This study was performed retrospectively by analyzing patients’ EMRs and was conducted at a single center (Inje University Sanggye Paik Hospital). Patient privacy and data confidentiality were maintained throughout the research process. The Institutional Review Board of the corresponding author’s affliated Inje University approved the study (IRB No. 2020-01-008-001). The informed consent was waived.

Eligible study participants were patients diagnosed with ACS (unstable angina, ST elevation MI, and non-ST elevation MI) who underwent CABG or PCI and visited our cardiac rehabilitation clinic after discharge. Patients who met the following criteria were included in the study: (1) patients with ACS who underwent CABG or PCI between January 2009 and December 2013; (2) participants of a CR program, regardless of whether it was CBCR or HBCR; and (3) patients who underwent regular checkup and follow-up CPX tests for at least 5 years. The exclusion criteria were as follows: (1) patients who had been treated or underwent surgery for heart disease other than coronary artery disease (e.g., cardiac valve disease); (2) inability to ambulate due to physical problems (paresis induced by cerebral stroke, spinal cord injury, amputation, severe pain, dyspnea, etc.); and (3) incomplete EMR. Patients who wanted to participate in the CR program in the hospital setting were assigned to the CBCR group and those who could not participate in the hospital CR program were assigned to the HBCR group unless they were categorized as “high-risk” according to the risk classification for exercise training, as established by the American Association of Cardiovascular and Pulmonary Rehabilitation [13]. The CBCR group participated in the CR program at least three times in the hospital over a 6-to 8-week period. The HBCR group were educated in the CR exercise method immediately after discharge and performed self-exercises near their residence. The subject selection flow is shown in Fig. 1.

According to the contents of the EMRs, CR exercise was performed under supervision at a single CR center using a standardized protocol. The patients were instructed to visit our CR clinic for preliminary examination, including a CPX test, during their first visit. A real-time recording 12-channel electrocardiography (CASE; GE-Marquette, Waukesha, WI, USA), respiratory gas analyzer (Quark CPET; COSMED Co., Rome, Italy), automatic blood pressure and pulse monitor (TANGO M2; SunTech Medical Inc., Morrisville, NC, USA), and treadmill (T-2100; GE-Marqutte) were used in the CPX test. To assess changes in the cardiopulmonary capacity in the CPX test, variables such as maximal oxygen consumption (VO_2max), maximal metabolic equivalents (METs_max), oxygen consumption at anaerobic threshold (VO_2AT), submaximal rate pressure product (RPP_{at_stage3}), submaximal rate of perceived exertion (RPE_{at_stage3}), resting and maximal heart rate (HR_rest and HR_max, respectively), and resting systolic and diastolic blood pressures (SBP_rest and DBP_rest, respectively) were measured.

An exercise program based on the results of the CPX test was prescribed for the patients in the CBCR group and performed in accordance with the modified Bruce protocol within a week of discharge for the PCI cases or 4 weeks after CABG surgery. The patients attended a supervised exercise program consisting of 18 sessions, with 2–3 sessions per week, for over 2 months in a hospital setting. The exercise intensity of each session was increased in a stepwise manner on the basis of the target heart rate. The target heart rate was supposed to be 60%–85% of the heart rate reserve value, which can be calculated using the maximum and stable heart rates obtained from the CPX test. Each exercise session was divided into a 10-minute warm-up, 30-minute prescribed exercise, and 10-minute cooldown period. After completing the program, we advised the patients to participate in self-exercise programs based on their target heart rate and rate of perceived exertion measured in the follow-up CPX test.

An exercise program was prescribed for the patients in the HBCR group based on their CPX test results within a week of discharge for the PCI cases or 4 weeks after CABG surgery. The patients chose to participate in home-based exercise programs with exercise education through one or two monitoring exercises after the first CPX test in the hospital. The main reasons were time conflicts and the long distance of their residents from the hospital. However, patients who had a high risk of exercise-related cardiovascular events were not permitted to join the HBCR group. HBCR training consisted of following the exercise program with a frequency of 5–6 non-consecutive days per week and a duration of approximately 1 hour per day. The warm-up consisted of 5 minutes of light stretching, and 3–5 minutes of easy cycling, followed by an exercise session of a maximum of 40 minutes and 5–10 minutes cooldown period. The type of prescribed exercise included fast walking, power walking, cycling, and jogging, depending on the individual’s exercise capacity and systemic conditions. The initial work intensity was 60%, increasing to 85% of the maximum heart rate reserve or functional capacity. To patients were trained to check their radial pulses by counting how many times the radial artery beat every 10 seconds. The self-exercise prescriptions were the same as those for the CBCR group.

All the patients who participated in the CR program were educated on risk factor management and smoking cessation and were provided obesity [14] and nutrition consultations. In addition, the patients in both groups visited the CR clinic regularly several times during the first year after the ACS event and then every year to undergo follow-up CPX tests. To increase patient compliance with self-exercises, when the patients came for their follow-up CPX tests, their attending physician motivated them to continue their exercise, and re-education was conducted after checking the previous self-exercise methods. In addition, prescriptions for exercise methods, including the updated target heart rate based on the CPX test results were distributed to patients in the form of hand-outs; these efforts ensured that self-exercise compliance was maintained as high as possible. Study outcomes were measured using the CPX test at baseline and follow-up (3, 6, and 12 months and every year thereafter) after initiating CR. The extent of improvement and the pattern of cardiopulmonary capacity (VO_2max, METs_max, etc.) were compared between the two groups at baseline, 12 weeks, 6 months, 1 year, 3 years and 5 years.

Data were analyzed using SPSS version 25 (IBM SPSS, Armonk, NY, USA). To compare the baseline characteristics of the two groups, the Student t-test was used. A paired t-test was performed to compare the parameters, including VO_2max, VO_2AT, METs_max, RPP_{at_stage3}, and RPE_{at_stage3}, before and after the CR program. To determine the presence of an association between time and the METs_max of the various groups, a repeated-measure analysis of variance (ANOVA) model was performed. For analysis over time, the variables representing CRF (e.g., METs_max) were analyzed using a two-way repeated measures ANOVA. Statistical significance was defined as p<0.05.

The comparison of the demographic data between the two groups did not reveal any significant differences (p>0.05) (Table 1). However, the exercise capacity (VO_2max) at baseline in the CBCR group was significantly lower than that in the HBCR group (26.4±6.4 mL/kg/min vs. 30.4±7.8 mL/kg/min; p=0.02). METs_max at baseline was also significantly lower in the CBCR group than in the HBCR group (7.6±1.8 vs. 8.9±2.6, p=0.01) (Table 2). The patients were predominantly male in both groups. The CBCR group included 37 patients (28 men and 9 women) with a mean age of 63.5±9.3 years. The HBCR group included 41 patients (35 men and 6 women) aged 62.1±9.6 years. The mean body mass index (kg/m²) at baseline in the HBCR group was higher than that in the CBCR group, but the difference between the two groups was not significant (26.2±3.2 vs. 24.8±3.6; p=0.09).

Likewise, the changes in the CPX results from baseline in the CRF state (METs_max, VO_2max) and myocardial oxygen demand (e.g., RPP) for the CBCR and HBCR groups are presented in Table 2. In the follow-up CPX tests in the CBCR group, both VO_2max and METs_max increased significantly after 12 weeks (30.2±7.6 and 8.6±2.2 mL/kg/min, respectively), and peaked at 1 year (32.0±11.0 and 9.2±3.1 mL/ kg/min, respectively) after baseline (p<0.05) (Table 2, Fig. 2). In the HBCR group, both VO_2max and METs_max increased significantly and peaked at 6 months (33.2±7.4 and 9.5±2.1 mL/kg/min, respectively) after baseline, and were well maintained up to 1 year (32.1±8.3 and 9.4±2.4 mL/kg/min, respectively) after baseline (p<0.05) (Table 2). After 1 year, both VO_2max and METs_max showed a downward trend up to 5 years in both groups. However, in the HBCR group, the values at 5 years (VO_2max 29.0±7.0; METs_max 8.3±2.0) were lower than the baseline values but not significantly (p>0.05) (Fig. 2).

In the CBCR and HBCR groups, the RPP_{at_stage3}/100 was significantly lower at 12 weeks (131±28 and 133±31 mmHg∙bpm, respectively) than at baseline (148±38 and 146±40 mmHg∙bpm, respectively). The RPP_{at_stage3}/100 values at 5 years returned to near-baseline values in both groups (CBCR 148±39; HBCR 142±32) (Table 2).

No significant difference in VO_2AT at baseline was found between the two groups (CBCR 15.3±4.5 mL/kg/min; HBCR 15.7±3.4 mL/kg/min; p>0.05). In the CBCR group, the VO_2AT at 6 months (18.2±4.6 mL/kg/min) and 5 years (17.9±4.4 mL/kg/min) were significantly higher than the baseline values (p<0.05). In the HBCR group, the VO_2AT at 12 weeks (18.0±4.5 mL/kg/min) and 5 years (19.9±5.1 mL/kg/min) were also significantly higher than the baseline values (p<0.05) (Table 2, Fig. 2).

Based on the METs_max value in the baseline CPX test, we classified the patients into the low, average, and high CRF groups regardless of the CR program they participated in [14-17]. The low CRF group included patients with ≤6 METs (n=11); the average CRF group included patients with 6–8 METs (n=31); and the high CRF group included patients with ≥8 METs (n=36) (Fig. 3, Table 3).

The METs_max values in the low CRF group increased significantly and peaked at 1 year (7.2±2.2 mL/kg/min) after baseline (p<0.05). After 1 year, the METs_max values showed a downward trend to 5 years after baseline; however, the value at 5 years (6.3±1.4 mL/kg/min) was still significantly higher than that at baseline (p<0.05). In the average CRF group, the METs_max value at 12 weeks (8.0±1.5 mL/kg/min) was significantly higher than the baseline value (7.1±0.6 mL/kg/min) and peaked at 1 year after baseline (8.2±1.8 mL/kg/min; p<0.05). In the high CRF group, the METs_max value at 1 year (11.0±2.6 mL/kg/min) was not significantly different from the baseline value (10.2±1.9 mL/kg/min) but showed an upward trend for the first year after baseline, like in the other groups. After 1 year, the METs_max values in the high CRF group decreased; however, the value at 5 years after baseline (9.2±1.6 mL/kg/min) remained high (>8 METs).

Although not shown in Table 3, two patients in the low CRF group improved and moved to the high CRF group after 5 years, and three patients in the low CRF group improved and moved to the average CRF group after 5 years. More than half of the patients in the low CRF group (n=6) remained as such after 5 years (≤6 METs). However, none of the patients in the high CRF group whose METs_max decreased could be classified in the low CRF group after 5 years. The METs_max values of most patients in the high CRF group maintained a relatively adequate CRF level (>8 METs) even after 5 years.