This was a prospective, longitudinal, randomized and controlled study. Patients with uncomplicated AMI admitted to the Acute Coronary Disease Unit of the Heart Institute of the University of Sao Paulo Medical School (InCor/HCFMUSP), who agreed to participate in the study, were randomized to two groups (TRAINING and CONTROL), in the proportion 1:1 according to the following protocol: Time 0: still hospitalized, all patients, after due explanations, signed the informed consent for inclusion in the study. Time 1: all participants returned one month after AMI and were submitted to an echocardiogram and cardiopulmonary exercising testing (CPX). After this, participants randomized to the TRAINING group were included in a supervised physical training program, twice a week, for four months, at the Rehabilitation Lab, while individuals randomized to the CONTROL group received usual recommendations for physical activity at home. Briefly, patients were instructed to engage in aerobic exercises of low-to-moderate intensity, with a minimum duration of 30 minutes, at least three times a week in the first two months. In the third month, they were instructed to increase the intensity to a moderate level 3-5 times a week, 30 to 60 minutes. No specific monitoring was carried out. Supervised training program consisted of two 60-minute exercise sessions per week. Each exercise session consisted of 5 minutes of stretching exercises, 40 minutes of cycling on a stationary bicycle, 10 minutes of local strengthening exercises, and five minutes of cool down with stretching exercises. The exercise intensity was established by heart rate levels that corresponded to the anaerobic threshold and respiratory compensation point. Time 2: at the end of the fourth month (fifth month after the AMI event), all patients repeated the echocardiogram and CPX (supplementary data I).

Inclusion criteria were: age over 18; hospitalization for spontaneous AMI with or without ST-segment elevation, established according to the 3^rdUniversal Definition of AMI;³¹ clinically and hemodynamically stable patients; LVEF > 0.40 and Killip classes I or II. The exclusion criteria were: any condition that contraindicated physical activity; regular physical activity practitioners prior to the event (confirmed by interview for inclusion); LVEF ≤ 40%; Killip classes III or IV; irregular heart rhythm (such as atrial fibrillation, frequent premature atrial or ventricular contractions); limited echocardiographic window for analysis.

The present study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional ethics and scientific committees. All the patients signed the informed consent form before inclusion in the study.

The echocardiogram was performed by a single experienced operator blind to the allocation group. Commercially available echocardiographic equipment (Vivid E9; GE Medical Systems, Milwaukee, WI, USA) was used, equipped with linear broadband transducers with a frequency of 5-2 MHz. Measurements, and analysis of valve flow and LV diastolic function were performed according to current guidelines.^{32 - 35}

For image acquisition for speckle tracking analysis, the machine was adjusted to acquire three cardiac cycles, with frame rate between 40-80 frames/s. Images were obtained in the parasternal short axis (PSAX) view in its three main levels: basal (mitral valve), mid (papillary muscles) and apical. LV apical window was composed by: longitudinal (APLAX), four (A4C) and two chambers (A2C) views.

Off-line analyses were performed using the EchoPAC software, version v20.1 (GE Medical Systems, Milwaukee, WI, USA). The parameters evaluated were: strain (ε,%) and strain rate (SR) (ε’,s^-1); maximum basal and apical rotations (LVrot,^o) and their peak absolute rotational velocities (LVrot-v,^o/s), LV twist (LVtw,^o) and torsion (LVtor,^o/cm), and their velocities (LVtw-v,^o/s; LVtor-v,^o/s.cm). LV twist was calculated as the absolute difference between peak systolic apical and basal rotations (LVtw = apical LVrot - basal LVrot) and torsion as the twist normalized to LV longitudinal length.²⁴ By convention, baseline rotation values were negative, and apical, positive ( Figure 1 ).²⁴ In this paper, the negative strain values will be expressed as modulus (positive) for a better understanding.

Figure 1

The CPX was performed on an electromagnetic cycle ergometer (Medfit 400L, Medical Fitness Equipment, Maarn, Netherlands), following a ramp protocol, with a speed of 60 revolutions per minute (rpm) and workload increments from 10w to 20w per minute, until achieving physical exhaustion. Participants were connected to a ventilator (SensorMedics Corp, CA, USA), and pulmonary ventilation, oxygen (O₂) and carbon dioxide (CO₂) concentrations were measured with calculation of oxygen consumption (VO₂) and CO₂production. When the patient achieved exhaustion, the anaerobic threshold and respiratory compensation point (i.e. points when blood lactate starts to increase rapidly due to high exercise intensity and tissue anaerobiosis) were determined, both used to prescribe the intensity of physical training.³⁶

Continuous data are presented as mean ± standard deviation (SD) or median and interquartile ranges (IQR), according to normality of distribution assessed by the Shapiro-Wilk test. Categorical data were presented as numbers and percentages. To calculate the sample size, the mean value of global longitudinal strain of 15.9% (± 2.3) was considered.³⁷ For an expected gain of 10% in longitudinal strain in the TRAINING group, considering the effect size of this intervention (d = 0.65), a power (1 - β) 0f 80% to detect differences and alpha index of 0.05, it was necessary to include 36 patients in each group (a sample size of 72 participants). Randomization of the sample was made through the website “randomization.com”. Comparisons of continuous variables between groups were performed using the Student’s t-test (in case of normal distribution) or Mann-Whitney test (in case of non-normal distribution) for independent samples. For categorical variables, the Chi-square test (χ²) or Fisher’s exact test was used, as appropriate. The within-group comparison of means was performed using the paired Student’s t test. Finally, comparison of the “delta” (Δ) of values obtained from the follow-up of each group was done by the unpaired Student’s t test.

All tests were 2-tailed and a p value <0.05 was considered statistically significant. Statistical analysis was done using the SPSS v.25.0 software for Macintosh (SPSS Inc, Chicago, IL). Inter and intra-observer analyses and Bland-Altman plots are disposed in the supplementary data II.

From 2016 to 2019, a total of 76 patients was enrolled. Of these, 23 were excluded for several reasons, with poor adherence to exercise as the main reason ( Figure 2 ). The final study population consisted of 53 participants. After randomization, 27 individuals were included in the TRAINING group and 26 in the CONTROL group.

Figure 2

The clinical characteristics of the population are shown in Table 1 . As can be seen, there was no significant differences between the groups regarding age, sex, variables related to coronary artery disease (CAD), in-hospital treatment, risk and severity of the ischemic event.

Table 1

Variables	Control (n = 26)	Training (n = 27)	p
Demographics
Age, y	59±11	60±9	0.73
Gender M / F, n (%)	19 (73%) / 7 (27%)	20 (74%) / 7 (26%)	0.93
BMI, kg/m2	27.4±3.9	27.2±3.9	0.84
Hypertension, n (%)	17 (65%)	13 (48%)	0.21
Diabetes Mellitus, n (%)	10 (38%)	6 (22%)	0.84
Smoker, n (%)	5 (19%)	10 (37%)	0.15
Sedentarism, n (%)	26 (100%)	27 (100%)	1.00
Dyslipidemia	8 (31%)	9 (33%)	0.84
Previous AMI, n (%)	4 (15%)	3 (11%)	0.70
Treatment
Aspirin, n (%)	26 (100%)	27 (100%)	1.00
Clopidogrel, n (%)	24 (92%)	27 (100%)	0.24
Ticagrelor, n (%)	2 (8%)	0 (0%)	0.24
β-blocker, n (%)	17 (65%)	23 (85%)	0.09
ACEI/ARB	16 (61%)	21 (78%)	0.20
Event
STEMI, n (%)	14 (54%)	16 (59%)	0.51
NSTEMI, n (%)	12 (46%)	11 (41%)	0.69
Troponin, ng/mL (median, IQR)	22 (3-50)	32 (11-50)	0.35
Risk Scores
TIMI STEMI score	3.3±1.8	2.7±1.8	0.38
TIMI NSTEMI score	3.7±1.3	3.2±1.1	0.31
GRACE (median, IQR)	129 (120-159)	139 (121-151)	0.74
Treatment
Fibrinolysis, n (%)	4 (15%)	6 (22%)	0.73
Primary PTCA, n (%)	8 (31%)	8 (30%)	0.93
Pharmaco-invasive strategy, n (%)	3 (12%)	2 (7%)	0.66
BMS, n (%)	19 (73%)	23 (85%)	0.23
DES, n (%)	5 (19%)	1 (4%)	0.10

Data expressed as mean ± SD, median (IQR, interquartile range; 95% CI) or number and percentage (%). Categorical data presented as numbers and percentages. BMI: body mass index; ACEI: angiotensin I converting enzyme inhibitor; ARB: angiotensin II receptor blocker; STEMI: ST-segment elevation myocardial infarction; TIMI: Thrombolysis in Myocardial Infarction Group; NSTEMI: Non-ST segment elevation myocardial infarction; GRACE: Global Registry of Acute Coronary Events Score; PTCA: percutaneous transluminal coronary angioplasty; BMS: bare-metal stent; DES: drug-eluting stent. P = NS for all comparisons (Student’s t, Chi-square or Fisher’s exact tests).

The average adherence of patients in the TRAINING group to the exercise program was 28.0±6.8 sessions (minimum of 15 and maximum of 39 sessions), reaching an overall adherence of 88%.

Table 2 provides the CPX main data. After the exercise program period, the TRAINING group showed a significant increase in exercise duration, maximum workload achieved and peak VO_2.However, at the end of the four-month follow-up, no differences were observed in the variations of the parameters (delta [Δ]) measured between the groups (supplementary data III).

Table 2

Variables	Control			Training
	Basal	4M	p*	Basal	4M	p*
Echocardiogram
LA, mm	37.3±3.7	36.6±3.4	0.02	38.7±3.4	38.8±3.3	0.63
LVEDD, mm	50.2±4.3	50.0±4.3	0.21	50.9±3.1	51.3±3.5	0.38
LVESD, mm	33.1±5.0	31.9±3.9	0.05	33.6±3.0	34.6±4.6	0.11
LVEDV, ml	99.9±23.2	99.5±23.5	0.66	94.3±15.2	92.2±16.4	0.40
LVESV, ml	39.6±11.1	42.0±10.2	0.11	42.3±11.5	42.6±9.1	0.85
LVMI. g/m2	90.7±18.6	89.7±18.0	0.28	92.7±16.7	92.2±16.4	0.88
LVEF, %	61.3±5.7	60.0±5.1	0.19	60.0±5.9	59.6±5.6	0.70
WMSI	1.19±0.19	1.16±0.19	0.19	1.18±0.16	1.17±0.18	0.39
E Vel, cm/s	81±24	81±22	0.89	74±0.18	74±18	0.95
e’ Vel cm/s	7.1±1.6	7.2±1.8	0.80	6.9±1.7	6.9±1.8	0.89
E/e’ ratio	12.1±5.3	11.8±4.6	0.65	11.5±3.9	11.6±4.9	0.92
E/A ratio	1.21±0.52	1.13±0.39	0.34	1.09±0.41	1.20±0.46	0.16
CPX
Exercise duration, secs	491±93‡	593±93	0.006	524±95‡	636±131	0.001
Basal HR, bpm	73±14	70±11	0.58	70±10	64±10	0.002
Max HR, bpm	133±22	134±15	0.84	127±19	128±19	0.78
Basal SBP, mmHg	124±11	119±14	0.35	120±18	117±14	0.51
Peak SBP, mmHg	169±18†	176±16	0.41	184±27†	178±18	0.41
Maximal exercise workload, W	109±43	119±48	0.11	126±49	151±65	0.007
Peak VO2, ml/kg/min	20.6±4.6	21.7±4.9	0.12	21.7±5.2	23.4±5.9	0.04
AT, ml/kg/min	12.3±2.3	13.7±2.4	0.08	12.3±3.2	14.3±3.4	0.11
VE, L/min	59.5±21.3	59.4±19.5	0.97	62.7±18.7	69.2±20.4	0.93

Data expressed as mean ± SD. LA: left atrium; LVEDD: left ventricular end diastolic diameter; LVFSF: left ventricular end systolic diameter; LVMI: left ventricular mass index; LVEDV: left ventricular end-diastolic volume; LVESV:left ventricular end-systolic volume; LVEF: left ventricular ejection fraction; WMSI: wall motion score index; E Vel: transmitral E wave velocity; e’ Vel:septal e’ wave velocity; A Vel: transmitral A wave velocity. HR indicates heart rate; SBP: systolic blood pressure; VO2: oxygen consumption; AT: anaerobic threshold; VE: minute volume. *Paired Student´s t test. ‡p=0.02, unpaired Student’s t test, for comparison with CONTROL basal. †p=0.04, unpaired Student’s t test, for comparison with.

Echocardiographic data are shown in Table 2 . A slight difference in the left atrial dimensions in the CONTROL group was noted at four months. No other significant differences were found between groups, including comparison of variations in all of these parameters.

First, the relatively small sample size, the short duration of the exercise program, the small number of training sessions for patients in the TRAINING group (twice a week), and the low adherence of some patients to the program may have decreased the power of the study to demonstrate possible intergroup differences. Subjects were encouraged to attend the sessions, and those who attended less than one session per week, or stopped the sessions, were excluded. A sample size of 72 participants, as calculated, was not achieved, and this is also a limitation. In addition, besides exercise, no other health instructions, such as nutritional or psychological support were given, which could have affected the final result. No data concerning chronic coronary syndromes or peripheral artery disease were collected, which may have had an impact on exercise. Finally, not all data derived from the CPX, including data on the respiratory exchange ratio, were addressed in this study - these could have revealed some differences between groups.

Another important point is the subjectivity of the echocardiogram which can lead to quantification bias. Small variations in the acquisition at LV apical level, which do not have an anatomical marker, can result in distorted values. Images were acquired from the apical short-axis view when the visibility of the myocardial segments was clear. Additionally, we created another criterion to ratify the correct acquisition, which was the visualization of at least a trend of a counterclockwise rotation of these segments, which would be physiologically expected. Finally, it is important to remind that the examiner was blinded to the patient’s group allocation.