Results
Baseline data of exercise and control group are presented in
Table 1. Kolmogorov-Smirnov tests showed a Gaussian
distribution of systolic and diastolic daytime, nighttime, and
24-hour ABP in both exercise and control groups. One
subject in the exercise group discontinued the exercise
program because of need for abdominal surgery independent
of the study. One patient in both the exercise and the control
groups reported a change in antihypertensive medication at
the follow-up examination, and data were excluded from
analysis. Thus, there were 2 dropouts in the exercise group
and 1 in the control group. The mean follow-up period was
9.82.0 weeks in the exercise group and 10.22.0 weeks in
the control group (P value not significant). The exercise
program was well tolerated by all of the patients. Mean
training lactate concentration was 1.60.6 mmol/L, corresponding
with a mean training heart rate of 100.311.6 per
minute. Four patients did not redo the treadmill stress test. In
these cases, the results of baseline stress tests were excluded
2
Dimeo et al Exercise in Resistant Hypertension 655
from analysis. At the follow-up examination, all of the
patients of both exercise and control groups stated that there
was no change of compliance in the course of the study. Table
2 presents ABP data, office blood pressure, vascular compliance,
and cardiac index in training and control groups before
and after the observation period. An ANCOVA model adjusted
for baseline values, age, and diabetes mellitus was used
to analyze the impact of the exercise training on the cardiovascular
system. The exercise program significantly reduced
daytime systolic and diastolic ABPs by 5.911.6 and
3.36.5 mm Hg, respectively (P0.03 each; Table 2).
Nighttime systolic and diastolic ABPs were numerically
lower after the exercise program, but changes did not reach
significance (3.817.1 and 1.98.2 mm Hg; P0.05
each; Table 2). Twenty-four— hour systolic and diastolic
ABPs were significantly decreased by 5.412.2 (P0.03)
and 2.85.9 mm Hg (P0.01). Systolic and diastolic office
blood pressures were numerically but not significantly reduced
by the exercise training (6.615.7 and 2.78.0
mm Hg; P value not significant), with changes 1mmHgin
the control group. Pulse wave analysis revealed no alterations
of large and small artery elasticity in the exercise and control
groups in the course of the observation period (P value not
significant for each; Table 2). Cardiac index at rest remained
unchanged by the exercise program (P value not significant).
The exercise program led to an increase in physical performance,
as illustrated by an increase of maximal oxygen
uptake from 22.85.7 to 24.35.1 mL/kgmin (P0.01;
Table 2) and by a right shift of lactate curves (Figure 2). The
exercise program allowed a significant increase in mean
maximal workload level of the Bruce protocol from 5.71.7
to 6.61.4 (P0.01) with no change in the control group
(4.31.6 to 4.01.6). Perceived exertion as measured by the
Borg scale decreased in the exercise group without significant changes in the control group (Figure 2). The exercise program
significantly lowered both systolic and diastolic blood pressures
on exertion (Figure 2). A correlation analysis of the
difference of systolic daytime ABP in the exercise group with
the mean reduction of lactate during treadmill stress tests
revealed a Pearson coefficient of 0.08 (P0.97), indicating
that there was no significant correlation of the extent of
improvement of physical performance and the reduction of
systolic daytime ABP. Accordingly, there was no significant
correlation between decrease of systolic daytime ABP and
change of maximal oxygen uptake (Pearson coefficient,
0.19; P0.44).