Relative Impairments in Hemodynamic Exercise Reserve Parameters in Heart Failure With Preserved Ejection Fraction

Abstract

OBJECTIVES

The aim of this study was to compare the relative impairment in different exercise hemodynamic reserve parameters in patients with heart failure with preserved ejection fraction (HFpEF) and control patients using a study-level meta-analysis.

BACKGROUND

A cardinal manifestation of chronic HFpEF is severely decreased exercise capacity. Developing effective therapies for exercise intolerance in HFpEF requires optimal understanding of the factors underlying exercise intolerance.

METHODS

Data were included from 17 unique cohorts that measured peak oxygen uptake and hemodynamic or echocardiographic parameters during exercise in patients with HFpEF and control subjects in this meta-analysis. Standardized mean differences (SMDs) in the exercise reserve (exercise – resting) measures of hemodynamic or echocardiographic parameters between the HFpEF and control groups were pooled in a random-effects meta-analysis.

RESULTS

The pooled analysis included 910 patients with HFpEF and 476 control subjects. In pooled analysis, patients with HFpEF had significantly lower peak oxygen uptake (SMD: −2.13; 95% confidence interval [CI]: −2.68 to −1.57). Among hemodynamic exercise reserve parameters, the largest impairment was observed in chronotropic response reserve (change in heart rate from rest to peak exercise; SMD: −1.87; 95% CI: −2.44 to −1.29), followed by exaggerated increase in pulmonary capillary wedge pressure with exercise (SMD: 1.78; 95% CI: 1.46 to 2.09). Significant abnormalities were also observed in the arteriovenous oxygen difference reserve and stroke volume reserve between the HFpEF and control groups.

CONCLUSIONS

The most consistent and severe hemodynamic reserve abnormalities observed in patients with HFpEF were impairment in chronotropic reserve and exaggerated increase in pulmonary capillary wedge pressure with exercise. These may be important targets for therapeutic strategies to improve exercise tolerance in patients with HFpEF.

Heart failure with preserved ejection fraction (HFpEF) accounts for nearly 50% of all heart failure in the community, and its prevalence, relative to heart failure with reduced ejection fraction (EF), continues to increase (1,2). Exercise intolerance, measured objectively as reduced peak oxygen uptake (VO₂), is the primary manifestation of HFpEF and is associated with poor quality of life and a greater risk for adverse clinical outcomes (3,4). Several pharmacological agents have failed to improve exercise capacity and quality of life in patients with HFpEF (5–7). This highlights the need to better understand the complex pathophysiology that underlies exercise intolerance in these patients. In accordance with the Fick principle (VO₂ = stroke volume [SV] × heart rate [HR] × arteriovenous [AV] oxygen difference), reduced peak VO₂ may be the result of central and/or peripheral factors that result in reduced O₂ delivery or use by the exercising muscles. Multiple studies have identified distinct hemodynamic impairments in HFpEF with exercise (8–27). Some studies have attributed decreased peak VO₂ to central hemodynamic impairments with reduced cardiac output and exaggerated left ventricular (LV) filling pressures (8,12,14,17,19,26,27). Others have identified abnormalities in both cardiac output and peripheral mechanisms with skeletal muscle microvascular or mitochondrial dysfunction that result in decreased peripheral AV O₂ difference during maximal exercise in HFpEF (9,11,13). Accordingly, we analyzed pooled study-level data from such studies to assess the relative contribution of different physiological factors to reduced exercise capacity among HFpEF and control group participants that could be potential targets for future therapeutic interventions.

METHODS

SEARCH STRATEGY.

We performed a search of published research for this study in accordance with the Meta-Analysis of Observational Studies in Epidemiology protocol (28). We performed a search of the MEDLINE database for English-language studies published between January 1, 1990, and August 30, 2017, that compared resting and exercise cardiovascular hemodynamic parameters in patients with versus those without HFpEF (see the Online Appendix for details of the search strategy). Additional manual searches were performed through the references cited in the original reports and relevant review papers.

STUDY SELECTION.

All reports were screened by title and abstracts by 2 independent reviewers (B.P. and A.P.) for potential inclusion. We included studies of adult participants (>18 years of age) that presented the following information: 1) well-defined HFpEF and control groups; 2) well-described method of diagnosing clinical HFpEF; 3) EF in the HFpEF group ≥50%; 4) a detailed method of evaluating peak exercise performance; and 5) available data on hemodynamic parameters and/or LV dimensions assessed noninvasively or invasively at rest and exercise. Studies without control groups with exercise hemodynamic or LV volume data were excluded. All discrepancies regarding study inclusion were adjudicated by the senior author (J.D.B.). If 2 or more studies included overlapping patient samples, we included data from the study with the larger sample size and/or with more detailed hemodynamic or echocardiographic assessment data. Furthermore, for overlapping study populations, hemodynamic data from smaller cohorts were included in the analysis if they were not available in the larger cohort study. Study quality was assessed using the Newcastle-Ottawa quality assessment tool, which summarizes 8 aspects of each study, with a highest score of 9 (29).

DATA EXTRACTION AND RISK FOR BIAS ASSESSMENT.

Data collection was performed independently by 2 authors (A.P. and B.P.) using a standardized data collection form. The following data were collected from each study: investigator, year of publication, geographic location, mean age, proportion of women, inclusion and exclusion criteria for HFpEF and control groups, baseline clinical characteristics, echocardiographic characteristics, anthropometric data (body mass index [BMI] and body surface area [BSA]), stress test protocol used, and hemodynamic parameters and/or LV volumes at rest and exercise. Any discrepancies between reviewers were resolved through discussion and final consensus. The main hemodynamic and Fick parameters at rest and peak exercise of interest for our study included pulmonary artery pressure (PAP), pulmonary capillary wedge pressure (PCWP), HR, indexed LV end-diastolic volume (LVEDV), indexed SV, AV O₂ difference, LV EF, and indexed systemic vascular resistance (SVR). In studies with only graphical reporting of outcomes, the quantitative values were estimated from the graphs. BSA-indexed outcomes were included in the analysis. In studies without indexed volumes but with BSA reported, average volumes were divided by BSA to obtain average index volumes.

STATISTICAL ANALYSIS.

We performed a pairwise meta-analysis of all outcomes using the DerSimonian and Laird random-effects approach, incorporating both within- and between-study heterogeneity (30). In our analysis, we examined for differences in changes in the hemodynamic parameters listed earlier from rest to peak exercise between the 2 study groups. Standard deviation for change from rest to exercise for study outcomes was determined on the basis of the previously reported methods using correlation coefficients (24). Pooled estimates for each outcome are reported as standardized mean difference (SMD) between the HFpEF and the control groups, along with their respective 95% confidence intervals (CIs). SMD was reported to account for differences in protocols used for hemodynamic assessment across studies and to allow direct comparisons of the magnitude of difference between the HFpEF and control groups for different hemodynamic parameters. We also assessed for the relative effect sizes using previously suggested thresholds: small, medium, and large effect sizes defined by SMDs of 0.2 to 0.5, 0.5 to 0.8, and ≥0.8, respectively (31). Weighted mean differences were also reported where necessary for clinical relevance. A pooled analysis was performed with included studies stratified into a priori–defined subgroups according to the stress test type (supine vs. upright) to account for impact of posture on these hemodynamic parameters. Studies with semirecumbent exercise testing were included in the upright group for the stratified analysis. We assessed for statistical heterogeneity by examining the direction of the association for each study relative to the pooled effect and by computing the I² statistic. Values of I² >50% indicate heterogeneity (32). To assess if the differences observed between the HFpEF and control groups were confounded by differences in baseline patient age and BMI in included studies, we performed a random-effects meta-regression of aggregate data incorporating the between group differences in mean age and BMI of study participants as independent variables in the model and examined for their association with the study outcomes. As recommended, the meta-regression analyses were limited to study outcomes reported by close to 10 or more studies (33). Small study effects (or publication bias) were evaluated by examining for funnel-plot asymmetry and Egger’s regression test (34).

All analyses were performed using Stata version 14.1 (StataCorp, College Station, Texas). The level of significance was set at a p value of 0.05.

RESULTS

CHARACTERISTICS OF INCLUDED STUDIES.

We included 910 participants with clinical HFpEF and 476 control group participants from 17 unique cohorts (8–24) across 20 studies in our meta-analysis (8–27) (Online Figure 1). Five studies were done in Europe, 3 in Australia or Asia, and 9 in the United States. The diagnoses of HFpEF were based on objective clinical signs and symptoms, with an EF cutoff of ≥50% in all included studies (Online Table 1). One-half of the included cohorts used objective evidence of diastolic dysfunction at rest or exercise to define HFpEF (n = 8). Among the control participants, 5 studies included healthy participants without traditional risk factors such as hypertension and diabetes. Only 3 studies reported including patients with atrial fibrillation, who represented a minority of the included participants in these studies (4% of control subjects and 14% of patients with HFpEF). Nine studies included control group participants who underwent clinically indicated invasive cardiopulmonary evaluation for dyspnea or exercise intolerance (Online Table 1). Supine exercise testing was performed in 5 cohorts, 11 studies performed upright or semiupright exercise testing and hemodynamic assessment, and 1 study included both supine and upright assessment. A symptom-limited exercise testing protocol was used in all included studies. Only invasive hemodynamic assessment was performed in 3 cohorts, and only noninvasive hemodynamic assessment was performed in 8 cohorts; 6 used both invasive and noninvasive methods to measure the hemodynamic parameters (Online Table 1). Six studies reported withholding all cardiac medications or at least beta-blockers prior to hemodynamic assessments. All included studies performed peak VO₂ and exercise hemodynamic assessments at the same time.

Study-level and pooled baseline characteristics of HFpEF and control participants across included studies are presented in Table 1 and Online Table 2, respectively. Patients with HFpEF were older, were more commonly women, and had a greater burden of cardiovascular risk factors such as hypertension, diabetes, and obesity. Resting hemodynamic parameters in the HFpEF and control groups of the included studies are shown in Online Table 3.

TABLE 1.

Basic Clinical Characteristics of Control and Heart Failure With Preserved Ejection Fraction Groups Across Included Studies

				Age (yrs)	Female (%)	BMI (kg/m2)	HTN (%)	DM (%)
First Author (Year) (Ref. #)	Position	Exercise	Control/HF	Control/HF	Control/HF	Control/HF	Control/HF	Control/HF
Abudiab et al. (2013) (8)	Supine/upright	Cycle	73/109	59/67	75/72	29.1/33.2	64/82	16/33
Bhella et al. (2011) (9)	Upright	Treadmill	13/11	70.2/73.0	46.2/63.6	25.7/33.6	0/100	0/55
Borlaug et al. (2010) (10)	Supine	Cycle	23/32	47/65	65/72	27.3/32.0	57/72	22/16
Borlaug et al. (2016) (20)	Supine	Cycle	24/50	61/70	46/54	27.2/34.4	63/94	21/36
Dhakal et al. (2015) (11)	Upright	Cycle	24/48	55/63	38/60	27.6/33.7	37/60	0/25
Ennezat et al. (2008) (12)	Semirecumbent	Cycle	25/25	72/74	76/76	28/28	91/84	27/16
Haykowsky et al. (2011) (13)	Upright	Cycle	25/48	68/69	52/85	25.0/30.6	0/81	0/17
Henein et al. (2013) (21)	Semirecumbent	Cycle	14/17	65/68	50/64.7	25.3/28.3	NA	0/5.9
Kasner et al. (2015) (22)	Semirecumbent	Cycle	26/52	48/55	50/48	24.6/27.3	0/54	0/19
Kitzman et al. (1991) (14)	Upright	Cycle	10/7	61/65	40.0/57.1	NA	0/57.1	NA
Kosmala et al. (2016) (15)	Upright	Treadmill	60/207	62.8/63.7	67/73	27.6/29.6	90/89	26.7/33.3
Maeder et al. (2012) (19)	Supine	Cycle	8/10	61/69	37.5/10.0	25.0/31.3	NA	NA
Obokata et al. (2017) (23)*	Supine	Cycle	71/96/99	47/47/49	58/64/64	25.4/26.0/40.8	72/80/82	13/15/33
Phan et al. (2009) (16)	Semirecumbent	Cycle	20/37	63/67	50/76	26/30	0/73	0/11
Santos et al. (2015) (17)	Upright	Cycle	31/31	65/65	26/26	28.8/32.5	56/77	13/29
Shimiaie et al. (2015) (18)	Semirecumbent	Cycle	14/16	50.9/57.2	35.7/12.5	27.0/26.6	13/46	13/60
Tschöpe et al. (2005) (24)	Supine	Cycle	15/15	51/55	27/47	26.1/26.3	27/47	0/12

^*Obokata et al. (23) included 3 study groups: 1 control group and 2 HFpEF groups (obese and nonobese).

BMI = body mass index; DM = diabetes mellitus; HF = heart failure; HFpEF = heart failure with preserved ejection fraction; HTN = hypertension; NA = not available.

DIFFERENCES IN EXERCISE HEMODYNAMIC RESERVE BETWEEN PATIENTS WITH HFpEF AND CONTROL SUBJECTS.

In random-effects pooled analysis, stratified by the cardiopulmonary exercise testing position (supine vs. upright vs. both), patients with HFpEF had significantly lower peak VO₂ compared with control group participants (SMD: −2.13; 95% CI: −2.68 to −1.57) (Figure 1). Among measures of exercise hemodynamic reserve, compared with the control group, patients with HFpEF had significantly blunted augmentation in exercise HR, indexed SV, LV EF, and AV O₂ difference from rest to peak exercise (Figures 2 and 3). Furthermore, patients with HFpEF also had significantly greater increases in PCWP and PAP and less decline in indexed SVR with exercise (Figures 2 and 3). There were no significant differences in exercise-associated change in indexed LVEDV between the 2 groups (Figure 3). Significant heterogeneity was observed across the pooled studies for each study outcome. Similar findings were also noted on meta-analysis using weighted mean differences (Online Table 4).

FIGURE 1. Forest Plots Showing the Pooled Standardized Mean Differences Between Patients With Heart Failure With Preserved Ejection Fraction and Control Subjects for Peak Oxygen Uptake.

CI = confidence interval; HFpEF = heart failure with preserved ejection fraction; SMD = standardized mean difference.

FIGURE 2. Forest Plots Showing the Pooled Standardized Mean Differences Between Patients with Heart Failure with Preserved Ejection Fraction and Control Subjects for Changes From Resting to Exercise in Heart Rate, Pulmonary Capillary Wedge Pressure, Indexed Stroke Volume, and Arteriovenous Oxygen Difference.

aVO₂ = arteriovenous oxygen; PCWP = pulmonary capillary wedge pressure; SVI = indexed stroke volume; other abbreviations as in Figure 1.

FIGURE 3.

Forest Plots Showing the Pooled Standardized Mean Difference Between Heart Failure with Preserved Ejection Fraction and Control Group Patients for Changes From Rest to Peak Exercise in Indexed Systemic Vascular Resistance, Mean Pulmonary Arterial Pressure, Indexed Left Ventricular End-Diastolic Volume, and Left Ventricular Ejection Fraction

LVEDVI = indexed left ventricular end-diastolic volume; LVEF = left ventricular ejection fraction; PAP = pulmonary artery pressure; SVRI = indexed systemic vascular resistance; other abbreviations as in Figure 1.

The severity of impairment in different exercise hemodynamic reserve factors in HFpEF was compared using pooled SMD estimates (Figure 4). In pooled analysis of the HFpEF compared with the control group, the largest impairment was observed in chronotropic response reserve at peak exercise (SMD: −1.87; 95% CI: −2.44 to −1.29), followed by the exaggerated increase in PCWP from rest to peak exercise (SMD: 1.78; 95% CI: 1.46 to 2.09). Among other hemodynamic parameters, major abnormalities were also observed in AV O₂ difference reserve (SMD: −1.61; 95% CI: −2.41 to −0.82), indexed SV reserve (SMD: −1.25; 95% CI: −1.81 to −0.70), indexed SVR response from rest to peak exercise (SMD: 1.18; 95% CI: 0.59 to 1.77), EF change with exercise (SMD: −1.27; 95% CI: −1.62 to −0.91), and mean PAP response from rest to peak exercise (SMD: 1.01; 95% CI: 0.40 to 1.63) among patients with HFpEF compared with control subjects. Meta-regression analysis did not show a significant effect of age or BMI on the observed differences in exercise reserve factors between the HFpEF and control groups (Online Table 5).

FIGURE 4.

Comparison of Pooled Standardized Mean Difference Estimates for Changes in Different Hemodynamic Parameters from Rest to Exercise Between Heart Failure with Preserved Ejection Fraction and Control Groups

A-VO₂ diff. = arteriovenous oxygen difference; HR = heart rate; other abbreviations as in Figures 1 to 3.

STUDY QUALITY ASSESSMENT.

Study quality assessment using the Newcastle-Ottawa Scale demonstrated an average score of 8.4 (with a maximum score of 9 representing the highest quality). Fifteen studies scored 8 or better, with 8 studies scoring 9 (Online Table 6). There was evidence of significant small-study effects, suggestive of limited data from small studies with small effect sizes (Online Figure 2).

DISCUSSION

The novel finding of this pooled analysis is that the most severe exercise reserve abnormalities observed in patients with HFpEF, as determined by the pooled SMD, were impairment in chronotropic reserve at peak exercise and exaggerated increase in PCWP with exercise. Other major physiological limitations associated with HFpEF were impairment in AV O₂ difference reserve, indexed SV and EF reserve, and lesser decline in indexed SVR in response to exercise. Furthermore, no significant differences were observed in exercise-related change in indexed LVEDV between HFpEF and control group patients. Taken together, the findings of our study provide valuable insights into the key abnormalities that underlie the pathophysiology of exercise intolerance in HFpEF.

We observed significant impairments in central cardiac reserve with reduced chronotropic reserve response to exercise. It is noteworthy that among cardiac reserve limitations, abnormality in chronotropic response was much more severe and pronounced than SV abnormalities. This could be predicted from the Fick equation, because in healthy persons during exercise, HR increases approximately 2.5-fold, whereas SV increases by only about 30% (35). Thus, SV increases much less with exercise than does HR. In addition, the variability in measurement of SV is much greater than that of HR. From the earliest reports, chronotropic incompetence has been the most consistently reported finding in exercise studies of HFpEF (14,26,36). Indeed, HR response is a dominant contributor to the increase in cardiac output during exercise (35). Chronotropic incompetence has also been uniformly reported as a major contributor to reduced peak VO₂ in patients with heart failure with reduced EF (37). The mechanisms of chronotropic incompetence in HF are not well understood. An early trial using atrial pacemaker devices to correct chronotropic incompetence in patients with HFpEF was terminated because of slow enrollment (38), but another trial is ongoing (RAPID-HF [Rate-Adaptive Atrial Pacing In Diastolic Heart Failure]; NCT02145351). Conversely, agents that suppress HR increase have been reported to result in worsening of peak VO₂ in HFpEF (39).

Exaggerated increase in PCWP with exercise was another key central hemodynamic abnormality noted in our study. This may be related to increased LV stiffness or left atrial dysfunction. Interestingly, we did not observe a significant difference in exercise-related change in LVEDV between HFpEF and control group participants. These findings suggest that the Frank-Starling response is preserved in patients with HFpEF with similar increases in end-diastolic volume in response to increased venous return with exercise compared with healthy control subjects. However, this occurs at significantly higher PCWP and thus was also associated with higher PAP. These may contribute to symptoms of shortness of breath and exercise intolerance. Along these lines, a recent study demonstrated that improvement in peak exercise LV filling pressures mediates increase in peak VO₂ in patients with HFpEF that are treated with spironolactone (40).

Impairment in SV reserve with no differences in LVEDV indicates that there is, on average, a limited ability to reduce LV end-systolic volume with exercise in HFpEF. This could be related to abnormal contractile reserve, high afterload, or both. Along these lines, we observed significantly reduced EF and lesser decline in SVR in response to exercise in patients with HFpEF, indicating that abnormalities in both systolic and vasodilator reserve are also important contributors.

We also observed significant impairment in the ability to increase AV O₂ difference at peak exercise in patients with HFpEF. The mechanism underlying this impairment is not well understood. However, multiple abnormalities that could be responsible for reduced peak AV O₂ difference in HFpEF have been reported, including reduced capillary density, shift in skeletal muscle fiber type, reduced mitochondrial density and function, reduced skeletal aerobic metabolism and phosphate kinetics, and impaired endothelial function (13,41–46). Other factors, such as early exercise termination in patients with HFpEF because of dyspnea or inadequate distribution of cardiac output to skeletal muscle in preference to other organs, may also contribute to the blunted increase in AV O₂ difference with exercise in patients with HFpEF. Importantly, these abnormalities have also been reported in heart failure with reduced EF, and reduced AV O₂ difference has been shown to contribute to lower peak VO₂ in heart failure with reduced EF (47,48).

Our study findings may have important implications for future investigations of exercise intolerance in patients with HFpEF. HFpEF is increasingly seen as a multisystem disorder, with multiple comorbidities, and with significant variability in the underlying pathophysiological abnormalities and associated phenotypes. Notwithstanding this heterogeneity, we observed consistent patterns of abnormalities in patients with HFpEF, including chronotropic incompetence, exaggerated PCWP response to exercise, and impaired AV O₂ difference reserve. Each of these has potentially important implications for the design of future interventions to improve exercise capacity in HFpEF. The therapeutic implications of chronotropic incompetence and past trials that worsened this finding and ongoing trials were discussed earlier (39). Inorganic nitrite has recently been shown to reduce PCWP and PAP at rest and even to greater extent during exercise (49,50). Similarly, interatrial shunt devices (REDUCE LAP-HF [Reduce Elevated Left Atrial Pressure in Patients With Heart Failure] trial) have demonstrated significant improvement in peak VO₂ in patients with HFpEF by reducing exercise PCWP (51,52). Future studies are needed to determine if a tailored approach to HFpEF treatment targeting specific hemodynamic abnormalities observed in a patient would be associated with better long-term clinical outcomes. Future studies are also needed to determine if the key hemodynamic parameters identified in our study can be used to assess treatment efficacy, disease progression, and overall prognosis.

We observed significant heterogeneity in our pooled analysis for most study outcomes. This may be related to several factors. Because of the heterogeneous nature of the disease process, it is likely that the HFpEF and control group patients from individual cohorts had differences in clinical and hemodynamic characteristics. There were also differences in the hemodynamic assessment (invasive vs. noninvasive) and exercise protocols (upright vs. supine and cycle ergometer vs. treadmill) used in the studies. These differences were addressed with analyses stratified by posture, and the use of statistical tools appropriate for pooling data in this setting, namely the use of random-effects models, and SMDs to determine pooled estimates. Furthermore, visual inspection of plots suggests heterogeneity in magnitude of effect size, as opposed to direction, which was consistent across cohorts, for most study outcomes. Overall, the significant differences in hemodynamic parameters between the HFpEF and control groups observed in our pooled analysis despite a high degree of heterogeneity suggests a greater confidence in the results of our study and reflects favorably on the generalizability of the study findings.

STUDY LIMITATIONS.

First, because of the observational nature of included studies, there is a potential for unmeasured or residual confounding in our study results. Second, the study participants in many included studies were not randomly selected, and thus there is a potential for selection bias. Third, there may be significant differences in hemodynamic mechanisms between patients with HFpEF that may underlie the heterogeneity in the study-level pooled analysis. We could not assess concordance or discordance of mechanisms between patients with HFpEF because of the lack of individual-level data across studies. Fourth, we observed evidence for small-study effects, suggesting that negative data from small studies may theoretically not have been published (or studies were not performed). However, in our variance-weighted, random-effects model with robust effect sizes, the lack of data from such studies is not expected to affect the findings of our study. Finally, because we pooled cross-sectional studies, causality in the relationship between hemodynamic impairments and exercise intolerance cannot be established.

CONCLUSIONS

The findings of our study suggest that impairment in chronotropic response reserve and exaggerated increase in PCWP with exercise may be the most consistent and severe physiological abnormalities in HFpEF, with lesser contributions from impairment in AV O₂ difference reserve, SV reserve, contractile reserve, and attenuated decline in SVR with exercise. Further studies are needed to determine if therapies designed to address these hemodynamic impairments may improve long-term outcomes in patients with HFpEF.

PERSPECTIVES.

COMPETENCY IN MEDICAL KNOWLEDGE:

Impairment in chronotropic response to exercise and exaggerated increase in PCWP with exercise are the most severe hemodynamic abnormalities in patients with HFpEF.

TRANSLATIONAL OUTLOOK:

Future studies are needed to determine if targeting these hemodynamic impairments with novel therapeutic strategies may improve exercise capacity and long-term clinical outcomes in patients with HFpEF.

ABBREVIATIONS AND ACRONYMS

AV

arteriovenous

BMI

body mass index

BSA

body surface area

CI

confidence interval

EF

ejection fraction

HFpEF

heart failure with preserved ejection fraction

HR

heart rate

LV

left ventricular

LVEDV

left ventricular end-diastolic volume

PAP

pulmonary arterial pressure

PCWP

pulmonary capillary wedge pressure

SMD

standardized mean difference

SV

stroke volume

SVR

systemic vascular resistance

VO₂

oxygen uptake