of anxiety when exposed to stress. Cardiovascularly, voluntary production of this ventilatory pattern appears to bring about significant reduced parasympathetic tone and increased sympathetic dominance, which are expressed in augmented heart rate and cardiac output, muscle vasodilation, decreased blood flow and oxygen supply to the heart and brain, reduced [respiratory sinus arrythmia] and baroreceptor responsiveness, and increased likelihood of major ECG abnormalities. (p. 293)
Stress causes a tendency for enhanced ventilation with upper chest breathing patterns that can become habitual in some people. This conclusion is supported by a variety of studies of stress reviewed by Grossman, by more recent studies of stress (Freeman, Conway & Nixon, 1986) and by studies of hyperventilation (Lum, 1976; Magarin, 1982).
The increased ventilation in response to stress presumably is in anticipation of physical activity such as a fight or flight reaction (Grossman, 1983). However, when little physical activity follows, a tendency to breath in excess of metabolic needs results. In this context, the tendency to hyperventilate in response to stress in a civilized society is not surprising. The degree of hyperventilation and associated symptoms vary from mild to severe depending on dispositional and situational factors (Bass & Gardner, 1985; Clark & Hemsley, 1982; Freeman, Conway, & Nixon, 1986; Wientjes, Grossman & Defares, 1984).
As noted in the introduction, over breathing causes a variety of effects, including increased heart rate, arousal and anxiety, and various clinical symptoms due to decreased blood flow to the brain and heart. Note that voluntary hyperventilation usually induces increased arousal and anxiety in normal subjects1 (Clark & Hemsley, 1982; Grossman, 1983; Thyer, Papsdorf, & Wright, 1984). The reduced carbon dioxide concentration in the blood is a key physiological factor underlying these effects. The decreased blood flow to the heart and the heart rhythm abnormalities can pose a significant risk for those with cardiovascular problems.
Thoracic breathing is symptomatic of habitual or chronic hyperventilation and may be a potentiating factor (Freeman, Conway & Nixon, 1986; Lum, 1976). Lum (1976) reported that over 99 percent of the 640 patients he had seen for chronic hyperventilation were thoracic rather than diaphragmatic breathers. He suggested that some people with a tendency to respond to stress with thoracic breathing become habitual over breathers. The result is that:
The chronic hyperventilator lives much nearer the frontier of hypocapnic [low carbon dioxide] symptoms and any small additional stress, whether psychological or physical, may push him over into symptoms which add to the stress while leaving him mentally and physically less able to cope. Thus the vicious cycle may be triggered. (Lum, 1976, p. 214)
Based on clinical experience and very limited published data, Fried (1987, p. 8) estimated that incidence of habitual hyperventilation in the general population may be 10 to 15 percent and perhaps over 20 percent.
Diaphragmatic breathing appears to lead to advantageous physiological and psychological effects through autonomic nervous system activity. Grossman (1983) concluded:
A slow, large-tidal-volume, predominantly abdominal pattern of ventilation . . . is associated at the psychological level with emotional stability, sense of control over the environment, calmness, a high level of physical and mental activity, and relative absence of perceived or objective stressors. Short-term modification of breathing pattern toward this type seems to cause a reduction of subjective and physiological indices of anxiety under conditions of stress; long-term modification seems to produce—with certain clinical populations—a diminution of psychological difficulties, e.g. neurotic tendencies, chronic anxiety responses, and psychosomatic symptoms. Cardiovascularly, this breathing pattern appears independently to produce relatively high [parasympathetic] tone and low sympathetic activation, which manifest as low heart rate, increased supply of blood and oxygen to the heart and the brain, and enhanced [respiratory sinus arrythmia] and baroreceptor responsiveness. (p. 292)
Of particular relevance, Grossman noted that the four studies investigating the effect of paced slow respiration in stressful situations "uniformly indicate that mere voluntary changes of respiration rate by subjects under stressful circumstances serve to modify the subjective perception of anxiety" (p. 292). A more recent study by Cappo and Holmes (1984) also supports this conclusion. Several studies have also found paced slow respiration reduces autonomic reactivity as measured by skin resistance (but not heart rate) (Cappo & Holmes, 1984; Harris, Katkin, Lick, & Habberfield, 1976; McCaul, Solomon, & Holmes, 1979).
Similarly, slow diaphragmatic breathing has consistently proven successful therapy for persons with hyperventilation stress responses that reached clinical severity (Bonn, Readhead, & Timmons, 1984; Clark, Salkovskis & Chalkley, 1985; Grossman, de Swart, & Defares, 1985; Hegel, Abel, Etscheidt, Cohen-Cole, & Wilmer, 1989; Hibbert & Chan, 1989; Kraft & Hoogduin, 1984; Lum, 1976). Grossman (1983) cites various studies suggesting that slow breathing causes blood carbon dioxide concentrations to be in the upper normal range, which promotes psychophysiological effects generally opposite to those of hyperventilation.
Grossman (1983) notes that respiratory sinus arrythmia (increased heart rate during inspiration) is a useful index of parasympathetic tone and is largest during slow deep breathing. He also cites evidence suggesting that normal parasympathetic tone promotes good health and may serve a protective function for the heart, whereas decreased parasympathetic tone may be related to heart disorders. He further suggests that the relative balance between the parasympathetic and sympathetic nervous systems may be important in determining responses to stress. More recent studies support the hypothesis that parasympathetic dominance has protective value for the cardiovascular system (Beere, Glagov, & Zarins, 1984; Jennings & Follansbee, 1985; Muranaka, et al., 1988).
Diaphragmatic breathing has traditionally been considered the most efficient mode of quiet breathing (e.g., Miller, 1954; Sharp et al., 1974). Because tidal volume is typically larger in diaphragmatic breathing, the proportion of ventilation wasted as dead space is minimized. In addition, enhanced ventilation to the lower lungs increases efficiency of gas exchange because gravitational forces cause much higher blood flow in the lower lungs (West, 1988). Diaphragmatic-abdominal breathing can cause higher air flow to the lower lungs than thoracic breathing (Fixley, Roussos, Murphy, Martin, & Engel, 1978; Roussos et al., 1977; Sampson & Smaldone, 1984); however, this effect was not found in other studies (Bake, Fugl-Meyer, & Grimby, 1972; Grassio, Bake, Martin & Anthonisen, 1975; Grimby, Oxhoj, & Bake, 1975; Sackner, Silva, Banks, Watson, & Smoak, 1974) and apparently depends on details of respiratory muscle action and perhaps experimental methodology (see, Roussos et al., 1977).
Pressure on the abdominal viscera from diaphragmatic motion also contributes to venous blood return to the heart (Grossman, 1983; Permutt & Wise, 1986), which is an important determinant of cardiac output and efficiency (Guyton, 1986).
Diaphragmatic breathing has historically been recommended for persons with chronic obstructive lung disease (Barach, 1955; Frownfelter, 1987; Miller, 1954). However, efforts to quantify the benefits have given mixed results (Jones, 1974; Rochester & Goldberg, 1980). A detailed review of the literature is needed, but is outside the scope of the present paper. Potential psychophysiological and psychological benefits should be considered in addition to the usual measures of lung function.
Nasal breathing is the best means of warming and humidifying inhaled air in preparation for the lungs. Available information on the function and evolution of the human nose is consistent with a primal purpose of conserving moisture and heat (Cole, 1988; Franciscus & Trinkaus, 1988). In a temperate climate, the estimated energy expenditure to condition inhaled air can be equivalent to about one sixth of a person's daily energy output; however, about 30 to 40 percent of this energy is recovered by exhaling through the nose (Cole, 1982, 1988). Higher efficiencies of heat and moisture recovery occur in cold and/or dry environments.
The nose also filters incoming air (Guyton, 1986, p. 477), has irritant receptors that trigger protective reflexes (Widdicombe, 1986), and, of course, provides the sense of smell. The resistance to air flow in nasal breathing may be an efficient passive means of slowing air flow to provide adequate gas exchange at low ventilation rates (Hairfield, Warren, Hinton, & Seaton, 1987; Jackson, 1976; McCaffrey and Kern, 1979a). Nasal breathing is the normal and preferred mode of quiet respiration.
The intriguing hypothesis that nasal respiration plays an important role in controlling brain temperature may have important implications for brain functioning and psychological states (Dean, 1988; Zajonc, Murphy, & Inglehart, 1989). However, the basic mechanisms and effects of brain cooling have not yet been resolved (Wheeler, 1990).
Role of abdominal muscles. Both the yoga and scientific literatures have focused on comparing diaphragmatic-abdominal breathing with thoracic breathing, but have little discussion of the specific role of the abdominal muscles. The abdominal muscles may shift the expiratory end volume, alter the rib cage shape, or play no role. Differing use of the abdominal muscles may be a factor in the inconsistent replications of certain respiratory findings. One yoga master recommends that abdominal muscles not be used once diaphragmatic breathing is established (Samskrti & Veda, 1985 p. 10). Research may be of particular value on the following topics:
1. Abdominal breathing may help stretch and relax the diaphragm in persons who manifest stress by excessive tonic diaphragm contraction. Some individuals have increasing contraction and immobilization of the diaphragm as stressful topics are discussed (Faulkner, 1941; Holmes, Goodell, Wolf, & Wolff, 1950, p. 49; Wolf, 1947). The prevalence of excessive tonic diaphragm tension, both acute and chronic, and the effects of abdominal pressure on the diaphragm in these cases may merit further investigation.
2. Diaphragmatic and abdominal breathing cause rhythmic pressure on and movement of the abdominal organs, which could affect the functioning of those organs. In fact, a yoga breathing exercise of pulling in the abdominal muscles during exhalation is claimed to create perfect digestion (Rama, 1988, pp. 191-192). Digestion and slow diaphragmatic breathing are both associated with parasympathetic activity and therefore may be both autonomically and mechanically coupled. Similarly, the stress response of thoracic breathing with a relatively inactive diaphragm may provide minimal mechanical stimulation of the abdominal organs and appears consistent with reduced gastrointestinal activity during sympathetic arousal and anticipated physical activity. The potential interaction between the gastrointestinal system and the respiratory system deserves investigation, particularly with regard to the effects of psychological factors such as stress.
3. The use of abdominal muscles to drive end expiration below relaxed expiratory position (FRC) may lead to less efficient gas exchange and to lower cardiac efficiency, particularly in older persons and persons with lung impairment. The small airways in the lower lung tend to close with exhalation below FRC (Collet, Roussos, & Macklem, 1988). These airways reopen only when pressure is sufficient to overcome surface tension. Until inspiration exceeds the needed pressure, air is distributed to the upper lung, resulting in inefficient gas exchange. For normal young people some airways are closed at residual capacity (maximum possible exhalation), but most are open. With age, lower airway closure increases and may occur with normal exhalation (i.e., at FRC). In addition, respiratory actions that increase plural pressure (pressure in the thoracic cavity surrounding the heart and lungs) tend to decrease venous return to the heart (Permutt & Wise, 1986) and thus reduce cardiac efficiency. Exhalation below FRC increases plural pressure (Collett, Roussos, & Macklem, 1988) and, therefore, may reduce cardiac efficiency.
Improved experimental controls. Studies on the effects of breathing mode have rarely considered (a) the subjects' responses to the experimental procedure, and (b) individual differences in pattern of breathing and chronic stress level. Troyer and Loring (1986, p. 473) note that normal subjects are well known to adopt a more thoracic breathing mode during respiration experiments. This result is not surprising in light of the evidence that anxiety leads to a tendency for thoracic breathing. Likewise, the variation in the tendency to hyperventilate suggests that individual differences are very important factors. Studies that find thoracic breathing prevalent in a quiet breathing condition (e.g., Sharp, Goldberg, Druz, & Danon, 1975) raise questions about the effects of the experimental procedure and subject pool. Careful attention should be given to subject pool and the subjects' reactions to experimental procedures and personnel.