Home Links Articles & Archives What is Ayurveda? What is Yoga therapy? What is Yoga? History Publications Teaching Schedule Blog Facebook Bio

Biofeedback Research and Yoga

Scientific studies of meditation and other forms of contemplative experience have only recently become a subject of scientific interest within the last half century. In 1931 Kovoor Behanan, an Indian graduate student in psychology at Yale, was awarded a Sterling Fellowship to undertake what has since been recognized as the first empirical study of yoga and meditation. Supported in this research by Walter Miles, an eminent professor of psychology, Behanan wrote a book about yoga that described quantitative studies of his own yogic breathing. During 72 days of experiments at Yale, he found that one breathing exercise, or pranayama, increased his oxygen consumption by 24.5%, a second by 18.5%, and a third by 12% (Behanan, 1937, Miles, 1964). This study helped stimulate interest in meditation research by showing that the physiological effects of yoga could be examined in the laboratory (Behanan, 1937). Unlike many tales by travelers to the East, Behanan's straightforward, well-observed account of his laboratory research was free of exaggeration and mystification.

Behanan also studied Indian yogis. He was guided in this work by Swami Kuvalayananda, who promoted yoga research at a center for meditation practice he founded in the 1920s at Lonavla, a hill station near Bombay. Kuvalayananda developed a system of physical culture that included asanas and pranayamas, and he established a yogic therapy for many afflictions. His work was supported by several Indian states, two provincial governments of British India, Indian health agencies, and American foundations. For many years, the results of his laboratory research were published in a quarterly journal, Yoga Mimamsa, which also provided instruction on postures, breathing exercises, and other disciplines. Many people interested in yoga research visited Lonavla, among them psychologists Basu Bagchi of the University of Michigan Medical Center and M. A. Wenger of UCLA, who gave new impetus to meditation studies in the 1950s. From the 1920s into the 1960s, Swami Kuvalayananda did much to promote the scientific study of yoga.

In 1935 a French cardiologist, Therese Brosse, took an electrocardiograph to India and studied yogis who said they could stop their heart. According to Brosse's published report, readings produced by a single EKG lead and pulse recordings indicated that the heart potentials and pulse of one of her subjects decreased almost to zero, where they stayed for several seconds (Brosse, 1946). Her finding was criticized, though, by Wenger, Bagchi, and B. K. Anand in their later, more thorough studies of yogic adepts (see below). Brosse also studied a yogi who was buried for ten hours, and described other examples of self-control she had witnessed. Like Behanan and Swami Kuvalayananda, she helped promote the idea that yogic feats could be studied with scientific instruments.

The instrumented study of yogic functioning was expanded by Bagchi, Wenger, and Anand. Anand was then chairman of the Department of Physiology at the All-India Institute of Medical Sciences in Delhi. Their landmark studies during the late 1950s were reported in American scientific journals. Along with studies of Zen masters by Akira Kasamatsu and Tomio Hirai in Japan (see below) the Indian studies gave new momentum to meditation research. For five months in 1957, Bagchi and Wenger traveled through India with an eight-channel electro-encephalograph and accessory instruments to record respiration, skin temperature, skin conductance, and finger blood-volume changes. During their trip they established experiments in Calcutta, Madras, Lonavla, and New Delhi, and conducted further tests in homes and a mountain retreat (Bagchi and Wenger, 1957; Wenger and Bagchi, 1961; Wenger et al., 1961; Bagchi, 1969). Among the subjects they examined, one could perspire from his forehead upon command in his freezing Himalayan retreat; a second could regurgitate at will to cleanse himself (Wenger & Bagchi, 1961). Three others altered their heartbeats so that they could not be heard with a stethoscope, though EKG and plethysmographic records showed that their hearts were active and their pulses had not disappeared. [28] In tests to compare relaxation in a supine position with seated meditation, Bagchi and Wenger found that four yoga students had faster heart rates, lower finger temperatures, greater palmar sweating, and higher blood pressure during meditation, though their respiration rates were reduced. Five yogis given similar tests exhibited even faster heart rates, lower finger temperatures, greater palmar conductance, and higher blood pressures during meditation than the students, though their breathing was slower. Such differences suggested that for these yogis meditation was an active rather than a passive process (Wenger and Bagchi, 1961).

Bagchi and Wenger also studied the effects of breathing exercises and found that some of their subjects, especially experienced ones, could produce bidirectional changes in every autonomic variable that the experimenters measured. Though the two psychologists found that their subjects exhibited some dramatic physiological changes, they were cautious in drawing conclusions about yogic claims in general. "Direct voluntary control of autonomic functions is probably rare among yogis," they wrote. "When such control is claimed, intervening voluntary mechanisms are usually employed." They made this qualification, however: "We have met many dedicated yogis who described experiences to us that few Western scientists have heard of and none has investigated. It is possible that the mere presence of a foreigner precludes optimum results"(Wenger and Bagchi, 1961).

Other researchers have confirmed the discovery by Bagchi and Wenger that some subjects exhibit more than one pattern of physiological activity during their yogic practices. N. N. Das and H. Gastaut studied seven Indian yogis, who registered no muscular electrical activity during periods of complete immobility though their heart rates accelerated in almost perfect parallel with accelerations of their brain waves during moments of ecstasy. The most accomplished among these seven subjects, moreover, exhibited "progressive and very spectacular modifications" in their EEG records during their deepest meditations, including recurrent beta rhythms of 18-20 cycles per second in the Rolandic area of the brain, a generalized fast activity of small amplitude as high as 40-45 cycles per second with occasional amplitudes reaching 30 to 50 microvolts, and the reappearance of slower alpha waves after samadhi, or ecstasy, ended. In summarizing their study, Das and Gastaut concluded that:

The modifications [we] recorded during very deep meditation are much more dramatic than those known up till now, which leads us to suppose that western subjects are far from being able to attain the yogi state of mental concentration.

It is probable that this supreme concentration of attention . . . is responsible for the perfect insensibility of the yogi during samadhi; this insensibility, accompanied by immobility and pallor often led people to describe this state as sleep, lethargy, anesthesia, or coma. The electroencephalographic evidence here described contradicts such opinions and suggests that a state of intense generalized cortical stimulation is sufficient to explain such states without having to invoke associated processes of diffuse or local inhibition (Das and Gastaut, 1955)

Das and Gastaut's conclusion does not contradict the widespread findings of subsequent meditation studies that many or most meditators experience the trophotropic or relaxation response described by E. Gellhorn, W. Kiely, Herbert Benson, and other researchers (Gellhorn and Kiely, 1972; and Benson, 1975). Most subjects in meditation studies do not experience yogic ecstasy and so do not exhibit the cortical excitement that Das and Gastaut observed. Furthermore, different kinds of religious practice produce different types of experience accompanied by different types of physiological change. Kasamatsu and Hirai's Zen masters, for example, exhibited high-amplitude alpha and theta waves, not beta waves, during their deepest meditations (see below).

Further evidence that contemplative practice produces different physiological profiles was provided by B. K. Anand, G. S. Chhina, and Baldev Singh, who found that four yogis exhibited persistent alpha activity with increased amplitude during trance. These four yogis exhibited no alpha-wave blocking when they were bombarded with loud banging, strong lights, and other sensory stimuli, and two of them showed persistent alpha activity while holding their hands in ice-cold water for forty-five to fifty-five minutes (Anand, Chhina, and Singh, 1961a). The yogis in this experiment exhibited physiological differences during meditation from at least two other groups of accomplished meditators. They did not exhibit alpha blocking in response to strong stimuli, in contrast to the Zen masters studied by Kasamatsu and Hirai (see below). Nor did they exhibit the beta waves that appeared on the EEGs of Das and Gastaut's subjects. The difference from the Zen masters probably resulted from a basic difference in focus between the two groups, the yogis having withdrawn their attention from external stimuli, whereas the Zen masters remained aware of their external environment. Their difference from Das and Gastaut's yogis, on the other hand, might have been due to differences between their styles of meditation, the conditions of the experiments, or the qualities of their experience. The strong stimuli Anand gave his subjects, for example, may well have prevented the more ecstatic absorptions experienced by Das and Gastaut's yogis. The published reports of the Das-Gastaut and Anand-Chhina-Singh experiments do not provide enough detail to fully explain their different results, but they remind us that there are different kinds of contemplative experience. Roland Fischer, Julian Davidson, and other researchers have proposed some ways in which internal states might be correlated with different physiological profiles (Fischer, 1971; and Davidson, 1976).

In a study published in 1958, the Indian researchers G. G. Satyanarayanamurthi and B. P. Shastry described a yogi whose heart kept beating for thirty seconds even though his radial pulse could not be felt and his heart could not be heard with a stethoscope. This yogi's EKG showed no abnormalities, moreover, and finger plethysmography showed that his pulse was present though greatly reduced. The two researchers claimed that fluoroscopy conducted while the yogi was lying down showed that for several 30-second periods the beating of his heart was just a "flicker along the left border below the pulmonary conus and in the apical segment of the left ventrical." They concluded that he achieved this control through the Valsalva maneuver. [29]

Elmer and Alyce Green, with their colleagues at the Menninger Foundation in Topeka, Kansas, also observed exhibitions of yogic heart control. Their subject, Swami Rama, while sitting perfectly still, produced an atrial flutter of 306 beats per minute that lasted for sixteen seconds. During a fibrillation of this kind, a section of the heart oscillates rapidly while its chambers do not fill and its valves do not work properly, but Swami Rama gave no sign that the maneuver caused him any pain or heart damage. The swami also produced an IIF difference between the left and right sides of his right palm. While he did this, the left side of his palm turned pink and the right side gray (Green and Green, 1977).

Yogis frequently use abdominal contractions to slow their heart rate rather than intervening more directly through the central nervous system. Curiously, though, an earlier study had examined a man with no yogic training at all who could stop his heart without such maneuvers, simply by relaxing and "allowing everything to stop." By this procedure, he could induce a gradual slowing of his pulse until he started to faint, at which point he would take a deep breath. When EKG tests showed that his heartbeat did indeed disappear, the doctor who examined him concluded that the man's cardiac arrest was induced through some mechanism which, although under voluntary control, is not known to the patient himself. Careful observation did not reveal any breath-holding or Valsalva maneuver. Apparently the patient simply abolished all sympathetic tone by complete mental and physical relaxation (McClure, 1959).

Like heart stopping, the live burial of yogis has excited the interest of several researchers. A physician, Rustom Jal Vakil, published an account in the British journal Lancet of such a confinement that was witnessed by some 10,000 people near Bombay in February 1950. According to Vakil, an emaciated sadhu named Ramdasji sat cross-legged in a subterranean 216-cubic-foot cubicle and remained there for sixty-two hours. His pulse remained steady at eighty beats per minute; his blood pressure was 112/78; and his respiratory rate fluctuated from eight to ten breaths per minute. Though he had some scratches and cuts, Vakil wrote, Ramdasji appeared "none the worse for his grueling experience.'' (Vakil, 1950).

In June 1956, a more closely observed study of yogic confinement was conducted under the auspices of the All-India Institute of Mental Health in Bangalore with a Hatha yogi, Krishna Iyengar. Hoenig, a psychiatrist from the University of Manchester, witnessed the experiment and described it in a review of yoga research published in 1968 (Hoenig, 1968). According to Hoenig's report, a pit some two by three by four feet was dug on the institute's grounds and covered with wire meshing, a rubber sheet, and cotton carpet. An electrode junction box connected to an EEG and an EKG was placed in the pit along with instruments to measure temperature and concentration of gas. The yogi was confined for nine hours. When he was released he immediately walked about the grounds, according to Hoenig's firsthand account, and demonstrated athletic feats including a headstand with his legs in the lotus position. The percentage of carbon dioxide in the air in his enclosure, which was 1.34% at the beginning of the experiment, was only 3.8% at the end, lower than would normally be expected. Iyengar's heart rate gradually slowed from 100 to 40 beats a minute in recurring twenty- to twenty-five-minute cycles, but his EKG record did not register any other abnormality and the cycles did not coincide with his breathing or brain-wave patterns. The yogi's EEG showed a normal waking record for the full nine hours, characterized by a stable alpha rhythm of 50 microvolts with no evidence of sleep or interference caused by physical movement. From these records, the experimenters concluded that their subject lay motionless and wide awake, without the active cognition that would have reduced or eliminated his alpha rhythm. Iyengar said he had maintained the shavasana, or corpse pose, using ujjaya breathing while remembering the names of God. He was surprised that his heart had speeded and slowed, and could not explain why it had done so. It beat normally, however, after the experiment.

Because the earthen pits used in most yogic confinements leak oxygen and carbon dioxide, Anand, Chhina, and Singh tested a yogi named Ramanand in an airtight glass and metal box, once for eight hours and again for ten hours. The yogi's average oxygen use during the first experiment decreased from the basal rate of 19.5 liters per hour to 12.2, and during the second experiment to 13.3 liters per hour. His carbon dioxide output went down during both experiments. Ramanand, moreover, did not exhibit any rapid breathing or speeded heart rate as the oxygen in his box diminished and carbon dioxide increased. "Sri Ramanand Yogi could reduce his oxygen intake and carbon dioxide output to levels significantly lower than his requirements under basal conditions," Anand and his colleagues wrote. "It appears from this study that [he] could voluntarily reduce his basal metabolic rate on both occasions he went into the box." [30]

During a remarkable experiment reported by L. K. Kothari and associates, a yogi was buried for eight days in an earthen pit and connected by leads to an EKG in a nearby laboratory. After the pit was boarded up, the subject's heart rate sometimes went as high as fifty beats per minute, until a straight line appeared on the EKG tracing when the yogi had been in the pit for twenty-nine hours. There had been no slowing of his heart immediately before the straight line appeared, nor any sign of electrical disturbance, but the experimenters proceeded with certainty that their subject had not died. Suspecting that their EKG leads had been deliberately or accidentally disconnected, they checked their machine and continued to monitor its tracings. To their astonishment, it started to register electrical activity some seven days later, about a half hour before the yogi's scheduled disinterment. "After some initial disturbance," they wrote, "a normal configuration appeared. The [speeded heart rate] was again there but there was no other abnormality." When the pit was opened, the yogi was found sitting in the same posture he had started in, but in a stuporous condition. In accounting for his remarkable EKG record, the experimenters argued that a disconnection of the EKG lead would have produced obvious markings on the tracings in their laboratory, as they found when they tried to simulate ways in which the yogi might have tinkered with it.

Furthermore, the yogi was ignorant about such machines, and the pit was completely dark. If the machine had malfunctioned in some way they could not ascertain, it seemed an extraordinary coincidence that it started again just a half hour before their subject's scheduled release. Apparently, the yogi was operating with some kind of internal clock that did not depend upon the daily cycles of light and darkness, for the most likely cause of the straight line on his EKG tracing was a dramatic decrease in the activity of his heart. Kothari and his colleagues finally could not account for this remarkable cardiac record (Kothari et al., 1973).

Studies of Zen Buddhist Monks

In a study that attracted much attention among meditation and biofeedback researchers during the 1960s, Akira Kasamatsu and Tomio Hirai, physicians at the University of Tokyo, studied the EEG changes exhibited during meditation by Zen teachers and their disciples (forty-eight in all) from Soto and Rinzai centers in Japan. For experimental control, they studied the EEGs of twenty-two subjects with no experience at meditation. They made EEG recordings; recorded their subjects' pulse rates, respiration, and galvanic skin response; and tested their responses to sensory stimuli during meditation. The recordings on the Zen monks were made during a weeklong retreat, or sesshin, at a Zendo, except for a few tests at the experimenters' laboratory. The Zen teachers and their most experienced students exhibited a typical progression of brain-wave activity during meditation, which Kasamatsu and Hirai divided into four stages:

  • Stage 1: Characterized by the appearance of alpha waves in spite of opened eyes.
  • Stage 2: Characterized by an increase in amplitude of persistent alpha waves.
  • Stage 3: Characterized by a decrease in alpha frequency.
  • Stage 4: Characterized by the appearance of rhythmical theta trains (Kasamatsu and Hirai, 1966).
Not all four stages were evident in every Zen practitioner, nor in any of the controls, but a strong correlation existed between the number of stages a given student exhibited and that student's length of time in Zen training. This correlation was supported by a Zen teacher's evaluation of each student's proficiency. The teacher ranked the students in three levels, without seeing their EEG records, and his rankings correlated well with Kasamatsu and Hirai's assessment of their EEGs.

The Kasamatsu-Hirai study also revealed significant differences between four Zen masters and four control subjects in their response to repetitive click stimuli. Like the Zen masters, the controls exhibited a blocking of alpha when a click sound first occurred, but they gradually became habituated to such stimuli so that their brain-wave activity no longer responded when a click was made. The Zen masters, however, did not become habituated, but continued to exhibit blocking as long as the stimuli continued. This finding indicates that Zen practice promotes a serene, alert awareness that is consistently responsive to both external and internal stimuli (Kasamatsu et al., 1957; Hirai, 1960; and Kasamatsu and Hirai, 1963).

Difficulties of Research with Religious Adepts

Though people testified under oath before the Congregation of Rites that they had seen Saint Teresa of Avila or Saint Joseph of Cupertino defy gravity, no scientific studies have recorded instances of levitation. There are at least three possible reasons for this lack of evidence. First, of course, it might be that levitation has never happened. Second, the contemplative traditions might have lost their power to evoke the phenomenon. Third, levitation might only occur during rare and spontaneous ecstasies that cannot be programmed to meet the requirements of a scientific experiment. Superordinary lifting from the ground, if it in fact occurs, would require an improbable set of circumstances which a scientist would be lucky to witness. Levitation, like other holy powers, would have to be caught "in the wild." In a laboratory, with wires attached to his head and a thermometer up his rectum, a yogi or lama is unlikely to exhibit a capacity that is rare in any case. In studies of extraordinary functioning there is a trade-off between robust results and scientific precision. Uninhibited by recording machines and safety rules, for example, the Maharaja Runjeet Singh could bury Haridas for forty days. More recent studies of yogic confinement, however, have been constrained by procedural controls and humane considerations.

Furthermore, there is often a disjunction between a scientist's attitude toward exceptional powers and an adept's ideas about them. Elmer Green, for example, described differences he had with the healer Jack Schwartz in interpreting Schwartz's intuitive diagnosis of illness. According to Schwartz, the question was:

Are the auras one sees always radiatory patterns of energy from the human body . . . or are they automatic mental projections of one kind or another that are used psychologically to interpret a "knowing"? Sometimes when we "know" something in this way we tend to "see" it in the same way that we see a memory (Green and Green, 1977, p. 240).

Green was sympathetic to Schwartz, however, realizing that a scientist's constant doubt can inhibit or destroy a psychic's intuitions. This fundamental difference between scientists and psychics, Green wrote:

Need not cause problems if each takes time to understand the framework in which the other necessarily operates. If the psychic tries to pull apart every perception in order to find out if it is incorrect, so as to better determine the "truth," what is most likely to be pulled apart is the faculty of "seeing." The talent for perceiving might well fade away. On the other hand, if scientists stopped trying to find alternate explanations for the facts, they might get lost in a maze of [incoherent] ideas. For both scientists and mystics, however, the area of facts rather than interpretations is common ground. Excluding the opinions of fanatics, most of the arguments that we are aware of between the two camps have revolved around interpretations. Because psychics almost always have idiosyncratic factors in their frames of reference, scientists often do not understand them. And psychics do not understand what seems to them to be a destructive attitude on the part of scientists (Green and Green, 1977, p. 242).

Sympathy between scientists and adepts was evident in Swami Kuvalayananda's projects noted above, and in other experimenter-subject teams described in the preceding pages. Even the stern mutual challenge between Haridas and Maharaja Runjeet Singh exhibited an exemplary, if somewhat perverse, cooperation. Productive study of extraordinary functioning requires understanding between accomplished subjects and imaginative experimenters.

Contemporary Meditation Research

Meditation research increased dramatically during the 1970s and 1980s, particularly in the United States. This burgeoning effort was stimulated in part by the studies of yogis and Zen masters noted in the previous section, and in part by the publication of landmark studies by Herbert Benson and Keith Wallace in Science, the American Journal of Physiology, and Scientific American between 1970 and 1972 (Wallace, 1970; Wallace et al., 1971b; Wallace and Benson, 1972). The Transcendental Meditation Society supported much of this work, though its enthusiastic claims and advertising efforts caused doubts among some researchers about the highly favorable outcomes in studies it sponsored (Shapiro, 1982). These doubts led to further research, which has either contradicted, tempered, or confirmed the TM-sponsored claims. Since the early 1970s, more than a thousand studies of meditation have been reported in English-language journals, books, and graduate theses. The range of outcomes included in this research has grown considerably since the studies of yogis and Zen masters by Bagchi, Wenger, Kasamatsu, and Hirai. Cardiovascular, cortical, hormonal, and metabolic changes, several behavioral effects, and alterations of consciousness resulting from meditation have been explored in recent years. The medical instrumentation, psychological tests, and methods of analysis used in such experiments have been improved, and the range of subject populations has been enlarged to include different kinds of subject groups. This growth in sophistication of method is gradually improving our scientific understanding of meditation in ways that complement the insights contained in the traditional contemplative literature. However, the overall picture of results on the subject of meditation produced by modern research remains uneven. Some effects have appeared consistently, but others have not.

The apparent inconsistencies defining the effects of meditation can be accounted for in various ways. Some physiological processes, perhaps, are unaffected by meditation, no matter how proficient or experienced the meditator might be; or perhaps they are affected to an insignificant degree. For some changes, such as amino acid concentrations in the blood, there has not been enough research to establish a consistent picture, partly because there has not been as much interest in these variables as in the effect of meditation on blood pressure, heart rate, and other indices that have an obvious bearing on health. Taking blood samples during meditation, moreover, is harder to accomplish than recording blood pressure or skin responses.

Individual differences also present a special problem for understanding the results of meditation studies, because subject populations have included people of both sexes, all ages, various levels of education, and different kinds of social background. Many subjects have been college students with no previous experience at meditating; others have been recent converts to religious groups; but only a few have been highly skilled in spiritual practice. The incentives to concentrate during experimental sessions have also varied. Some subjects have wanted success for religious or other reasons, while others seem not to have been well motivated. And differences between meditation styles also complicate the results of such research. Though most studies have used some type of quiet concentration, some have used active methods such as rapid breathing. Julian Davidson, Roland Fischer, and others have distinguished between two classes of meditation, those that relax and those that excite, associating their effects with the trophotropic and ergotropic conditions of the central nervous system modeled by Gellhorn and Kiely (Davidson, 1976; Fischer, 1971, 1976; Gellhorn and Kiely, 1972).

The results of scientific research on the subject of meditation are accumulating now, forming a publicly accessible body of empirical data that can serve generations to come. Unfortunately, however, these data are derived mainly from beginning practitioners of meditation, and taken as a whole do not reflect the richness of experience described in traditional contemplative teachings. They are also limited by the conventional scientific insistence that results be repeatable. Certain important experiences occur only rarely in meditation, and a science that disregards them loses important empirical results. For these reasons, contemporary research does not illumine the full range of experience described in the contemplative scriptures and the oral traditions from which they come. Modern studies give us only a first picture of the foothills, with a few glimpses of the peaks. Still, what they give us corresponds in several ways with traditional accounts.

Back to articles

Contact about this site