Patients were recruited on entry to a phase II outpatient cardiac rehabilitation program and therefore were not undergoing previous exercise training. Inclusion criteria included a recent CAD event that was defined as the patient having at least one of the following: angiographically documented stenosis ≥50% in at least one major coronary artery; myocardial infarction, percutaneous coronary intervention, or coronary artery bypass graft surgery; and positive exercise stress test determined by a positive nuclear scan or symptoms of chest discomfort accompanied by ECG changes of >1 mm horizontal or down sloping ST-segment depression. Twenty-seven men and three women with documented CAD were recruited from the Cardiac Health and Rehabilitation Centre at the Hamilton Health Sciences General Site (Ontario, Canada). Three men and one woman dropped out due to reasons unrelated to the exercise interventions. Three patients had changes to their beta-blockers, and one patient was put on a calcium channel blocker during the study period. Therefore, 22 patients were included in the study. Exclusion criteria included smoking within 3 months, noncardiac surgical procedure within 2 months, myocardial infarction or coronary artery bypass graft within 2 months, percutaneous coronary intervention within 1 month, New York Heart Association class II-IV symptoms of heart failure, documented valve stenosis, documented severe chronic obstructive pulmonary disease, symptomatic peripheral arterial disease, unstable angina, uncontrolled hypertension, uncontrolled atrial arrhythmia or ventricular dysrhythmia, and any musculoskeletal abnormality that would limit exercise participation. The study protocol was reviewed and approved by the Hamilton Health Sciences/Faculty of Health Sciences Research Ethics Board, conforming to the Declaration of Helsinki on the use of human subjects, and written informed consent was obtain from patients before participation. Participant characteristics, medical history, and medications are presented in Table 1.
This study used a factorial repeated-measure design. Brachial artery endothelial-dependent function, which was assessed using flow-mediated dilation (FMD), was the primary outcome. Therefore, to ensure equivalent FMD values between exercise groups, patients were randomized into END or HIT after pretraining assessments, based on their relative FMD. Assessments were performed at baseline (pretraining) and after 12 wk of exercise training (posttraining). Timing of sessions was different between participants, but within-subject sessions were scheduled at the same time of day. Before testing sessions, participants were instructed to fast for at least 8 h; to abstain from exercise for 24 h, caffeine and alcohol consumption for 12 h; and to take all medications and vitamins as usual, except for nitroglycerin (NTG), which was withheld on testing days. All testing was performed in a temperature-controlled room (23.1°C ± 1.2°C).
Height (cm) and weight (kg) were measured, and body mass index was calculated. Seated blood pressure was measured in triplicate using an automated oscillometric device (Dinamap Pro 100; Critikon LCC, Tampa, FL) after 10 min of quiet rest. The first was considered a calibration measure, so brachial blood pressure was determined from the average of the second and third measures. Heart rate was measured throughout the testing session using a single-lead (CC5) ECG (model ML 123; ADInstruments Inc., Colorado Springs, CO) and is reported as the average value from a 5-min sample collected after 10-min of supine rest.
Cardiorespiratory Fitness Assessment
Fitness was assessed using a medically supervised graded exercise test to exhaustion on a cycle ergometer (Ergoline, Bitz, Germany). Heart rate was monitored throughout the test using a 12-lead ECG (MAC 5500; General Electric, Freiburg, Germany). After a brief warm up, participants cycled at approximately 70 rpm for 1 min at a workload of 100 kpm. After the first minute, the workload was increased 100 kpm every minute until exhaustion. Expired gas was analyzed using a semiautomated metabolic cart (Vmax 229; SensorMedics Corporation, Yorba Linda, CA), and oxygen consumption was determined at peak (V·O2peak) and anaerobic threshold (respiratory quotient = 1.0) from breath-by-breath samples averaged for 20 s. One participant did not complete the posttraining test due to a musculoskeletal injury (unrelated to the exercise training); therefore, analysis was performed on 21 patients (END = 10, HIT = 11).
Brachial Artery Assessments
All brachial artery endothelial function assessments were conducted in the supine position after 20 min of quiet rest. Brachial artery endothelial-dependent function was assessed using the FMD test, based on previously established guidelines.[6,32] Duplex ultrasound (Vivid Q; GE Healthcare, Horten, Norway) was used to capture simultaneous images of the right brachial artery (13 MHz) at a frame rate of 7.7 frames per second and blood velocity measurements (4 MHz) throughout the FMD protocol. Baseline preocclusion images and velocities were collected 3–5 cm proximal to the antecubial fossa for 30 s. A pneumatic cuff positioned on the forearm distal to the antecubial fossa was inflated using a rapid cuff inflator (model E20 and AG101; Hokanson, Bellevue, WA) to an occlusion pressure of 200 mm Hg. After 5 min of occlusion, the cuff was released and duplex postocclusion images, and velocities were collected for 3 min. Duplex images were stored in Digital Imaging and Communications in Medicine (DICOM) format for offline analysis. End-diastolic frames, determined by the R-spike of the ECG trace, were extracted and stacked in a new DICOM file using commercially available software (Sante DICOM Editor, Version 3.0.12; Santesoft, Athens, Greece). Brachial artery diameters were determined using edge-detection software (Artery Measurement System; Image and Data Analysis, Gothenburg, Sweden). Preocclusion diameters were determined from the average of the 30-s sample. Postocclusion diameters were averaged in rolling five-cycle bins. Peak postocclusion diameter was defined as the maximum five-cycle average. Absolute FMD was calculated as the difference between the peak postocclusion diameter and the preocclusion diameter. Relative FMD is the ratio of the absolute FMD to the preocclusion diameter. We previously reported day-to-day intraclass correlations coefficients in a similar population of 0.89 for absolute and relative FMD.
Blood velocity raw audio signals were continuously analyzed by an external spectral analysis system (model Neurovision 500M TCD; Multigon Industries, Yonkers, NY) to determine continuous intensity weighted mean blood velocity (MBV). The MBV was sampled at 100 kHz during the FMD tests using commercially available hardware (Powerlab model ML795; ADInstruments) and analyzed offline using LabChart 7 Pro for Windows (Powerlab ML 795; ADInstruments). MBV signals were corrected for the angle of insonation (all ≤70°). Blood flow was calculated by multiplying brachial artery cross-sectional area by MBV. Postocclusion reactive hyperemic blood flows and MBV were averaged into five-cycle bins to align with the five-cycle diameter bins. Peak reactive hyperemic blood flow is reported as the maximum value during the 3-min postocclusion period. Shear rate for each bin was calculated by multiplying the MBV bin by 8 and by dividing it by the corresponding reactive hyperemic five-cycle bin. The reactive hyperemic stimulus until peak dilation was quantified as shear rate area under the curve (AUC).
Endothelial-independent function was assessed 10 min after cuff release using a 0.4-mg sublingual spray of NTG. Longitudinal B-mode images (8 MHz) of the right brachial artery were collected before the administration of NTG (pre-NTG; 10 cardiac cycles) and for 10 cardiac cycles every minute post-NTG up to 10 min at a frame rate of 22.9 frames per second. End-diastolic frames were extracted and stacked, and peak-NTG diameter was determined from the average of 10 cardiac cycles at each minute. NTG results are expressed in absolute and relative units. One participant declined the NTG assessment due to a history of severe headaches; therefore, NTG results are reported for 21 patients.
Exercise Training Protocols
Participants attended two supervised sessions per week for 12 wk at the Cardiac Health and Rehabilitation Centre at the Hamilton Health Sciences General Site (Hamilton, Ontario). Each session involved a 10-to 15-min standardized warm-up and cooldown involving light aerobic exercise and dynamic stretching. Exercise intensities for each protocol were based on the peak power output (PPO) achieved during the pretraining exercise stress test. The prescription for END was based on the Canadian Association of Cardiac Rehabilitation exercise guidelines and involved continuous cycling at 58% of PPO (range 51%–65%). Participants progressed from 30 min (weeks 1–4) of cycling to 40 min (weeks 5–8) to 50 min (weeks 9–12). The protocol for HIT was based on previous research in adults with type 2 diabetes and involved ten 1-min cycling intervals at 89% PPO (range, 80%–104%) separated by 1-min intervals at 10% PPO. Rather than increasing duration, workload was increased every 4 wk to elicit the heart rate responses achieved during 89% pretraining PPO. Consequently, patients were training at 102% pretraining PPO during weeks 5–8 and 110% pretraining PPO for weeks 9–12. In addition to the supervised training sessions, patients were instructed to perform lower limb exercise at least one additional day per week, using similar exercise durations and intensities as their exercise protocol. Unsupervised sessions were tracked using Polar heart rate monitors (RS300X; Polar Electro Inc., Lake Success, NY).
Statistical analyses were performed using the Statistical Package for the Social Sciences (Version 11.5; SPSS, Chicago, IL). Factorial (END versus HIT group) repeated-measures (pre- versus posttraining) analyses of variance were used to compare indices of cardiorespiratory fitness, brachial artery endothelial function, and resting hemodynamics. Group differences in pretraining characteristics and exercise training data were compared using independent t-tests. Data are presented as mean ± SD, with P ≤ 0.05 considered statistically significant.
Med Sci Sports Exerc. 2013;45(8):1436-1442. © 2013 American College of Sports Medicine