Maximal exercise oxygen pulse as a predictor of mortality among male veterans referred for exercise testing

Abstract

Background

Maximal oxygen pulse (O₂ pulse) mirrors the stroke volume response to exercise, and should therefore be a strong predictor of mortality. Limited and conflicting data are, however, available on this issue.

Methods

Nine hundred forty-eight participants, classified as those with cardiopulmonary disease (CPD) and those without (non-CPD), underwent cardiopulmonary exercise testing (CPX) for clinical reasons between 1993 and 2003. The ability of maximal O₂ pulse and maximal oxygen uptake (peak VO₂) to predict mortality was investigated using proportional hazards and Akaike information criterion analyses. All-cause mortality was the endpoint.

Results

Over a mean follow-up of 6.3 ± 3.2 years, there were 126 deaths. Maximal O₂ pulse, expressed in either absolute or relative to age-predicted terms, and peak VO₂ were significant and independent predictors of mortality in those with and without CPD (P < 0.04). Akaike information criterion analysis revealed that the model including both maximal O₂ pulse and peak VO₂ had the highest accuracy for predicting mortality. The optimal cut-points for O₂ pulse and peak VO₂ (< 12; ≥ 12 ml/beat and < 16; ≥ 16 ml/(kg · min) respectively) were established by the area under the receiver-operating-characteristic curve. The relative risks of mortality were 3.4 and 2.2 (CPD and non-CPD, respectively) among participants with both maximal O₂ pulse and peak VO₂ responses below these cut-points compared with participants with both responses above these cut-points.

Conclusion

These results indicate that maximal O₂ pulse is a significant predictor of mortality in patients with and without CPD. The addition of absolute and relative O₂ pulse data provides complementary information for risk-stratifying heterogeneous participants referred for CPX and should be routinely included in the CPX report.

Keywords

exercise mortality oxygen pulse

Introduction

Over the last two decades, the cardiopulmonary exercise test (CPX) has become increasingly recommended in clinical practice because of its impact on the clinical decision-making process [1]. The information obtained from CPX helps to identify unexplained exercise intolerance, determine disability, make decisions regarding therapeutic interventions, and to estimate prognosis [2, 3]. Among the ventilatory expired gas variables obtained during exercise, maximal oxygen uptake (peak VO₂) remains the most frequently applied variable in both research and clinical settings because of its widely recognized value for clinical assessment and for estimating prognosis [4, 5]. Despite the well-established value of peak VO₂, a large body of research has been recently published regarding the prognostic value of other CPX variables. Responses such as the slope of the relationship between ventilation and carbon dioxide production (VE/VCO₂ slope) [6, 7], the oxygen uptake efficiency slope [8], and maximal oxygen pulse (O₂ pulse) [9–11] have been shown to have complementary and even superior prognostic value to that of peak VO₂.

O₂ pulse (defined by dividing VO₂ by heart rate) is generally considered a reflection of cardiovascular efficiency during exercise. As VO₂ is the product of cardiac output and the arterio-venous O₂ content difference [C(a-v)O₂], O₂ pulse is in effect the oxygen consumed per heart beat. Although a few previous studies have investigated the prognostic value of maximal O₂ pulse in healthy participants and patients with heart failure [9–11], conflicting results have been observed. In addition, previous findings have not incorporated potentially important information such as the degree of chronotropic incompetence, use of β-blockers, and body proportions. Thus, the primary purpose of this study was to investigate the prognostic value of maximal O₂ pulse among male adult participants with and without cardiopulmonary disease (CPD) referred for exercise testing for clinical reasons. A secondary purpose was to determine whether maximal O₂ pulse has prognostic value that complements other exercise test responses and whether chronotropic incompetence, β-blockers and body weight change this association.

Methods

Population

The study sample consisted of 948 men who underwent maximal CPX for clinical reasons at the Veterans Affairs Palo Alto Medical Center between 1993 and 2003. Detailed clinical history, current medications, and risk factors were recorded prospectively on computerized forms at the time of the tests. We defined participants with CPD as those with a history of angiographically documented coronary artery disease, myocardial infarction, coronary bypass surgery, coronary angioplasty, typical angina, claudication, atrial fibrillation, stroke, or an abnormal result on an exercise test suggestive of coronary artery disease (horizontal or downsloping ST-segment depression ≥ 1.0 mm, exercise-induced angina, or both). Nine percent of the population (88 participants) had a history of mild pulmonary disease and were included in the CPD group, which totalled 442 participants. The remaining 506 participants were classified as non-CPD. All patients with a diagnosis of heart failure were excluded from the study.

Exercise testing

After providing written informed consent, all the participants underwent symptom-limited exercise testing using an individualized ramp treadmill protocol [12]. Before testing, a questionnaire was used to set an individualized treadmill ramp rate such that test duration was targeted to achieve the recommended range of 8-12 min [13]. Standard criteria for termination were used, including moderately severe angina, greater than 2.0 mm abnormal ST depression, a sustained drop in systolic blood pressure, or serious rhythm disturbances. The Borg 6-20 perceived exertion scale was used to estimate the degree of exertion [14]. An electrocardiogram was continuously obtained throughout the exercise test. Blood pressure was obtained manually every 2 min. Heart rate was measured supine, standing, during each minute of exercise, and at maximum exercise from an electrocardiogram. Chronotropic incompetence was defined as the inability to achieve 80% of predicted heart rate reserve, using the regression equation that best fits the population [(174-0.54 (age)] [15]. Thus, the equation used to define percentage heart rate reserve was {(maximum heart rate-resting heart rate)/[(174-0.54 × age) − (resting heart rate)] × 100}. No test was classified as indeterminate, medications were not stopped for testing, and age-predicted maximum target heart rates were not used as endpoints. The use of handrails during exercise was discouraged. The exercise tests were performed, analyzed, and reported using a standard protocol incorporating a computerized database with all definitions and measurements prospectively defined [16].

Oxygen uptake and oxygen pulse assessments

Respiratory gas exchange data were analyzed breath-by-breath and expressed every 10 s using rolling 30 s averages. Peak VO₂ was expressed as the highest 30-s average value obtained during the CPX. Maximal O₂ pulse was calculated by dividing peak VO₂ by heart rate, at the time peak VO₂ was achieved and was expressed in milliliters per beat. In addition, this value was expressed as a percentage of age-predicted value, which corresponds to the ratio between the predicted peak VO₂ and predicted maximal heart rate. This approach has the advantage of incorporating body weight through the use of relative predicted peak VO₂ in ml/(kg-min) [17].

Follow-up

The study endpoint was all-cause mortality. The California Health Department Service and Social Security Death Indices were used to ascertain the vital status of each patient as of 7 December 2007.

Statistical analysis

Clinical and exercise variables comparing CPD and non-CPD participants and survivors versus nonsurvivors were compared using χ² tests for categorical variables and unpaired t-tests for continuous variables. The relationship among maximal O₂ pulse, percentage-predicted O₂ pulse, peak VO₂, and all-cause mortality was evaluated using the Kaplan-Meier method. A Cox proportional hazard model was used to determine which variables were independently associated with time to death. Optimal cut-points for maximal O₂ pulse, percentage-predicted O₂ pulse, and peak VO₂ were defined by area under the receiver-operating-characteristic curves, and used in the Cox regression models. The Akaike information criterion (AIC) and weights were used to select the models with highest accuracy for predicting mortality [18]. AIC is a technique designed to optimize model selection by ranking several competing models in terms of their goodness of fit. AIC weights compare multiple candidate models, and determine which most accurately describes the data. Statistical analyses were performed using NCSS (NCSS Inc., Kayesville, Utah, USA).

Table 1

Baseline demographic and clinical characteristics


	All (n = 948)	CPD (n = 442)	Non-CPD (n=506)	P

Demographic characteristics
Age (years)	57 ± 12	59 ± 11	54 ± 12	< 0.001
Body mass index (kg/m²)	28.3 ± 4.9	28.4 ± 5.1	28.2 ± 4.8	0.47
Medications (%)
β-blockers	16	23	10	< 0.001
Calcium channel blockers	19	27	13	< 0.001
Nitrates	10	21	−	−
ACE inhibitors	17	19	15	0.06
Anticoagulant	33	43	24	< 0.001
Statins	4.9	6.6	3.4	0.02
Risk factors (%)
Hypertension	49	57	42	< 0.001
Dyslipidemia	38	42	33	0.003
Obesity	31	32	30	0.48
Diabetes	12	15	8.9	0.005
Smoking history	71	77	65	< 0.001
Medical history (%)
Typical angina	16	34	−	−
Claudication	4.1	8.8	−	−
Stroke	4.0	8.6	−	−
Myocardial infarction	11	23	−	−
Atrial fibrilation	2.6	5.7	−	−
Pulmonary disease	9.3	20	−	−
Valvular disease	3.5	4.1	3.0	0.35
Procedures (%)
CABG	7.0	15	−	−
PCI	5.7	12	−	−

Plus-minus values are means ± SD. ACE, angiotensin-converting enzyme; CABS, coronary artery bypass graft; CPD, cardiopulmonary disease group; Non-CPD, noncardiopulmonary disease group; PCI, percutaneous coronary intervention.

Results

During a mean (± SD) follow-up period of 6.3 ± 3.2 years (median of 6.7 years; range 0.2-12.6), a total of 126 (13%) participants died, yielding an average annual mortality rate of 2.2%.

Demographic characteristics

Compared with the non-CPD group, participants with CPD were older, had a greater prevalence of dyslipidemia, hypertension, diabetes, more extensive use of medicines, more cardiovascular interventions, and were more likely to have smoked (Table 1).

Exercise test responses

Demographic characteristics and exercise test responses for participants who survived and those who died between the CPD and non-CPD groups are presented in Table 2. Maximum heart rate, peak VO₂, maximal O₂ pulse, and the percentage-predicted O₂ pulse were significantly higher in survivors than in nonsurvivors in both the groups.

Maximal oxygen pulse, peak oxygen uptake, and mortality

The optimal cut-points for maximal O₂ pulse, percentage-predicted O₂ pulse, and peak VO₂ were less than 12; ≥ 12 ml/beat, less than 76%; ≥ 76%, and less than 16; ≥ 16 ml/(kg·min), respectively.

After adjusting for potential confounders including age, medications, risk factors, medical history, and chronotropic incompetence, maximal O₂ pulse significantly predicted all-cause mortality in both CPD [hazard ratios (HR): 2.91, 95% confidence interval (CI): 1.70-5.01, P = 0.01] and non-CPD (HR: 1.80, 95% CI: 1.01-3.20, P = 0.04) groups when the participants with low O₂ pulse (< 12 ml/beat) were compared with participants with high O₂ pulse (≥ 12 ml/beat). Results of the Kaplan-Meier analysis are shown in Fig. 1a.

Table 2

Characteristics and exercise test responses among participants who survived and participants who died


	CPD				Non-CPD

Variable	All (n = 442)	Survived (n = 370)	Died (n = 72)	P	All (n = 506)	Survived (n = 452)	Died (n = 54)	P

Demographic characteristics
Age (years)	59 ± 11	59 ± 11	64 ± 12	< 0.001	54 ± 12	53 ± 12	63 ± 12	< 0.001
Body mass index (kg/m²)	28.4 ± 5.1	28.6 ± 5.0	27.5 ± 5.3	0.04	28.2 ± 4.8	28.3 ± 4.8	26.8 ± 4.9	0.01
Height (cm)	176.5 ± 7.2	176.6 ± 7.3	176.1 ± 6.5	0.59	177.1 ± 7.1	177.1 ± 7.2	177 ± 7.0	0.72
Weight (kg)	88.6 ± 16.8	89.2 ± 16.3	85.5 ± 18.7	0.04	88.6 ± 16.7	89.0 ± 16.6	84.6 ± 17.0	0.04
Resting values
Heart rate (beats/min)	76 ± 13	75 ± 13	76 ± 14	0.27	76 ± 14	76 ± 14	75 ± 13	0.29
Blood pressure (mmHg) Diastolic	83 ± 12	83 ± 12	81 ± 13	0.10	84 ± 11	84 ± 11	81 ± 12	0.02
Systolic	136 ± 20	135 ± 19	140 ± 24	0.02	133 ± 19	133 ± 19	135 ± 23	0.26
Maximal values
Heart rate (beats/min)	136 ± 23	138 ± 23	127 ± 21	< 0.001	150 ± 23	152 ± 23	135 ± 23	< 0.001
Heart rate reserve (%)	92 ± 34	94 ± 33	82 ± 35	0.002	110 ± 33	112 ± 32	96 ± 35	< 0.001
Blood pressure (mmHg) Diastolic	87 ± 16	87 ± 16	85 ± 15	0.20	86 ± 15	86 ± 15	86 ± 20	0.45
Systolic	177 ± 23	179 ± 28	172 ± 28	0.02	185 ± 26	185 ± 26	179 ± 27	0.07
Peak VO₂ (ml/(kgmin))	21.9 ± 7.6	22.7 ± 7.7	18.0 ± 5.9	< 0.001	26.8 ± 9.0	27.5 ± 9.1	21.0 ± 5.8	< 0.001
O₂ pulse (ml/beat)	14.2 ± 4.5	14.6 ± 4.5	12.0 ± 4.0	< 0.001	15.6 ± 4.5	15.9 ± 4.5	13.2 ± 4.1	< 0.001
Predicted O₂ pulse achieved (%)	81 ± 23	82 ± 23	72 ± 19	< 0.001	86 ± 24	86 ± 24	78 ± 23	0.007

Plus-minus values are means ± SD. CPD, cardiopulmonary disease group; Non-CPD, noncardiopulmonary disease group. P values are for comparison between the participants who survived and those who died in each group.

An age-adjusted multivariate Cox regression analysis, controlled for all potential confounders, revealed that the percentage-predicted O₂ pulse achieved was also a significant predictor of all-cause mortality in both CPD (HR: 1.98, 95% CI: 1.16-3.39, P = 0.01) and non-CPD (HR: 2.46, 95% CI: 1.03-4.15, P = 0.008) groups, when the participants with low percentage-predicted O₂ pulse (< 76%) were compared with participants with high percentage-predicted O₂ pulse (≥ 76%). When both groups were considered together in the Cox model, every 10% increase in the percentage-predicted O₂ pulse achieved yielded an improvement of 11% in survival (HR: 0.89, 95% CI: 0.80-0.99, P =0.04).

There was no interaction between the use and nonuse of β -blockers and predictive value of maximal O₂ pulse. The maximal O₂ pulse showed a similar trend for predicting all-cause mortality when the Cox regression model was adjusted for β-blocker use, when those taking β-blockers were assessed separately and when they were excluded from the analysis.

After adjusting for potential confounders, peak VO₂ was a significant predictor of all-cause mortality in both CPD (HR: 1.91, 95% CI: 1.12-3.26, P = 0.017) and non-CPD (HR: 2.52, 95% CI: 1.08-5.86, P = 0.03) groups, when the participants with low peak VO₂ (< 16ml/(kg-min)) were compared with participants with high peak VO₂ (≥ 16ml/(kg-min)). Results of the Kaplan-Meier analysis for peak VO₂ are shown in Fig. 1b.

AIC weights showed a higher accuracy for predicting mortality when maximal O₂ pulse and peak VO₂ were included together in the Cox model (79%), compared with models including peak VO₂ (4%) or maximal O₂ pulse (17%), separately.

The complementary predictive value of maximal O₂ pulse and peak VO₂ are shown in Table 3. Participants with both maximal O₂ pulse and peak VO₂ abnormal responses had 3.4-fold and 2.2-fold higher (CPD and non-CPD, respectively) risk of mortality than those with both normal responses. Results of the Kaplan-Meier analysis for complementary predictive value of O₂ pulse and peak VO₂ are shown in Fig. 2.

Discussion

The results of this study add to the existing body of research showing the prognostic value of maximal O₂ pulse in CPD and non-CPD participants referred for exercise testing for clinical reasons. The results support the concept that maximal O₂ pulse provides complementary information to the prognostic value of peak VO₂ and also that maximal O₂ pulse predicts risk when expressed as a percentage of the age-predicted value achieved.

Maximal oxygen pulse and mortality

According to the modified Fick equation, O₂ pulse is numerically equal to the product of stroke volume and C(a-v)O₂. Normally, O₂ pulse increases with incremental exercise because of increases in both stroke volume and oxygen extraction. At a submaximal level, C(a-v)O₂ generally plateaus and remains relatively constant as exercise progresses. Thus, the pattern of further change in O₂ pulse will reflect the pattern of change in stroke volume, and will be reduced in any condition that reduces stroke volume [2, 19]. Several studies have evaluated O₂ pulse and left ventricular function in patients who have undergone both CPX and nuclear ventriculography, and demonstrated that an impaired O₂ pulse response to exercise is associated with abnormalities in left ventricular function because of myocardial ischemia or infarction [20, 21]. Considering the well-established prognostic value of impaired left ventricular function [22–25], O₂ pulse should therefore be a strong predictor of mortality.

Table 3

Age-adjusted Cox regression for the combined maximum oxygen pulse and oxygen uptake exercise test responses


		All			CPD			Non-CPD

Exercise testing result	Description	HR (95% CI)	P	% Event free	HR (%95 CI)	P	% Event free	HR (%95 CI)	P	% Event free

Peak VO₂ ≥ 16 and O₂ pulse ≥ 12	All responses normal	1 (ref)	−	88	1 (ref)	−	87	1 (ref)	−	88
Peak VO₂ <16 or O₂ pulse < 12	1 of 2 abnormal	1.83 (1.19-2.80)	0.006	74	2.40 (1.34-4.31)	0.003	70	1.28 (0.65-2.49)	0.47	79
Peak VO₂ < 16 and O₂ pulse < 12	All abnormal	2.73 (1.73-4.32)	< 0.001	59	3.37 (1.82-6.23)	< 0.001	56	2.23 (1.06-4.69)	0.03	65

Peak VO₂ is expressed in ml/(kgmin). Maximal O₂ pulse is expressed in ml/beat. CPD, cardiopulmonary disease group; Non-CPD, noncardiopulmonary disease group; O₂ pulse, oxygen pulse; Peak VO₂, maximal oxygen uptake; ref, reference group; 95% CI, 95% confidence interval.

Few studies have evaluated the prognostic value of O₂ pulse [9–11]. Laukkanen et al. [10] studied 1596 men who underwent CPX and made a follow-up for a mean of 14 years. After excluding those with cardiovascular disease and those taking β-blockers, the risks for cardiovascular and all-cause mortality were 2.4-fold and 1.8-fold higher, respectively, when participants with a relatively low maximal O₂ pulse (< 13.5 ml/beat) were compared with participants with a high O₂ pulse (> 17.8 ml/beat). Similarly, in our study, the risk of mortality was 1.8 times higher in the non-CPD group for those with a low maximal O₂ pulse. Our results, however, extend earlier findings, in that maximal O₂ pulse also predicts all-cause mortality in CPD participants, among whom a nearly three-fold higher risk was observed in those with a low maximal O₂ pulse.

Potential confounders of maximal oxygen pulse

Our approach was novel, in that we assessed the influence of potential confounding variables such as chronotropic incompetence, body weight, and β-blockers on the ability of O₂ pulse to predict mortality. Chronotropic incompetence has been shown to predict mortality and coronary heart disease risk even after adjusting for age, exercise capacity, and cardiovascular risk factors [26, 27]. Recently, we observed that those who were unable to achieve 80% of predicted heart rate reserve had a nearly three-fold higher risk of cardiovascular mortality [15]. Therefore, considering the influence of heart rate on O₂ pulse, adjustments for chronotropic incompetence should be considered in studies assessing the predictive value of maximal O₂ pulse. The present results indicate that even after adjusting for chronotropic incompetence, maximal O₂ pulse remained a strong multivariate predictor of all-cause mortality.

We also observed that the percentage-predicted O₂ pulse was significantly different between survivors and nonsurvivors and also between CPD and non-CPD participants (P < 0.001) (Table 2). Moreover, a novel finding was the observation that participants who achieved less than 76% of predicted maximal O₂ pulse had more than two times the risk of all-cause mortality. Out of 126 participants who died, 81 (64%) did not achieve 76% of predicted O₂ pulse. These results confirm previous findings showing not only the value of predicted O₂ pulse in identifying circulatory impairment but also that it is an important predictor of all-cause mortality in both CPD and non-CPD participants.

An additional observation we made was that the results were not greatly affected by β-blocker therapy. This is important, in that previous studies have excluded patients on β-blockers, assuming that the prognostic applications of maximal O₂ pulse would not apply to patients taking these medications. Although our sample of patients taking β-blocking agents was limited (154, of whom 26 died), the trend observed suggests that maximal O₂ pulse predicts risk regardless of the β-blocker used.

Combined predictive value of maximal oxygen pulse and peak oxygen uptake

In contrast to the results of Laukkanen et al. [10], our findings support the concept that maximal O₂ pulse provides information that complements peak VO₂ in predicting mortality. We tested this hypothesis through the use of AIC weight analysis and also by combining exercise responses for both maximal O₂ pulse and peak VO₂ in an age-adjusted multivariate Cox model (Table 3). AIC weights consider model selection and quantify a model's predictive accuracy, based on the assumption that more variables in a given model would provide better predictive accuracy than fewer variables. Stated differently, the AIC weights attempt to find the model that best explains the relationships of data with minimum parameters [18, 28]. In our study, the AIC weights showed that the model including both maximal O₂ pulse and peak VO₂ had the highest predictive accuracy for mortality. After adjustments for all potential confounders, the risks for mortality when both maximal O₂ pulse and peak VO₂ responses were abnormal were 3.4-fold and 2.2-fold higher (CPD and non-CPD, respectively), compared with those whose responses were both normal (Table 3). This might, in part, be explained by the fact that O₂ pulse is more dependent on left ventricular function than peak VO₂, asit incorporates the heart rate response to exercise.

Limitations

As this was a Veterans Affairs cohort, our findings were limited to men. In addition, we had information only on death from any cause; therefore, further studies are needed to assess the prognostic value of maximal O₂ pulse for cardiovascular death.

Conclusion

Our findings have shown that maximal O₂ pulse is an independent predictor of all-cause mortality after adjusting for potential confounding factors including medications, risk factors, and clinical history. In addition, we demonstrated that the age-predicted O₂ pulse achieved was an independent predictor of mortality. These results support the concept that maximal O₂ pulse complements peak VO₂ in predicting the risk of mortality. The simple addition of absolute or relative O₂ pulse to CPX results provides added information on risk stratification and should be routinely included in the test report.

Footnotes

Acknowledgements

Dr Oliveira was supported by CAPES (Brazil)-BEX-3853-06-3. Potential conflicts of interest or disclaimers: none declared.

References

1 Palange

Ward

Carlsen

Casaburi

Gallagher

Gosselink

. Recommendations on the use of exercise testing in clinical practice. Eur Respir J 2007; 29:185–209.

2 American Thoracic Society, American College of Chest Physicians. ATS/ACCP statement on cardiopulmonary exercise testing. Am J Respir Crit Care Med 2003; 167:211–277.

Myers

Applications of cardiopulmonary exercise testing in the management of cardiovascular and pulmonary disease. Int J Sports Med 2005; 26 (Suppl 1):S49–S55.

4 Arena

Myers

Williams

Gulati

Kligfield

Balady

. Assessment of functional capacity in clinical and research settings: a scientific statement from the American Heart Association Committee on Exercise, Rehabilitation, and Prevention of the Council on Clinical Cardiology and the Council on Cardiovascular Nursing. Circulation 2007; 116:329–343.

5 Gibbons

Balady

Bricker

Chaitman

Fletcher

Froelicher

. ACC/AHA 2002 guideline update for exercise testing: summary article: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Update the 1997 Exercise Testing Guidelines). Circulation 2002; 106:1883–1892.

Arena

Myers

Aslam

Varughese

Peberdy

MA.

Peak vo2 and ve/vco2 slope in patients with heart failure: a prognostic comparison. Am Heart J 2004; 147:354–360.

Arena

Myers

Guazzi

The clinical and research applications of aerobic capacity and ventilatory efficiency in heart failure: an evidence-based review. Heart Fail Rev 2007; 13:245–269.

8 Arena

Myers

Hsu

Peberdy

Pinkstaff

Bensimhon

. The minute ventilation/carbon dioxide production slope is prognostically superior to the oxygen uptake efficiency slope. J Card Fail 2007; 13: 462–469.

9 Cohen-Solal

Barnier

Pessione

Seknadji

Logeart

Laperche

. Comparison of the long-term prognostic value of peak exercise oxygen pulse and peak oxygen uptake in patients with chronic heart failure. Heart 1997; 78:572–576.

10.

Laukkanen

Kurl

Salonen

Lakka

Rauramaa

Peak oxygen pulse during exercise as a predictor for coronary heart disease and all cause death. Heart 2006; 92:1219–1224.

11.

Lavie

Milani

Mehra

MR.

Peak exercise oxygen pulse and prognosis in chronic heart failure. Am J Cardiol 2004; 93:588–593.

12.

12 Myers

Buchanan

Smith

Neutel

Bowes

Walsh

. Individualized ramp treadmill. Observations on a new protocol. Chest 1992; 101:236S–241S.

13.

Myers

Herbert

Ribisl

Froelicher

VF.

A nomogram to predict exercise capacity from a specific activity questionnaire and clinical data. Am J Cardiol 1994; 73:591–596.

14.

Borg

Borg's perceived exertion and pain scales. Champaign, IL: Human Kinetics; 1998.

15.

Myers

Tan

Abella

Aleti

Froelicher

VF.

Comparison of the chronotropic response to exercise and heart rate recovery in predicting cardiovascular mortality. Eur J Cardiovasc Prev Rehabil 2007; 14: 215–221.

16.

Shue

Froelicher

VF.

Extra: an expert system for exercise test reporting. J Non-Invasive Test 1998; II-4:21–27.

17.

Wasserman

Hansen

Sue

Whipp

BJ.

Principles of the exercise testing and interpretation. 3rd ed. Baltimore: Lippincott Williams & Wilkins; 1999.

18.

Wagenmakers

Farrell

AIC model selection using Akaike weights. Psychon Bull Rev 2004; 11:192–196.

19.

19 Munhoz

Hollanda

Vargas

Silveira

Lemos

Hollanda

. Flattening of oxygen pulse during exercise may detect extensive myocardial ischemia. Med Sci Sports Exerc 2007; 39:1221–1226.

20.

Klainman

Fink

Lebzelter

Krelbaumm

Kramer

MR.

The relationship between left ventricular function assessed by multigated radionuclide test and cardiopulmonary exercise test in patients with ischemic heart disease. Chest 2002; 121:841–845.

21.

Klainman

Kusniec

Stern

Fink

Farbstein

Contribution of cardiopulmonary indices in the assessment of patients with silent and symptomatic ischemia during exercise testing. Int J Cardiol 1996; 53:257–263.

22.

22 The Multicenter Postinfarction Research Group. Risk stratification and survival after myocardial infarction. N Engl J Med 1983; 309:331–336.

23.

Kambara

Nakagawa

Kinoshita

Kawai

Long-term prognosis after myocardial infarction: univariate and multivariate analysis of clinical characteristics in 1000 patients. Kyoto and Shiga Myocardial Infarction (KYSMI) Study Group. Clin Cardiol 1993; 16:872–878.

24.

24 Sanz

Castaner

Betriu

Magrina

Roig

Coll

. Determinants of prognosis in survivors of myocardial infarction: a prospective clinical angiographic study. N Engl J Med 1982; 306:1065–1070.

25.

25 Zaret

Wackers

Terrin

Forman

Williams

Knatterud

. Value of radionuclide rest and exercise left ventricular ejection fraction in assessing survival of patients after thrombolytic therapy for acute myocardial infarction: results of Thrombolysis In Myocardial Infarction (TIMI) phase ii study. The TIMI Study Group. J Am Coll Cardiol 1995; 26:73–79.

26.

Lauer

Francis

Okin

Pashkow

Snader

Marwick

TH.

Impaired chronotropic response to exercise stress testing as a predictor of mortality. JAMA 1999; 281:524–529.

27.

Lauer

Okin

Larson

Evans

Levy

Impaired heart rate response to graded exercise. Prognostic implications of chronotropic incompetence in the Framingham heart study. Circulation 1996; 93:1520–1526.

28.

Akaike

Data analysis by statistical models. No To Hattatsu 1992; 24:127–133.