Subjects
The subjects were 35 male volunteers, with ages ranging from 20 and 46 years (M = 27.5, SD = 4.7 yr.) and with an average of 30.2 months (SD = 25.7) of experience with yoga practice. All the subjects had completed a residential course in yoga (i.e., for five and a half months) at the residential yoga center. Subsequently, they were staying on at the center for varying durations to participate in other activities and to carry on practicing yoga, which included both shavasana and "guided relaxation". As it was not possible to record autonomic variables and oxygen consumption simultaneously, with the equipment used by us, these two assessments were made in separate sessions. For some subjects this number of sessions was too cumbersome, hence one assessment (i.e., autonomic variables or oxygen consumption) was made. Hence the autonomic variables were recorded in 15 subjects before, during, and after Guided relaxation and similarly for Supine Rest; in 25 subjects, the oxygen consumption was recorded before and after Guided Relaxation and before and after Supine Rest. There were 5 subjects who had recordings of both autonomic variables and oxygen consumption.

Design
Subjects were studied in two separate relaxation sessions, viz., guided relaxation and supine rest. The two sessions were on different days at the same time of the day. For half the subjects, alternately, the guided relaxation session was on the first day with the supine rest session the next time. The order was reversed for the remaining subjects. Each session took 20 minutes and consisted of 3 periods, viz., Before (5 minutes), Test (10 minutes), and After (5 minutes). For analysis, the Test period was divided as two, During 1 and During 2, each of 5 minutes. The subjects were sitting at ease before and after the Test periods, and supine during the Test periods This change in posture (i.e., sitting before and after, compared with supine, during), was chosen as the supine posture is one the positions chosen for relaxation by yoga practitioners, and being in the supine position would hence have been similar to supine rest.

Assessment
The oxygen consumption was recorded with a closed circuit Benedict-Roth apparatus (INCO, Ambala, India) using the standard method (Mountcastle, 1980). The subject breathed into an oxygen tank from which exhaled carbon dioxide was excluded by absorption in sodium hydroxide. The subjects were asked to breathe into the mask, which covered their nose and mouth. Recordings were made before and after, but not during test periods.

A 4-channel polygraph (Medicaid Systems, Chandigarh, India) was used to record the electrocardiogram (EKG), respiration, finger plethysmogram and skin conductance level. EKG was recorded using standard limb lead I configuration. The EKG was digitized using a 12 bit analog-to-digital convertor (ADC) at a sampling rate of 500 Hz. The data recorded were visually inspected off-line and only noise free data were included for analysis (Raghuraj, Ramakrishnan, Nagendra, & Telles, 1998). The R waves were "identified" and used to obtain a point event series of successive R-R intervals, from which the beat to beat heart rate series was computed.

Skin conductance level was recorded using Ag/AgCl disk electrodes with electrode gel, placed in contact with the volar surfaces of the distal phalanges of the index and middle fingers of the left hand. A low-level DC preamplifier was used and a constant voltage of 0.5V was passed between the electrodes. Respiration was recorded using a nasal thermistor attached to the more patent nostril. Finger plethysmogram was recorded placing the photoplethysmograph on the volar surface of the distal phalanx of the index finger of the right hand.

Guided relaxation technique
The guided relaxation technique lasts for 10 minutes and is done in 5 phases of step-wise relaxation, detailed below (Nagendra & Nagarathna, 1988). (i) Relaxing from the tip of the toes to the waist, mentioning each part of the body specifically, followed by chanting the syllable "A", the first part of the syllable "Om" (Chinmayananda, 1984), (ii) relaxing from the waist to the neck, followed by chanting the syllable "U", the middle part of "Om". (iii) Relaxing the head and neck, followed by chanting "M", the last part of "Om". (iv) Letting the body `collapse" on the ground with a feeling of "letting go", chanting AUM. (v) Allowing oneself to feel apart from one"s physical body with a feeling of expansion by merging with a limitless expanse such as the sky or ocean. This is practiced slowly and with instructions about relaxation and awareness of physical and mental sensations. Throughout the practice the eyes are closed.

Supine rest
During supine rest, the subject lies supine with the legs apart, arms away from the sides of the body, and eyes closed. This session lasts 10 minutes, as for guided relaxation. In traditional yoga practice, all yoga postures are meant to be practiced with relaxation (i.e., not contracting muscles during the posture) and with a mental state which is calm and "expanded". To reach the latter state, the subject initially visualizes a limitless expanse such as the sky or the ocean. While this is true for all yoga postures (e.g., those practiced while standing or seated), it is especially true for shavasana (the supine rest posture). Hence subjects could be expected to anticipate feeling relaxed in this posture.

Data Extraction
The end expiratory points of the respirogram obtained from Benedict-Roth apparatus were joined as a slanting line, the slope of which gave the difference between initial and final volumes of oxygen in the tank in a given period, which was approximately 3-4 minutes in most cases.

The following data were extracted from the polygraph records: the breath rate (in cycles per minute) was calculated by counting the breath cycles in 60 second epochs, continuously. The skin conductance level (in micro siemens) and finger plethysmogram amplitude (in mm) were sampled at 20 second intervals. Values averaged across each of the periods (before, test, after) of a session, were used for analysis.

Frequency domain analysis of heart rate variability data was carried out for the 5-minute recordings Before, During the Test period (During 1 and 2) and After the sessions. The mean heart rate was obtained from this record. The mean values were removed from the heart rate series to obtain the heart rate variability values. The heart rate variability power spectrum was obtained using Fast Fourier Transform. The power in heart rate variability series in the following specific frequency bands was studied, viz., the very low frequency band (0 - 0.05 Hz), low frequency (LF) band (0.05 - 0.15 Hz), and high frequency (HF) band (0.15 - 0.50 Hz). The low frequency and high frequency values were expressed as normalized units, which represent the relative value of each power component in proportion to the total power minus VLF component (LF norm = LF/(total power-VLF) X100; HF norm = HF/(total power-VLF) X100) (Task force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, 1996).

ANALYSIS
The t-test for paired data was used to assess the significance between Before values and those recorded During (During 1, During 2) and After. In the case of oxygen consumption and breath volume, comparisons were between Before and After values. The Wilcoxon paired signed ranks test was used to study changes in LF/HF, for subjects with baseline LF/HF >0.5 (Group I) and for those with baseline LF/HF £ 0.5 (Group II), separately