Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Introduction
The autonomic nervous system plays an important role in
regulating the cardiovascular system. Both high sympathetic and low parasympathetic tones have been shown to be
associated with increased cardiovascular mortality.1,2 Heart
rate recovery (HRR), defined as the fall in heart rate through
the first minute after exercise, is a measure of the capacity
of the cardiovascular system to reactivate the parasympathetic nervous system after exercise3,4 and has been found to
be an independent predictor of all-cause mortality in various
populations.57 Previous studies have shown that HRR measured at predischarge exercise stress test (EST) predicts
Methods
Study population
This prospective cohort study consisted of consecutive patients who were enrolled into our cardiac rehabilitation
program from 1996 to 2007. All patients had confirmed
diagnoses of acute MI based on World Health Organization
criteria.14 A total of 446 patients who were able to perform
1547-5271/$ -see front matter 2010 Published by Elsevier Inc. on behalf of Heart Rhythm Society.
doi:10.1016/j.hrthm.2010.03.023
930
symptom-limited EST on treadmill, completed phase 1 and
phase 2 of the program with good attendance over 70%, no
pacemaker implantation, and no change in medications between the two ESTs were recruited. Among these patients,
60 were excluded because of loss to follow-up (n 55) or
missing data (n 5). As a result, 386 patients were included
in the analysis.
Study design
Among the 386 patients recruited, 334 underwent an 8-week
exercise training program in phase 2 of our program, which
consisted of twice-weekly 45-minute scheduled sessions of
supervised exercise at an intensity based on individual assessment. The remaining 52 patients did not receive any
exercise training because they were included into the control arm of our prior randomized study conducted between
1997 and 2002.15 Nevertheless, all 386 patients received the
same magnitude of education, counseling, and advice from
our nurses, physiotherapists, occupational therapists, dietitians, and clinical psychologists during cardiac rehabilitation.
Patients demographic data, cardiovascular risk factors,
and medications were recorded at program entry. Left ventricular ejection fraction (LVEF) was measured by transthoracic echocardiography. All patients were prospectively followed-up in our outpatient clinics or until they left the
cohort through death. Information on deaths during the
follow-up period after recruitment were retrieved from medical records and discharge summaries from our hospital as
well as other institutions. Patients who were lost to follow-up were contacted by phone. In addition, survival data
were obtained from the Births and Deaths General Register
Office. The survival rate, cause of death, and clinical characteristics of these patients were examined. Sudden cardiac
death was defined as death from an unexpected circulatory
arrest occurring within 1 hour of symptom onset.16 Cardiac
death was defined as death due to MI, intractable heart
failure, or malignant arrhythmia and included cases of sudden cardiac death.
Exercise testing
Symptom-limited ESTs were conducted in phase 1 (before
exercise training) and at the end of phase 2 (after exercise
training) using the modified Bruce protocol. Treadmill was
terminated within 10 seconds after peak exercise, and patients were allowed to assume a sitting position immediately. Resting heart rate was defined as heart rate before
EST during sitting. Peak heart rate was defined as heart rate
achieved at peak exercise. Heart rate increment was defined
as heart rate increase from baseline to peak (i.e., peak heart
rate resting heart rate). HRR was defined as heart rate
decrease through the first minute of recovery (i.e., peak
heart rate heart rate at 1 minute of recovery phase). Duke
treadmill score (DTS) was calculated using the following
formula: (Duration of exercise [min]) (5 Maximal
ST-segment deviation [mm]) (4 treadmill angina index), where index 0 no angina, 1 nonlimiting angina,
Statistical analysis
Continuous variables are expressed as mean SD, and
categorical variables are presented in frequency tables. Statistical comparisons were performed using Students t-test
or Pearson Chi-square test, as appropriate. LVEF 30%
was chosen as a predictive variable of cardiac death, in line
with the Multicenter Automatic Defibrillator Implantation
Trial II (MADIT II).23 For all other variables, cutoff points
were obtained from receiver operating characteristic curve
analyses. Hazard ratio (HR) and 95% confidence interval
(CI) of each variable to predict cardiac death were determined by multivariate Cox regression model using P .1
for inclusion. P .05 was considered significant. Statistic
analyses were performed using SPSS software (version
16.0, SPSS, Inc., Chicago, IL, USA).
Results
Table 1 lists baseline clinical demographic features of the
study population. A total of 334 (86.5%) patients received
exercise training. More patients in the exercise training
group received antiplatelets or warfarin, statins, and revascularizations before enrollment compared to patients without exercise training. Their changes in EST parameters and
LVEF during follow-up are listed in Table 2. Peak heart
rate, heart rate increment, and HRR increased significantly
after completion of phase 2 in patients who received exercise training but not in those without training. In both
groups, exercise capacity and LVEF improved significantly
over time compared to baseline. Although LVEF increased
similarly in the two groups (5.1% 11.4% vs 6.2%
14.0%, P 0.53), the increase in exercise capacity was
more profound in patients who received exercise training
(1.9 0.1 METs vs 1.4 0.2 METs, P .001). Although
there were significant changes in the prevalence of patients
in different DTS risk groups between phase 1 and phase 2 in
both patients with and those without exercise training (Table 2), exercise training did not modify the change in DTS
risk between the two phases (Figure 1).
After 78.5 40.5 months of follow-up, 40 (10.4%)
patients died of cardiac events, of which 26 (6.7%) were
sudden cardiac deaths. Baseline clinical characteristics of
patients with and those without cardiac death are listed in
Table 3. Compared with survivors, patients who suffered
from cardiac death were older, were more likely to be
diabetic, developed acute pulmonary edema on presentation, received angiotensin-converting enzyme inhibitors/angiotensin receptor blockers, and had lower initial LVEF, but
they were less likely to receive coronary revascularizations,
Hai et al
Table 1
931
Exercise training
(n 334)
No exercise training
(n 52)
63.8 11.2
256/78
321/13
124 (37.1)
122 (36.5)
178 (53.3)
226 (67.7)
29 (8.7)
36 (10.8)
46.5 9.9
65.0 10.2
41/11
52/0
24 (46.2)
25 (48.1)
30 (57.7)
28 (53.8)
4 (7.7)
6 (11.5)
46.3 10.1
85 (25.4)
21 (6.3)
255
333
266
266
276
.40
.86
.23
1
.13
.65
.059
1
.81
.90
14 (26.9)
4 (7.7)
(76.3)
(99.7)
(79.6)
(79.6)
(82.6)
23
48
38
41
30
P value
.87
.76
.001*
.001*
.28
.86
.001*
(44.2)
(92.3)
(73.1)
(78.8)
(57.7)
exercise training, and statins (Table 3, all P .05). However, there were no significant differences in other demographic data and cardiovascular risk factors, incidence of
malignant arrhythmias during index hospitalizations, or use
of antiplatelets/warfarin or beta-blockers between the two
groups.
Univariate analysis demonstrated that diabetes mellitus,
acute pulmonary edema on presentation, phase 1 or phase 2
LVEF 30%, and use of angiotensin-converting enzyme
inhibitors/angiotensin receptor blockers were significantly
associated with cardiac death, whereas use of statins, revas-
Table 2
Exercise stress test parameters and mean left ventricular ejection fraction
With exercise training (n 334)
Phase 1
71.2
121.7
50.5
17.5
5.4
46.5
187
91
3
53
Phase 2
13.1
22.8
20.8
10.0
3.2
9.9
(56.0)
(27.2)
(0.9)
(15.9)
71.9
125.4
53.4
19.0
7.2
52.1
167
94
7
66
12.8
23.3
22.0
12.3
3.4
13.4
(50.0)
(28.1)
(2.1)
(19.8)
Phase 1
.26
.001*
.006*
.011*
.001*
.001*
71.3
118.9
47.6
19.1
4.3
46.3
.001*
.001*
.001*
.001*
20
22
2
8
Phase 2
16.3
25.6
23.2
14.4
2.7
10.1
(38.5)
(42.3)
(3.8)
(15.4)
72.3
124.4
52.2
20.9
5.1
53.2
22
18
3
9
P value
15.6
21.7
21.4
15.1
3.1
12.1
(42.3)
(34.6)
(5.8)
(17.3)
.61
.068
.060
.37
.001*
.003*
.001*
.001*
.001*
.10
932
Figure 1
Change in Duke treadmill risk score between phase 1 and
phase 2 with reference to exercise training. Improved risk improved
from higher-risk to lower-risk group; same risk no change in risk group;
worsened risk worsened from lower-risk to higher-risk group.
Table 3
Cardiac death
(n 40)
No cardiac death
(n 346)
66.8 12.0
29/11
39/1
15 (37.5)
23 (55.0)
22 (55.0)
23 (57.5)
4 (10.0)
6 (15.0)
43.0 10.3
63.6 11.0
268/78
334/12
133 (38.4)
124 (35.8)
186 (53.8)
231 (66.8)
29 (8.4)
36 (10.4)
46.9 9.8
19 (47.5)
4 (10.0)
16
38
27
37
21
26
(40.0)
(95.0)
(67.5)
(92.5)
(52.5)
(65.0)
80 (23.1)
21 (6.1)
262
343
277
270
285
308
(75.7)
(99.1)
(80.1)
(78.0)
(82.4)
(89.0)
P value
.019*
.55
1
.26
.01*
1
.29
.76
.42
.028*
.002*
.31
0.001*
.086
.10
.037*
.001*
.001*
Hai et al
Table 4
933
Age
Sex
Smoker
Diabetes mellitus
Hypertension
Hypercholesterolemia
Renal impairment
Respiratory disease
Phase 1 LVEF 30%
Phase 2 LVEF 30%
Acute pulmonary edema
Ventricular tachycardia/ventricular fibrillation
Revascularizations
Antiplatelet/warfarin
Beta-blocker
Angiotensin-converting enzyme inhibitor/
angiotensin receptor blocker
Statin
Exercise training
Phase 1 resting heart rate 65 bpm
Phase 1 heart rate increment 45 bpm
Phase 1 heart rate recovery 12 bpm
Phase 1 exercise capacity 4 METs
Phase 1 Duke treadmill score (risk group)
Low risk
Intermediate risk
High risk
Phase 2 heart rate increment 45 bpm
Phase 2 heart rate recovery 12 bpm
Phase 2 exercise capacity 4 METs
Phase 2 Duke treadmill score (risk group)
Low risk
Intermediate risk
High risk
Multivariate predictors
HR
95% CI
P value
1.03
0.86
0.59
2.64
0.94
1.00
0.83
1.19
2.75
4.10
3.40
1.91
0.45
0.31
0.61
3.92
1.001.06
0.431.72
0.271.29
1.414.95
0.501.75
0.531.88
0.322.51
0.502.84
1.265.97
1.6710.06
1.836.34
0.685.37
0.240.88
0.071.28
0.311.18
1.2112.71
.10
.66
.19
.002*
.84
.99
.89
.70
.011*
.002*
.001*
.22
.018*
.10
.14
.023*
0.40
0.43
3.04
2.25
1.56
2.90
0.210.75
0.220.83
1.287.26
1.164.36
0.822.96
1.386.10
0.77
1.42
0.47
2.76
4.00
6.51
0.40 1.50
0.732.77
0.00 161.33
1.455.25
2.147.48
3.1813.33
0.53
2.10
0.50
0.26 1.09
1.01 4.36
0.073.67
.004*
.011*
.012*
.017*
.18
.005*
.44
.30
.46
.002*
.001*
.001*
HR
95% CI
P value
2.28
1.015.19
.049*
1.99
4.70
1.76
0.745.32
1.3416.46
0.803.89
.17
.016*
.160
1.08
0.402.92
.877
2.80
0.6312.45
.177
0.36
0.59
5.37
1.79
0.160.80
0.221.11
1.3321.61
0.565.70
.012*
.088
.018*
.325
0.49
0.141.71
.261
0.80
2.49
3.63
0.302.08
1.105.63
1.1711.28
.641
.028*
.026*
.11
.10
.49
CI confidence interval; HR hazard ratio; LVEF left ventricular ejection fraction; METs metabolic equivalents.
*P .05.
Discussion
Our results demonstrated exercise training improved HRR,
LVEF, and exercise capacity in patients with recent MI. In
this study, none of the parameters during baseline phase 1
EST, except resting heart rate, predicted cardiac death during long-term follow-up. In contrast, left ventricular dysfunction with LVEF 30%, impaired HRR 12 bpm, and
Table 5
Parameter
Phase
Phase
Phase
Phase
Phase
Phase
1
1
1
2
2
2
AUC
95% CI
P value
Sensitivity (%)
Specificity (%)
PPV (%)
NPV (%)
0.62
0.66
0.66
0.66
0.66
0.75
0.570.67
0.610.70
0.610.71
0.610.71
0.610.71
0.700.79
.0144*
.0002*
.0001*
.0001*
.0001*
.0001*
80.0
67.5
77.5
62.5
52.5
75.0
36.71
54.9
50.0
62.1
77.8
72.0
12.7
14.8
15.2
16.0
21.4
23.6
94.1
93.6
95.1
93.5
93.4
96.1
AUC area under the curve; CI confidence interval; METs metabolic equivalents; NPV negative predictive value; PPV positive predictive value.
*P .05.
934
Table 6
HRR 12 bpm
(n 99)
HRR 12 bpm
(n 287)
P value
68.4 9.1
73/26
96/3
39 (39.4)
48 (48.5)
64 (64.6)
61 (61.6)
13 (13.1)
18 (18.2)
46.1 9.8
49.9 12.5
62.4 11.3
224/63
277/10
109 (38.0)
99 (34.5)
144 (50.2)
193 (67.2)
20 (7.0)
24 (8.4)
46.6 9.9
53.1 13.4
.001*
.38
.83
.08
.013*
.013*
.31
.059
.007*
.67
.047*
37 (37.4)
7 (7.1)
62 (21.6)
18 (6.3)
65
96
70
80
73
249
(65.7)
(97.0)
(70.7)
(80.8)
(73.7)
(86.8)
213
285
234
227
233
85
(74.2)
(99.3)
(81.5)
(79.1)
(81.2)
(85.9)
73.5
40.3
3.6
12.2
70.4
53.5
5.8
19.6
15.0
18.1
2.1
9.6
43 (43.4)
39 (39.4)
0 (0)
75.2 14.9
38.2 15.4
5.1 2.6
33 (33.3)
36 (36.4)
2 (2.0)
.002*
.78
.014*
.08
.023*
.72
.12
.82
12.9
21.0
3.2
10.3
.076
.001*
.001*
.001*
164 (57.1)
74 (25.8)
5 (1.7)
70.9 12.4
58.5 21.4
7.5 3.5
.014*
.005*
.19
.005*
.001*
.001*
156 (54.4)
76 (26.5)
8 (2.8)
.005*
.003*
.82
Hai et al
Figure 2
935
tant mechanism by which exercise training decreases mortality, as patients who had restoration of autonomic control
after exercise training, as evidenced by HRR improved to
12 bpm, had better survival during long-term follow-up.
Assessment of autonomic control of the cardiovascular
system after MI provides completely distinct information
from conventional prognostic markers.4 Several different
indexes have been used to assess cardiac autonomic function.4 Although widely studied, parameters such as heart
rate variability and baroreflex sensitivity are of limited clinical value due to their intrinsic methodologic and interpretive difficulties.30,31 In contrast, ESTs are routinely performed for risk stratification after MI, and simple autonomic
parameters obtained from ESTs are easy to interpret. Furthermore, unlike chronotropic index and heart rate increment, which also depend on myocardial perfusion and physical strength, HRR is a more precise measurement of the
cardiovascular autonomic balance.32 In particular, our study
failed to demonstrate any correlation between heart rate
increment and cardiac death, in contrast to findings from
previous studies on healthy volunteers.32,33 The reason for
this discrepancy remains unclear but can be explained partly
by the fact that 80% of our patients were treated with
beta-blockers, which blunted heart rate response during
exercise but either enhanced or had no effect on HRR.34,35
DTS failed to predict survival for our patients, which is
in contrast with findings of previous validation studies of
MI populations.17,19 Although the reason remains unclear, it
is possible that a significant proportion of our patients had
indeterminate DTS due to baseline abnormal ECG, which
reduces the power of DTS to predict clinical outcomes in
this study. On the other hand, HRR appears to be superior to
Study limitations
First, our study recruited patients who completed phase 1
and phase 2 of the cardiac rehabilitation program with good
attendance and were able to perform symptom-limited EST
on treadmill. As a result, high-risk populations of patients
who could not complete the program due to repeated hospitalizations or could not perform treadmill tests due to poor
cardiovascular function were excluded. Indeed, the cardiac
death rate was 21% and sudden cardiac death rate was
12.3% among patients excluded (n 211), which were
much higher than that in the population studied. Therefore,
generalization of our results to this group of patients should
be made with caution. Second, this study was not designed
to evaluate the benefit of exercise training. Although this
study included patients who were randomized to the control
arm in our previous trial, the nonrandomized nature of the
present study and the severe imbalance between treatment
and control groups preclude analysis of the effect of exercise training. Nevertheless, previous randomized studies
have already proved that HRR improves with exercise training,9,11 and our study only addressed the effect of changes
936
in HRR on survival. Third, although this was a prospective
observational study, the endpoints were not prespecified but
were derived from post hoc analysis. Finally, only data on
HRR at 1 minute were prospectively collected for all patients in this study; whether HRR measured at 2 minutes or
later would provide better prognostic information remains
unclear.
16.
17.
Conclusion
Our study demonstrated that exercise training improved
HRR and that patients with persistently impaired HRR 12
bpm after rehabilitation had higher cardiac mortality during
long-term follow-up, suggesting that modulation of cardiovascular autonomic control with exercise training may contribute to the long-term beneficial effects of a cardiac rehabilitation program. Its clinical application as a therapeutic
target for exercise prescription and medication titration
needs to be further addressed in future studies.
References
1. Curtis BM, OKeefe JH Jr. Autonomic tone as a cardiovascular risk factor: the
dangers of chronic fight or flight. Mayo Clin Proc 2002;77:4554.
2. Schwartz PJ, La Rovere MT, Vanoli E. Autonomic nervous system and sudden
cardiac death. Experimental basis and clinical observations for post-myocardial
infarction risk stratification. Circulation 1992;85(1 Suppl):I77I91.
3. Freeman JV, Dewey FE, Hadley DM, Myers J, Froelicher VF. Autonomic
nervous system interaction with the cardiovascular system during exercise. Prog
Cardiovasc Dis 2006;48:342362.
4. Frenneaux MP. Autonomic changes in patients with heart failure and in postmyocardial infarction patients. Heart 2004;90:1248 1255.
5. Nishime EO, Cole CR, Blackstone EH, Pashkow FJ, Lauer MS. Heart rate
recovery and treadmill exercise score as predictors of mortality in patients
referred for exercise ECG. JAMA 2000;284:13921398.
6. Cole CR, Foody JM, Blackstone EH, Lauer MS. Heart rate recovery after
submaximal exercise testing as a predictor of mortality in a cardiovascularly
healthy cohort. Ann Intern Med 2000;132:552555.
7. Cole CR, Blackstone EH, Pashkow FJ, Snader CE, Lauer MS. Heart-rate
recovery immediately after exercise as a predictor of mortality. N Engl J Med
1999;341:13511357.
8. Nissinen SI, Makikallio TH, Seppanen T, et al. Heart rate recovery after exercise
as a predictor of mortality among survivors of acute myocardial infarction. Am J
Cardiol 2003;91:711714.
9. Tiukinhoy S, Beohar N, Hsie M. Improvement in heart rate recovery after
cardiac rehabilitation. J Cardiopulm Rehabil 2003;23:84 87.
10. Adams BJ, Carr JG, Ozonoff A, Lauer MS, Balady GJ. Effect of exercise
training in supervised cardiac rehabilitation programs on prognostic variables
from the exercise tolerance test. Am J Cardiol 2008;101:14031407.
11. Streuber SD, Amsterdam EA, Stebbins CL. Heart rate recovery in heart failure
patients after a 12-week cardiac rehabilitation program. Am J Cardiol 2006;97:
694 698.
12. Malfatto G, Facchini M, Sala L, Branzi G, Bragato R, Leonetti G. Relationship
between baseline sympatho-vagal balance and the autonomic response to cardiac
rehabilitation after a first uncomplicated myocardial infarction. Ital Heart J
2000;1:226 232.
13. Lucini D, Milani RV, Costantino G, Lavie CJ, Porta A, Pagani M. Effects of
cardiac rehabilitation and exercise training on autonomic regulation in patients
with coronary artery disease. Am Heart J 2002;143:977983.
14. Nomenclature and criteria for diagnosis of ischemic heart disease. Report of the
Joint International Society and Federation of Cardiology/World Health Organization task force on standardization of clinical nomenclature. Circulation 1979;
59:607 609.
15. Yu CM, Lau CP, Chau J, et al. A short course of cardiac rehabilitation program
is highly cost effective in improving long-term quality of life in patients with
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.