Exercise capacity and VO2 in heart failure

INTRODUCTION — Limitation of exercise capacity is one of the cardinal manifestations of heart failure (HF), varying directly with the severity of the disease. Thus, decreased maximal exercise capacity is associated with decreased patient survival and impairment of quality of life. Exercise training may have a variety of benefits, including an improvement in quality of life, reduced hospitalization, and improved survival. (See "Cardiac rehabilitation in patients with heart failure".)

The methods used to measure exercise capacity, the factors that limit exercise capacity in HF, and the use of measurement of peak exercise capacity (also called functional exercise testing) to predict prognosis will be reviewed here. This discussion is generally consistent with the 2000 recommendations published by the European Society of Cardiology and the 2014 update including the expanding role of cardiopulmonary testing [1,2]. This discussion is also consistent with the 2010 clinician’s guide to cardiopulmonary testing and the 2012 clinical recommendations for specific populations, including HF, published by the American Heart Association (AHA) [3,4]. Cardiopulmonary exercise testing and the effects of exercise training in patients with HF are discussed separately. (See "Cardiopulmonary exercise testing in cardiovascular disease" and "Cardiac rehabilitation in patients with heart failure".)

MEASUREMENT OF EXERCISE CAPACITY — Peak exercise capacity is defined as "the maximum ability of the cardiovascular system to deliver oxygen to exercising skeletal muscle and of the exercising muscle to extract oxygen from the blood" [5]. As a result, exercise tolerance is determined by three factors: pulmonary gas exchange (including ventilation); cardiovascular performance, including the peripheral vascular tree; and skeletal muscle metabolism including oxygen extraction. (See "Exercise physiology".)

While the New York Heart Association (NYHA) class is in fact a clinical estimation of activity tolerance, it is a highly subjective one and fails to discriminate or reflect smaller changes in function and important differences within each class level. Hence, a more objective measure is needed for prognosis testing beyond a clinical opinion (ie, peak VO₂).

Peak VO2 — Exercise capacity can be quantitated clinically by measurement of oxygen uptake (VO₂), carbon dioxide production (VCO₂), and minute ventilation [6]. These parameters are measured during exercise with rapidly responding gas analyzers capable of breath-by-breath determination of O₂ and CO₂ concentrations in inspired and expired air, respectively. The maximal oxygen uptake (VO₂max) eventually reaches a plateau despite increasing workload (figure 1). Not surprisingly, VO₂max has a strong linear correlation with both cardiac output and skeletal muscle blood flow (figure 2) [7] (see "Cardiopulmonary exercise testing in cardiovascular disease"). The peak VO₂ divided by the heart rate (peak VO₂/HR) is called the oxygen pulse and is an indirect measure of stroke volume.

The relationship between cardiac output and O₂ uptake (or oxygen consumption) also forms the basis for the Fick equation used to measure cardiac output:

Ventilatory threshold — The ventilatory threshold (VT), formerly referred to as the anaerobic threshold (AT), is another index used to estimate exercise capacity. It is defined as the point at which minute ventilation increases disproportionately relative to VO₂, a response that is generally seen at 60 to 70 percent of VO₂max (figure 1). The VT is a reflection of the disproportionate increase in lactic acid production by working muscles. It can be used to distinguish between noncardiac (pulmonary or musculoskeletal) and cardiac causes of exercise limitation, since patients who fatigue prior to reaching VT are likely to have a noncardiac problem [6].

Six-minute walk test — The six-minute walk test is a useful, simple, and inexpensive test that correlates with VO₂max and outcomes and has been used in clinical trials [8-11]. This submaximal exercise test measures the distance ambulated on a level hallway surface during six minutes as described in the 2002 American Thoracic Society (ATS) statement on the six-minute walk test.  

EXERCISE CAPACITY IN HF — Exercise capacity is reduced even in mild HF. The cardiac output may be relatively normal at rest, but is usually unable to increase adequately with even mild exertion [7]. As in normal subjects, peak VO₂ in HF is directly related to peak exercise cardiac output and muscle blood flow (figure 2). However, the inability to appropriately increase cardiac output results in an insufficient increase in perfusion to exercising muscle, which can cause early anaerobic metabolism, muscle fatigue, and eventual muscle wasting [12]. Since HF patients often do not attain a true VO₂ max, the term "peak VO₂" is often used.

Several factors contribute to the inadequate response to exercise in patients with HF [13]. A review of exercise intolerance in HF emphasized factors in addition to diminished cardiac reserve that contribute to exercise intolerance [14]:

●The inotropic and chronotropic response to catecholamines is reduced; this defect is due at least in part to downregulation of beta receptors and the lower workload achieved by patients with HF [15,16].

●Augmentation of stroke volume via the Starling mechanism may be limited by diastolic dysfunction (changes in the pressure/volume curve) or pericardial constraint.

●In contrast to people without HF, exercise in patients with HF is associated with an elevation in the pulmonary wedge pressure, reflecting high left atrial pressure. This can exacerbate pulmonary congestion, thereby causing dyspnea and limiting exercise capacity.

●Inadequate ventilation and O₂/CO₂ diffusion across the blood-alveolar barrier coupled with insufficient respiratory muscle strength to maintain the higher demands of activity.

●Impaired O₂ carrying capacity of the blood and impaired O₂ diffusion and extraction in skeletal muscle. This parameter may explain exercise fatigue in patients with HF and anemia.

●Both biochemical and functional abnormalities in skeletal muscle are often present, limiting muscle metabolic capacity (figure 3).

●The presence of pulmonary hypertension and increased pulmonary vascular resistance can reduce the cardiac output response to exercise, impairing exercise capacity [17].

●Concurrent valve disease (eg, mitral regurgitation) may be present, limiting the increase in forward output when the regurgitation becomes severe.

Patients with HF with preserved ejection fraction (HFpEF) can have as severe exercise intolerance as patients with HF with reduced ejection fraction (HFrEF). Central (reduced O₂ transport) and peripheral factors contribute to exercise intolerance in both types of HF. Some data suggest that peripheral factors may be particularly important in patients with HFpEF [18]. The prognostic value of peak VO₂ testing in patients with HFpEF is discussed below. (See 'Type of HF' below.)

Pulmonary dysfunction — Reduced respiratory muscle endurance may be present as part of the generalized skeletal myopathy in patients with HF [19]. This abnormality may contribute to the symptoms of fatigue and dyspnea on exertion.

The diaphragm shows a different adaptation from skeletal and respiratory muscle. There is a shift from fast to slow fibers with an increase in oxidative capacity and a decrease in glycolytic capacity [20]. These changes are similar to those seen in the limb muscles that occur with endurance training, suggesting that they result from the increased work of breathing.

There are also changes in pulmonary function in HF. Even in the absence of pulmonary congestion, HF is associated with impaired pulmonary diffusion [21] and an exaggerated increase in minute ventilation in response to exercise, out of proportion to the increase in carbon dioxide production [22,23]. Hyperventilation is primarily due to ventilation/perfusion mismatching, the severity of which is related to the severity of the HF [23]. The combination of increased ventilation and normal gas exchange (unless limited by pulmonary congestion) means that arterial hypoxia does not generally occur during exercise.

USE OF PEAK VO2

Predictive value — In patients with HF who require an objective measure of exercise capacity, we prefer the measurement of peak VO₂ (ie, maximal oxygen uptake). This parameter is a surrogate marker for the maximal cardiac output that an individual can achieve. Patients with a peak VO₂ ≤10 mL/kg/min have the worst prognosis (normal value with exercise is above 20 mL/kg/min) (figure 4).

The predictive value of peak VO₂ is valid only in patients whose exercise capacity is limited by HF. In patients with other limiting comorbidities, additional or alternative exercise parameters may be helpful.

In addition, since peak VO₂ is influenced by age, sex, and body weight, the percent predicted value may be a more reliable indicator of prognosis [24,25]. Values of 50 percent or more are associated with an excellent short-term prognosis.

The potential utility of percent predicted peak VO₂ was illustrated in a study of exercise testing in 181 ambulatory patients with HF who were being considered for transplantation [25]. Survival in patients who achieved ≤50 percent predicted VO₂max was significantly lower than in those who achieved more than 50 percent predicted VO₂max at one (74 versus 98 percent) and two years (43 versus 90 percent). Use of the percent predicted VO₂max provided additional prognostic information to the absolute peak VO₂.

Use — In addition to its utility to predict prognosis in patients with HF, the peak VO₂ is an important component in the evaluation of patients for cardiac transplantation. (See "Heart transplantation in adults: Indications and contraindications".)

Limitations — The prognostic value of peak VO₂ has several limitations:

●The data correlating peak VO₂ and outcomes were derived before the use of therapies that improve survival in patients with advanced HF, such as beta blockers. With such therapies, peak VO₂ still predicts survival, but a peak VO₂ cut point lower than 14 mL/kg/min may be warranted as an indication for referral for transplantation. The ISHLT Guidelines provide guidance for patients on beta blockers suggesting a peak VO₂ cutoff of <12 mL/kg/min in contrast to <14 mL/kg/min for those not on a beta blocker (Class I, Level of Evidence: B) [26]. However, peak VO₂ must be considered in the setting of the patient’s age and sex, given that age is the most powerful predictor of peak VO₂ [27]. (See "Heart transplantation in adults: Indications and contraindications", section on 'Peak VO2'.)

●The predictive value of peak VO₂ is accurate only when exercise capacity is limited by HF. Factors that can prematurely terminate the test must be excluded, including significant peripheral muscular deconditioning, peripheral artery disease, arthritis, angina pectoris, or low patient motivation. In such situations, patients are unable to achieve a true VO₂max. The proper interpretation of the peak VO₂ requires that the patient achieve the ventilatory threshold (VT), indicating that the level of exercise performed exceeded that which can be supported by the cardiovascular system on an aerobic basis and is sufficient to render a test interpretation.

●Peak VO₂ may be less useful in females than in males. The influence of sex on both peak VO₂ and its prognostic value was illustrated in a series of 594 patients with advanced HF (28 percent females) [28]. Although females were younger (49 versus 53 years in males) and had better systolic function (mean left ventricular ejection fraction [LVEF] 29 versus 25 percent), the mean peak VO₂ was lower in females than in males (14.0 versus 16.6 percent, respectively). Despite the lower peak VO₂, females had a higher one-year transplant-free survival (94 versus 81 percent). A better measure of exercise capacity in females may be the percent predicted peak VO₂ for age and body weight. In the HF-ACTION trial, the peak VO₂ for females did correlate with New York Heart Association (NYHA) class but was considerably lower than that for the males [29].

●The absolute peak VO₂ is also influenced by such factors as age and body weight. The percent predicted peak VO₂ corrects for these factors and may be a more accurate determinant of prognosis [24,25]. HF centers have been using the percent predicted peak VO₂ to assess prognosis.

●Exercise oscillatory ventilation is often observed in patients with severe HF and may make the determination of peak VO₂ more challenging. However, oscillatory ventilation itself may be a marker of more severe disease and worse prognosis [30-32].

Type of HF — Cardiopulmonary testing cannot reliably differentiate HFrEF from HFpEF. Studies shown that CPX variables are equally abnormal in HFpEF patients compared with HFrEF patients when matched by clinical characteristics and that the VO₂ can predict outcomes in this population, including readmissions [33,34]. In addition, there seems to be a correlation between peak VO₂ and other physiologic variables, such as N-terminal pro-B-type natriuretic peptide [33]. However, in HFpEF patients, the source of impaired VO₂ may be more strongly related to abnormal arteriovenous O₂ difference with reduced maximal oxidative capacity in skeletal muscle [35-37].

While the multiple sources of exercise intolerance have been extensively studied in HFrEF, those of HFpEF are less well-accepted or studied. Large database cohorts are showing a strong association with obesity and low levels of leisure-time physical activity with HFpEF. Limitations of exercise capacity in HFpEF may include poor skeletal muscle function and fat/muscle ratio, greater prevalence of right ventricular dysfunction, and both systemic and pulmonary hypertension. Obesity as a phenotype of the HFpEF syndrome is also associated with poor ventilation from reduced respiratory muscle strength [14,38].

Additional predictors — The prognostic value of peak VO₂ may be enhanced by additional measures of the physiologic response to exercise:

●Systolic blood pressure – The predictive value of peak exercise systolic blood pressure was illustrated in a three-year follow-up study of 500 patients with HF referred for transplantation [39]. On multivariate analysis, peak exercise systolic blood pressure and the percent predicted VO₂max were the two most important predictors for the combined end point of death or listing for transplant. The three-year survival rate for patients with a peak exercise VO₂ ≤14 mL/min/kg who were unable to reach a peak exercise systolic blood pressure of 120 mmHg was 55 percent compared with 83 percent for those able to achieve this blood pressure (figure 5).

●Cardiac output – The incremental predictive survival value of the cardiac output response to maximal exercise was illustrated in the following studies:

In a study 185 patients with advanced HF, a "normal" cardiac output response to exercise provided incremental information above the absolute VO₂ determination in predicting survival [40]. This was especially true in patients with VO_2max responses below 10 mL/min/kg.

In another study, two hundred and nineteen patients underwent exercise testing while being monitored with a Swan-Ganz catheter [41]. Peak exercise stroke work index was the most powerful prognostic variable; the two-year survival for those with a stroke work index ≤30 g/m² was 54 percent compared with 91 percent for those with an index >30 g/m² (p <0.0001). Among patients with a relatively low peak VO₂, those who had a normal cardiac output response to exercise had an excellent two year survival (87 versus 58 percent in the others) (figure 6).

However, invasive hemodynamic exercise determinations are not practical for routine evaluation. Furthermore, a smaller study found that only 6 percent of patients had a normal response to exercise, suggesting that this approach may not be cost-effective [42].

●Ventilatory threshold – It has been suggested that the VT might be more predictive than the peak VO₂ because it is less prone to error. In one report of 223 consecutive patients with HF, a VT <11 mL/kg/min was more predictive of six-month mortality than a peak VO₂ ≤14 mL/kg/min (odds ratio [OR] 5.3 versus 3.4) [43]. The risk is almost doubled when a low VT was combined with impaired ventilatory efficiency (OR 9.6).

●Stress induced LV dilation – One noninvasive technique that may be helpful, particularly in patients with an idiopathic dilated cardiomyopathy who have a borderline VO₂max (10 to 14 mL/kg/min) is dobutamine echocardiography. An increase in LV end-diastolic diameter and wall stress in such patients is predictive of mortality [44].

●Ventilatory response – There has been increased interest in using the VE/VCO₂ slope for prognosis. During cardiopulmonary exercise testing, a close linear relationship exists between the production of CO₂ (VCO₂) and minute ventilation (VE). The slope of the regression line relating CO₂ production and minute ventilation (VE/VCO₂) can be used to describe the ventilatory response to exercise. Some studies have now shown that the combination of the VE/VCO₂ slope added to the peak VO₂ may better predict outcome. Others have reported that the ratio is superior to peak VO₂ alone [2]. There is general agreement that the measurement of VE/VCO₂ during the entire exercise time is more physiologic. The VE/VCO₂ slope is discussed further below. (See 'Ventilatory efficiency' below.)

OTHER PREDICTIVE EXERCISE PARAMETERS

Six-minute walk test — In patients with HF who cannot perform a peak VO₂ test, we typically measure exercise capacity with a six-minute walk test (6MWT) (table 1). The 6MWT does not require the use of specialized equipment, can be performed in patients who cannot exercise on a treadmill or cycle, and provides prognostic information. The 6MWT is particularly useful in patients with New York Heart Association (NYHA) class III to IV HF symptoms in whom the test result is reproducible and correlates with self-reported health status (eg, activities of daily living). These characteristics of the 6MWT are illustrated by the following studies:

●In a retrospective analysis of 440 patients from a randomized controlled trial with NYHA class III or IV HF symptoms, the 6MWT was associated with mortality (HR/100 m 0.58; 95% CI 0.5-0.68) and hospitalization (HR 0.85; 95% CI 0.46-0.90) [10].

●Similarly, in a series of 476 patients from a single referral center, the distance walked at baseline was an independent predictor of two-year survival [11].

●In HF-ACTION, changes in patient-reported health status by the Kansas City Cardiomyopathy Questionnaire (KCCQ) were weakly correlated with changes in functional capacity. A 5-point improvement in KCCQ was associated with a 112 m improvement in six-minute walk distance and a 2.5 mL/kg/min increase in peak VO₂ [45].

Exercise duration — In HF ACTION, the most powerful predictor of the primary end point (death or all cause hospitalization) or death at median 2.5-year follow-up was exercise duration on a standardized baseline cardiopulmonary exercise test (Modified Naughton protocol) [46].

Ventilatory efficiency — A possible alternative in patients who cannot achieve VO₂max is measurement of the ventilatory efficiency, ie, ventilation-to-carbon dioxide production ratio in early exercise [47-49]. During cardiopulmonary exercise testing, a close linear relationship exists between the production of CO₂ (VCO₂) and minute ventilation (VE). The slope of the regression line relating CO₂ production and minute ventilation (VE/VCO₂) can be used to describe the ventilatory response to exercise. In patients with HF, the VE/VCO₂ slope is easier to obtain than parameters of maximal exercise capacity and is a better predictor of outcome than VO₂max, NYHA class, or LVEF [43,49,50].

HF is associated with an increase in VE due to increased dead space ventilation because of poor ventilation/perfusion matching, and an increase in VCO₂ relative to VO₂ resulting from bicarbonate buffering of lactic acid [51]. A VE/VCO₂ regression line slope of >34 is associated with reduced cardiac output during exercise, increased pulmonary artery wedge pressures, and reduced survival [43,47]. In one report, patients with a VE/VCO₂ slope above 34 had more severe HF with significantly reduced survival during 18-month follow-up (69 versus 95 percent for those with lower values) (figure 7). In addition, an elevated VE/VCO₂ slope adds incremental prognostic value to both VO₂max [47,49] and the anaerobic threshold [43], while a value above 44.7 adds incremental prognostic value to a low chronotropic index, which correlates with the peak heart rate achieved with exercise (figure 8) [52].

An increase in VE/VCO₂ slope is also predictive of outcome in patients with preserved exercise capacity. Among 123 patients with a VO₂max ≥18 mL/kg/min, the three year survival was significantly lower in those with a VE/VCO₂ ≥34 (57 versus 93 percent for VE/VCO₂ ≤34) [53].

Oxygen uptake efficiency parameters — The oxygen uptake efficiency slope (OUES) is a parameter derived from VO₂ and VE [54]. This parameter appears insensitive to the duration of exercise and correlates with outcomes in patients with advanced HF [55]. These features suggest that OUES may be more useful than peak VO₂, particularly in patients with limited exercise capacity for reasons other than HF.

A preliminary study of 508 patients with HF with LVEF <35 percent found that the highest plateau of oxygen uptake efficiency (OUEP) during incremental cardiopulmonary exercise was a stronger predictor of mortality at six months than VO₂ or six-minute walk distance [56]. However, further validation of this finding is required.

Exercise hemodynamics — We do not recommend the routine use of exercise hemodynamics [41].

EXERCISE TRAINING IN HF — A separate issue is the use of exercise training to improve symptoms in selected patients with HF (see "Cardiac rehabilitation in patients with heart failure"). Training does not improve cardiac function at rest, as estimated from LVEF, baseline cardiac output, or wedge pressure [57]. It can, however, have the following benefits in both males and females [58]:

●Increased peak VO₂, peak cardiac output, and leg blood flow during exercise [57,59]. These improvements may be related to an increase in peak early diastolic filling rate of the LV at rest and during exercise [60].

●Improved muscle energetics so that oxygen utilization becomes more efficient, allowing a similar amount of work to be performed at a lower heart rate, rate-pressure product, and minute ventilation (indicating improved gas exchange) [61,62].

●Reversal of the abnormalities in mitochondrial density and ultrastructure and on fiber type distribution in skeletal muscle seen with HF. For example, one study prospectively randomized 18 patients with HF to an ambulatory training or to a control group that was physically inactive [63]. After six months, the patients who exercised had a significant increase in mitochondrial density and enhanced oxidative enzyme activity and a concomitant reshift to type I slow-twitch fibers. These changes appear to be unrelated to changes in peripheral perfusion.

●Partial reversal of endothelial dysfunction with restoration of endothelium-mediated flow-dependent dilation, possibly due to enhanced endothelial release of nitric oxide [64].

●In patients with an ischemic cardiomyopathy, there is an improvement in both thallium activity within the myocardium and the contractile response of dysfunctional myocardium to low dose dobutamine; this correlates with an increase in collaterals [65].

●Reduction in plasma B-type natriuretic peptide values. (See "Natriuretic peptide measurement in heart failure".)

As a result, more exercise can be performed before reaching the anaerobic threshold (figure 9). In one study, for example, exercise training in stable patients with a mean LVEF of 20 percent led to an increase in peak VO₂ from 13.2 to 15.6 mL/kg/min [57]. Another study reported that the improvement in VO₂ was not related to baseline central hemodynamics, but was greater in patients with a lower baseline VO₂ [66].

Thus, all three determinants of peak exercise - cardiac efficiency, pulmonary gas exchange, and skeletal muscle metabolism - may be improved by exercise training. The ability to perform more work with less energy expenditure may have the following benefits for the patient.

●There may be an improvement in symptoms such as dyspnea and fatigue.

●The lower rate-pressure product may allow patients with coronary disease to perform their daily tasks with fewer symptoms and less disability.

●There may be a reduction in sympathetic tone and an increase in vagal tone at rest, thereby restoring autonomic cardiovascular control towards normal [67]. To the degree that these changes reduce systemic vascular resistance and cardiac afterload, they may lead to an improvement in cardiac performance.

There was previously no proof that improvement in peak VO₂, as can be induced by exercise training, is associated with improvements in outcomes. In HF-ACTION, every 6 percent increase in peak VO₂, adjusted for other significant predictors, was associated with a 5 percent lower risk of the primary end point, a 4 percent lower risk of the secondary end point of time to cardiovascular mortality or cardiovascular hospitalization, an 8 percent lower risk of cardiovascular mortality or HF hospitalization, and a 7 percent lower all-cause mortality [68].

Furthermore, there is sufficient evidence of an improvement in symptoms and exercise capacity that we recommend that all stable patients with symptomatic HF participate in an exercise training program [57]. (See "Cardiac rehabilitation in patients with heart failure".)

The impact of exercise programs on outcomes was studied in the HF ACTION trial, which is discussed separately. (See "Cardiac rehabilitation in patients with heart failure".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Heart failure in adults" and "Society guideline links: Stress testing and cardiopulmonary exercise testing".)

SUMMARY — In clinical practice, many physicians follow the symptoms and signs of patients with heart failure (HF) as a guide to their course and prognosis. Nevertheless, serial assessment of maximal exercise capacity in patients with HF every three to six months is an important method to evaluate both the therapeutic response to pharmacologic intervention as well as prognosis. Determination of functional capacity is now an outpatient performance measure as published by the American Heart Association/American College of Cardiology Performance Measures for Chronic Heart Failure [69].

Patients with mild HF (New York Heart Association [NYHA] class I or II) can be evaluated by repetitive treadmill exercise testing without a metabolic cart assessment for maximal oxygen consumption (ie, peak VO₂) (table 2); in this setting, total exercise duration can be used as an objective measure of functional capacity.

By comparison, patients who present with or progress to moderate to severe HF (NYHA class III or IV) should be referred for VO₂ testing. The specific method used to measure VO₂ is less important as long as the test is performed and interpreted in a consistent manner. Improved exercise capacity is often associated with an increase in peak VO₂ in these patients [57].