Effects of Training and of Nitroglycerin

Using a Davidon-Fletcher-Powell nonlinear function minimization algorithm," the function 7) is fit to the observed control pressure-volume relationships. The output of this computer subprogram is a set of estimates of the three parameters a,, a2, and a,. Using these parameters in 7), the analysis program computes the predicted pressure at each actually observed cubed circumference, using the minimum cubed circumference observed in the control period as C, the baseline circumference. The squared difference between this predicted pressure and the end-diastolic pressure actually observed gives an estimate of the variance of the observed pressures about the predicted pressure volume curve, when summed over all points observed in the control period. A similar procedure gives the estimated variance about the same predicted pressure volume curve in each ischemic period. The ratio of the estimated ischemic variance about the predicted pressure volume curve to the control variance about the same curve yields the statistic, F," which is used to test the hypothesis that the ischemic pressure-volume observations come from the same underlying pressure-circumference curve as the control period observations. In all cases, the computed F was consistent with the hypothesis (P > 0.01, single-tailed test) that the observations made during ischemia were samples from the same pressure-volume curve as that generated from the control observations. Thus, it can be inferred that ischemia does not alter any of the parameters (a,, a2, a,) which characterize the predicted pressure-volume relationship. Consequently, the parameters characterizing the pressurevolume relationship (a,, a,, a,) must remain unaltered by ischemia. Since left ventricular end-diastolic volume stiffness is determined by these parameters, it is concluded that end-diastolic volume stiffness is unaltered by ischemia.