Академический Документы
Профессиональный Документы
Культура Документы
Background Exercise-based cardiac rehabilitation (CR) remains an underused tool for secondary prevention post
myocardial infarction (MI). In part, this arises from uncertainty regarding the efficacy of CR, particularly with respect to
reinfarction, where previous studies have failed to show consistent benefit. We therefore undertook a meta-analysis of
randomized controlled trials (RCTs) to (1) estimate the effect of CR on cardiovascular outcomes and (2) examine the effect of CR
program characteristics on the magnitude of CR benefits.
Methods We systematically searched MEDLINE as well as relevant bibliographies to identify all English-language RCTs
examining the effects of exercise-based CR among post-MI patients. Data were aggregated using random-effects models.
Stratified analyses were conducted to examine the impact of RCT-level characteristics on treatment benefits.
Results We identified 34 RCTs (N = 6,111). Overall, patients randomized to exercise-based CR had a lower risk of
reinfarction (odds ratio [OR] 0.53, 95% CI 0.38-0.76), cardiac mortality (OR 0.64, 95% CI 0.46-0.88), and all-cause
mortality (OR 0.74, 95% CI 0.58-0.95). In stratified analyses, treatment effects were consistent regardless of study periods,
duration of CR, or time beyond the active intervention. Exercise-based CR had favorable effects on cardiovascular risk factors,
including smoking, blood pressure, body weight, and lipid profile.
Conclusions Exercise-based CR is associated with reductions in mortality and reinfarction post-MI. Our secondary
analyses suggest that even shorter CR programs may translate into improved long-term outcomes, although these results need to
be confirmed in an RCT. (Am Heart J 2011;162:571-584.e2.)
Despite guidelines recommending the use of cardiac exercise-based secondary prevention CR program. 3,4 The
rehabilitation (CR) programs for patients with ST- reason for such low participation is likely multifactorial, 4
segment elevation myocardial infarction (MI) 1 and non but 1 important obstacle is the infrastructure to support
ST-segment elevation MI/unstable angina, 2 participation prolonged participation in these programs. Another
in these programs continues to be low; only 10% to 20% barrier to usage is likely the absence of large randomized
of patients who survive an acute MI participate in an controlled trials (RCTs) evaluating its efficacy and
insufficient data on what features of CR programs result
in the greatest efficacy.
Individual RCTs studying the efficacy of exercise-based
From the aDivisions of Cardiology and Clinical Epidemiology, Jewish General Hospital/ CR have provided conflicting results, and the most recent
McGill University, Montreal, Quebec, Canada, bDepartment of Medicine, McGill meta-analysis (published in 2004) found that CR de-
University Health Center, Montreal, Quebec, Canada, cDivision of Epidemiology and
Community Health, School of Public Health, University of Minnesota, Minneapolis, MN, and
creased all-cause and cardiac mortality but had no
d
Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, beneficial effects on reinfarction. 5 In addition, consider-
Montreal, Quebec, Canada. able heterogeneity among programs exists, and the
Dr Eisenberg is a National Researcher of the Quebec Foundation for Health Research. factors that define effective CR on a program-level
Submitted April 10, 2011; accepted July 21, 2011.
Reprint requests: Mark J. Eisenberg, MD, MPH, FAHA, FACC, Professor of Medicine, Divisions
including optimal duration 6are currently unclear. 7
of Cardiology and Clinical Epidemiology, Jewish General Hospital/McGill University, 3755 Given these areas of uncertainty, we undertook a meta-
Cte Ste-Catherine Road, Suite H-421, Montreal, Quebec, Canada H3T 1E2. analysis of RCTs to (1) accurately estimate the effect of CR
E-mail: mark.eisenberg@mcgill.ca
on cardiovascular outcomes and (2) conduct stratified
0002-8703/$ - see front matter
2011, Mosby, Inc. All rights reserved. analyses to examine the effect of RCT-level characteristics
doi:10.1016/j.ahj.2011.07.017 on the benefits of CR.
American Heart Journal
572 Lawler, Filion, and Eisenberg October 2011
Figure 1
Excluded (n = 1,800)
- Not relevant (n = 1,795)
- Review article (n = 3)
- Editorial (n = 1)
- Study design (n = 1)
Excluded (n = 107)
- CAD but not exclusively post-MI (n = 53)
- Follow-up inadequate (n = 23)
- No non-exercising controls group (n = 12)
- Not randomized (n = 6)
- Not reporting outcomes of interest (n = 5)
40 articles from 34 RCTs included in meta-analysis - Exercise intervention inadequate (n = 5)
- Inadequate reporting of methods (n = 1)
- Review (n = 1)
- Study design article (n = 1)
Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram of RCTs included in the meta-analysis. CAD indicates coronary
artery disease.
constructed and visually assessed for publication bias. All Study and patient characteristics
analyses were conducted using Stata version 11.1 (StataCorp
The characteristics of the 34 identified RCTs are shown
LP, College Station, TX).
in Table I. All 34 RCTs were open-label RCTs. The exercise
No extramural funding was used to support this work. The intervention varied considerably and, in some cases, was
authors are solely responsible for the design and conduct of this not clearly defined. Furthermore, the duration of time
study, all study analyses, and drafting and editing of the paper. between the index MI and commencement of the CR
program varied between and within studies considerably.
Overall, 19 RCTs used exercise-only CR, 14 used exercise-
Results based CR as part of a comprehensive secondary preven-
Search results tion program, and 1 used both in 2 independent
Our search strategy identified 2,169 potentially rele- intervention arms. 46 Twenty RCTs used aerobic-only
vant articles of which 369 were reviewed as abstracts exercise programs, none used anaerobic-only (ie, resis-
and 147 as full articles (Figure 1). Thirty-four RCTs were tance training) programs, 7 used a mixed aerobic and
included in the final analysis. 10,11 , 15-52 One study anaerobic or undirected program, and 7 RCTs did not
reported 15-year follow-up data 27 ; these protracted specify the type of exercise used. Duration of exercise-
follow-up data were not included, as they represented based CR programs ranged from 2 weeks (the prespeci-
an outlier, and instead, data from the original report fied minimum for inclusion in the present study) to
were included in the analyses. 11 3 years. Of note, only 1 RCT (Oya et al 44, N = 28 patients)
American Heart Journal
574 Lawler, Filion, and Eisenberg October 2011
Table I. Study and patient characteristics of RCTs examining the effect of exercise-based CR on cardiovascular outcomes and modifiable
cardiovascular risk factors
Study Patients Exercise intervention
Ballantyne 1982 Scotland 42 52.8 5.8 100 Canadian Air 6.0 No Home 6.0
et al 15 Force 5BX plan
Bengtsson 16 1983 Sweden 171 Interval training 3.0 Yes Hospital/group 14.0
(jogging,
calisthenics,
cycling) for 30 min
at 90% baseline
peak HR, 2/wk
Bertie 1992 England 110 52.4 1.3 Pulse-monitored 1.0 No Hospital/group 24.0
et al 17 group exercise
sessions, 2/wk
Bethell and 1990 England 229 53.7 7.5 100 Circuit training, 3.0 No Hospital/group 3.0
Mullee 18 3/week
Carlsson 1997 Sweden 168 62.1 75 Interval training 3.0 Yes Hospital/group 12.0
et al 19 with cycling and
jogg ing f or 40
min, 2-3/wk
Carson 1982 England 303 51.5 0.7 100 Circuit training 3.0 No Hospital/group 25.2
et al 20 2/wk
DeBusk 1994 United 585 57.0 8.0 79 Aerobic exercise 12.0 Yes Home 12.0
et al 21 States for 30 min per
day, 5/wk
supervised by a
RN case manager
Dugmore 1999 England 124 55.3 98 Aerobic training 12.0 No Hospital/group 60.0
et al 22 3/wk
Fridlund 1991 Sweden 178 Multimodal CR, 6.0 Yes Hospital/group 12.0
et al 23 including contact
with physiotherapist
for 2 h, 1/wk
Giallauria 2006 Italy 40 68.4 2.5 83 Bicycle ergometer 3.0 No Hospital/group 3.0
et al 24 for 30 min at 60%
of the VO2 peak
achieved at
baseline, 3/wk
Giannuzzi 1993 Italy 103 50.5 8.5 100 Bicycle ergometer 6.0 No Mixed 6.0
et al 26 for 30 min at 80%
baseline peak HR,
3/wk for 2 m
supervised, then
same at home
unsupervised
Giannuzzi 1997 Italy 80 53.5 8.5 95 Bicycle ergometer 6.0 No Mixed 6.0
et al 25 for 30 min at 80%
baseline peak HR,
3/wk for 2 m
supervised and
then same at home
unsupervised
Heldal 2000 Norway 37 53.0 7.8 100 Cycling or running 1.0 No Hospital/group 6.0
et al 28 for 2 h per session
at 85% baseline
peak HR, 5/wk
Heller 1993 Australia 450 58.5 8.0 71 Patients given 6.0 Yes Home 6.0
et al 10 written materials
directing a
walking plan
American Heart Journal
Volume 162, Number 4
Lawler, Filion, and Eisenberg 575
Table I (continued)
Table I (continued)
Myers 2000 CHE and 25 55.5 6.1 100 Walking for 1 h 2.0 Yes Hospital/group 12.0
et al 38,39 2001 United 2/d, and cycling
States 45 min 1/wk
Naughton 2000 United 641 100 Aerobic exercises 2.0 No Hospital/group 36.0
et al 40 States at HR 75%-85%
baseline peak HR,
3/wk
Oldridge 1998 Canada 201 52.8 9.5 88 Treadmill, 2.0 Yes Hospital/group 12.0
et al 41-43 1995 stationary cycling,
1991 arm bike for
50 min at 65%
baseline peak HR,
2/wk
Oya 1999 Japan 28 58.6 7.0 93 Cycle ergometer 0.5 No Hospital/group 3.0
et al 44 for 30 min, 2/d
PRECOR 1991 France 121 50.0 100 Cycle ergometer 1.5 Yes Hospital/group 24.0
group plus walking for
25 min, 3/wk
Sivarajan 1982 United 258 56.3 8.4 Progressive 3.0 No Mixed 6.0
et al46 States walking and
(Group B1)# calisthenics
Sivarajan program modified 3.0 Yes Mixed 6.0
et al46 weekly depending
(Group B2)# on performance
Specchia 1996 Italy 256 52.9 7.5 91 Bicycle ergometer 1.0 No Hospital/group 34.5
et al 47 for 30 min
progressing to 75%
baseline peak
maximal work
capacity, with
calisthenics 5/wk
Stern 1983 United 71 85 Mixed aerobic 3.0 No Hospital/group 12.0
et al 48 States exercise for 1 h at
85% baseline peak
HR, 3/wk
Taylor 1988 United 210 100 Patient- 5.75 Yes Mixed 6.5
et al 49,50 1986 States individualized
exercise program
by RN
Vermeulen 1983 Holland 98 49.2 4.1 100 Multimodal CR 1.5 Yes Hospital/group 60.0
et al 51 program including
physical
intervention
Zheng 2008 China 60 Bicycle ergometer 6.0 No Hospital/group 6.0
et al 52 for 30 min at
baseline peak HR,
3/wk
CHE, Switzerland, CPX, cardiopulmonary exercise test; HR, heart rate; RN, registered nurse; /wk, number of times per week; VO2 is a measure of oxygen consumption.
Most programs included brief warm-up and cool-down periods before and after the exercise intervention. Determination of baseline peak heart rate was, in all cases, done with
cardiopulmonary testing after MI.
The Canadian Air Force 5BX (Five Basic Exercises) plan is an incremental calisthenics and aerobic exercise program.
Corresponds to fairly light to somewhat hard.
www.theheartmanual.com.
SD not provided.
Miller et al37 included 4 intervention groups: home group with 23 weeks of exercise (1A), home group with 8 weeks of exercise (1B), center-based group with 23 weeks, and
center-based group with 8 weeks (2B).
# Sivarajan et al46 included 2 groups: exercise-only CR (group B1) and comprehensive including exercise CR (group B2).
American Heart Journal
Volume 162, Number 4
Lawler, Filion, and Eisenberg 577
Figure 2
Forest plot of effect of exercise-based CR on reinfarction. Data were pooled using random effects models. Exercise-based CR significantly reduces
reinfarction among MI survivors.
Table II. Effect of exercise-based CR on overall cardiovascular outcomes and outcomes stratified by duration of exercise intervention, length of
follow-up beyond the intervention period, sample size, and date of publication
Cardiovascular
All-Cause mortality mortality Cardiac mortality Reinfarction
All Studies 31 0.74 (0.58-0.95) 18 0.61 (0.40-0.91) 22 0.64 (0.46-0.88) 18 0.54 (0.38-0.76)
Stratified Analyses
Exercise duration of 0.5 to 3 m 20 0.71 (0.51-1.01) 9 0.83 (0.31-2.22) 12 0.59 (0.34-1.03) 13 0.69 (0.43-1.11)
Exercise duration of N3 m 11 0.77 (0.54-1.09) 9 0.57 (0.37-0.88) 10 0.66 (0.44-0.98) 5 0.40 (0.24-0.66)
Pre-1995 19 0.77 (0.59-1.01) 9 0.70 (0.48-1.01) 7 0.58 (0.38-0.89) 12 0.58 (0.40-0.85)
1995 or later 8 0.59 (0.32-1.11) 9 0.49 (0.26-0.91) 7 0.83 (0.26-2.65) 3 0.33 (0.14-0.82)
Small studies 15 0.54 (0.24-1.21) 9 0.81 (0.29-2.25) 11 0.59 (0.29-1.17) 9 0.48 (0.28-0.81)
Large studies 16 0.77 (0.59-0.99) 9 0.57 (0.37-0.89) 11 0.65 (0.45-0.94) 9 0.58 (0.37-0.93)
b1 m between end of CR 13 0.81 (0.57-1.15) 10 0.57 (0.36-0.90) 11 0.67 (0.45-1.00) 4 0.55 (0.27-1.10)
and end of follow-up
1-12 m between end of CR 12 0.95 (0.56-1.61) 6 1.25 (0.32-4.85) 6 1.25 (0.32-4.85) 7 0.45 (0.21-0.98)
and end of follow-up
N12 m between end of 6 0.54 (0.35-0.85) 2 0.51 (0.15-1.76) 5 0.51 (0.29-0.92) 7 0.56 (0.35-0.90)
CR and end of follow-up
Comprehensive CR 20 0.62 (0.43-0.91) 13 0.70 (0.29-1.70) 16 0.57 (0.34-0.97) 13 0.63 (0.39-1.02)
Exercise-only CR 11 0.85 (0.61-1.17) 5 0.58 (0.37-0.92) 6 0.68 (0.45-1.01) 5 0.45 (0.27-0.74)
Distinct trial arms were analyzed as separate studies referenced to a duplicate control group and were not double counted).
Cardiovascular mortality was defined as mortality due to cardiac, cerebrovascular, or peripheral vascular diseases.
American Heart Journal
578 Lawler, Filion, and Eisenberg October 2011
Figure 3
Forest plot of effect of exercise-based CR on cardiac mortality. Data were pooled using random effects models. Exercise-based CR significantly
reduces cardiac mortality among MI survivors.
used a CR program shorter than 4 weeks, whereas all based CR were also at significantly lower risk of cardiac
other RCTs used a minimum program duration of death (OR 0.64, 95% CI 0.46-0.88) (Figure 3), cardiovas-
1 month. Follow-up duration ranged from 3 months (the cular death (OR 0.60, 95% CI 0.40-0.76), and all-cause
prespecified minimum for inclusion) to 5 years. Twenty- mortality (OR 0.74, 95% CI 0.58-0.95) (Figure 4). No
seven RCTs were single-center RCTs, and 8 were multi- significant difference in revascularization was observed
center RCTs. Twelve studies were published in 1990 or (OR 0.92, 95% CI 0.68-1.25), although these data were
earlier, 17 were published during 1991 to 2000, and inconsistently reported, and our pooled estimates are
5 were published between 2001 and June 2010. Based on accompanied by a wide 95% CI.
the Jadad scale, 5 RCTs were classified as high quality,
whereas the remainder was classified as low quality. Stratified analyses
The total number of patients randomized was 6,111. The impact of several study-level covariates on treat-
Most patients randomized were men (88.9%). The mean ment benefits was examined using stratified meta-
age of participants at the time of enrollment was analyses. To investigate whether there was evidence of
54.7 years. Only 5 RCTs reported that patients were benefit among shorter duration exercise-based CR pro-
eligible for exercise-based CR after percutaneous coro- grams, we performed stratified analyses of outcomes
nary intervention as treatment for patients' index MI, and using a program duration of 3 months as a cutoff. When
3 studies reported including subjects after coronary this was done, the effect sizes were similar among those
artery bypass graft. studies with a program duration 3 months as compared
with those b3 months for the outcome examined
Overall cardiovascular outcomes (Table II), although the effects in the shorter duration
Overall, patients randomized to exercise-based CR programs did not reach statistical significance.
were at significantly lower risk of reinfarction than To examine if benefits on reinfarction and mortality
those randomized to control (OR 0.53, 95% CI 0.38, 0.76) persisted beyond the period of active treatment, we
(Figure 2; Table II). Patients randomized to exercise- analyzed data based on duration of follow-up beyond the
American Heart Journal
Volume 162, Number 4
Lawler, Filion, and Eisenberg 579
Figure 4
Forest plot of effect of exercise-based CR on all-cause mortality. Data were pooled using random effects models. Exercise-based CR significantly
reduces all-cause mortality among MI survivors.
period of exercise intervention. There were consistent smoking among those randomized to exercise-based CR
beneficial effects seen for reinfarction, cardiac mortality, (Online Appendix 2). Similarly, effects on systolic and
and all-cause mortality across these follow-up categories diastolic blood pressure were generally more favorable in
(Table II), suggesting that the benefits of a CR program the intervention arms compared with the control arms
post-MI persist beyond the period of active intervention. (Online Appendix 2). Total cholesterol showed a trend
There was no suggestion of difference between the toward reduction in the intervention arm of several RCTs,
overall findings among small versus large studies using whereas there was no evidence of a decrease in the
the median sample size as the cutoff (Table II). The corresponding control arms (Online Appendix 3).
impact of date of publication was investigated by Changes in body weight were minimal in both groups
stratifying RCTs into time-related strata; the beneficial (Online Appendix 3).
effects of CR on cardiac outcomes persisted across study
periods (Table II). Finally, studies were stratified by Sensitivity analyses
comprehensive versus exercise-only CR, which showed Results of sensitivity analyses that excluded zero-event
persistent benefit among outcomes in both strata RCTs (in either one or both arms) were consistent with
(Table II). those of our primary analyses (data not shown).
Furthermore, exclusion of the RCT by Heller et al, 10
Cardiovascular risk factors which did not confirm the diagnosis of MI, did not
Several RCTs examined the effect of CR on modifiable appreciably alter the results. Initial stratification by
cardiovascular risk factors, including blood pressure, publication date was done using decades as cutoffs.
smoking cessation, lipids, and weight loss. Overall, RCTs When we reanalyzed data using 1995 as a cutoff, similar
found a more favorable reduction in the prevalence of results were obtained (Table II). Finally, given that the
American Heart Journal
580 Lawler, Filion, and Eisenberg October 2011
the specific objective to investigate at what point post-MI chose this approach because mortality rates have
optimal benefits occur. changed considerably over time, and limiting analysis to
There is biologic basis for the hypothesis that exercise nonfatal events, therefore, may bias the measured effect
improves outcomes post-MI independent of its effects of of exercise-based CR on reinfarction risk. Another
risk factors. Specifically, experimental evidence impli- possible explanation for these divergent findings is our
cates exercise post-MI as an important determinant of more rigorous selection criteria, limiting our analysis to
contractility and myofilament Ca 2+ sensitivity 59,60 as well peer-reviewed journal articles, where the previous meta-
as a potentially important regulator of the renin-angio- analysis also included data from abstracts and nonpeer-
tensin-aldosterone system post-MI 61 and myocardial reviewed sources (eg, thesis dissertations). Our results
fibrosis and remodeling. 62 Overall, these studies suggest also include expanded secondary analyses and several
that exercise favorably modulates several physiologic and newer RCTs. With the inclusion of these more recent
pathologic processes at play post-MI when important RCTs, the consistency of results over time suggests that
remodeling changes are underway. there is still a role for exercise-based CR in the presence
Several included RCTs were conducted before the era of newer evidence-based MI treatment. Our secondary
of current evidence-based treatment and secondary analyses offer a platform for hypothesis generation and
prevention strategies, including coronary angioplasty will hopefully motivate and direct further study. Ulti-
with and without intracoronary stenting, statins, mately, a modern well-powered RCT could answer many
-blockers, dual antiplatelet therapy, and ACE inhibitors. of the questions that remain herein, but regrettably, it is
When we investigated the potential effect of publication unlikely that such a study will be undertaken given that it
date, a proxy for concomitant therapies, on the benefits would be of questionable ethics to randomize patients
of CR using stratified analyses, these analyses revealed no into a nonexercising group with the breath of evidence in
evidence that the beneficial effects of CR were changing support of exercise-based CR.
over time. However, with the advent and acceptance of Our study has several potential limitations. First,
these other evidence-based secondary preventative because of the nature of the intervention, all included
therapies, it is clear that the effect sizes of the benefits RCTs were open label. We attempted to control for this
seen in our study are unlikely to represent the effect sizes by focusing on objective end points, including mortality
that would be expected from exercise-based CR in and reinfarction, and used sample size and duration of
modern practice. These concerns are not limited to CR, follow-up time as proxies for study quality in our
as evidence for several commonly used secondary secondary analyses. Furthermore, although not always
prevention therapies is derived from data collected noted, some RCTs reported that outcomes were adjudi-
some years ago in the absence of the full modern cated by investigators blinded to the subjects' random-
compliment of therapies. Nonetheless, although also true ization. Second, because of the limited number of RCTs
for other therapies, this may be particularly true for examining each outcome and their relatively small
exercise-based CR. In addition, it is important to number of events, some of our subgroup analyses had
emphasize that these results are drawn from a population wide 95% CIs. Third, approximately 89% of included
of patients who were predominantly male and who were patients were men. As discussed above, additional
generally younger than the average age of presentation studies are required among women and older patients
for patients with MI currently. These demographic post-MI. Fourth, our inclusion criteria were relatively
characteristics highlight a shortcoming of the CR broad to permit our search strategy to capture all relevant
literature, and as it is becoming clear in many other articles, as this is a field dominated by numerous small
fields of cardiology as well, women and older patients trials. As such, there was heterogeneity in study design,
represent very important demographics in whom populations investigated, and intervention examined,
dedicated study is very much needed. including variability in the type and quality of CR
Two early meta-analyses from the 1980s suggested program. We therefore used random effects models to
benefit of exercise-based CR among patients after MI. 63,64 account for both within- and between-study variability.
In 2003, one of these groups updated their analysis and Furthermore, through our stratified analyses, we inves-
showed statistically significant reductions in cardiac and tigated potential sources of heterogeneity and their
all-cause mortality. 5 Our findings are consistent with influence on the effects of CR programs on clinical
these previous results but include a significant reduction outcomes. With specific respect to the inclusion criteria
in reinfarction. This would be expected to be the allowing RCTs using b4 weeks of CR, we would point out
underlying cause of the observed reductions in cardiac that only 1 RCT used an exercise duration b4 weeks, 44
and all-cause mortalities but was not found in this and this study contributed only N = 28 patients to the
previous meta-analysis. Likely, this arose because in total pool of 6,111 patients studied, and hence, it is
previous meta-analyses, the risk of reinfarction was unlikely that changing the minimum program duration to
determined only for nonfatal reinfarction, whereas we 4 weeks would have significantly affected our results.
used a composite of fatal and nonfatal reinfarction. We Finally, data for cardiovascular risk factors were often not
American Heart Journal
582 Lawler, Filion, and Eisenberg October 2011
presented in a poolable manner. Consequently, these 4. Leon AS, Franklin BA, Costa F, et al. Cardiac rehabilitation and
data were systematically reviewed but not formally secondary prevention of coronary heart disease: an American Heart
pooled. These data were also generally restricted to Association scientific statement from the Council on Clinical
Cardiology (Subcommittee on Exercise, Cardiac Rehabilitation, and
subjects who attended follow-up visits and thus are
Prevention) and the Council on Nutrition, Physical Activity, and
presented using a modified intention-to-treat approach,
Metabolism (Subcommittee on Physical Activity), in collaboration with
which may lead to bias. Nonetheless, we have provided the American association of Cardiovascular and Pulmonary Reha-
these data for completeness. bilitation. Circulation 2005;111:369-76.
5. Taylor RS, Brown A, Ebrahim S, et al. Exercise-based rehabilitation
for patients with coronary heart disease: systematic review and meta-
Conclusion analysis of randomized controlled trials. Am J Med 2004;116:
Exercise-based CR reduces the risk of reinfarction and 682-92.
cardiac, cardiovascular, and all-cause mortality when 6. Reid RD, Dafoe WA, Morrin L, et al. Impact of program duration and
used as secondary prevention post-MI. Even if more contact frequency on efficacy and cost of cardiac rehabilitation:
results of a randomized trial. Am Heart J 2005;149:862-8.
modest benefits are obtained when CR is used in
7. Thompson DR, Clark AM. Cardiac rehabilitation: into the future.
everyday current practice, our results suggest that
Heart 2009;95:1897-900.
exercise-based CR is efficacious for secondary preven- 8. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for
tion. Our examination of RCT-level characteristics systematic reviews and meta-analyses: the PRISMA statement. BMJ
suggests that even short-term CR programs post-MI 2009;b2535:339.
could be of benefit and that the overall observed benefits 9. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports
persisted beyond the period of active intervention. These of randomized clinical trials: is blinding necessary? Control Clin Trials
results encourage future RCTs to compare the efficacy of 1996;17:1-12.
short- and long-term CR programs for the secondary 10. Heller RF, Knapp JC, Valenti LA, et al. Secondary prevention after
acute myocardial infarction. Am J Cardiol 1993;72:759-62.
prevention of cardiovascular events.
11. Kallio V, Hamalainen H, Hakkila J, et al. Reduction in sudden deaths
by a multifactorial intervention programme after acute myocardial
infarction. Lancet 1979;2:1091-4.
Acknowledgements 12. Teo KK, Yusuf S, Furberg CD. Effects of prophylactic antiarrhythmic
We would like to thank Tara Dourian for her help with drug therapy in acute myocardial infarction. An overview of
data abstraction. results from randomized controlled trials. JAMA 1993;270:
1589-95.
13. Pfeffer MA, Braunwald E, Moye LA, et al. Effect of captopril on
Disclosures mortality and morbidity in patients with left ventricular dysfunction
The authors declare no conflict of interest. after myocardial infarction. Results of the survival and ventricular
enlargement trial. The SAVE Investigators. N Engl J Med 1992;327:
669-77.
14. Ambrosioni E, Borghi C, Magnani B. The effect of the angiotensin-
References converting-enzyme inhibitor zofenopril on mortality and morbidity
1. Antman EM, Anbe DT, Armstrong PW, et al. ACC/AHA guidelines
after anterior myocardial infarction. The Survival of Myocardial
for the management of patients with ST-elevation myocardial
Infarction Long-Term Evaluation (SMILE) Study Investigators. N Engl J
infarctionexecutive summary: a report of the American College of
Med 1995;332:80-5.
Cardiology/American Heart Association Task Force on Practice 15. Ballantyne FC, Clark RS, Simpson HS, et al. The effect of
Guidelines (Writing Committee to Revise the 1999 Guidelines for the moderate physical exercise on the plasma lipoprotein subfractions
Management of Patients With Acute Myocardial Infarction). Circu- of male survivors of myocardial infarction. Circulation 1982;65:
lation 2004;110:588-636. 913-8.
2. Anderson JL, Adams CD, Antman EM, et al. ACC/AHA 2007 16. Bengtsson K. Rehabilitation after myocardial infarction. A controlled
guidelines for the management of patients with unstable angina/ study. Scand J Rehabil Med 1983;15:1-9.
non ST-elevation myocardial infarction: a report of the American 17. Bertie J, King A, Reed N, et al. Benefits and weaknesses of a cardiac
College of Cardiology/American Heart Association Task Force on rehabilitation programme. J R Coll Physicians Lond 1992;26:
Practice Guidelines (Writing Committee to Revise the 2002 147-51.
Guidelines for the Management of Patients With Unstable Angina/ 18. Bethell HJ, Mullee MA. A controlled trial of community based
Non ST-Elevation Myocardial Infarction): developed in collabora- coronary rehabilitation. Br Heart J 1990;64:370-5.
tion with the American College of Emergency Physicians, the 19. Carlsson R, Lindberg G, Westin L, et al. Influence of coronary nursing
Society for Cardiovascular Angiography and Interventions, and the management follow up on lifestyle after acute myocardial infarction.
Society of Thoracic Surgeons: endorsed by the American Associ- Heart 1997;77:256-9.
ation of Cardiovascular and Pulmonary Rehabilitation and the 20. Carson P, Phillips R, Lloyd M, et al. Exercise after myocardial
Society for Academic Emergency Medicine. Circulation 2007;116: infarction: a controlled trial. J R Coll Physicians Lond 1982;16:
e148-304. 147-51.
3. Suaya JA, Shepard DS, Normand SL, et al. Use of cardiac 21. DeBusk RF, Miller NH, Superko HR, et al. A case-management system
rehabilitation by Medicare beneficiaries after myocardial infarction for coronary risk factor modification after acute myocardial
or coronary bypass surgery. Circulation 2007;116:1653-62. infarction. Ann Intern Med 1994;120:721-9.
American Heart Journal
Volume 162, Number 4
Lawler, Filion, and Eisenberg 583
22. Dugmore LD, Tipson RJ, Phillips MH, et al. Changes in cardiorespi- one-year follow-up with magnetic resonance imaging. Am Heart J
ratory fitness, psychological wellbeing, quality of life, and vocational 2000;139:252-61.
status following a 12 month cardiac exercise rehabilitation pro- 40. Naughton J, Dorn J, Oberman A, et al. Maximal exercise systolic
gramme. Heart 1999;81:359-66. pressure, exercise training, and mortality in myocardial infarction
23. Fridlund B, Hogstedt B, Lidell E, et al. Recovery after myocardial patients. Am J Cardiol 2000;85:416-20.
infarction. Effects of a caring rehabilitation programme. Scand J 41. Oldridge N, Guyatt G, Jones N, et al. Effects on quality of life with
Caring Sci 1991;5:23-32. comprehensive rehabilitation after acute myocardial infarction. Am J
24. Giallauria F, Lucci R, De LA, et al. Favourable effects of exercise Cardiol 1991;67:1084-9.
training on N-terminal probrain natriuretic peptide plasma levels in 42. Oldridge N, Gottlieb M, Guyatt G, et al. Predictors of health-related
elderly patients after acute myocardial infarction. Age Ageing 2006; quality of life with cardiac rehabilitation after acute myocardial
35:601-7. infarction. J Cardiopulm Rehabil 1998;18:95-103.
25. Giannuzzi P, Temporelli PL, Corra U, et al. Attenuation of unfavorable 43. Oldridge N, Streiner D, Hoffmann R, et al. Profile of mood states and
remodeling by exercise training in postinfarction patients with left cardiac rehabilitation after acute myocardial infarction. Med Sci
ventricular dysfunction: results of the Exercise in Left Ventricular Sports Exerc 1995;27:900-5.
Dysfunction (ELVD) trial. Circulation 1997;96:1790-7. 44. Oya M, Itoh H, Kato K, et al. Effects of exercise training on the
26. Giannuzzi P, Tavazzi L, Temporelli PL, et al. Long-term physical recovery of the autonomic nervous system and exercise capacity after
training and left ventricular remodeling after anterior myocardial acute myocardial infarction. Jpn Circ J 1999;63:843-8.
infarction: results of the Exercise in Anterior Myocardial Infarction 45. P.RE.COR Group. Comparison of a rehabilitation programme, a
(EAMI) trial. EAMI Study Group. J Am Coll Cardiol 1993;22: counselling programme and usual care after an acute myocardial
1821-9. infarction: results of a long-term randomized trial. P.RE.COR. Group.
27. Hamalainen H, Luurila OJ, Kallio V, et al. Reduction in sudden deaths Eur Heart J 1991;12:612-6.
and coronary mortality in myocardial infarction patients after 46. Sivarajan ES, Bruce RA, Lindskog BD, et al. Treadmill test responses to
rehabilitation. 15 year follow-up study. Eur Heart J 1995;16: an early exercise program after myocardial infarction: a randomized
1839-44. study. Circulation 1982;65:1420-8.
28. Heldal M, Sire S, Dale J. Randomised training after myocardial 47. Specchia G, De SS, Scire A, et al. Interaction between exercise
infarction: short and long-term effects of exercise training after training and ejection fraction in predicting prognosis after a first
myocardial infarction in patients on beta-blocker treatment. A myocardial infarction. Circulation 1996;94:978-82.
randomized, controlled study. Scand Cardiovasc J 2000;34:59-64. 48. Stern MJ, Gorman PA, Kaslow L. The group counseling v exercise
29. Holmback AM, Sawe U, Fagher B. Training after myocardial therapy study. A controlled intervention with subjects following
infarction: lack of long-term effects on physical capacity and myocardial infarction. Arch Intern Med 1983;143:1719-25.
psychological variables. Arch Phys Med Rehabil 1994;75:551-4. 49. Taylor CB, Houston-Miller N, Haskell WL, et al. Smoking cessation
30. Karvetti RL, Knuts LR. Effects of comprehensive rehabilitation on after acute myocardial infarction: the effects of exercise training.
weight reduction in myocardial infarction patients. Scand J Rehabil Addict Behav 1988;13:331-5.
Med 1983;15:11-6. 50. Taylor CB, Houston-Miller N, Ahn DK, et al. The effects of exercise
31. La Rovere MT, Bersano C, Gnemmi M, et al. Exercise-induced training programs on psychosocial improvement in uncomplicated
increase in baroreflex sensitivity predicts improved prognosis after postmyocardial infarction patients. J Psychosom Res 1986;30:581-7.
myocardial infarction. Circulation 2002;106:945-9. 51. Vermeulen A, Lie KI, Durrer D. Effects of cardiac rehabilitation after
32. Lee BC, Chen SY, Hsu HC, Su MY, et al. Effect of cardiac myocardial infarction: changes in coronary risk factors and long-term
rehabilitation on myocardial perfusion reserve in postinfarction prognosis. Am Heart J 1983;105:798-801.
patients. Am J Cardiol 2008;101:1395-402. 52. Zheng H, Luo M, Shen Y, et al. Effects of 6 months exercise training on
33. Lee BC, Hsu HC, Tseng WY, et al. Effect of cardiac rehabilitation on ventricular remodelling and autonomic tone in patients with acute
angiogenic cytokines in postinfarction patients. Heart 2009;95: myocardial infarction and percutaneous coronary intervention.
1012-8. J Rehabil Med 2008;40:776-9.
34. Lewin B, Robertson IH, Cay EL, et al. Effects of self-help post 53. Chow CK, Jolly S, Rao-Melacini P, et al. Association of diet, exercise,
myocardial-infarction rehabilitation on psychological adjustment and and smoking modification with risk of early cardiovascular events
use of health services. Lancet 1992;339:1036-40. after acute coronary syndromes. Circulation 2010;121:750-8.
35. Marchionni N, Fattirolli F, Fumagalli S, et al. Improved exercise 54. Witt BJ, Jacobsen SJ, Weston SA, et al. Cardiac rehabilitation after
tolerance and quality of life with cardiac rehabilitation of older myocardial infarction in the community. J Am Coll Cardiol 2004;44:
patients after myocardial infarction: results of a randomized, 988-96.
controlled trial. Circulation 2003;107:2201-6. 55. Lloyd-Jones D, Adams R, Carnethon M, et al. Heart disease and
36. Marra S, Paolillo V, Spadaccini F, et al. Long-term follow-up after a stroke statistics2009 update: a report from the American Heart
controlled randomized post-myocardial infarction rehabilitation Association Statistics Committee and Stroke Statistics Subcommittee.
programme: effects on morbidity and mortality. Eur Heart J 1985;6: Circulation 2009;119:e21-181.
656-63. 56. Patel NB, Balady GJ. The rewards of good behavior. Circulation
37. Miller NH, Haskell WL, Berra K, et al. Home versus group exercise 2010;121:733-5.
training for increasing functional capacity after myocardial infarc- 57. Menzin J, Wygant G, Hauch O, et al. One-year costs of ischemic
tion. Circulation 1984;70:645-9. heart disease among patients with acute coronary syndromes:
38. Myers J, Gianrossi R, Schwitter J, et al. Effect of exercise training on findings from a multi-employer claims database. Curr Med Res Opin
postexercise oxygen uptake kinetics in patients with reduced 2008;24:461-8.
ventricular function. Chest 2001;120:1206-11. 58. Thompson DR, Lewin RJ. Coronary disease. Management of the post
39. Myers J, Goebbels U, Dzeikan G, et al. Exercise training and myocardial infarction patient: rehabilitation and cardiac neurosis.
myocardial remodeling in patients with reduced ventricular function: Heart 2000;84:101-5.
American Heart Journal
584 Lawler, Filion, and Eisenberg October 2011
59. de Waard MC, van d V, Bito V, et al. Early exercise training normalizes 62. Xu X, Wan W, Powers AS, et al. Effects of exercise training on cardiac
myofilament function and attenuates left ventricular pump dysfunction in function and myocardial remodeling in post myocardial infarction
mice with a large myocardial infarction. Circ Res 2007;100:1079-88. rats. J Mol Cell Cardiol 2008;44:114-22.
60. Schober T, Knollmann BC. Exercise after myocardial infarction 63. O'Connor GT, Buring JE, Yusuf S, et al. An overview of randomized
improves contractility and decreases myofilament Ca2+ sensitivity. trials of rehabilitation with exercise after myocardial infarction.
Circ Res 2007;100:937-9. Circulation 1989;80:234-44.
61. Wan W, Powers AS, Li J, et al. Effect of post-myocardial infarction 64. Oldridge NB, Guyatt GH, Fischer ME, et al. Cardiac rehabilitation
exercise training on the renin-angiotensin-aldosterone system and after myocardial infarction. Combined experience of randomized
cardiac function. Am J Med Sci 2007;334:265-73. clinical trials. JAMA 1988;260:945-50.
American Heart Journal
Volume 162, Number 4
Lawler, Filion, and Eisenberg 584.e1
Description
Online of MEDLINE
Appendix literature search
1. Description conducted
of MEDLINE on June
literature 13,conducted
search 2010 on June 13, 2010
Search no. Search description No. of results
Online
Effect ofAppendix 2. Effect
exercise-based CR onofsmoking
exercise-based CR on
and blood smoking
pressure and blood
among pressure
patients among patients post-MI
post-MI
Smoking (n/N) Blood pressure (systolic/diastolic) (Mean SD)
Study Year Baseline Follow-up Baseline Follow-up Baseline Follow-up Baseline Follow-up
Data are only presented for studies that reported data for smoking or blood pressure at follow-up. Duration of follow-up for each trial is described in Table I.
Cardiovascular risk factor follow-up data are generally presented for those who attended the follow-up clinic visit; consequently, these data are presented using a modified intention-
to-treat approach.
This study included 2 intervention groups (group B1, exercise-only CR; group B2, comprehensive CR including exercise) and 1 duplicate control group.
American Heart Journal
584.e2 Lawler, Filion, and Eisenberg October 2011
Effect ofAppendix
Online exercise-based CR onofbody
3. Effect weight and
exercise-based CRtotal cholesterol
on body weightamong patients
and total post-MI
cholesterol among patients post-MI
Body weight (Mean SD)(kg) Total cholesterol (Mean SD) [mg/dL]
Study Year Baseline Follow-up Baseline Follow-up Baseline Follow-up Baseline Follow-up
Data are only presented for studies that reported follow-up data for body weight or total cholesterol. Duration of follow-up for each trial is described in Table I.