DISCUSSION

This type of a detailed and differential study has not been reported previously on the subject of changes of oxygen consumption during the subtypes of kumbhak-pranayamic breathing. Miles1 reported data using the oxygen estimation method similar to the one used in the present research, on a single subject who practised a slow breathing, kumbhak- pranayama (Ujjayi type, but time ratios not specified), and also two non-kumbhak types of rapidly (80/min) breathing varieties of pranayamas, namely, the kapalabhati and bhastrika. In the single subject study, all the three types of Pranayama caused an increase in oxygen consumption varying from +12 to +35 per cent. Rao2 also reported data on one subject, showing a 7-9 per cent increase in oxygen consumption during Ujjayi type of pranayama having time ratios of 1:4:2. The present results obtained from larger number of subjects differed from those of single subject studies1,2, since the data revealed for the first time that there could be effects of not only increase but also decrease in oxygen consumption in different sub- types of the Ujjayi pranayamas. Hence our study points out the need for a change in the perspective of understanding of effects of pranayamas.

The present study offers an indication, on the basis of the data of several test sessions conducted on a larger number of subjects, that the Ujjayi types of pranayamic practice could cause either an increase or a decrease in the oxygen consumption, depending on variations in the duration of kumbhak or the ratios of durations of different phases of the respiratory cycle.

Although the oxygen consumption was altered during both the pranayamic types, it returned after the Pranayama to pre-pranayamic levels. There was no evidence of increased consumption of oxygen following the long kumbhak pranayama, thereby indicating that no debt had been incurred during such a pranayama or that the low oxygen consumption would have matched to a lowered metabolic state during the long kumbhak pranayama. Also, there seems to be no gross lowering of PO2, or rise of Pco2, since no effects expected of arterial anoxemia and respiratory chemoreceptor stimulation were apparent as the post-pranayamic ventilation was at the same level as pre-pranayamic level. Contrary trends of oxygen consumption (implying corresponding trends to be occurring in the status of metabolism) observed during the two types of pranayamas may be related to the differences in their effects on skeletal muscular relaxation or activity, on variations of autonomic nervous system discharges altering regional blood flows and metabolism, and even on adrenal cortico-medullary secretions. The two types of Pranayamic procedures having differing patterns of consciously controlled execution of respiratory rhythms with different time ratios of their phases, and differing in experiences of resulting sensations, could have led to either increase or decrease of arousal or relaxation states and associated shifts in muscular activities and metabolism. The differences in oxygen consumption in the two pranayamas may also be partly related to changes in brain blood flow and metabolism due to differences in intensities of regional activities of neurons in the cerebrum under differences in the nature of conscious control of the respiratory rhythm, and in the experiencing of sensations of breathing in the two pranayamas.

ACKNOWLEDGEMENT
Authors thank the Indian Council of Medical Research, New Delhi, for financial support to a major study on yoga, of which this communication forms a part.