Aerobic interval training improves VO_2peak in coronary artery disease patients; no additional effect from hyperoxia

Abstract

Objectives. To investigate whether hyperoxic aerobic interval training improves training quality in coronary artery disease patients. Design. Twenty-one stable coronary artery disease patients were recruited to hyperoxic (n=10) and normoxic (n=11) groups (age: 62.4±6.8 years). Patients underwent 30 supervised 4×4 minutes interval training sessions using treadmill walking, at 85–95% of peak heart rate. Results. Arterial saturation was significantly increased by 3% at pretest from normoxic to hyperoxic testing conditions. Peak oxygen uptake and stroke volume increased significantly by 16% and 17% (p<0.05) and by 16% and 18% (p<0.05) in the hyperoxic and normoxic training groups respectively. No difference was revealed between groups for peak oxygen uptake and stroke volume. Blood volumes were unchanged from pre to post training. Peak oxygen uptake measured in normoxia and hyperoxia in the hyperoxia training group revealed no difference. Conclusion. The present study shows that breathing 100% oxygen enriched air during aerobic interval training in stable coronary artery disease patients does not improve peak oxygen uptake above the level attained with normoxic training.

• 100% oxygen
• Endurance training
• 4×4 minutes intervals
• maximal oxygen uptake
• stroke volume

Endurance training is an effective means in prevention and rehabilitation of coronary artery disease and VO_2max identified as the single best predictor of mortality 1. Our research group has previously shown that in short-term interventions high intensity aerobic interval training is superior to moderate intensity training for increasing aerobic power in coronary artery disease (CAD) patients 2. Reduced stroke volume has been reported at rest and maximal exercise in coronary artery disease patients in whom exercise induced myocardial ischemia may reduce the myocardial contractile function 3. Improvements in VO_2max after aerobic interval training has been linked to improved maximal stroke volume 4. Increased left ventricular ejection fraction and remodeling of the left ventricle associated with increased VO_2max have been noted in heart failure patients after interval training 5. A few training studies have investigated the effect of hyperoxic training on healthy subjects, showing a 8–9% increase in exercise load at the same heart rate during hyperoxic training, but none have found significant effect on VO_2max after 5- and 6 weeks of hyperoxic endurance training 6, 7. These finding are in contrast to the well documented increase seen in VO_2max and performance during acute inspiration of hyperoxic gas 8–10. In CAD patients oxygen breathing increased the angina threshold allowing the heart to do more work before the development of coronary insufficiency 11, and increased exercise performance 11–14. Some studies recommend use of oxygen during physical activity, especially to patients with anginal pain and ischemic ST depression after exercise. Aerobic interval training in hyperoxia may increase the exercise power output from normoxic training without raising the already high training heart rate, with the potential of increasing the training outcome 6, 7. Both duration and intensity of the training intervention in hyperoxia are crucial. No hyperoxic training study has to our knowledge been performed with CAD patients, who may have reduced ability for oxygen delivery to the working muscles through reduced myocardial contractility and stroke volume if ischemic 3. If hyperoxia increases the training work load at the normoxic training heart rate, it might increase the training quality, improving stroke volume and VO_2peak to a greater extent than normoxic exercise. The purpose of this study was to investigate whether breathing 100% oxygen enriched air during aerobic interval training in cardiovascular disease patients improves the training outcome compared to aerobic interval training in normoxic conditions.

Methods

Twenty-one clinically diagnosed stable coronary artery disease (CAD) patients were recruited and randomly allocated to a hyperoxic training group (HT) n = 10, and a normoxic training group (NT) n = 11 from the St. Olav University Hospital of Trondheim. Physical descriptions of patients are shown in Table I. Inclusion criteria were stable CAD, angina pectoris class I-III in the Canadian Cardiovascular Society Classification (CCS), ischemia in exercise electrocardiogram, or angiographically documented cardiovascular disease. Exclusion criteria were unstable angina pectoris, myocardial infarction during the last month, percutaneous coronary intervention (PCI) during the last month, left ventricular ejection fraction below 40%, complex ventricular arrhythmias, and orthopedic or neurological limitations to exercise. The following number of patients used the listed medication; beta-blockers 13, antiplatelet agents 15, statins 13, angiotensin-converting-enzyme inhibitors 3, long-acting nitrates 2, and diuretics 2. No change in medication was reported during the study.

Table I.  Physical characteristic of the subjects at inclusion.

	Hyperoxic training group (n = 8)	Normoxic training group (n = 10)
Men/women	6/2	7/3
Age (years)	61.1±7.1	63.6±6.5
Stature (cm)	175.5±11.1	173.4±9.2
Body mass (kg)	82.3±13.2	79.3±12.2
Body mass index (kg•m^−2)	27.2±2.8	26.3±2.6
Systolic blood pressure (mmHg)	130±27	139±20
Diastolic blood pressure (mmHg)	77±10	84±8
Coronary artery disease
Myocardial infarction	1	3
Percutaneous coronary intervention	4	4
Coronary artery bypass surgery	5	3

The study protocol was approved by the regional committee for medical research ethics, and was accomplished according to the declaration of Helsinki. Written consent was obtained from the subjects. Two patients dropped out of the study while one patient was excluded due to repeated non- cardiac illness affecting training quality.

Exercise testing were performed pre and post the exercise training period. VO_2peak were tested in both normoxia (21% oxygen) and hyperoxia (65% oxygen) in the HT in a random order 2 days apart while VO_2peak were tested in normoxia (21% oxygen) in the NT. Respiratory testing (V-max Spectra, SensorMedics, USA) were performed during treadmill walking at 3–5 km per hour (Technogym, Italy). The treadmill inclination was raised (1–3%/min) until subjects reached exhaustion, and the average of the three continuous highest 10 seconds measurements determined VO_2peak. Criteria for exhaustion were an RQ value above 1.0 and a Borg scale value above 15. RQ was above 1.0 in all patients and Borg scale values above 15 were observed in 14 patients. In addition, the authors did a subjective evaluation of the level of exhaustion through observations of ventilation, walking action and facial expressions in patients at the end stage of the test.

Heart rate (HR) was measured by a heart rate monitor (Polar Sport, Finland), while arterial oxygen saturation (SpO₂) was recorded by pulsoximetry in the HT (Critcare Systems INC, USA). A capillary blood sample taken immediately after the tests was analyzed for lactate using an YSI 1500 sport tester (YSI Incorporated, USA). Patients self reported exercise exertion through the 6–20 Borg scale for ratings of perceived exertion 15.

Cardiac output (CO) was measured during treadmill walking at 80% of VO_2peak through acetylene breathing according to the methods previously described by Helgerud et al. 4. During steady state walking patients did one complete inspiration and expiration of acetylene gas mixture where inspiration and expiration values of gases were used to calculate cardiac output. The test method has previously been validated and a coefficient of variation of 7.6% was found 16. Blood- and plasma volume was measured before and after training in the HT using the Evans blue dye dilution technique 17, according to procedures previously described 4. The questionnaires SF36 and Macnew were distributed to the patients for measurements of quality of life before and after the training periods.

Interval training

Patients completed 30 interval sessions with treadmill walking in three weekly sessions during 10 weeks after the initial testing. Subjects in the hyperoxic and the normoxic training group had a compliance of respectively 29.6±0.7 and 29.1±1.1 training sessions. One hundred percent oxygen enriched air was distributed to the patients in the HT during exercise. Gas was distributed to the patients from a Douglas bag connected to a gas tank, and patient breathed through a face mask and a three way valve system. No discomfort was noted due to training with a facial mask, and consequently all subjects were able to adhere properly to the training intensity. After 5 minutes warm up, patients continued with 4 times 4 minutes of interval training, with 3 minutes active breaks in-between each interval. During the 4 minutes intervals patients trained at 85–95% of HR_peak, while during active breaks intensity was at 60–70% of HR_peak. To compensate for increased VO_2peak capacity, treadmill speed and grade were increased several times during the study, to make sure the patients trained at 85–95% of their HR_peak at all times. HR and SpO₂ were recorded during exercise and patients reported self reported exercise stamina through the 6–20 Borg scale for ratings of perceived exertion 15.

All training sessions were supervised by an exercise physiologist. Two subjects in each training group experienced angina pain during the start of training sessions early in the study, but were able to continue training without having to use nitroglycerine. No other cardiac related incidents were reported during the study.

All values are expressed as mean±standard deviation (SD). Changes within groups were determined by the Wilcoxon signed ranks test. Differences between the HT and the NT response to training were calculated by using the Mann-Whitney U-test. A two-tailed p < 0.05 was accepted as statistically significant for all tests.

With a power of 0.80 and a two sided α value of 0.05 and an expected difference in training work load of 8% the calculated number of subjects needed in each group was 7.

Results

VO_2peak, performance, peak ventilation, peak cardiac output and stroke volume increased significantly from pre to post training in both training groups (Figure 1, Table II, Table III) together with scores for physical (10%) and social (9%) quality of life in the HT, and total (9%), physical (13%) and social score (10%) in the NT measured by the Macnew questionnaire. No difference was found between groups in the improvement in VO_2peak, performance, peak cardiac output and stroke volume from pre to post the training intervention. Peak oxygen saturation was significantly increased by 3% (p < 0.05) from normoxic to hyperoxic testing in the HT (Table II, Table IV). VO_2peak and performance in the HT was equal in hyperoxic and normoxic testing both before and after training (Table II, Table IV). Peak heart rate and lactate increased significantly from pre to post training in the HT, and the HT had a significantly greater increase in peak ventilation than the NT, while the opposite were true for perceived exertion through the Borg scale (Table II). All other measures, including total blood volume were not significantly different (Table V).

Figure 1.  Percentage change in VO_2peak (L • min⁻¹) and peak stroke volume (mL • stroke⁻¹) from pre and post training for the hyperoxic and normoxic training groups presented as mean±SE. Significant difference within groups from pre to post training; * = p < 0.05.

Table II.  Peak metabolic data in normoxia before and after training.

	Hyperoxic training group (n = 8)		Normoxic training group (n = 10)
	Before	After	Before	After
Oxygen uptake
L•min^−1	2.11±0.35	2.44±0.48*	2.17±0.53	2.53±0.61*
mL•kg^−1•min^−1	25.9±4.2	29.9±3.9*	27.3±4.6	31.8±5.0*
mL•kg^−0.75•min^−1	77.7±11.3	89.6±11.6*	81.4±14.4	94.8±16.1*
Heart rate (beats•min^−1)	141±19	152±19*	163±19	160±19
Ventilation (BTPS) (L•min^−1)	72.6±14.4	87.3±19.2*^#	86.1±19.8	92.7±24.2*
Respiratory exchange ratio	1.11±0.08	1.20±0.10	1.14±0.06	1.15±0.07
Work load (watt)	123±27	159±37*	155±38	204±21*
Arterial oxygen saturation (%)	95±3	95±3	–	–
Lactate (mmol•L^−1)	3.61±1.01	6.90±1.57*	4.30±1.20	6.23±1.80
Borg scale	18±1**	18±1	15±2	17±1*^#

Data are presented as mean±SD for each variable.

* Significant difference between pre and post tests within exercise groups (p < 0.05).

** Significant difference between groups at pre test time point.

^# Significant difference between groups (p < 0.05).

Table III.  Peak cardiac function at 80% of VO_2peak work load before and after training.

	Hyperoxic training group (n = 8)		Normoxic training group (n = 10)
	Before	After	Before	After
Heart rate (beat•min^−1)	112±11	125±14*	127±19	133±19
Cardiac output (L•min^−1)	11.0±2.1	14.1±2.3*	12.1±2.6	14.9±3.5*
Stroke volume (mL•beat^−1)	98.9±18.9	114.2±21.3*	95.5±20.8	112.2±21.1*

* Significant difference between pre and post tests within exercise groups (p < 0.05).

Table IV.  Peak metabolic data from hyperoxic testing in the hyperoxic training group (n = 8).

	Before	After
Oxygen uptake
L•min^−1	2.07±0.35	2.41±0.43
mL•kg^−1•min^−1	25.3±4.0	29.7±4.0
mL•kg^−0,75•min^−1	76.0±11.3	88.9±11.0
Heart rate (beats•min^−1)	140±21	151±17
Ventilation (BTPS) (L•min^−1)	55.5±10.4	69.5±10.4
Respiratory exchange ratio	0.96±0.03	1.15±0.14^†
Work load (watt)	121.0±34.3	158.7±34.2
Arterial oxygen saturation (%)	98±2	98±3 †
Lactate (mmol•L^−1)	2.50±1.06	5.71±2.14
Borg scale	18±2	18±2

Data are presented as mean±SD for each variable.

† Significant differences between normoxic and hyperoxic tests (p < 0.05).

Table V.  Blood volume before and after training in the hyperoxic training group (n = 8).

	Before	After
Blood volume (L)	4.28±0.67	4.25±0.71
Plasma volume (L)	2.81±0.36	2.87±0.42
Red cell mass (L)	1.47±0.34	1.39±0.32*
Blood volume (mL•kg^−1)	52.29±8.74	52.23±7.60
Plasma volume (mL•kg^−1)	34.47±5.69	35.21±4.94
Red cell mass (mL•kg^−1)	17.81±3.69	17.01±3.26*
Hematocrit (%)	39.1±3.5	37.3±3.1*

* Significant difference between pre and post tests (p < 0.05).

The Borg rate of perceived exertion after each training session was 16.3±0.7 for the HT and 14.4±1 for the NT. One patient in the HT showed an elevated exercise ST segment after the training period and was referred for further examination at the cardiac unit at the University hospital.

Discussion

The most important finding in this experiment was that stable coronary artery disease patients breathing 100% oxygen enriched air during interval training did not show a superior training effect over patients training in normoxic conditions. Although arterial oxygen saturation was significantly improved in the hyperoxic pretest compared to the normoxic pretest in the hyperoxic training group, hyperoxic training showed no additional effect on peak VO₂, performance, cardiac output or stroke volume compared to normoxic training. In addition no acute effect of hyperoxia was detected on VO_2peak in the hyperoxic training group before or after the training period.

Hyperoxic breathing in cardiovascular disease patients may improve oxygen delivery to the myocardium and the skeletal muscles. The use of hyperoxic gas has been found to restore electrocardiographic abnormalities in cardiovascular disease patients 12, and may protect the myocardium from the hypoxic effect of exercise and increase the angina threshold 11. The negative effect of ischemia with reduced myocardial contractility and stroke volume during high intensity training may thereby be prevented. In the present study, a 3% increase in arterial oxygen saturation may not be sufficient to overcome a potential ischemic exercise restriction during interval exercise, or an oxygen desaturation of 95% may not have been low enough to cause myocardial insufficiencies in normoxia effecting aerobic performance. A significant increase in SpO₂ from 95 to 98% has been reported to be sufficient to increase exercise performance in chronic heart failure patients 14, so our patients may be at borderline in terms of getting effect from oxygen supplementation on hemoglobin oxygen saturation and training load. In addition the present study may not have succeeded in recruiting a severe enough ischemic patient population to gain effect from increased myocardial oxygen delivery. Due to treatment options like PCI and coronary bypass surgery fewer severe angina patients seems to be available as volunteers, and the level of angina might be milder. Hyperoxic supplementation permits the ischemic heart to carry out more work before coronary insufficiency develop 11 and may therefore be effective in a severe ischemic patient group with greater limitation in cardiovascular function during peak exercise.

A great number of studies have found VO_2max and performance increased in acute hyperoxia, alongside an increased hemoglobin oxygen saturation and arterial oxygen content 8–13, however all data do not point in the same direction in terms of heart patients. When studying the acute effect of hyperoxia in coronary artery disease patients, some investigations did not detect improvements in VO_2peak and performance. In one investigation, haemoglobin oxygen saturation was not increased in hyperoxia compared to normoxia, however, a trend of reduction in leg blood flow was detected implying that hyperoxia did not improve muscle oxygen delivery, thereby explaining the lack of effect on leg oxygen uptake and performance 18. In another study of heart failure patients haemoglobin oxygen saturation was increased in hyperoxia however no improvements in VO_2max and performance was noted 19. The present study displays similar results with equal VO_2peak in hyperoxia and normoxia both before and after the hyperoxic training intervention. The lack of improvement in VO_2peak in hyperoxia may be a result of decreased leg blood flow. In a study by Russell and colleagues 18, a trend towards reduced leg blood flow was observed despite of no change in haemoglobin oxygen saturation. In the present study, in which SpO₂ was significantly increased during hyperoxic testing, one might suggest that leg blood flow may significantly be reduced resulting in no additional oxygen delivery to the working skeletal muscles serving as an explanation for absence of additional oxygen consumption in hyperoxia and the lack of effect from hyperoxic training. This could explain the deficiency of accumulative effect of hyperoxia on VO_2peak and stroke volume over time with aerobic interval training despite increased haemoglobin oxygen saturation and the theoretical possibility of increased oxygen delivery to the working muscles.

Increased training work load has been reported during hyperoxic training in previous studies 6, 7. In those training studies in healthy subjects hyperoxic exposure enabled the subjects to increase the training work load by 8–9% at the normoxic training heart rate. Despite increased training work load hyperoxic training was not found to be superior in terms of VO_2max and performance enhancement 6, 7. This is in line with the present study where no difference was found between the effects of interval training in hyperoxia and normoxia. In the study by Ploutz-Snyder et al. 7 the lack of improvement in VO_2max after hyperoxic training may be explained by the relatively low training intensity (70% of maximal heart rate), while intensity levels above 85% of peak heart rate was used by Perry et al. 6 and in the present study, and thereby should be optimal for detecting any effects of hyperoxic training on stroke volume and VO_2peak 2, 5.

The notion in this experiment was that the ability of CAD patients’ muscles to increase the a-vO₂ difference as shown in a study of intermittent claudication patients 20 would enable the patients to utilize the extra 3% arterial oxygen saturation during training leading to a cumulative positive effect on both workload and training response during the 30 interval sessions. As this did not happen, other explanations may be that the a-vO₂ difference did not change fast enough to pick up the advantage of extra oxygen. However, great changes in a-vO₂ difference has been shown in previous experiments 10 and may not be the most likely cause. The most probable reason for the lack of improved training effect from a higher arterial oxygen saturation is that hyperoxic exercise, despite higher blood oxygen carrying capacity, does not increase the exercise induced load on the heart compared to normoxic training conditions. Thereby the main limiting factor for VO_2peak, namely, the stroke volume of the heart changed to the same degree in the two training groups. One may speculate that the improvement in stroke volume in the two groups improves the blood and oxygen supply to the working muscles to a similar extent as the improvement in the muscles ability to utilize oxygen since no additional affect of hyperoxia was detected at post testing. Since the difference between normoxic and hyperoxic training is arterial oxygen content and not cardiac output or shear stress in the blood supply chain, these findings seem to support Wagner's 21 notion that the oxygen supply is of the greater importance than the demand for oxygen in terms of explaining training induced changes.

Despite the lack of difference in the training response between normoxic and hyperoxic training, a substantial improvement in VO_2peak following 30 interval sessions using 4x4 minutes interval training at 85 to 95% of maximal heart rate was found. The 16–17% improvement in VO_2peak confirms that aerobic interval training at 85–95% of HR_peak is highly effective for improving VO_2peak and stroke volume in coronary artery disease patients 2, 22.

No significant change in total blood-or plasma volume was observed from pre to post training in the HT in the present study (Table V). Red blood cell mass was, however, significantly decreased after exercise. The findings in the present study differ from the data of the studies reporting long-term changes in plasma volume with aerobic exercise 23, but in line with others having found no change 4, 24. A reduced red cell mass could decrease the oxygen carrying capacity of the blood, and thereby could not explain the improved VO_2peak in this experiment.

In the present study the HT group displayed a significant improvement in peak heart rate and lactate concentration from pre to post training. Both the HT and the NT training groups improved peak ventilation from pre to post training as expected from improvements in VO_2peak. The HT group did however improve ventilation to a greater extent than the NT group. Improvements in these variables in the HT group may be a result of a greater motivation to push through to exhaustion during VO_2peak post testing. Part of the increased heart rate may be due to atria fibrillation during exercise and not a true difference in exertion during pre and post testing. In the NT group significant increase in self reported perceived exertion from pre to post training may be a factor of patients adjusting their perception of exercise strain over the course of the training intervention. Despite differences between the groups at the pretesting point in the Borg scale, no significant difference in R values was noted between the groups or in any of the exercise groups between the pre and the post tests. This implies that the level of strain and hyperventilation was equal between the tests. Therefore, change in the perceived exertion was not an effect of greater level of exhaustion in the post test, but a factor of the patients’ interpretation of the scale.

In the present study breathing 100% oxygen enriched air during high intensity aerobic interval training improves VO_2peak to the same extent as ambient air training in stable coronary artery disease patients with mild to moderate coronary ischemic response to exercise.

Acknowledgements

We gratefully acknowledge the assistance of MSc Siri Bjørgen, MSc Vigdis Schnell Husby and research nurse Aud Hiller. The study was supported by grants from The Norwegian University of Science and Technology and St. Olav University Hospital.