Comparison of Supervised and Unsupervised Exercise Training After Coronary Bypass Surgery

RICHARD STEVENS, DHSc, and PETER HANSON, MD

Functional capacity and cardiovascular responses to serial graded treadmill testing (GXT) were com- pared in 180 patients who performed prescribed unsupervised exercise and 24 patients who were referred for supervised exercise after coronary ar- tery bypass surgery (CABS). The groups were men similar in age range, number of bypass grafts, pre- operative left ventricular impairment and number of days hospitalized. All patients received similar predischarge exercise monitoring and began a progressive home walking or cycling program. Initial GXT (Tl) was performed 44 f 9 days postopera- tively. Both groups were instructed to continue prescribed exercise at 75 to 85% maximal heart

rate (HR) for 30 to 40 minutes 3 days (supervised) or 5 days (unsupervised) per week. The second GXT (T2) was performed 115 f 27 days after CABS. In each group there were significant (p <O.Ol) in- creases in exercise capacity and HR from Tl to T2. However, there were no significant differences in maximal exercise capacity and HR between groups at Tl or T2. Improvement in functional capacity was not influenced by therapeutic p blockade. These findings indicate that prescribed unsupervised ex- ercise can be performed safely and results in similar functional improvements compared with supervised exercise after uncomplicated CABS.

(Am J Cardiol 1984;53:1524-1528)

Functional and symptomatic improvements after cor- onary artery bypass surgery (CABS) have been well documented. Most studies report improvement in ex- ercise tolerance, left ventricular ejection fraction, ischemic electrocardiographic changes and decreases in angina1 symptoms with exercise.1-5 The additional effects of exercise training after CABS has also been reported. These studies also indicate a significant in- crease in measured exercise capacity or oxygen con- sumption in trained compared with nontrained CABS patients.cml’ Although supervised exercise training is recommended for CABS patients, most smaller com- munities lack facilities and personnel to support a tra- ditional cardiac rehabilitation program. Unsupervised exercise is frequently the only available alternative for many CABS patients.

Recent studies have shown that unsupervised pre- scribed exercise is safely performed during the initial 11 weeks after uncomplicated myocardial infarction (MI) and improvement in exercise capacity is compa- rable to patients with MI enrolled in a supervised ex-

From the Cardiology Section, Department of Medicine, University of Wisconsin, Clinical Science Center, Madison, Wisconsin. This study was supported in part by a gift from The Rennebohm Foundation to the University of Wisconsin Medical School. Madison, Wisconsin. Manu- script received September 20, 1983; revised manuscript received January 23, 1984, accepted January 26, 1984.

Address for reprints: Peter Hanson, MD, University of Wisconsin, Clinical Science Center, Madison, Wisconsin 53792.

1524

ercise program.r2J3 However, there are no equivalent data that document unsupervised exercise training after CABS.14J5 This study compares the effectiveness of unsupervised and supervised exercise training programs in a series of CABS patients during the initial 3 to 4 months of postoperative recovery.

Methods

Patients: We reviewed the results of postoperative exercise testing and training in 515 consecutive male patients who underwent CABS at the University Hospital from June 1979 to December 1982. A study group of 204 patients was selected by the following criteria: age range 40 to 69 years; completion of serial graded exercise testing (GXT) at 4 to 6 weeks and 12 to 16 weeks after CABS; and continuous participation in prescribed exercise training programs during the study period. Three hundred eleven patients were excluded: 52 because of age; 92 because of combined cardiac surgery (valve replace- ment and repair of ventricular aneurysm) or for contraindi- cations to exercise testing; and 167 patients were not seen in the follow-up period because of geographic reasons, physician referral preference or other nonmedical factors.

Patients gave informed consent for all exercise tests and for participation in supervised exercise programs.

Predischarge evaluation: Routine cardiac rehabilitation began during the initial 48 to 72 hours after CABS and in- cluded progressive assisted ambulation, monitored treadmill or cycle ergometer exercise and patient education classes. Exercise intensity was initially maintained in a range of 2 to 3 metabolic units (METS) where 1 MET is equal to the oxygen consumption at rest: 3.5 ml Os/kg/min.16

June 1, 1984 THE AMERICAN JOURNAL OF CARDIOLOGY Volume 53 1525

TABLE I Predischarge Characteristics of Patients After Coronary Artery Bypass Surgery

Age Deciles (yr) Study

Groups 40-49 50-59

Total Population U s U s U S

ige (yr) 204 55 55 180 24 56 42 45 55 84 53 5: f7.7 f7.5 f8.9 f2.6 f2.5 f2.9 f2.5

Hospital days postop 7.6 7.7 6.6 6.9 6.3 7.0 6.8 f4.4 54.6 f0.9 f2.5 fl.7 f2.2 60.4

Arteries bypassed 3.4 3.4 3.3 3.5 2.6’ 3.3 3.7 fl.1 fl.1 fl.O 51.2 61.3 fl.O 60.9

Patients with left 56 56 54 60 29” 57 87 ventricular impairment (“/)

60-69

U S

:4 ;: f3.1 f3.0

9.4 6.6 f7.3 f0.5

3.3 3.5 fl.1 f0.7

52 64

l p <0.05 (unpaired t test). Values are mean f 1 standard deviation. S = supervised; U = unsupervised.

At the time of discharge all patients were instructed to walk or cycle on an exerciser at an intensity that produced a heart rate (HR) increase of 20 beats/min above standing-resting HR. For patients receiving /3-adrenoreceptor blockade therapy (p blockade), an increase in HR 10 beats above standing-resting HR was assigned. Both training HRs were found to coincide with a perceived exertion rating of 12 to 13 on the Borg scale of 6 to 20.17 The initial duration of exercise was 5 minutes twice daily and progressed by adding 1 minute daily, for a total of 15 to 20 minutes twice daily.

Follow-up exercise testing: For the first 4 to 6 weeks after discharge, all 204 study patients were encouraged to follow the predischarge exercise prescription. At the end of this period, 180 patients returned by referral to the University Hospital for serial outpatient evaluation and instruction for unsuper- vised exercise training. A second group of 24 were referred by their primary physician to 1 of the 2 medically supervised cardiac rehabilitation programs in Madison, Wisconsin.

Treadmill exercise testing for both groups was conducted with modified Balke treadmill protocols that used approxi- mately l-MET increases in work intensity every 2 minutes.16 Only touch contact with handrails was permitted during treadmill testing. The electrocardiogram was monitored continuously and blood pressure measured every minute by auscultation.

Patients were encouraged to exercise to an end point of subjective fatigue that corresponded to a perceived exertion level of 17 to 18 on the Borg scale. The usually accepted cri- teria for test termination due to abnormal symptoms, blood pressure responses or dysrhythmias were observed. Maximal exercise capacity was determined from the treadmill speed and grade completed during the final 2-minute stage.16

Patients who could not perform treadmill exercise were evaluated with a cycle ergometer exercise protocol that used 150-kpm increases in work intensity every 2 minutes. Moni- toring techniques and end points were similar. MET values were calculated from the maximal number of kilopond meters adjusted for body weight.l”

Exercise prescription and training: Exercise prescrip- tions for both groups were determined from maximal HR and METS achieved on the initial exercise tests (Tl). The training HR range was calculated as 75 to 85% of maximal HR. Training MET range was 60 to 75% of maximal MET value. After completing GXT, patients continuing with unsupervised exercise were monitored for 4 to 5 minutes on a treadmill or bicycle at the calculated training MET and HR levels to confirm that subjective tolerance and cardiovascular responses corresponded to the calculated exercise prescription.

Unsupervised exercise was performed by vigorous walking at a pace of 3.0 to 4.0 mph or with a stationary cycle exerciser for 30 to 40 minutes 5 days/week. Patients were instructed to monitor training pulse rate and perceived exertion level (12 to 14) to maintain assigned exercise intensity. Outdoor exer- cise was encouraged when weather conditions were suitable. Many patients used local shopping malls for exercise walks.

Supervised exercise programs were conducted by qualified exercise specialists 3 days/week. Each l-hour session included warm-up, dynamic and strength activities and cool-down. Dynamic training lasted for 30 to 40 minutes. Electrocardio- graphic monitoring was used for the initial 5 to 10 sessions and subsequently as indicated. Periodic adjustments in training intensity were made according to individual patient pro- gression. The overall patient attendance rate for supervised exercise sessions was 70%.

Medications: No attempt was made to alter cardiovascular medications during exercise testing or training. All patients were maintained on digoxin (0.125 to 0.25 mg/day) and pro- pranolol(l0 to 20 mg/day) during the immediate 3 to 4 weeks after CABS to control supraventricular dysrhythmias. Digoxin was usually discontinued; however, the decision to maintain /3 blockade during the subsequent 2 to 3 months varied with individual patients and their physicians.

Data analysis: Data were entered and stored on the Wis- consin Storage and Retrieval System (WISAR).18 Data analysis utilized standard descriptive techniques. For com- parisons, the following statistical tests were used: chi-square (categorical clinical characteristics), unpaired t test (between group comparisons), and paired t test (within group com- parisons). A value (Y 10.05 was considered significant. Sta- tistical comparisons were also determined for 3 age deciles (40 to 49,50 to 59 and 60 to 69) within and between groups.

Results

Predischarge comparisons (Table I): Predischarge hospitalization characteristics of each group are sum- marized in Table I. There were no significant differences between the supervised and unsupervised groups for age, hospital days after CABS, number of vessels by- passed, and percent of patients with angiographic evi- dence of left ventricular wall motion impairment (graded mild or moderate) on preoperative cardiac catheterization. Similar results were obtained for the 3 age deciles except for the supervised subgroup age 40 to 49 years, which had significantly fewer vessels

1526 EXERCISE TRAINING AFTER CORONARY BYPASS SURGERY

TABLE II Exercise Test Results at Follow-Up*

Age Deciles (yr) Study

Groups 40-49 50-59 60-69 Total

Population U S U S LJ S U S

204 180 5: 42 54 11 61 62 71

:3 :: :7 57 55

44 44 45 44 38 43 49 45 47 f9 f9 514 It8 f5 zt9 117 f9 f15

METS 7.3 7.3 7.5 8.3 8.5 7.4 8.0 6.3 6.6 f1.8 il.8 f1.9 f1.8 f1.9 f1.5 f2.1 f1.6 f1.5

RPP (X102) 248 248 242 263 269 252 225 231 235 f58 f57 f65 f56 f50 f55 f97 f59 f53

Max HR fbeats/min) 135 136 132 145 142 137 130 127 128 619 f19 It19 f16 f18 f18 f27 f21 f13

Max SBP (mm Hg) 181 181 179 180 183 182 167 181 183 f30 f30 f32 f29 f29 f30 f39 f31 f30

l There were no significant differences between groups. CABS = coronary bvnass suroery: HR = heart rate: Max = maximal; METS = metabolic units (rest = 1 MET); RPP = max HR X max SBP;

S = supervised; SBP G systolicblood pressure; U = unsupervised. Values are mean f 1 standard deviation.

bypassed and less impairment of ventricular wall motion.

Initial follow-up evaluation (Table II): Initial exercise tests (Tl) was performed at 44 f 9 days after CABS. There were no significant differences in METS, maximal HR, maximal systolic blood pressure (SBP), rate-pressure product, days after CABS or percent of patients receiving p blockade therapy between the su- pervised and unsupervised groups and their age de- tiles.

Final follow-up evaluation (Table III): The second exercise tests (TZ) were performed 115 f 27 days after the initial tests (Tl). There were no significant differ- ences in METS, maximal HR, rate-pressure product, days after CABS or percentage of patients on p block- ade. The supervised group had significantly lower maximal SBP, which appears to be due to the pre- dominantly lower maximal SBP in the 60 to 69-year age decile.

The mean increase in exercise capacity from Tl to T2 for each group was 1.4 f 1.5 METS. The absolute MET level achieved was greatest for age decile 40 to 49 years and declined with increasing age. However, the relative increase from Tl to T2 was remarkably constant across all age deciles.

Effect of beta blockade (Fig. 1): The potential ef- fects of 0 blockade on training responses in the unsu- pervised group was determined by comparing patients who continued to receive P-blocker therapy (n = 87) with those who discontinued P-blocker therapy (n = 67) during the training period. Patients who received P-blocker therapy were treated with 40 to 80 mg/day of propranolol or the equivalent dosage of nadolol or me- toprolol. No significant differences in Tl or T2 MET capacity were found between the p blockade subgroup and the untreated subgroup or the 3 age deciles of each subgroup.

Effect of left ventricular impairment (Fig. 2): The role of preoperative left ventricular impairment on ex- ercise training was also evaluated in the unsupervised group. Patients with impaired ventricular wall motion (n = 101) showed equal improvement in exercise ca-

pacity than patients with normal ventricular wall mo- tion (n = 79). However, in age decile 40 to 49 years, pa- tients with ventricular impairment had significantly lower initial and final exercise capacity although the relative change from Tl to T2 was equal to the subgroup with normal ventricular wall motion.

Morbidity and mortality: No deaths were reported in either the supervised or unsupervised training groups during the study period.

In the unsupervised group abnormal GXT responses occurred in 12 patients at Tl (angina, 3 patients; dyspnea, 6; and complex ventricular ectopy, 3) and 8 patients during T2 (angina, 3; dyspnea, 5). One unsu- pervised patient sustained an uncomplicated posterior

11 I- Meant I SEM

10 -

cn g- t- w > 8-

7-

6-

*INS

/ I

lncreose T, * P4.001

to T2

Op- Blockade

0 Nonp-Blockade

I 1 I I I I I 1

TI T2 TI T2 TI T2 TI T2

Total 40-49 50-59 60-69

Group Years

FIGURE 1. Effect of @-adrenergic blockade on exercise capacity of the unsupervised training group. There are no significant differences in maximal exercise capacity (MET) level achieved between patients receiving and those not receiving p blockade at initial (Tl) and follow-up (T2) exercise tests. SEM = standard error of the mean.

June 1, 1984 THE AMERICAN JOURNAL OF CARDIOLOGY Volume 53 1527

TABLE III Exercise Test Results at Follow-Up

Study Groups

Total Population U S

! blockade (%) 204 51 55 180 24 42 Days after CABS 115 115 118

f27 f26 f33 METS 8.7 8.7 9.0

zt2.1 f2.1 fl.6 RPP (X102) 277 278 266

f64 f62 581 Max HR (beatsimin) 144 144 147

zt22 f21 626 Max SBP (mm Hg) 192 194 177’

f30 f29 f31

40.-49 ~--

U S

42 55 :3 116 111 f22 f31

9.7 9.8 62.5 fl.3

294 306 f56 f83

153 160 f19 f29

193 189 f28 f28

Age Deciles (yr)

50-59

U S

z: !O 114 115 f27 f28

8.8 9.1 fl.8 f1.8

286 272 f59 f90

146 149 fl9 f25

196 179 f28 f34

60-69

U S

54 54 :: 114 124 f27 f38

7.7 8.3 fl.8 fl.4

255 238 f65 f69

134 138 f21 f23

192 168+ f32 f31

l p CO.01 (unpaired t test); + p <0.05 (unpaired t test). Values are mean f 1 standard deviation. Abbreviations as in Tables I and II.

wall myocardial infarction shortly after an apparently normal GXT. The supervised group had no major ab- normalities during GXT or Tl or T2.

The statistical summaries for the unsupervised group were not influenced by inclusion or exclusion of the symptom-terminated GXT data.

Discussion

Our data indicate that CABS patients who follow a prescribed unsupervised exercise program achieve sig- nificant improvements in functional capacity and car- diovascular responses to exercise testing. The im- provements were equal in relative change and absolute magnitude to a clinically comparable group of CABS patients referred to a supervised exercise program for a similar duration and pattern of training.

Comparison with previous studies: Selected re- ports of exercise training in patients with coronary heart disease (CABS and MI) are summarized in Table IV. These studies were chosen for similarity of patient age, pretraining exercise capacity, and the methods and duration of exercise training. Our results compare fa- vorably with supervised exercise studies in CABS pa- tients that report a 19 to 29% increase in exercise ca- pacity after 3 to 6 months of training. This response appears to be independent of the time of initiation of training after CABS.7-g

Unsupervised and supervised exercise training have been studied in patients who have had MI. DeBusk et all2 compared 28 patients referred to a supervised program with 12 patients assigned to unsupervised home bicycle exercise training, monitored by a tele- phone-transmitted electrocardiogram 3 times weekly. Both groups showed similar increases in exercise ca- pacity over a period of 2 months. In addition, a third group of patients with MI who were advised to follow ad lib activity showed a significant but lesser improve- ment in exercise capacity over the same period.

To our knowledge there are no previous studies comparing supervised and unsupervised exercise training in CABS patients. Serial exercise testing data from a multicenter study of CABS patients reported by Weiner et al* show a significant increase in exercise

capacity from before CABS (5.4 METS) to 6 months after CABS (7.5 METS) with a further increase at 19 to 36 months (8.4 METS) and 37 to 48 months (9.3 METS). These results probably reflect the composite effects of spontaneous exercise, physician advised ex- ercise and referral to supervised exercise programs.

Effect of beta blockade: We found no significant effect of fl blockade on exercise capacity in the unsu- pervised group. Increases in maximal HR, maximal SP and maximal RPP in both groups from Tl to T2 may be attributed to fewer patients who remained on p blockade. Our results agree with controlled studies that indicate normal exercise training effects can be achieved in patients with heart disease maintained on small-dose /3 blockade, despite a reduced training heart rate.lg

I I meon? I SEM

IO

** p<.oo1 I I I I 1 1 1 1

TI T2 TI T, TI T2 TI T2

Total 40-49 50-59 60-69 Group

Years

FIGURE 2. Exercise capacity (METS) of unsupervised exercise subgroups with normal and impaired preoperative left ventricular (LV) wall motion. Both subgroups show equal increases in MET capacity from Tl to T2. However, the absolute MET capacity was significantly de- creased in the age range 40 to 49 years with impaired LV wall motion. SEM = standard error of the mean.

1526 EXERCISE TRAINING AFTER CORONARY BYPASS SURGERY

TABLE IV Summary of Supervised and Unswervised Trainina Studies

Author Clinical No. of Age

Characteristics Pts (yr)

Functional Capacity Training Conditions (METS)

Time from A Post CABS/MI Format Duration Tl T2 (%)

Oldridge” CABS 6 49 16 wk S 4mo 7.1 9.0 27 1978

DeBusk12 M”I 28 5552 3 wk u” 2mo 6.8 10.4 53 1979 3 wk 2mo 7.3 10.3

Hartung7 CL :: 53 18 mo ss 3-6 mo 7.3 9.1 t: 1981 IO

z: 9mo 3-6 mo 8.0 10.3 29

Dornang CABS 210 8 wk S 3 mo 6.3 7.5 19 1982

Present CABS 180 6 wk U 2-3 mo 7.1 8.5 study CABS 24 z: 7 wk S 2-3 mo 7.4 9.0 Z8

CABS = coronary artery bypass surgery; MI = myocardial infarction; S = supervised; Tl = pretraining treadmill test; T2 = posttraining treadmill test; U = unsupervised; n = change.

Limitations of study: Several limitations of our study are apparent. Our data were derived by compar- ing the outcome of 2 referral options determined by primary physicians. The patients were not prospectively assigned to the 2 training groups and the number of supervised patients was substantially smaller than the number of unsupervised patients. However, comparison of preoperative and discharge characteristics, (age, number of vessels bypassed, left ventricular impairment and discharge day after CABS) indicates no important intergroup differences attributable to unequal numbers or nonrandom assignment to supervised or unsuper- vised exercise. In addition, the study population did not differ significantly in these characteristics compared with the 167 patients who were not included in this study because of geographic reasons or physician re- ferral pattern.

We could not document directly the exercise com- pliance of the unsupervised group. However, a recent survey of our patient group (Eyherabide A, unpublished thesis) revealed an 80% compliance to the discharge and follow-up prescribed exercise. This reported compliance compares favorably with the supervised exercise group that had a known attendance rate of at least 70% and with published adherence data from secondary pre- vention exercise programs.20

Finally, we did not follow a nonexercise group to de- termine the outcome of spontaneous recovery from CABS. We would anticipate a modest degree of im- provement in exercise capacity in patients who are ac- tive in household, farming or other occupation tasks. However, Oldridge reported no change in exercise ca- pacity in patients who engaged in ad lib recreation during recovery from CABS.‘l

Acknowledgmenti We gratefully acknowledge the coop- eration of Josephine Will, RN, Cardiac Rehabilitation Coor- dinator, Madison General Hospital, for providing follow-up data on supervised training patients. We also thank A. J. Liedtke, MD, for constructive criticism of the manuscript.

and Margaret Jorgensen and Sandy Yost for secretarial support.

References

1. Rahkntoola SH. Postoperative exercise responses in the evaluation of the ohvsioloaic status after coronarv bvoass suroetv. Circulation 1982:65:suooI

2.

3.

4.

5.

6.

7.

6.

9.

10.

11.

12.

13.

14.

, I. I _ ll:ll-108-11-114. Clausen JP. Circulatory adjustments to dynamic exercise and effects of physical training in normal subjects and in patients with coronary artery disease. Prog Cardiovasc Dis 1976;18:459-495. Reeves TJ. Medical management of the patient with angina pectoris: an overview of the problem. Circulation 1982;65:suppl ll:ll-3-11-13. Vlletstra RE. Chesebro JH. Frve RI.. Wallace RB. lmorovement of left ventricular exercise hemodyn&mic fbnction after aort&coronary artery bypass grafting. J Thorac Cardiovasc Surg 1981;81:85-91. Roberts AJ. Soies SM. Mevers SN. Moran JM. Sanders JH. Lichtenlhal PR, Mlch&i ii. Ea;ly aid long-berm imprdvement in lefi ventricular $.s;ymance following coronary bypass surgery. Surgery 1980;88:467-

Mui;ay GC, Belier GH. Cardiac rehabilitation following coronary artery surgery. Am Heart J 1983;105:1009-1018. Hartung GH, Rangel R. Exercise training in post-myocardial infarction pa- tients: comparison of results with high risk coronary and post-bypass pa- tients. Arch Phys Med Rehabil 1981;82:147-150. Weiner DA, McCabe CH, Roth RL, Cutler SS, Berger RL, Ryan TJ. Serial exercise testing after coronary bypass surgery. Am Heart J 1981;lOl: 149-154. Doman J, Rolko AF, Greenfield C. Factors affecting rehabilitation following aortocoronary bypass procedures. Can J Surg 1982;25:677-880. Frlck MH, Harjola PT, Valle M. Persistent improvement after coronary bypass surgery: ergometric and angiographic correlations at 5 years. Cir- culation 1983;67:491-496. Oldrldge NB, Nagle FJ, Balke B, Corllss RJ, Kahn DR. Aortocoronary by- pass surgery: effects of surgery and 32 months of physical conditioning on treadmill performance. Arch Phys Med Rehabil 1978;59:268-275. DeBusk RF, Houston N, Haskell W, Fry 0, Parker M. Exercise training soon afler myocardial infarction. Am J Cardiol 1979;44:1223-1229. Haskell WL, Savln W, Oldrldge N, DeBusk RF. Factors influencing esti- mated oxygen uptake during exercise testing soon after myocardial in- farction. Am J Cardiol 1982;50:299-303. Williams RS, Miller H, Kolsch FP, Rlblsl P, Graden H. Guidelines for un- supervised exercise in atients with ischemic heart disease. J Cardiac Rehabil 1981:1:213-21 c:

15. Council on Scjentific Affairs (AMA). physician-supervised exercise programs in rehabilitation of patients with coronary heart disease. JAMA 1981; 245:1463-1466. American College of Sports Medicine. tiuidelines tor tiraaea txerclse Testing and Exercise Prescription. Philadelphia: Lea & Febiger, 1980. Borg G, Linderholm H. Perceived exertion and pulse rate during raded exercise in various age groups. Acta Med Stand 1967;472: 194-2 8 6. Harberg J (ed). WISAR: Wisconsin Storage and Retrieval System. Madison, Wisconsin Clinical Cancer Center. University of Wisconsin, 1981. Pratt CM, Welton DE, Squires WG, Kirby TE, Hartung OH, Miller RR. Demonstration of training effect during chronic P-adrenergic blockade in patients with coronary artery disease. Circulation 1981;64:1125-1129.

20. OldrIdge NB. Compliance and exercise in primary and secondary prevention of coronary heart disease-a review. Prev Med 1982;11:56-70.

16.

17.

16.

19.