Interest has been revived recently in the relative importance of changes in heart rate and stroke volume during exercise in man. Rushmer and Smith have examined the literature and concluded that an increase in stroke volume is not an essential feature of the ventricular response in normal human subjects (1, 2). Contrary views, particularly in the older literature, were attributed to differences in the technics used to measure cardiac output and in the degree of training of the subjects (1). However, in the studies reviewed by Rushmer, the subjects had exercised in various postures. This could have accounted for much of the apparent discrepancy (3), a point which Rushmer has now emphasized (4).The present investigation demonstrates the importance of the posture of the subject in the interpretation of changes in stroke volume from rest to exercise. The stroke volume increases only slightly once exercise has begun, regardless of the severity of the exercise or the position assumed by the subject.
METHODSSeven healthy "untrained" men, aged 25 to 35 years, were studied. Cardiac output was measured by the indicator-dilution method. A cardiac catheter was placed with its tip in the superior vena cava. The indicator, cardio-green dye,1 was given in a dose of 7.5 mg by sudden injection from a specially calibrated syringe. The syringe was attached to the catheter throughout the experiment and was filled before each injection from a supply syringe connected to it by a three-way tap. Blood was sampled continuously for each dilution curve from the radial artery through a Wood oximeter at * This investigation was supported in part by a grant from the Minnesota Heart Association. rates of 28 to 50 ml per minute. The dilution curve was recorded on a photokymograph. Not more than 40 ml was withdrawn for any curve. On completion of each curve the blood was reinfused.Samples of blood were withdrawn from the radial artery before and after each experiment. Known amounts of dye were added to 10 ml aliquots of blood and the resulting mixtures were drawn through the cuvet oximeter to permit calibration of tIle dilution curves. The calibrations for the two samples invariably were close (Figure 1, upper panel) and were reproducible from day to day. Calibrations for blood from different subjects were similar. Figure 1