The CVAR for each variable as a function of exercise stage is presented in Figures 1 through 4 for all patients. As can be seen, for all but one (systolic BP, Fig 1) of the variables, there was a tendency for CVAR to be greater at rest and the beginning stages of exercise than it was at the end of exercise. This trend was particularly evident for Vo2 and Vco2 (Fig 4). It also can be seen that CVAR was not a function of functional class; for each stage of exercise, the range of CVAR is essentially identical for the three classes.
The range of CVAR for HR at rest was 4.9 to 15.2 percent (Fig 1). After the first two stages of exercise the upper limit of this range fell below 9 percent in all but two patients: one in class A, the other in class C. The CVAR for this class A patient remained around 10 percent throughout the test while that for the class С patient increased from a value of 7.7 percent at rest to 18.0 percent in stage 2; thereafter it declined so that at the end of exercise (stage 4) the value was 7.7 percent. The range of CVAR for systolic BP remained between 1 and 15 percent throughout the test (Fig 1).
FIGURE 1. Coefficient of variation vs exercise stage for systolic blood pressure and heart rate for functional class A, B, and С heart failure patients (see text for definition of functional class).
The CVAR responses for respiratory rate and VT are presented in Figure 2 and that for VE in Figure 3. As can be seen, the range of CVAR for respiratory rate at rest is rather large (ie, 4.9 to 20.8 percent). Also, the upper value of the CVAR range remains high for all the stages, never falling below 12.0 percent. The CVAR associated with VT, on the other hand, is below 10.0 percent for the exercise stages beyond stage 3. The range of CVAR for VE is similar to that obtained for respiratory rate. For rest and stages 1 to 4 the CVAR is between 2.5 and 21.5 percent and for the other stages CVAR lies between 1.3 and 14.7 percent.
Table 3—Coefficient of Variation
In Figure 4 CVARs associated with Vo2 and Vco2 are presented. In general, CVAR was high for both of these variables during rest and stage 1. In fact one of the class С patients had values at rest of 37.2 percent for Vo2 and 33.5 percent for Vco2. However, in stage 1 the CVAR for this patient fell to levels that were below 9 percent. Marked declines in CVAR in stage 1 were similarly seen in most of the other patients. Thereafter, the range of CVAR generally became narrower with its upper level becoming (ie, beyond stage 4) less than 9 percent for Vo2 and 13.5 percent for Vco2.
FIGURE 2. Coefficient of variation vs exercise stage for respiratory rate and tidal volume for functional class A, B, and С heart failure patients.
In every test, all patients crossed their AT. The Vo2max was attained in at least one of the tests in patients 1 and 7, 50 to 65 percent of the tests in patients 2, 5, and 13, and 75 to 100 percent of the tests in the remainder of patients. For classes A, B, and C, the average duration of exercise was 17, 13, and 9 minutes, respectively; the average Vo2max was 26.5, 17.4, and 13.8 ml/min/kg, respectively; and the average AT was 19.0, 13.2, and 10.7 ml/min/kg, respectively. The CVARs associated with these variables are given for each patient in Table 3. Also listed are the CVARs for maximum HR and systolic BP. Of note is the fact that the C\AR for exercise duration increases with the severity of heart failure.
FIGURE 3. Coefficient of variation vs exercise stage for minute ventilation for functional class A, B, and С heart failure patients.
Time Trend Analysis
There were a few instances in which systematic differences over time were found. In four functional class A patients (ie, 1A, 2A, 4A, and 6A) and one class В (9B) patient significant correlation coefficients were obtained between the slope of the Vco2-work relation and the time the test was administered. In four patients (1A, 2A, 4A, and 9B) the trend was for the slope to decrease in subsequent exercise tests while in patient 6A the trend was opposite. Also, in one or two patients, significant positive ( + ) or negative (—) correlation coefficients between the slopes of the HR- work (—,4A), VE-work (+,7B and – ДЗС), systolic BP-work (— ,9B) VT-work (-,9B), and RR-work (+ ,10B) relations and the time the test was adminis¬tered were obtained. Finally, in a few patients, there was a time trend in some of the end-exercise variables (Table 3).
FIGURE 4. Coefficient of variation vs exercise stage for o2 uptake and co2 production for functional class A, B, and С heart failure patients.