Cardiopulmonary exercise testing in the evaluation of unexplained dyspnea

INTRODUCTION — Cardiopulmonary exercise testing (CPET) provides an integrated evaluation of the cardiorespiratory system during exercise. Dyspnea, or shortness of breath, is one of the most common complaints by patients and is characterized by general breathing discomfort that has different meanings to different people. The American Thoracic Society and other expert groups recognize dyspnea as a very complex phenomenon that arises from multiple physiologic, psychologic, social, and environmental factors [1].

The evaluation of dyspnea can be complex and generally requires a history (eg, smoking, pattern and triggers of symptoms), complete physical examination, laboratory testing, electrocardiogram, and pulmonary function tests [2]. While the five most common causes of dyspnea are asthma, chronic obstructive pulmonary disease (COPD), interstitial lung disease (ILD), heart disease, and obesity/deconditioning [3], the wide variety of possibilities makes it challenging to develop a diagnostic strategy that is both time and cost effective, as well as complete [4,5]. A CPET can often be helpful in narrowing down the diagnostic possibilities.

The use of CPET in the evaluation of dyspnea will be reviewed here. Other topics, such as an approach to the evaluation of dyspnea, the physiology of exercise, the role of exercise testing in children and adolescents, and exercise testing in the management of heart failure, are discussed separately.

●(See "Approach to the patient with dyspnea".)

●(See "Exercise physiology".)

●(See "Exercise testing in children and adolescents: Principles and clinical application".)

●(See "Cardiopulmonary exercise testing in cardiovascular disease".)

DEFINITIONS

V_E – Minute ventilation, volume of air exhaled (L/minute)

V_EMAX – Maximum minute ventilation at peak exercise (L/minute)

VO₂ – Measurement of oxygen uptake (L/minute)

VCO₂ – Measurement of carbon dioxide output from the lungs (L/minute)

VCO₂/VO₂ – Respiratory exchange ratio (RER; measured at the mouth)

V_E/VCO2 – Ratio of minute ventilation (L/minute) to rate of carbon dioxide clearance (L/minute)

P_ETO₂ – End-tidal oxygen tension (mmHg)

P_ETCO₂ – End-tidal carbon dioxide tension (mmHg)

V_T – Tidal volume (L)

V_D/V_T – Proportion of dead space ventilation (V_D) per tidal volume breath

O₂ pulse – VO₂/heart rate (HR; L/beat)

OUES – Oxygen uptake efficiency slope (slope of VO₂ versus logV_E)

Anaerobic threshold – The point at which lactate accumulation exceeds lactate clearance during exercise, thought to represent the shift to anaerobic metabolism (measured as a percentage of predicted maximum VO₂)

Chronotropic index – [HR_peak-HR_rest]/[(220-age)-HR_rest], where HR_peak and HR_rest are the heart rate at peak exercise and at rest, respectively

PREPARATION FOR CPET

Indications — CPET can be helpful in a number of settings, most commonly in the evaluation of exercise intolerance.

●Undiagnosed exercise intolerance (eg, dyspnea, fatigue) (see "Approach to the patient with dyspnea")

●Assessment of exercise tolerance in patients with known pulmonary or cardiac disease (see "Cardiopulmonary exercise testing in cardiovascular disease")

●Risk stratification in patients with severe heart failure to determine appropriateness for advanced or invasive therapies (see "Exercise capacity and VO2 in heart failure")

●Preoperative evaluation for lung resection (see "Preoperative physiologic pulmonary evaluation for lung resection", section on 'Integrated cardiopulmonary exercise testing')

●Assessment of impairment for disability claim (see "Evaluation of pulmonary disability", section on 'Exercise tests')

Contraindications — The following are considered contraindications to CPET for most patients:

●Uncontrolled hypertension or arrhythmia

●Unstable angina, myocardial infarction in preceding four weeks, or active heart failure

●Pulmonary embolism in preceding four weeks

●Poorly controlled asthma

●Acute myocardial or pericardial inflammatory disease

●Severe cardiac outflow tract obstruction for which surgical intervention is indicated

Pretest evaluation — In addition to obtaining informed consent, a simple pretest evaluation is important to determine baseline patient characteristics that may be important for the final interpretation of CPET results. Such an evaluation should include review of medications (eg, beta-blockers that may affect heart rate response or inhaled beta-agonists that may affect ventilatory response) and a directed physical exam (eg, baseline cardiac and lung exam) so that any changes with exercise may be detected. Baseline spirometry should be obtained to detect any underlying obstruction or possible restrictive disease and to establish the maximum voluntary ventilation (MVV) for later calculation of breathing reserve. Most CPET software programs use forced expiratory volume in one second (FEV₁) x 35 or 40 as an estimate of MVV, although it can also be measured directly in motivated patients.

Most patients with unexplained dyspnea should have had prior work-up for cardiopulmonary disease and anemia, including resting echocardiography, a cardiac stress test, complete pulmonary function testing (spirometry, lung volumes, and diffusing capacity), exercise oximetry, computed tomography (CT) of the chest, and a complete blood count. (See "Approach to the patient with dyspnea".)

Pretest evaluation for coronavirus disease 2019 (COVID-19) through antigen or polymerase chain reaction (PCR) testing is appropriate as the high minute ventilation of CPET is thought to increase the risk of aerosol generation and COVID-19 (severe acute respiratory syndrome coronavirus 2; SARS-CoV-2) transmission by infected, but asymptomatic patients [6-8].

Exercise testing equipment and protocols — CPETs are typically performed in a specialized testing unit with capability for continuous cardiopulmonary monitoring (ie, electrocardiogram, automated blood pressure monitoring, continuous pulse oximetry); a metabolic unit to measure respiratory rate, tidal volume, oxygen consumption, carbon dioxide production, and work output; and personnel to monitor the test, obtain arterial blood gases, and respond to emergencies.

For infection control purposes, CPET equipment should be fitted with single-use mouthpieces and disposable in-line bacterial and viral filters [6-8]. It is preferred that negative pressure rooms with at least six air exchanges per hour be used for testing.

The cycle ergometer and treadmill are the most common modalities of testing. Different protocols are available to allow the individual to reach their maximal power output within about 8 to 15 minutes with a symptom-limited end to the test. The protocols typically include incremental increases in the power output over predefined time intervals, such as an increase of 25 watts every three minutes (staged) or an increase of 5 to 25 watts every one minute (incremental), and may occur in a stepped or ramp-up format. Increments are often tailored to expected patient fitness; greater increments than those described may be used for testing in extremely fit individuals. Protocols for evaluating patients with heart failure are described separately. (See "Cardiopulmonary exercise testing in cardiovascular disease", section on 'Exercise test protocols'.)

Testing includes measurement of heart rate, blood pressure, pulse oximetry, electrocardiogram (ECG), ventilation, and concentrations of expired gases (oxygen [O₂] and carbon dioxide [CO₂]) both at rest (prior to starting exercise) and continuously throughout exercise. These parameters are typically reported at brief intervals (eg, every 30 seconds or sliding average over every seven breaths) during exercise. Borg ratings of dyspnea and/or fatigue (table 1) are usually obtained at baseline, and the patient is asked to rate their peak level of dyspnea and/or fatigue at the conclusion of the test. Patients are also asked to identify the primary reason they had to stop exercising (eg, shortness of breath, leg discomfort). Flow-volume loops may be performed during and after the test, which may be informative in patients where airflow obstruction is suspected of limiting exercise ability.

Arterial blood gases may be obtained at rest and at peak exercise to assess the effect of exercise on the alveolar-arterial oxygen difference (A-a O₂ difference) and document the development of metabolic acidosis after reaching the anaerobic threshold. In some laboratories, periodic arterial blood gases are obtained at short intervals of one to two minutes via an indwelling arterial line, although resting and peak values obtained by individual arterial sticks are typically sufficient.

Monitoring of all variables is typically continued one to five minutes into the recovery period, depending on local practice.

The data collected during the test are typically available to the interpreting clinician or physiologist in tables that display the data over time of the test (eg, every 30 seconds) and graphical displays that highlight cardiovascular and respiratory response patterns. A 9-panel plot is commonly used in final CPET reports (figure 1) [9].

The technique for continuous laryngoscopy during exercise for the evaluation of exercise-induced laryngeal obstruction is described separately. (See "Exercise-induced laryngeal obstruction", section on 'Continuous laryngoscopy during exercise'.)

OVERVIEW OF CARDIOPULMONARY EXERCISE TESTING (CPET)

General features — CPET is a comprehensive exercise test that is designed to allow assessment of the physiologic factors that limit maximal exercise capacity [4,5,10-17]. These factors may be related to the cardiovascular system, ventilatory and gas exchange response, or metabolic issues. The ability to exercise and reach a normal maximal exercise capacity is related to the normal functioning of each of these systems as an integrated whole. A full description of normal exercise physiology is provided separately. (See "Exercise physiology".)

CPET allows the measurement and identification of the patterns of the body’s responses to exercise. Since most dyspnea is associated with exertion, CPET provides the opportunity to simulate real-world conditions to cause dyspnea on exertion under controlled and monitored circumstances. The development of dyspnea can then be related to the normal or abnormal functioning of each of the body systems that contribute to maximal exercise capacity. Other clinical uses of CPET, such as the evaluation of pulmonary disability, preoperative assessment for lung resection, and monitoring exercise capacity in heart failure, are discussed separately. (See "Evaluation of pulmonary disability" and "Preoperative physiologic pulmonary evaluation for lung resection" and "Exercise capacity and VO2 in heart failure".)

Most metabolic carts generate a 9-panel display (figure 1). While the measurements are interrelated, the top row (panels 1 to 3) typically relates to metabolic and cardiovascular responses, the middle row (panels 4 to 6) to ventilatory and gas exchange responses, and the bottom row (panels 7 to 9) to mechanical responses. Individual measurements are discussed in more detail below.

Metabolic measurements

Maximum oxygen uptake — The maximum oxygen uptake (VO_2max) provides assessment of exercise capacity and differentiates normal from abnormal exercise tolerance. However, a true maximum VO₂ is often not reached, so the highest peak VO₂ is taken as the measure of exercise capacity, assuming a good effort on testing. VO₂ is measured at the mouthpiece, using inhaled and exhaled oxygen concentrations, and reported as L/min or mL/kg/min. Normal is typically considered to be >85 percent predicted. Reference values should represent the exercise protocol and patient population being tested in the exercise laboratory [18-20].

A low peak VO₂ in response to exercise suggests either a problem with O₂ delivery (due to cardiovascular, pulmonary, or circulatory etiologies) or a problem with peripheral utilization, which may be muscular in origin (figure 2). A low peak VO₂ may also reflect poor effort. Of note, it is not uncommon for there to be a discrepancy between the percent predicted peak VO₂ and the percent predicted peak work achieved. It seems reasonable to consider that the maximal metabolic capacity during exercise is assessed by the peak VO₂, whereas the maximal functional capacity during exercise is assessed by the peak work [21].

Respiratory exchange ratio — Overall metabolism can be assessed by the ratio of carbon dioxide output/oxygen uptake (VCO₂/VO₂), known as the respiratory exchange ratio (RER). This is the noninvasive measure of the respiratory quotient (RQ) that is occurring at the cellular level. After the change to anaerobic metabolism, more carbon dioxide is produced (from buffering of lactic acid) compared with oxygen consumed. Baseline RER is approximately 0.8, reaches 1.0 at the anaerobic threshold (AT), and exceeds 1.2 during the recovery phase of exercise. A sign of good effort on CPET is an RER >1.16, unless the patient is hyperventilating (most commonly due to anxiety), which can also increase the RER [18].

Anaerobic threshold — The AT, also known as the lactate threshold or the ventilatory threshold, represents the point at which lactate accumulation exceeds lactate clearance during exercise. The AT is thought to represent the shift to anaerobic metabolism during exercise. The gold standard for determining AT is measurement of lactic acid or trending of serum bicarbonate during exercise. AT normally occurs at more than 40 percent of maximal predicted VO₂ [18,22].

The most common noninvasive methods for determining the AT are by determining the VO₂ where the VCO₂ versus VO₂ slope increases (V-slope method), assessing the VO₂ at the nadir of the minute ventilation versus VO₂ relationship (V_E/VO₂), or evaluating where the plot of end-tidal oxygen tension (P_ETO₂) begins to rise (figure 3) [23]. The accuracy of less invasive approaches is somewhat controversial and can be prone to error that may vary by disease state. There is no one preferred method for determining AT, and we recommend using at least two if not all three methods to help determine the AT with more confidence.

Respiratory measurements — In normal individuals, exercise is not limited by respiratory factors, so evaluation of respiratory reserve can assess contributions of respiratory disease. Additional measurements of flow and volumes during and after exercise can assist in the diagnosis of respiratory diseases.

●Respiratory reserve – The maximum ventilation at peak exercise (V_Emax) is normally 60 to 70 percent of maximum voluntary ventilation (MVV), leaving a breathing reserve. The respiratory reserve can be described as the difference between the MVV and the V_Emax, expressed as a percentage ((MVV-V_Emax)/MVV x 100), such that a normal reserve is 30 to 40 percent. A reduced or absent respiratory reserve (or, equivalently, V_Emax/MVV approaching 100 percent) suggests that exercise is being limited by ventilatory capacity. This is usually due to respiratory or occasionally neuromuscular disease. (See "Exercise physiology", section on 'Breathing reserve index'.)

In performance athletes, there may be a reduced or absent breathing reserve after the maximum VO₂ has been exceeded, signifying a test that has met both the limits of cardiovascular and pulmonary physiologic responses. The patterns of such test results differ in that the maximum VO₂ in performance athletes typically exceeds 100 percent predicted whereas it is reduced in the case of ventilatory limitation to exercise.

●Breathing pattern – Assessment of breathing pattern entails assessment of both the tidal volume (V_T) and the respiratory frequency (f). Under normal conditions, the V_E increases early on primarily due to increases in V_T, which should at least double over baseline, and reach a peak of approximately 50 to 60 percent of forced vital capacity (FVC) or 70 percent of inspiratory capacity (IC) [18]. The respiratory frequency rises throughout exercise but accelerates at the approach of or just beyond the AT.

In patients with interstitial lung disease (ILD), the breathing pattern is typically characterized by rapid, shallower breaths, and this corresponds with disease severity. A similar pattern can be seen in chronic obstructive pulmonary disease (COPD), although patients with COPD are typically experiencing dynamic hyperinflation with this breathing pattern. Dynamic hyperinflation can be better determined by measuring inspiratory capacities (figure 4). (See "Dynamic hyperinflation in patients with COPD".)

Dyspnea associated with anxiety (aka, psychogenic dyspnea) is also frequently characterized by hyperventilation. Other patterns of dysfunctional breathing may also be seen (see below).

●Flow-volume measurements – Flow-volume loops can be measured during exercise and provide insight into ventilatory function during exercise [24]. Prior to exercise testing, a maximal flow-volume loop is measured, and then the tidal breathing loop is positioned within the maximal flow-volume loop based on the measured IC at rest. As exercise proceeds, the breathing loops are recorded in real time and periodically repositioned within the maximal flow-volume loop by recording intermittent ICs during exercise (figure 5). Normally, the maximal exercise flow-volume loop remains within the confines of the maximal flow-volume envelope. Flow limitation is considered to become significant if the volume over which the exercise expiratory loop overlaps the maximal expiratory loop is greater than approximately 20 percent of the exercise V_T [25]. Dynamic hyperinflation results in the end-expiratory lung volume (EELV) rising toward total lung capacity (TLC) rather than falling normally toward residual volume (RV).

Impaired inspiratory flow during exercise can suggest exercise-induced laryngospasm, a form of inducible laryngeal obstruction that frequently presents with unexplained exertional dyspnea. This can be further investigated with exercise laryngoscopy. (See "Inducible laryngeal obstruction (paradoxical vocal fold motion)".)

●Postexertional spirometry – When exercise-induced bronchoconstriction is suspected as a possible cause of unexplained dyspnea, spirometry can be measured periodically over the 30 minutes following a typical CPET [26]. A decrease in forced expiratory volume in one second (FEV₁) of more than 10 to 15 percent during this period compared with before exercise identifies exercise-induced bronchoconstriction, although 15 percent is more diagnostic. Typically, bronchodilators will be given after the 30-minute period if FEV₁ has not returned to normal; demonstration of bronchodilator reversibility of the FEV₁ impairment further supports the diagnosis. (See "Bronchoprovocation testing", section on 'Exercise challenge'.)

Cardiovascular measurements — Although noninvasive CPET is unable to measure cardiac output directly, several exercise measures indirectly assess cardiovascular performance.

●Heart rate – In normal exercise, there is a progressive heart rate response to near maximal predicted normal values (approximately 220 beats/min – age). The heart rate reserve (HRR) is the difference between the age predicted heart rate (HR) response and the measured HR at peak exercise, which is usually close to zero. There is little or no HRR at the end of exercise under normal conditions. Excess heart rate reserve suggests chronotropic incompetence, a cardiovascular etiology of impaired exercise performance. The chronotropic index is a useful metric of chronotropy; a value below 0.8 is the typical threshold used to define chronotropic incompetence.

●Blood pressure – Blood pressure (BP) is measured intermittently during exercise protocols. The normal response is an increase in both systolic and diastolic BP with exercise, but the diastolic BP should not rise more than 20 mmHg above baseline. An impaired rise in systolic BP (<20 to 30 mmHg) may result from aortic outflow obstruction (including aortic stenosis), severe left ventricular dysfunction, myocardial ischemia, and medications that reduce inotropy (eg, beta-blockers) [26]. A decrease in systolic BP below the resting pressure is a sign of severe left ventricular dysfunction and an indication to stop the exercise test.

An abnormally elevated diastolic BP (>90 to 100 mmHg) can lead to impaired tissue oxygen delivery, early anaerobic metabolism, and increased dyspnea [18].

●Electrocardiogram monitoring – Electrocardiogram (ECG) monitoring during exercise allows assessment of cardiac arrhythmia and ischemic changes. The development of paroxysmal atrial fibrillation or other supraventricular tachycardias during exercise is rare. However, ventricular ectopy during exercise is common and suggestive of cardiac disease [26]. The development of ST-segment depression during exercise, as well as the degree, time of appearance, duration, and number of leads involved, correlates with the severity of coronary artery disease or coronary vasospasm. (See "Exercise ECG testing: Performing the test and interpreting the ECG results", section on 'ECG abnormalities during exercise'.)

●Oxygen pulse – The O₂ pulse (VO₂/HR) reflects the delivery of O₂ per heartbeat. The value of the VO₂ divided by HR is equal to the stroke volume (SV) multiplied by the arterial-mixed venous O₂ difference ([a-v]O₂) (figure 2). A peak value of less than 80 percent predicted is generally considered abnormal.

A large percentage of the variation in O₂ pulse pattern reflects changes in SV during exercise. The appearance of a plateau in the slope of the VO₂/HR versus time below 80 percent predicted, particularly when the slope of HR versus time continues to rise, suggests impaired augmentation of SV (figure 6). Left ventricular dysfunction, right ventricular dysfunction, or pulmonary vascular disease (leading to right ventricular dysfunction) can demonstrate this exercise abnormality. (See 'Cardiovascular causes of dyspnea' below.)

Anemia, deconditioning, and metabolic or mitochondrial myopathies can also contribute to a low O₂ pulse due to reduced [a-v]O₂ difference (figure 2). Low O₂ pulse arising from these causes may result in a flatter VO₂/HR slope rather than the early plateau seen with failed SV augmentation [27]. (See 'Peripheral causes of dyspnea' below.)

Pulmonary disease can have a small impact on O₂ pulse by decreasing arterial oxygen saturation (contributing to decreased arterial oxygen content [C_aO₂]) or a larger effect in the setting of pulmonary vascular and right ventricular dysfunction.

Gas exchange measurements — Several of the measurements obtained during CPET yield information about the efficiency of gas exchange, which may be impacted by respiratory disease, pulmonary vascular disease, or heart failure.

●Exercise pulse oximetry – In normal individuals, there may be a normal fall in arterial partial pressure of oxygen (P_aO₂) at maximal exercise due to recruitment of poorly ventilated lung units, O₂ diffusion limitation, and low mixed venous O₂ [28]. However, it is also not uncommon to see a rise in PaO₂ compared with rest due to exercise-induced recruitment of atelectatic lung. A decrease of more than 5 percent in the pulse oximeter estimate of arterial saturation during exercise suggests abnormal exercise-induced hypoxemia. This finding generally results from increased blood flow through poorly ventilated lung units in pulmonary diseases or from worsened diffusion limitation due to lung disease, pulmonary vascular disease, or heart failure. Less commonly, it may be caused by intra- or extracardiac right to left shunt.

Pulse oximetry can be inaccurate during exercise. Most common inaccuracies are due to poor capillary perfusion during exercise (common in cardiovascular diseases) and motion artifacts [29]. An inaccurate pulse rate reading from the oximeter can sometimes identify these inaccuracies. Obtaining an arterial blood gas sample shortly after peak exercise can verify a fall in P_aO₂ and an increase in alveolar-arterial gradient. Postexercise alveolar-arterial gradient greater than 35 mmHg is likely abnormal [28].

●V_E/VCO₂ relationship – The efficiency of ventilation can be evaluated with respect to work rate, VO₂, or VCO₂, but the V_E/VCO₂ slope is the most widely studied and reproducible measure [26]. In those without disease, this slope is linear until very near maximal exercise, with approximately 23 to 25 L/min V_E required to eliminate 1 L/min of CO₂ production. At high levels of VCO₂ near peak exercise, there is often an increase in the V_E/VCO₂ slope due to excess respiratory drive from metabolic acidosis.

Diseases that cause inefficient ventilation (adequately ventilated but poorly perfused lung units), including heart failure, COPD (with dynamic hyperinflation), and pulmonary vascular disease, can manifest with large increases in the V_E/VCO₂ slope during exercise (figure 7). In general, a slope greater than 30 is considered abnormal, and more than 45 is a severe abnormality [18,26,28].

Changes in V_E/VCO₂ over time with increasing workload (VO₂) carry similar information, with higher values suggesting excess ventilation. The typical pattern is a fall in V_E/VCO₂ until near the AT, with a significant increase near the end of exercise due to metabolic acidosis. A nadir of greater than 32 to 34 is generally considered evidence of inefficient ventilation (figure 7) [18].

●P_ETO₂ and P_ETCO₂ – The end tidal measurements of O₂ and CO₂ also represent indirect measures of ventilation and perfusion. In general, excess ventilation compared with perfusion will increase P_ETO₂ and decrease P_ETCO₂. Similar to changes seen in V_E/VO₂ and V_E/VCO₂ with time/workload, P_ETO₂ decreases and P_ETCO₂ increases early prior to the AT, a trend which reverses late in exercise with increased ventilation caused by metabolic acidosis. Decreased resting P_ETCO₂ (<36 mmHg) can be seen with poor cardiac output, pulmonary vascular disease, or hyperventilation, but a fall in P_ETCO₂ (rather than the normal 3 to 8 mmHg rise) prior to the AT is more suggestive of pulmonary vascular dysfunction [30,31]. In contrast, a rising P_ETCO₂ throughout exercise, particularly after AT, can be seen in diseases with severe ventilatory impairment (eg, advanced COPD, neuromuscular diseases) [32].

●Oxygen uptake efficiency slope (OUES) – O₂ uptake efficiency slope (OUES) is the relationship between VO₂ and log V_E throughout exercise. Physiologically, OUES describes the relationship between ventilation and oxygen uptake in the body, which depends on peripheral metabolism, cardiac output, and effective pulmonary gas exchange [33]. It is linear and highly reproducible, with a higher slope indicating improved oxygen exchange (figure 8). It has been well studied in heart failure, where it is highly predictive of mortality when decreased below 1.47 L/min, but also correlates closely with VO_2max [33,34]. OUES is also impaired in pulmonary vascular disease [35,36] and metabolic disorders [37,38].

USING CPET TO DETERMINE THE CAUSE OF DYSPNEA

Approach — Each CPET yields a large amount of data, both individual numerical results and graphic displays. A careful examination of CPET results can generally help determine whether exercise capacity is reduced, and if so, whether exercise intolerance is associated with respiratory limitations or cardiovascular or circulatory limitations (algorithm 1). Additional examination of the data can provide additional clues that may guide further evaluation (table 2 and table 3).

●Assess peak performance – The first step is to determine whether the patient achieved a normal peak oxygen uptake (VO₂; >85 percent predicted) or peak workload (>80 percent predicted). A normal oxygen uptake or workload indicates normal exercise performance per age, which argues against a cardiopulmonary or metabolic cause of exertional dyspnea.

●Assess effort – For patients not achieving normal peak oxygen uptake or workload, we determine whether the test meets criteria for a maximal volitional effort. Maximal effort is indicated by a high respiratory exchange ratio (RER; indicating significant anaerobic metabolism) or a heart rate (HR) elevation to near maximal predicted levels. Accepted values for acceptable effort include an RER >1.16 or HR >90 percent of predicted [18]. Submaximal effort tests need to be evaluated with great caution, as many measurements are not interpretable in this setting.

●Assess ventilatory response to exercise – For patients who put forth a maximal effort but do not achieve normal peak oxygen uptake, the next step is to assess whether there is a normal ventilatory response to exercise. We first examine the breathing reserve. A low breathing reserve ([maximum voluntary ventilation (MVV) – maximum ventilation at peak exercise (V_Emax)]/MVV <30 percent) suggests that exercise may be limited by ventilation, which is abnormal. This limitation can result from poor respiratory mechanics (low forced expiratory volume in one second [FEV₁]), poor gas exchange (ie, physiologic dead space ventilation), or abnormal breathing patterns (rapid shallow breathing). Other abnormal ventilatory responses to exercise can include a respiratory rate >55, a blunted increase in tidal volume of less than two times baseline, a V_T/inspiratory capacity (IC) ratio ≥85 percent, an increase in end-expiratory lung volume (EELV) of ≥0.25 L, or evidence of postexercise bronchoconstriction. (See 'Ventilatory causes of dyspnea' below.)

●Assess cardiovascular response to exercise – After examining the ventilatory response, we investigate the cardiovascular response to exercise. This includes examination of blood pressure and HR responses, as well as any electrocardiogram (ECG) evidence of myocardial ischemia. Chronotropic insufficiency (failure to reach maximal predicted HR) is a common cause of reduced exertional cardiac output impaired exercise performance that can frequently be treated with downtitration of beta-blockers or calcium channel blockers. A reduced peak HR paired with a normal O₂ pulse suggests isolated chronotropic insufficiency. Impaired augmentation of stroke volume (SV) from heart failure or cardiac ischemia leads to a low anaerobic threshold (AT) (occurrence at <40 percent predicted VO₂) and a low O₂ pulse with an early plateau (figure 6). Oscillation in the V_E versus time curve (figure 9) is suspicious for poor forward flow and excessive circulation time. When more severe, heart failure can also lead to impaired gas exchange (V_E/VCO₂, P_ETCO₂, and O₂ uptake efficiency slope [OEUS]) due to poor perfusion of lung and tissues. (See 'Cardiovascular causes of dyspnea' below.)

●Examine gas exchange – For patients with either abnormal ventilatory responses or cardiovascular responses to exercise, an examination of gas exchange may be helpful in suggesting disease etiology or determining severity. For patients with pulmonary disorders, worsening gas exchange parameters such as exertional hypoxemia, increasing V_E/VCO₂ slope, lower P_ETCO₂ at rest or AT, or lower OUES suggest worsened V/Q mismatch and/or concomitant pulmonary vascular disease. For patients with findings suspicious for cardiovascular disease, these gas-exchange abnormalities correlate with more severe disease, pulmonary hypertension, and mortality. For patients without evidence of cardiac or pulmonary disease, a hyperventilatory breathing pattern can yield these results. (See 'Abnormal gas exchange associated with dyspnea' below.)

●Consider peripheral causes of dyspnea – Anemia, mitochondrial myopathies, dysautonomia, and deconditioning are additional causes of dyspnea due to poor blood oxygen content or extraction (figure 2) that typically mimic an abnormal cardiovascular response. Deconditioning, the most common, usually demonstrates a mildly reduced AT but relatively preserved O₂ pulse without significant gas exchange abnormalities (normal or near-normal V_E/VCO₂ slope, P_ETCO₂). Other conditions have typical disease patterns, but they may require invasive testing for confirmation. (See 'Peripheral causes of dyspnea' below.)

Ventilatory causes of dyspnea — A reduced breathing reserve ([MVV - V_EMAX]/MVV) <30 percent indicates a potential ventilatory limitation to exercise. Other abnormalities suggesting an abnormal ventilatory response include: a respiratory rate >55, a blunted increase in V_T of less than two times baseline, a V_T/IC ratio ≥85 percent, an increase in EELV of ≥0.25 L, or evidence of postexercise bronchoconstriction. Certain patterns of CPET results are associated with the major categories of lung disease (table 2):

●Obstructive airways disease – Asthma and chronic obstructive pulmonary disease (COPD) are the most common pulmonary causes of dyspnea [39]. The primary physiologic abnormality in both diseases is airflow limitation, but air trapping and hyperinflation are generally the consequences of airflow limitation that mediate dyspnea [40-42]. The importance of hyperinflation as a cause of dyspnea is supported by the finding that dyspnea is more strongly associated with IC (an indirect measure of hyperinflation) than with FEV₁ [43].

Among patients with obstructive airways disease, the following pattern is noted on CPET. First, V_T, respiratory rate, and V_E increase initially, as expected during exercise, but as the work load increases, dynamic hyperinflation limits the degree to which V_T may rise (figure 4) [44]. This results in a compensatory rise in respiratory rate, resulting in a rapid, relatively shallow breathing pattern. Since shallower breathing involves a disproportionate rise in V_D, the pattern is energy inefficient, and V_E must increase more than expected, reflected in an elevated V_E versus VCO₂ slope, nadir of the V_E/VCO₂ versus time curve, or fall in P_ETCO₂ [18], similar to the changes seen in patients with cardiovascular or pulmonary vascular disease (figure 7).

In addition, dynamic hyperinflation prevents V_T recruitment that is normally accomplished by exhaling closer to residual volume (RV), thus preventing the expected decrease in EELV (figure 4 and figure 5) [25,44]. Instead, V_T is recruited during inspiration toward total lung capacity (TLC), and as soon as an absolute threshold of inspiratory reserve volume (IRV; reflected by V_T/IC ~0.75) is reached, dyspnea becomes severe (figure 10) [40,45,46].

Arterial blood gas (ABG) analysis may reveal an increased arterial partial pressure of carbon dioxide (P_aCO₂), reflecting a failure to clear carbon dioxide as expected at peak exercise.

Emphysema may also lead to gas exchange abnormalities during exercise due to increasing blood flow to poorly ventilated lung (ie, worsened ventilation-perfusion matching). This can be detected based on worsening oxygenation, excessive widening of the A-a O₂ difference (>35 mmHg), or a failure to decrease dead space (as interpolated by noninvasive V_D/V_T) during exercise [18].

In asthma or COPD with variable obstruction, airflow limitation may worsen with exertional work, further limiting expiration, worsening hyperinflation, and resulting in excessive dyspnea. Exercise-induced bronchoconstriction can be diagnosed with a decline in postexercise FEV₁ of greater than 10 to 15 percent. (See "Bronchoprovocation testing", section on 'Exercise challenge'.)

●Restrictive disease – In restrictive disease (eg, chest wall abnormalities, interstitial lung disease [ILD], neuromuscular disease, obesity, pleural disease), similar breathing patterns are seen, but for different physiologic reasons [39,47]. In this case, the failure to augment V_T is due to lung volume restriction. The result is the adoption of rapid, shallow breathing, with the consequences of inefficient ventilation (including elevated V_E versus VCO₂ slope, or nadir of the V_E/VCO₂ versus time curve) described above. In general, patients with restrictive physiology will not demonstrate an increase in EELV and have a higher respiratory rate than that seen in obstructive disease.

If restriction is due, in part, to obesity, the exercise flow-volume loop may start off close to RV and may quickly encroach on the maximal flow-volume envelope, resulting in airflow limitation, further contributing to dyspnea. Other factors are also associated with dyspnea in obesity, including increased work of breathing, concomitant deconditioning, and psychologic factors (eg, perception of breathlessness, anxiety) [48].

In restrictive disease, there may also be additional derangements in gas exchange due to either recruitment of poorly ventilated units, shunting through diseased units, or concomitant pulmonary vascular disease, resulting in O₂ desaturation and/or excessive ventilation due to relatively high physiologic dead space [18,47].

●Dysfunctional breathing – Abnormal patterns of breathing, or so-called dysfunctional breathing due to fear, anxiety, or learned behaviors, may cause dyspnea in the absence of underlying pulmonary disease [49]. The most common pattern is rapid, shallow breathing, which leads to relatively high dead space fraction (V_D/V_T) and the need for increased V_E, thus causing ventilatory limitation. This pattern manifests with a low resting P_ETCO₂, an elevated V_E-VCO₂ slope and V_E/VCO₂ nadir in the absence of oxygen desaturation, reduced O₂ pulse, or abnormal spirometry or diffusing capacity.

Other patients may simply hyperventilate during exercise with development of a respiratory alkalosis and similar findings (except for P_ETCO₂, which will decrease only with exertion). This breathing pattern may be seen in association with anxiety or fear of exertion.

Another abnormal breathing pattern, sometimes seen in athletes, is persistent breathing at abnormally high functional residual capacity (FRC), which will mimic restrictive physiology but with an elevation in EELV and without abnormalities in oxygenation, spirometry, or diffusing capacity.

●Upper airway obstruction – Other pulmonary causes of dyspnea that may be discovered during CPET include exercise-induced laryngeal obstruction (EILO; also called paradoxical vocal fold motion) [50] and tracheobronchomalacia [51], both of which may manifest as abnormalities in the exercise flow-volume loops obtained during the test [25]. However, the sensitivity of exercise flow-volume loops is not optimal, and other diagnostic methods are generally needed. Descriptions of continuous laryngoscopy during exercise testing in the evaluation of EILO, and the use of bronchoscopy and dynamic CT to diagnose tracheomalacia are provided separately. (See "Exercise-induced laryngeal obstruction", section on 'Continuous laryngoscopy during exercise' and "Tracheomalacia in adults: Clinical features and diagnostic evaluation", section on 'Diagnostic criteria'.)

Cardiovascular causes of dyspnea — Cardiac output is the primary limiting factor that determines maximal exercise capacity under normal circumstances [52]; this is usually signaled by reaching the maximal predicted HR. Cardiac limitation is therefore either physiologic or pathologic, depending on the overall response to exercise. Reduced exertional cardiac output can arise from inability to achieve maximum HR (chronotropic incompetence) or impaired augmentation of SV.

Cardiac disease is a common cause of dyspnea and typically relates to ischemic heart disease, cardiomyopathy, or valvular heart disease, among others. In each case, cardiac performance is compromised, such that cardiac output becomes limited.

The CPET parameters used to characterize cardiac disease include HR, blood pressure, ECG, oxygen pulse (VO₂/HR), AT, and changes in V_E/VCO₂ and V_E versus time (table 3) [10]. (See 'Cardiovascular measurements' above.)

●Chronotropic incompetence – HR should rise linearly with VO₂, reaching a maximal value as estimated by 220 – age, or 90 percent of this value. If a patient has a submaximal exercise capacity in the setting of a submaximal HR but an otherwise maximal effort (RER >1.16), then cardiac response is the likely limiting factor. Submaximal HR with normal O₂ pulse suggests isolated chronotropic incompetence. Occasionally, there are other limiting factors (eg, reduced ventilatory reserve or hypoxemia) that prevent achievement of maximal HR. Impaired delivery of blood to the tissues will lead to a lower AT. At the end of exercise, the normal autonomic response is a rapid decrease in heart rate. Slower heart rate recovery (HRR), such as a fall in heart rate of less than 12 beats per minute (bpm) in the first minute after peak exercise, has been associated with increased risk of cardiovascular mortality [53].

●Impaired stroke volume augmentation – Many cardiac diseases result in impaired SV augmentation with peak exercise. Cardiac ischemia leads to reduced SV due to increased left ventricular stiffness and reduction in inotropy during exercise. Impaired contraction or relaxation of cardiac myocytes in systolic or diastolic heart failure, respectively, likewise reduce SV. Regurgitant or stenotic valves impede forward flow of blood during exertion.

O₂ pulse is a physiologic surrogate for SV because VO₂/HR is equal to SV∙(C_aO₂ - C_vO₂), based on Fick’s equation (figure 2). When there is a problem with SV, cardiac output is maintained by increasing HR. This yields an accelerated HR response at high workloads and a corresponding early plateau (below 80 percent predicted) in the maximal O₂ pulse (figure 6). The O₂ pulse may be limited due to dysfunction of the left ventricle (eg, HFpEF) or the right ventricle (eg, pulmonary arterial hypertension) during exercise. As with chronotropic incompetence, impaired delivery of blood to the tissues leads to an earlier AT.

Decreased cardiac output negatively impacts the efficiency of CO₂ clearance and oxygen uptake. Abnormal distribution of blood flow in cardiac failure results in relatively low perfusion in areas of the lung that are well ventilated, yielding less efficient clearing of CO₂ (measured by increasing physiologic dead space, lower P_ETCO₂, and elevated V_E-VCO₂ slope and V_E/VCO₂ nadir) (figure 7). Similarly, there may be a decrease in the OUES (<1.47), indicating less efficient O₂ uptake with any degree of ventilation (figure 8) [33,34,54-56].

Oscillation in the V_E versus time curve, known as exercise oscillatory ventilation (EOV), may be more specific than other measurements for intrinsic cardiac disease. This abnormality is thought to be related to excessive circulation time with resultant delayed feedback from arterial CO₂ chemoreceptors on ventilation rate (figure 9) [10,57]. (See "Exercise capacity and VO2 in heart failure".)

●Cardiac ischemia – In addition to impaired SV augmentation, ischemic changes, usually related to ST or T waves, may be seen on exercise ECG in patients who experience exertional dyspnea due to cardiac ischemia [58]. Many other ECG abnormalities may also be detected. For example, frequent premature ventricular contractions during exercise recovery have been associated with increased mortality, even in those without a known diagnosis of cardiac disease [59].

●Peripheral vascular disease – Peripheral vascular disease can also lead to exercise intolerance. Here, inability of the peripheral vessels to relax and dilate during exercise leads to inadequate O₂ delivery to the tissues, early onset of AT, and exertional dyspnea. The main clue to this during CPET is the excessive rise of diastolic blood pressure (DBP; >20 mmHg above baseline) [18].

Abnormal gas exchange associated with dyspnea — For either patients with ventilatory or cardiovascular limitations to exercise, we examine gas exchange parameters. Disproportionately poor gas exchange compared with other factors indicates impaired V/Q matching (eg. parenchymal destruction, intrapulmonary shunting), pulmonary vascular disease, or both. (See 'Gas exchange measurements' above.)

For patients with pulmonary vascular disease (eg, pulmonary hypertension, pulmonary veno-occlusive disease, thromboembolic disease) or other diseases causing abnormal ventilation-perfusion matching (eg, emphysema, ILD), the most easily identifiable sign of gas exchange abnormalities is the development of hypoxemia or desaturation with exercise (table 3) [60]. Additional signs of poor gas exchange are suggested by signs of ventilatory inefficiency. These include reduced P_ETCO₂ during exertion or an increase in V_E versus VCO₂ slope or elevated nadir of the V_E/VCO₂ curve (figure 7) [61]. The OUES has also been shown to be reduced in patients with pulmonary arterial hypertension and may predict a poor outcome (figure 8) [35].

Peripheral causes of dyspnea — Although most occult causes of unexplained dyspnea are cardiac or pulmonary in nature, a significant minority may be due to impaired oxygen storage or extraction in the tissues. These disorders tend to mimic cardiovascular exercise limitation and may require invasive testing for a definitive diagnosis (table 3).

●Anemia – Anemia results in reduced carrying capacity for oxygen, which impairs delivery to the tissues. Anemia leads to a reduced peak VO₂, an early AT, and low O₂ pulse without ventilatory limitation or chronotropic incompetence. Poor O₂ delivery may reduce the OUES, but other gas exchange factors are less likely to be affected. This mimics impaired SV augmentation of the left heart, but without impaired blood delivery to the lungs or pulmonary venous congestion. A complete blood count should always be performed within a month of CPET evaluation to make sure that anemia is not misinterpreted as cardiovascular dysfunction. Iron deficiency also has deleterious consequences on skeletal muscle leading to exercise intolerance beyond the effects of anemia, and intravenous repletion of iron has demonstrated large impacts on exercise capacity in patients in patients with heart failure [62].

●Deconditioning – Most sedentary patients will demonstrate a normal VO_2MAX at a decreased workload. However, lack of regular exercise may also lead to inefficient peripheral delivery and extraction of oxygen during exertion, termed deconditioning. Deconditioning is a common cause of dyspnea on exertion [4,14,15] that is reversible with a regular exercise regimen. Patients with deconditioning typically demonstrate mildly reduced peak VO₂, reduced AT, low O₂ pulse, and low workload with normal ventilatory and chronotropic responses as well as low-normal or normal gas exchange parameters. It is difficult to distinguish these abnormalities from mild cardiovascular limitations; the relatively normal V_E-VCO₂ slope and V_E/VCO₂ nadir may offer the best discriminative ability [63]. Exercise echocardiography directly assessing SV augmentation may be helpful if additional certainty is required.

Patients with specific underlying causes of dyspnea may also become deconditioned because of lack of exercise due to the dyspnea on exertion they experience. This is very common among patients with chronic heart or lung disease, which can lead to a mixed CPET result where both deconditioning and underlying disease are contributing to exercise intolerance.

Deconditioned patients frequently receive CPET testing as a work-up for dyspnea with exercise following a long period of sedentarism. In these cases, CPET can be reassuring because it can rule out major abnormalities and thereby give the patient and the provider some confidence in the continuing pursuit of an exercise program to overcome the deconditioned state.

●Metabolic or mitochondrial myopathies – Like deconditioning, metabolic and mitochondrial myopathies impair peripheral extraction of oxygen; unlike deconditioning, these abnormalities are not easily reversible and may be severe. Typical abnormalities include impaired VO_2MAX with an early AT, low O₂ pulse, and a decrease in the OUES (decreased slope of VO₂ versus logV_E) (figure 8), similar to what is seen in patients with cardiac limitation. O₂ pulse may be somewhat less likely to plateau in the setting of mitochondrial disease than in cardiac dysfunction [27]. More definitive distinguishing signs include verification of normal cardiac output, direct testing of C_aO₂-C_vO₂ difference, and serial lactate measurements, all of which require more invasive or specialized exercise testing. Molecular genetic studies or a muscle biopsy may be indicated [64].

●Dyspnea associated with long COVID-19 – One of the most prominent and troublesome symptoms associated with long COVID-19 is dyspnea on exertion. Several studies using CPET have now been performed trying to elucidate the mechanism involved. For most long COVID-19 patients, it appears that peripheral factors, including abnormalities of autonomic regulation, vascular endothelium, and muscle or mitochondrial function, rather than residual cardiopulmonary disease, play a causal role [65,66].

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: Dyspnea" and "Society guideline links: Pulmonary function testing" and "Society guideline links: Stress testing and cardiopulmonary exercise testing".)

SUMMARY

●Indication for CPET in unexplained dyspnea – The CPET is a valuable tool in determining the cause of dyspnea on exertion that is not otherwise apparent from routine evaluation, including history, physical examination, basic metabolic and hematologic analysis, pulmonary function testing, electrocardiogram (ECG), and heart or lung imaging. (See 'Overview of cardiopulmonary exercise testing (CPET)' above.)

●Principles of CPET interpretation – Understanding normal physiologic responses to exercise, and thereby being able to recognize abnormal responses, helps to identify the factors limiting exercise (table 2 and table 3). This information may pinpoint the etiology of dyspnea, or at least help distinguish the relative contributions of heart, lung, and metabolism to exertional intolerance. (See 'Overview of cardiopulmonary exercise testing (CPET)' above.)

●CPET features of diseases with abnormal ventilation – CPET features that indicate exercise limitation due to abnormal ventilation include the following (see 'Ventilatory causes of dyspnea' above):

•Low peak oxygen uptake (VO₂; <85 percent predicted), despite good effort

•Elevated minute ventilation (V_E) resulting in reduced breathing reserve

•Inability to fully recruit tidal volume (V_T; less than twice the baseline)

•Rapid, shallow breathing or other abnormal breathing pattern

•Elevated V_E versus carbon dioxide output (VCO₂) slope (>30 to 32) or nadir of V_E/VCO₂ curve (>32 to 34) (figure 7)

In those with obstructive disease, there may be additional features:

•Dynamic hyperinflation with low inspiratory capacity (figure 4)

•Abnormal flow-volume loops during exercise

•Decreased forced expiratory volume in one second (FEV₁) after exercise (exercise-induced bronchoconstriction)

●CPET features of cardiac disease – Cardiac causes of dyspnea can be difficult to differentiate from the normal cardiac limitation that determines maximal exercise capacity. The key is determining what other factors may be limiting exercise at the time that cardiac limitation occurs. CPET features seen in cardiac disease include (see 'Cardiovascular causes of dyspnea' above):

•Low peak VO₂ (<85 percent predicted), despite good effort

•Low anaerobic threshold (AT; <40 percent predicted maximal VO₂)

•Reduced maximal heart rate (HR) in chronotropic incompetence; otherwise, accelerated HR at high work loads

•Low O₂ pulse and early plateau (<80 percent predicted)

•Elevated V_E versus VCO₂ slope, or nadir of V_E/VCO₂ curve (figure 7)

•Decreased oxygen uptake efficiency slope (OUES; <1.47) (figure 8)

Additional features are more common in particular settings:

•ECG changes, in cardiac ischemia or arrhythmia

•Excessive blood pressure (BP) response, especially diastolic BP (>20 mmHg above baseline), in peripheral vascular disease

•Exercise oscillatory ventilation (EOV), in severe heart failure (figure 9)

●CPET features of pulmonary vascular disease – Pulmonary vascular disease (eg, pulmonary hypertension, chronic thromboembolic disease) results in prominent changes in gas exchange accompanied by impaired augmentation of cardiac output, resulting in a pattern that can sometimes be distinguished from other forms of cardiovascular or ventilatory limitation. CPET features of pulmonary vascular disease include the following (see 'Abnormal gas exchange associated with dyspnea' above):

•Low peak VO₂ (<85 percent predicted), despite good effort

•Drop in arterial partial pressure of oxygen (P_aO₂) or pulse oxygen (SpO₂) with exercise

•Widening of alveolar-arterial (A-a) difference >35

•Low AT (<40 percent predicted maximal VO₂)

•Low O₂ pulse and early plateau

•Failure of ratio of dead space (V_D) to V_T to decrease

•Elevated V_E versus VCO₂ slope, or nadir of V_E/VCO₂ curve (figure 7)

•Decreased resting P_ETCO₂ and P_ETCO₂ at AT

•Decreased OUES (figure 8)

●CPET features of diseases causing peripheral limitation to exercise – Although most occult causes of unexplained dyspnea are cardiac or pulmonary in nature, a significant minority may be due to impaired oxygen storage or extraction in the tissues. These disorders tend to mimic cardiovascular exercise limitation and may require invasive testing for a definitive diagnosis. CPET features of metabolic causes of dyspnea (eg, reduced oxygen carrying capacity due to anemia, early switch to anaerobic metabolism due to deconditioning, myophosphorylase deficiency, and mitochondrial myopathies) include (see 'Peripheral causes of dyspnea' above):

•Low peak VO₂ (<85 percent predicted), despite good effort

•Low AT (<40 percent predicted maximal VO₂)

•Low O₂ pulse

•Decreased OUES (figure 8)

•Otherwise relatively normal cardiovascular, ventilatory, and gas exchange responses