DISCUSSION
In the present study there was a significant decrease in breath frequency following IRT, and a decrease in finger plethysmogram amplitude after SR. There were no changes in measures of autonomic activity. The rate of respiration is highly sensitive to phasic changes in the psychological state (Lorig & Schwartz, 1990). The increase in breath rate has been correlated with experimentally evoked fear and anxiety (Ax, 1953), as well as before situations such as parachute jumping (Fenz & Jones, 1972). The reduction in breath rate after yoga based IRT is consistent with other reports of reduced breath rate related to the practice of yoga, both as an immediate effect (Wallace, Benson & Wilson, 1971) and after three months (Joseph et al., 1981). This suggests that IRT may help to reduce some aspects of physiological arousal, without modifying autonomic activity.

Similarly, the non-significant trend of decrease in oxygen consumption and increase in breath amplitude after IRT, also suggest a reduction of arousal. A decrease in finger plethysmogram amplitude is suggestive of increased peripheral vasoconstriction related to increased sympathetic vasomotor tone (Delius & Kellerova, 1971), which occurred after SR. This change may be due to the postural readjustment as the "post" recording was made while seated erect, compared to the supine position, during the test period.

In conclusion, yoga based IRT produced better physiological rest than SR, supporting the idea that a combination of stimulation (through isometric contractions) and relaxation may reduce arousal better than relaxation alone.