Renal Function and Peak Exercise Oxygen Consumption in Chronic Heart Failure With Reduced Left Ventricular Ejection Fraction

Abstract

Background: Chronic kidney disease is associated with sympathetic activation and muscle abnormalities, which may contribute to decreased exercise capacity. We investigated the correlation of renal function with peak exercise oxygen consumption (V̇O₂) in heart failure (HF) patients.

Methods and Results: We recruited 2,938 systolic HF patients who underwent clinical, laboratory, echocardiographic and cardiopulmonary exercise testing. The patients were stratified according to estimated glomerular filtration rate (eGFR). Mean follow-up was 3.7 years. The primary outcome was a composite of cardiovascular death and urgent heart transplantation at 3 years. On multivariable regression, eGFR was predictor of peakV̇O₂ (P<0.0001). Other predictors were age, sex, body mass index, HF etiology, NYHA class, atrial fibrillation, resting heart rate, B-type natriuretic peptide, hemoglobin, and treatment. After adjusting for significant covariates, the hazard ratio for primary outcome associated with peakV̇O₂ <12 ml·kg⁻¹·min⁻¹ was 1.75 (95% confidence interval (CI): 1.06–2.91; P=0.0292) in patients with eGFR ≥60, 1.77 (0.87–3.61; P=0.1141) in those with eGFR of 45–59, and 2.72 (1.01–7.37; P=0.0489) in those with eGFR <45 ml·min⁻¹·1.73 m⁻². The area under the receiver-operating characteristic curve for peakV̇O₂ <12 ml·kg⁻¹·min⁻¹ was 0.63 (95% CI: 0.54–0.71), 0.67 (0.56–0.78), and 0.57 (0.47–0.69), respectively. Testing for interaction was not significant.

Conclusions: Renal dysfunction is correlated with peakV̇O₂. A peakV̇O₂ cutoff of 12 ml·kg^–1·min^–1 offers limited prognostic information in HF patients with more severely impaired renal function. (Circ J 2015; 79: 583–591)

Chronic kidney disease (CKD) is highly prevalent in patients with chronic heart failure (HF), affecting nearly half of them.¹ In addition, a 13% rate of incident renal dysfunction over 6 months has been observed in systolic CHF patients,² even though progression to end-stage renal disease is rare.³ Renal dysfunction strongly affects morbidity and mortality risk, and estimated glomerular filtration rate (eGFR) is a well-known predictor of mortality.⁴ Furthermore, CKD may complicate the management of HF.⁵

Cardiopulmonary exercise testing (CPET) has several clinical applications in HF, including investigation of the pathophysiology of exercise intolerance and symptoms, evaluation of functional capacity and response to therapeutic interventions, prognostic evaluation, and prescription of exercise training. Risk stratification is a major clinical application of CPET. Peak exercise oxygen consumption (V̇O₂) is regarded as an important prognostic variable; although not formally tested, a cutoff value of 12 ml·kg^–1·min^–1 is currently used as a criterion to guide heart transplant listing in ambulatory patients with advanced HF receiving optimized medical treatment including β-blockers.^6,7 Peak V̇O₂ is, however, affected by age, sex, and anthropometric characteristics; thus, equations to adjust peak V̇O₂ for such variables have been developed.⁸ Because CKD is associated with sympathetic activation and skeletal muscle abnormalities,^9–11 which may contribute to reduced exercise capacity in HF patients, defining whether renal dysfunction influences peak V̇O₂ does appear to be a relevant issue.

In the present study, we sought to investigate the correlation of residual renal function with peak V̇O₂ and assess the effect of renal dysfunction on the prognostic value of peak V̇O₂ in HF patients with reduced left ventricular ejection fraction (LVEF).

Methods

The study population consisted of 2,938 ambulatory patients with HF and reduced LVEF included in the MECKI score database.¹² Selection criteria and methods are described in detail elsewhere.^12,13 Briefly, the patients were recruited and prospectively followed up in 14 Italian HF centers. The clinical, laboratory, ECG, echocardiographic, and CPET data were collected at enrollment. Inclusion criteria were: HF symptoms (NYHA functional classes I–III, stage C of ACC/AHA classification) and former documentation of LVEF <0.40, stable clinical condition with unchanged medications for at least 3 months, and no major cardiovascular treatment or intervention scheduled. Only patients who attained a peak exercise respiratory quotient >1.05 were included in the present study. Exclusion criteria were: history of pulmonary embolism, moderate-to-severe aortic and mitral stenoses, pericardial disease, severe obstructive lung disease, exercise-induced angina and significant ECG alterations, or the presence of any clinical comorbidity interfering with exercise performance. The primary outcome was a composite of cardiovascular death and urgent heart transplantation (HT) at 3 years.

Statistical Analysis

Details of data management were previously reported.^12,13 Data are reported here as mean±standard deviation for continuous variables or number (percentage) of patients for categorical variables. ANOVA test was used for comparison between groups and χ² test was used for comparing categorical variables. Skewed data are reported as median and interquartile range (IQR) and compared by the Kruskall-Wallis test. The eGFR was calculated using the 4-variable Modification of Diet in Renal Disease (MDRD) equation.¹⁴ Patients were stratified according to their eGFR into 4 clinically meaningful strata: <30, 30 to <45, 45–59, and ≥60 ml·min⁻¹·1.73 m⁻².¹⁵ Because only 79 patients (2.7%) were in the lowest eGFR stratum, they were grouped into the next stratum of 30 to <45 ml·min⁻¹·1.73 m⁻².

We used multivariable linear regression analysis to assess the association of eGFR with peak V̇O₂ as the dependent variable. The following variables were examined: age, sex, body mass index, etiology of HF, NYHA class, atrial fibrillation, diastolic (DBP) and systolic (SBP) blood pressures at rest, left ventricular end-systolic volume (LVESV), LVEF, resting heart rate, QRS duration, eGFR, B-type natriuretic peptide (BNP) concentration, sodium level, hemoglobin, and treatment with angiotensin-converting enzyme inhibitors (ACEIs), angiotensin type 1 (AT₁) receptor blockers, β-blockers, or cardiac resynchronization therapy (CRT). Because of collinearity between left ventricular end-diastolic volume and LVESV, the former was not included in the analysis. Because BNP values were available only for 1,297 patients, a separate regression analysis was performed. Because of a skewed distribution, SBP, LVESV, QRS duration, and peak V̇O₂ were logarithmically transformed.

Kaplan-Meier survival curves were applied for the categories of eGFR and peak V̇O₂ and compared using log-rank test. The area under the receiver-operating characteristic (ROC) curve (AUC) of peak V̇O₂ <12 ml·kg⁻¹·min⁻¹ for the primary outcome at 3 years was determined for the entire study population and each strata of eGFR. Because it has been suggested that peak V̇O₂ tends to lose its predictive value 3 years post-CPET,¹⁶ we also calculated the AUC of peak V̇O₂ <12 ml·kg⁻¹·min⁻¹ at 2 years for each strata of eGFR. The statistical significance of the interaction between eGFR and peak V̇O₂ in predicting the primary outcome was tested with the likelihood ratio test. The hazard ratio (HR) and 95% confidence interval (CI) for the primary outcome at 3 years associated with peakV̇O₂ <12 ml·kg⁻¹·min⁻¹ in each strata of eGFR was estimated using univariable and multivariable Cox proportional hazard models. Analyses were conducted using SAS v. 9.2 (SAS Institute Inc, Cary, NC, USA).

Results

Baseline characteristics by eGFR are reported in Table 1; 1,992 patients (67.8%) had an eGFR ≥60 ml·min⁻¹·1.73 m⁻², 574 (19.5%) in the range of 45–59 ml·min⁻¹·1.73 m⁻², and 372 (12.7%) <45 ml·min⁻¹·1.73 m⁻². The peak respiratory exchange ratio was comparable across the eGFR strata. Median (IQR) peak V̇O₂ was 15.2 (9.2–24.6) ml·kg⁻¹·min⁻¹ in the highest, 13.2 (8.2–20.4) ml·kg⁻¹·min⁻¹ in the intermediate, and 11.5 (7.5–17.9) ml·kg⁻¹·min⁻¹ in the lowest eGFR strata (P<0.0001). The proportion of patients with peak V̇O₂ <12 ml·kg⁻¹·min⁻¹ was 25.1%, 24.2%, and 50.7%, respectively (P<0.0001). The patients in the lowest eGFR strata were older, had higher NYHA class and BNP concentrations, lower LVEF and hemoglobin and sodium levels, higher minute ventilation/CO₂ production (V̇E/V̇CO₂) slope, and more frequently presented exercise oscillatory ventilation and atrial fibrillation (Table 1). Overall, 94.4% of the patients were on chronic treatment with ACEIs or AT₁-receptor blockers and 85% with β-blockers.

Table 1. Baseline Characteristics of Study Patients by Level of Renal Function

	All (n=2,938)	eGFR ≥60 (n=1,992)	eGFR 45–59 (n=574)	eGFR <45 (n=372)	All	eGFR ≥60 vs. eGFR 45–59	eGFR ≥60 vs. eGFR <45	eGFR 45–59 vs. eGFR <45
					P value	P value	P value	P value
Age (years), mean (SD)	60.6 (12.4)	57.9 (12.4)	65.5 (10.3)	67.2 (10)	<0.0001	<0.0001	<0.0001	0.006
Male sex, n (%)	2,488 (84.7)	1,702 (85.4)	480 (83.6)	306 (82.3)	NS	NS	NS	NS
BMI (kg/m2), mean (SD)	26.6 (4.3)	26.7 (4.3)	26.4 (4.1)	26.4 (4.4)	NS	NS	NS	NS
Etiology
Ischemic, n (%)	1,459 (49.7)	927 (46.5)	308 (53.7)	224 (60.2)	<0.0001	0.0026	<0.0001	0.05
Idiopathic, n (%)	1,230 (41.9)	907 (45.5)	206 (35.8)	117 (31.5)	<0.0001	<0.0001	<0.0001	NS
Valvular, n (%)	82 (2.8)	47 (2.4)	25 (4.4)	10 (2.7)	0.0376	0.0107	NS	NS
Other, n (%)	167 (5.6)	111 (5.9)	35 (6.1)	21 (5.6)	NS	NS	NS	NS
NYHA III class, n (%)	817 (27.8)	472 (23.7)	183 (31.9)	162 (43.6)	<0.0001	<0.0001	<0.0001	0.001
Atrial fibrillation, n (%)	446 (15.2)	254 (12.7)	117 (20.4)	75 (20.2)	<0.0001	<0.0001	0.0002	NS
Implanted ICD, n (%)	841 (29.1)	552 (28.3)	162 (28.4)	127 (34.6)	0.0461	NS	0.0145	NS
Implanted CRT, n (%)	329 (12.2)	193 (10.7)	72 (13.3)	64 (18.7)	0.0001	NS	<0.0001	0.0293
Implanted PM, n (%)	509 (17.5)	287 (14.5)	118 (20.7)	104 (28.2)	<0.0001	0.001	<0.0001	0.008
QRS duration (ms), median (IQR)	113 (90–140)	110 (90–140)	120 (90–140)	125 (100–151)	<0.0001	0.0249	<0.0001	0.0006
LVEF (%), mean (SD)	31.4 (9.5)	31.7 (9.3)	31.0 (10.2)	30.1 (9.3)	0.0055	NS	0.0025	NS
LVEDV (ml), mean (SD)	180.72 (72.14)	180.56 (72.1)	179.29 (72.18)	183.69 (72.35)	NS	NS	NS	NS
LVESV (ml), median (IQR)	115 (83.06–157)	114 (82.5–156)	114.7 (83.2–158)	123 (87.05–160.77)	NS	NS	NS	NS
Hemoglobin (g/dl), mean (SD)	13.52 (1.58)	13.74 (1.52)	13.34 (1.56)	12.62 (1.62)	<0.0001	<0.0001	<0.0001	<0.0001
eGFR (ml·min−1·1.73 m−2), mean (SD)	71.02 (22.86)	82.78 (17.1)	52.97 (4.34)	35.91 (7.56)	<0.0001	NS	NS	NS
Serum creatinine (mg/dl), median (IQR)	1.1 (0.92–1.34)	1.0 (0.88–1.11)	1.4 (1.3–1.5)	1.85 (1.7–2.11)	<0.0001	NS	NS	NS
Serum sodium (mmol/L), mean (SD)	139.44 (3.21)	139.57 (3.08)	139.43 (3.24)	138.74 (3.68)	<0.0001	NS	<0.0001	0.0013
Serum potassium (mmol/L), mean (SD)	4.3±0.48	4.27 (0.45)	4.33 (0.51)	4.4 (0.55)	<0.0001	0.017	<0.0001	0.0191
BNP (pg/ml) (n=1,297), median (IQR)	319 (120–800)	271 (101–677)	371 (140–842)	621 (273–1,300)	<0.0001	0.0011	<0.0001	<0.0001
CPET data
HR (bpm) at rest, mean (SD)	70 (12)	71 (12)	70 (12)	70 (13)	NS	NS	NS	NS
SBP (mm Hg) at rest, median (IQR)	120 (108.5–130)	120 (110–130)	120 (110–130)	115 (100–130)	0.0191	0.0413	NS	0.0072
AT, n (%)	2,613 (88.94)	1,791 (89.91)	510 (88.85)	312 (83.87)	0.003	NS	0.0006	0.0267
EOV, n (%)	499 (17)	288 (14.47)	129 (22.51)	82 (22.1)	<0.0001	NS	0.0002	NS
nsV̇O2 at AT (ml·kg−1·min−1), median (IQR)	9.72 (7.93–11.86)	10 (8.25–12.27)	9.35 (7.6–11.12)	8.55 (6.9–10.36)	<0.0001	<0.0001	<0.0001	<0.0001
V̇O2 at AT (% of peak), mean (SD)	67.5 (13.6)	66.4 (13.4)	69.2 (13.7)	71 (13.6)	<0.0001	<0.0001	<0.0001	NS
HR at AT (bpm), mean (SD)	97 (19)	99 (19)	94 (19)	90 (17)	<0.0001	<0.0001	<0.0001	0.0033
Work rate at AT (W), mean (SD)	51.08 (22.96)	53.45 (23.94)	47.78 (20.02)	42.84 (18.81)	<0.0001	<0.0001	<0.0001	0.0029
O2 pulse at AT (ml/bpm), mean (SD)	8.2 (2.87)	8.36 (2.89)	7.99 (2.75)	7.61 (2.85)	<0.0001	0.0139	<0.0001	NS
Peak HR (bpm), mean (SD)	121 (24)	125 (23)	116 (240)	108 (22)	<0.0001	<0.0001	<0.0001	<0.0001
ΔHR peak-rest, mean (SD)	50 (24)	54 (24)	45 (24)	36 (20)	<0.0001	0.0001	<0.0001	<0.0001
Peak V̇O2 (L/min), median (IQR)	1,080 (851–1,392.3)	1,162 (915–1,486)	996.15 (797.3–1,231)	873.2 (703.1–1,071)	<0.0001	<0.0001	<0.0001	<0.0001
Peak V̇O2 (ml·kg−1·min−1), median (IQR)	14.3 (11.7–17.5)	15.2 (9.2–24.6)	13.2 (8.2–20.4)	11.5 (7.5–17.9)	<0.0001	<0.0001	<0.0001	<0.0001
Peak V̇O2 <12 ml·kg−1·min−1, n (%)	826 (28.1)	500 (25.1)	139 (24.2)	187 (50.7)	<0.0001	NS	<0.0001	<0.0001
Peak V̇O2 (% of predicted), mean (SD)	54.98 (16.33)	56.95 (16.67)	52.99 (15.03)	47.48 (13.7)	<0.0001	0.0021	<0.0001	<0.0001
Peak work rate (W), median (IQR)	80 (60–103)	84 (61–110)	72.5 (60–97)	65 (50–80)	<0.0001	<0.0001	<0.0001	<0.0001
Peak O2 pulse (ml·beats−1·min−1), median (IQR)	9.3 (7.3–11.57)	9.61 (7.6–11.97)	8.82 (7–11.1)	8.32 (6.54–10.27)	<0.0001	<0.0001	<0.0001	0.0038
V̇E/V̇CO2 slope, median (IQR)	31 (27–37)	30 (26.61–35)	32.1 (28.53–37.99)	35.09 (30–41)	<0.0001	<0.0001	<0.0001	<0.0001
V̇O2/work slope (ml·min−1·W−1), median (IQR)	9.6 (8.4–10.91)	9.8 (8.6–11)	9.3 (8.3–10.75)	9 (8–10.35)	<0.0001	0.0027	<0.0001	NS
Peak RR, median (IQR)	31 (27–35.69)	31 (27–35)	31 (28–36)	31 (27–35.25)	NS	NS	NS	NS
Peak V̇E (L/min), median (IQR)	45.5 (37.2–56)	47.4 (38.5–57.8)	43.6 (36–52)	41.75 (33.1–50)	<0.0001	<0.0001	<0.0001	0.0041
Peak RER, median (IQR)	1.14 (1.09–1.21)	1.14 (1.09–1.21)	1.13 (1.09–1.2)	1.14 (1.09–1.21)	NS	NS	NS	NS
Weber class
A, n (%)	367 (12.5)	323 (16.2)	36 (6.3)	8 (2.2)	<0.0001	<0.0001	<0.0001	<0.0001
B, n (%)	906 (30.8)	702 (35.2)	149 (26)	55 (14.8)
C, n (%)	1,301 (44.3)	794 (39.9)	302 (52.6)	205 (55.1)
D, n (%)	364 (12.4)	173 (8.7)	87 (15.1)	104 (28)
Treatment
ACEI, n (%)	2,271 (77.5)	1,594 (80.1)	435 (76)	242 (65.4)	<0.0001	0.0335	<0.0001	0.0004
AT1-blocker, n (%)	495 (16.9)	302 (15.2)	105 (18.3)	88 (23.7)	0.0002	NS	<0.0001	0.047
β-blocker, n (%)	2,497 (85)	1,722 (86.5)	470 (81.9)	305 (82)	0.0059	0.0063	0.024	NS
Diuretic, n (%)	2,365 (80.5)	1,520 (76.3)	499 (86.9)	346 (93)	<0.0001	<0.0001	<0.0001	0.0031
K-sparing drugs, n (%)	1,577 (53.71)	1,031 (51.78)	334 (58.29)	212 (56.99)	0.009	0.0059	<0.0001	0.0004
Antiplatelet agents, n (%)	1,556 (53.03)	1,052 (52.89)	304 (53.05)	200 (53.76)	NS	NS	NS	NS
Oral anticoagulant therapy, n (%)	846 (28.81)	521 (26.17)	203 (35.43)	122 (32.8)	<0.0001	<0.0001	0.0084	NS
Digitalis, n (%)	684 (23.3)	441 (22.14)	152 (26.62)	91 (24.46)	NS	0.0252	NS	NS
Amiodarone, n (%)	714 (24.34)	414 (20.81)	171 (29.9)	129 (34.7)	<0.0001	<0.0001	<0.0001	NS

ACEI, angiotensin-converting enzyme inhibitor; AT, anaerobic threshold; AT₁, angiotensin type 1; BMI, body mass index; BNP, B-type natriuretic peptide; CPET, cardiopulmonary exercise testing; CRT, cardiac resynchronization therapy; eGFR, estimated glomerular filtration rate; EOV, exercise oscillatory ventilation; HR, heart rate; ICD, implantable cardioverter-defibrillator; IQR, interquartile range; LVEDV, left ventricular end-diastolic volume; LVEF, left ventricular ejection fraction; LVESV, left ventricular end-systolic volume; NYHA, New York Heart Association; PM, pace-maker; RER, respiratory exchange ratio; RR, respiratory rate; SBP, systolic blood pressure; SD, standard deviation; V̇CO₂, carbon dioxide production; V̇E, ventilation; V̇O₂, oxygen uptake.

In the multivariable linear regression analysis, eGFR was a significant predictor of log-transformed peak V̇O₂ (t value: 5.45; P<0.0001). Figure 1 shows the geometric mean peak V̇O₂ according to eGFR strata. Other significant variables were age (t value: −10.58; P<0.0001), sex (t value: 8.15; P<0.0001), body mass index (t value: −9.7; P<0.0001), etiology of HF (t value: −2.42; P=0.0157), NYHA class (t value: −7.12; P<0.0001), atrial fibrillation (t value: −2.65; P=0.0081), resting heart rate (t value: −3.26; P=0.0011), log QRS duration (t value: −2.78; P=0.0055), hemoglobin (t value: 5.64; P<0.0001), treatment with ACEIs (t value: 2.77; P=0.0056) or AT₁-blockers (t value: 2.7; P=0.007), and LVEF (t value: 5.85; P<0.0001). The multivariable linear regression model explained 42% of the variation in peak V̇O₂. When the analysis was restricted to the 1,297 patients with available BNP data, eGFR remained significantly correlated with log-transformed peak V̇O₂ (t value: 4.79; P<0.0001). The model incorporating BNP explained 48% of the variation in peak V̇O₂.

Figure 1.

Multivariate adjusted geometric mean peak V̇O₂ with 95% confidence intervals according to strata of estimated glomerular filtration rate (eGFR).

Mean follow-up was 3.7 years in the entire population and 3.7, 3.9, and 3.3 years, respectively, in the patient subgroups of eGFR. The primary outcome occurred in 521 cases (17.8%); 221 deaths occurred in the subgroup of patients with eGFR ≥60 ml·min⁻¹·1.73 m⁻² (11.1%), 131 among those with eGFR in the range of 45–59 ml·min⁻¹·1.73 m⁻² (22.8%), and 103 with eGFR <45 ml·min⁻¹·1.73 m⁻² (27.7%). Urgent HT had to be performed in 52 (2.6%), 9 with (1.6%) and 5 (1.3%) patients in the respective subgroups of eGFR.

Figure 2 shows survival curves in the 3 eGFR subgroups, stratified by peak V̇O₂, and Figure 3 shows the ROC curves of peak V̇O₂ <12 ml·kg^–1·min^–1 at 3 years in the entire study population and in each strata of eGFR. The comparison of AUC with 95% CI of peak V̇O₂ <12 ml·kg^–1·min^–1 for the primary outcome at 2 and 3 years in the entire population and in each strata of eGFR is shown in Table 2. There was no significant interaction between peak V̇O₂ and eGFR for the primary outcome, indicating that the association of peak V̇O₂ with the primary outcome did not depend on eGFR (P value for interaction 0.7716). Peak V̇O₂ <12 ml·kg^–1·min^–1 was associated with an unadjusted HR for the primary outcome at 3 years of 2.56 (95% CI: 1.98–3.31; P<0.0001) in patients with eGFR ≥60, 2.20 (95% CI: 1.51–3.21; P<0.0001) in those with eGFR of 45–59, and 2.35 (95% CI: 1.59–3.67; P=0.0002) in those with EGFR <45 ml·min⁻¹·1.73 m⁻². After adjusting for age, sex, body mass index, SBP, DBP, etiology of HF, NYHA class, atrial fibrillation, LVESV, LVEF, heart rate, QRS duration, sodium and hemoglobin levels, and treatments, the HR for the occurrence of the primary outcome at 3 years for patients who achieved a peak V̇O₂ <12 ml·kg⁻¹·min⁻¹ vs. those who achieved a peak V̇O₂ ≥12 ml·kg⁻¹·min⁻¹ was 1.75 (95% CI: 1.06–2.91; P=0.0292) in patients with eGFR ≥60, 1.77 (95% CI: 0.87–3.61; P=0.1141) in those with eGFR of 45–59, and 2.72 (95% CIs: 1.01–7.37; P=0.0489) in those with eGFR <45 ml·min⁻¹·1.73 m⁻².

Figure 2.

Kaplan Meyer survival curves in the 3 subgroups of estimated glomerular filtration rate (eGFR), stratified by peak V̇O₂.

Figure 3.

Receiver-operating characteristic curves of peak V̇O₂ <12 ml·kg^–1·min^–1 for the primary outcome at 3 years. Numbers in parentheses represent 95% confidence intervals. AUC, area under the curve; eGFR, estimated glomerular filtration rate.

Table 2. Area Under ROC Curve With 95% CIs for Peak V̇O₂ <12 ml·kg⁻¹·min⁻¹ in the Entire Patient Population and in Each Strata of eGFR

	2 years	3 years
All patients	0.64 (0.58–0.69)	0.63 (0.57–0.68)
eGFR (ml·min−1·1.73 m−2)
≥60	0.63 (0.55–0.72)	0.63 (0.54–0.71)
45–59	0.70 (0.59–0.81)	0.67 (0.56–0.78)
<45	0.58 (0.47–0.69)	0.57 (0.47–0.69)

CI, confidence interval; eGFR, estimated glomerular filtration rate; ROC, receiver-operating characteristic.

Discussion

Although the effect of renal dysfunction on morbidity and mortality in HF patients has been firmly established, its effect on exercise capacity has been poorly defined. In this study, we investigated the association of renal function with exercise capacity and the effect of impaired renal filtration rate on the prognostic accuracy of peak V̇O₂ in HF patients with reduced LVEF. There are 2 major findings of this study.

First, renal dysfunction, as assessed by eGFR, positively correlated with decreased peak V̇O₂, independent of other established factors influencing peak V̇O₂ such as age, sex, obesity, NYHA class, atrial fibrillation, hemoglobin, and treatments, including CRT. Peak V̇O₂ as well as other key CPET-derived variables, including V̇E/V̇CO₂, significantly worsened with declining renal function. These findings suggest that renal dysfunction may contribute to exercise intolerance in HF. A significant correlation, however, does not prove a cause-effect relationship and our study was not designed to investigate the mechanisms underlying the association between decreasing renal function and decreasing exercise capacity in HF. Thus, whether this association reflects a cause-effect relationship or merely a more advanced stage of HF remains to be elucidated. Nonetheless, it is tempting to speculate about some mechanisms potentially linking renal dysfunction to reduced exercise performance. In HF, chronic sympathetic activation results in maladaptive changes in target organs rich in α- and β-adrenergic receptors.^17–19 Chronic sympathetic activation leads to decreased responsiveness of the failing heart to catecholamines, thus limiting its ability to increase cardiac output during dynamic exercise, increased peripheral vascular resistance, and impaired skeletal muscle vasodilatation capacity during exercise resulting in muscle hypoperfusion.^17–23 In addition, chronic sympathetic activation may contribute to skeletal myopathy.^24–26 These central and peripheral abnormalities are causally linked to reduced exercise capacity in systolic HF.^27–32 Dysfunctional kidneys have a pronounced effect on sympathetic activity^9,10,33,34 and, as demonstrated by Grassi et al,³⁵ the magnitude of the adrenergic drive is proportional to the degree of renal dysfunction. In HF, the dysfunctional kidneys are both the target and source of central sympathetic drive,^{9,10,33,34,36–40} thus fuelling the vicious circle of sympathetic overactivity. In a landmark study, Petersson et al demonstrated that increased renal norepinephrine spillover has a pathophysiological and prognostic role in parallel with and independent of cardiac sympathetic drive in HF.³⁷ It is conceivable that renal dysfunction may act as an amplifier of sympathetic activation in HF,^17,33,41 thus potentiating the sympathetically mediated mechanisms underlying reduced exercise capacity. Additional mechanisms may be implicated. As CKD is a catabolic condition, it may potentially contribute to the exercise intolerance of HF by activating metabolic pathways leading to skeletal muscle wasting. Systemic inflammation, a prominent feature in both HF and CKD, may induce proteolysis in skeletal muscle.^11,29,42 Chronic metabolic acidosis is a common condition in moderate-to-severe CKD and may worsen the catabolic/anabolic imbalance in skeletal muscle.^11,42 In addition, elevated levels of angiotensin II may contribute to skeletal myopathy by enhancing protein degradation and myocyte apoptosis.⁴³ Hormonal disorders such as growth hormone- and insulin-resistance, oxidative stress, and uremic toxins also may contribute.^11,42

Second, there was no statistically significant interaction between the level of residual renal function and peak V̇O₂ in relation to prediction of the primary outcome. The discriminative value of the cutoff for peak V̇O₂ of 12 ml·kg^–1·min^–1 was modest, with the worst performance among the patients with more severely impaired renal function. The finding that peak V̇O₂ had a modest predictive accuracy in the overall population is consistent with prior data.⁴⁴ Among patients in the lowest eGFR strata, peak V̇O₂ <12 ml·kg^–1·min^–1 was associated with an adjusted HR of borderline statistical significance and had a low discriminative value. Because it has been suggested that peak V̇O₂ tends to lose its predictive value 3 years post-CPET,¹⁶ we also estimated the discriminative value of peak V̇O₂ <12 ml·kg^–1·min^–1 at 2 years. The analysis yielded similar results: peak V̇O₂ <12 ml·kg^–1·min^–1 remained poorly predictive in the lowest eGFR strata. It should be noted that the peak respiratory exchange ratio was comparable across eGFR strata, indicating an equivalent intensity of effort. These findings should, however, be interpreted in context. Only a relatively small proportion of patients had poor renal function (12.7%). Moreover, a potential referral bias should be considered; indeed, it is reasonable to hypothesize that patients with more severely impaired renal function were less frequently referred for CPET than those with mildly impaired or normal renal function. In the study by McCullough et al, renal dysfunction was associated with a “clustering of high risk features”.⁴⁵ Our data are consistent with that finding. Indeed, the patients with poor renal function (eGFR <45 ml·min⁻¹·1.73 m⁻²) were older and had a more advanced NYHA class, lower SBP and hemoglobin values, and higher BNP concentrations, and more frequently needed treatment with diuretics. It is likely that, in such patients, the prognostic weight of other risk markers outranks that of decreased exercise capacity. CPET is widely used to risk stratify HF patients. Although not formally tested, the cutoff for peak V̇O₂ of 12 ml·kg⁻¹·min⁻¹ is currently regarded as a key prognostic marker to guide transplant listing in optimally treated HF patients,^6,7 albeit recent observations suggest differently.⁴⁶ There is, however, evidence that a multiparametric approach to risk stratification incorporating measures of both exercise capacity and renal function can allow a more efficient use of key CPET-derived prognostic variables.¹²

Previous Studies

McCullough et al⁴⁵ observed that peak V̇O₂ significantly decreases as renal function declines. On logistic regression, only age and peak V̇O₂ were independently associated with an eGFR <30 ml·min⁻¹·1.73 m⁻². When eGFR was modeled as a continuous variable, lower LVEF, higher NHYA class, and race also were found to be statistically significant. Van Laethem et al⁴⁷ studied 79 HT patients, in whom eGFR was a strong independent predictor of decreased exercise capacity. An eGFR value of 53 ml·min⁻¹·1.73 m⁻² was the optimal cutoff for discriminating patients with a peak V̇O₂ < or >18 ml·kg⁻¹·min⁻¹. These studies support the concept that renal dysfunction may influence peak V̇O₂.

Study Limitations

Some limitations of the MECKI score database are already reported.^12,13 The patients were relatively young; mean age was 61 years, likely reflecting the inclusion of patients referred for CPET. Only 15% of the patients were female; pathophysiological and clinical sex-related differences have been noted in HF. Only HF patients with reduced LVEF were included; thus, the results cannot be extrapolated to patients with preserved EF.^48,49 Moreover, we evaluated patients able to perform a CPET, so patients with very severe HF were excluded. Similarly, patients with primary CKD as well as those requiring dialysis were also excluded. We focused on peak V̇O₂ instead of other CPET-derived prognostic variables, such as percent predicted peak V̇O₂ or V̇E/V̇CO₂, because peak V̇O₂ is the most widely used measure of exercise capacity in HF. Further studies addressing the influence of renal dysfunction on percent predicted peak V̇O₂ or V̇E/V̇CO₂ would be of interest. Peak V̇O₂ was dichotomized at 12 ml·kg⁻¹·min⁻¹. Dichotomization of continuous covariates may be arbitrary in statistics and potentially may result in loss of useful information; however, the cutoff of 12 ml·kg⁻¹·min⁻¹ is generally accepted as clinically relevant and meaningful.^6,7

Conclusions

Renal dysfunction is significantly correlated with peak V̇O₂, independently of other known factors influencing peak V̇O₂. Further mechanistic studies are warranted to understand the pathophysiological link between renal dysfunction and decreased exercise capacity in HF. In HF patients with poor renal function, peak V̇O₂ offers limited prognostic information.

Conflict of Interest: none declared.

Appendix

Other MECKI Score Group Members

Centro Cardiologico Monzino, IRCCS, Milano: Cesare Fiorentini, Anna Apostolo, Pietro Palermo, Mauro Contini, Erika Bertella, Valentina Mantegazza; Cardiologia Riabilitativa, Ospedali Riuniti, Ancona: Francesca Pietrucci; UOC Cardiologia Ospedale S. Spirito, Roma: Aessandro Ferraironi; Azienda Ospedaliera Sant’Andrea, “Sapienza” Università degli Studi di Roma, Roma: Matteo Casenghi; ISMETT (Mediterranean Institute for Transplantation and Advanced Specialized Therapies), Palermo: Francesco Clemenza; Cardiologia SUN, Ospedale Monaldi, Napoli: Teo Roselli, Andrea Buono, Raffaele Calabrò; “S. Maugeri” Foundation, IRCCS, Cassano Murge: Daniela Santoro, Saba Campanale, Domenica Caputo; “S. Maugeri” Foundation, IRCCS, Tradate: Donatella Bertipaglia, Raffaella Vaninetti; Ospedali Riuniti and University of Trieste: Marco Confalonieri, Elena Zambon, Emanuela Berton, Chiara Torregiani; Cardiology, University of Civil Hospital, Brescia: Livio Dei Cas, Valentina Carubelli; UOC Cardiologia, G da Saliceto Hospital, Piacenza: Simone Binno; Division of Cardiology, Salvatore Maugeri Foundation, IRCCS, Institute of Milan, Milan: Giovanni Marchese; Dipartimento Cardiologico “A. De Gasperis”, Ospedale Cà Granda- A.O. Niguarda, Milano: Fabrizio Oliva; Fondazione Gabriele Monasterio, CNR-Regione Toscana, Pisa: Luigi Pastormerlo.

References

• 1.    Yancy  CW,  Jessup  M,  Bozkurt  B,  Butler  J,  Casey  DE Jr,  Drazner  MH, et al. 2013 ACCF/AHA guideline for the management of heart failure: Executive summary: A report of the American College of Cardiology Foundation/American Heart Association Task Force on practice guidelines. Circulation 2013; 128: 1810–1852.
• 2.    de Silva  R,  Nikitin  NP,  Witte  KK,  Rigby  AS,  Goode  K,  Bhandari  S, et al. Incidence of renal dysfunction over 6 months in patients with chronic heart failure due to left ventricular systolic dysfunction: Contributing factors and relationship to prognosis. Eur Heart J 2006; 27: 569–581.
• 3.    Bosselmann  H,  Gislason  G,  Gustafsson  F,  Hildebrandt  PR,  Videbaek  L,  Kober  L, et al. Incidence and predictors of end-stage renal disease in outpatients with systolic heart failure. Circ Heart Fail 2013; 6: 1124–1131.
• 4.    Butler  J,  Chirovsky  D,  Phatak  H,  McNeill  A,  Cody  R. Renal function, health outcomes, and resource utilization in acute heart failure: A systematic review. Circ Heart Fail 2010; 3: 726–745.
• 5.    Acquarone  N,  Castello  C,  Antonucci  G,  Lione  S,  Bellotti  P. Pharmacologic therapy in patients with chronic heart failure and chronic kidney disease: A complex issue. J Cardiovasc Med (Hagerstown) 2009; 10: 13–21.
• 6.    Mancini  D,  Lietz  K. Selection of cardiac transplantation candidates in 2010. Circulation 2010; 122: 173–183.
• 7.    Mehra  MR,  Kobashigawa  J,  Starling  R,  Russell  S,  Uber  PA,  Parameshwar  J, et al. Listing criteria for heart transplantation: International Society for Heart and Lung Transplantation guidelines for the care of cardiac transplant candidates – 2006. J Heart Lung Transplant 2006; 25: 1024–1042.
• 8.    Guazzi  M,  Adams  V,  Conraads  V,  Halle  M,  Mezzani  A,  Vanhees  L, et al. EACPR/AHA Scientific Statement: Clinical recommendations for cardiopulmonary exercise testing data assessment in specific patient populations. Circulation 2012; 126: 2261–2274.
• 9.    Schlaich  MP,  Socratous  F,  Hennebry  S,  Eikelis  N,  Lambert  EA,  Straznicky  N, et al. Sympathetic activation in chronic renal failure. J Am Soc Nephrol 2009; 20: 933–939.
• 10.    Ewen  S,  Ukena  C,  Linz  D,  Schmieder  RE,  Bohm  M,  Mahfoud  F. The sympathetic nervous system in chronic kidney disease. Curr Hypertens Rep 2013; 15: 370–376.
• 11.    Adams  GR,  Vaziri  ND. Skeletal muscle dysfunction in chronic renal failure: Effects of exercise. Am J Physiol Renal Physiol 2006; 290: F753–F761.
• 12.    Agostoni  P,  Corra  U,  Cattadori  G,  Veglia  F,  La Gioia  R,  Scardovi  AB, et al. Metabolic exercise test data combined with cardiac and kidney indexes, the MECKI score: A multiparametric approach to heart failure prognosis. Int J Cardiol 2013; 167: 2710–2718.
• 13.    Agostoni  P,  Corra  U,  Cattadori  G,  Veglia  F,  Battaia  E,  La Gioia  R, et al. Prognostic value of indeterminable anaerobic threshold in heart failure. Circ Heart Fail 2013; 6: 977–987.
• 14.    Hillege  HL,  Nitsch  D,  Pfeffer  MA,  Swedberg  K,  McMurray  JJ,  Yusuf  S, et al. Renal function as a predictor of outcome in a broad spectrum of patients with heart failure. Circulation 2006; 113: 671–678.
• 15.   KDIGO (Kidney Disease: Improving Global Outcomes). 2012 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int Suppl 2013; 3: 1–150.
• 16.    Arena  R,  Myers  J,  Abella  J,  Pinkstaff  S,  Brubaker  P,  Kitzman  D, et al. Defining the optimal prognostic window for cardiopulmonary exercise testing in patients with heart failure. Circ Heart Fail 2010; 3: 405–411.
• 17.    Floras  JS. Sympathetic nervous system activation in human heart failure: Clinical implications of an updated model. J Am Coll Cardiol 2009; 54: 375–385.
• 18.    Parati  G,  Esler  M. The human sympathetic nervous system: Its relevance in hypertension and heart failure. Eur Heart J 2012; 33: 1058–1066.
• 19.    Lymperopoulos  A,  Rengo  G,  Koch  WJ. Adrenergic nervous system in heart failure: Pathophysiology and therapy. Circ Res 2013; 113: 739–753.
• 20.    White  M,  Yanowitz  F,  Gilbert  EM,  Larrabee  P,  O’Connell  JB,  Anderson  JL, et al. Role of beta-adrenergic receptor downregulation in the peak exercise response in patients with heart failure due to idiopathic dilated cardiomyopathy. Am J Cardiol 1995; 76: 1271–1276.
• 21.    Rundqvist  B,  Eisenhofer  G,  Elam  M,  Friberg  P. Attenuated cardiac sympathetic responsiveness during dynamic exercise in patients with heart failure. Circulation 1997; 95: 940–945.
• 22.    Notarius  CF,  Ando  S,  Rongen  GA,  Floras  JS. Resting muscle sympathetic nerve activity and peak oxygen uptake in heart failure and normal subjects. Eur Heart J 1999; 20: 880–887.
• 23.    Notarius  CF,  Millar  PJ,  Murai  H,  Morris  BL,  Floras  JS. Inverse relationship between muscle sympathetic activity during exercise and peak oxygen uptake in subjects with and without heart failure. J Am Coll Cardiol 2014; 63: 605–606.
• 24.    Middlekauff  HR. Making the case for skeletal myopathy as the major limitation of exercise capacity in heart failure. Circ Heart Fail 2010; 3: 537–546.
• 25.    Bacurau  AV,  Jardim  MA,  Ferreira  JC,  Bechara  LR,  Bueno  CR Jr,  Alba-Loureiro  TC, et al. Sympathetic hyperactivity differentially affects skeletal muscle mass in developing heart failure: Role of exercise training. J Appl Physiol (1985) 2009; 106: 1631–1640.
• 26.    Brum  PC,  Bacurau  AV,  Medeiros  A,  Ferreira  JC,  Vanzelli  AS,  Negrao  CE. Aerobic exercise training in heart failure: Impact on sympathetic hyperactivity and cardiac and skeletal muscle function. Braz J Med Biol Res 2011; 44: 827–835.
• 27.    Anker  SD,  Chua  TP,  Ponikowski  P,  Harrington  D,  Swan  JW,  Kox  WJ, et al. Hormonal changes and catabolic/anabolic imbalance in chronic heart failure and their importance for cardiac cachexia. Circulation 1997; 96: 526–534.
• 28.    Lang  CC,  Rayos  GH,  Chomsky  DB,  Wood  AJ,  Wilson  JR. Effect of sympathoinhibition on exercise performance in patients with heart failure. Circulation 1997; 96: 238–245.
• 29.    Piepoli  MF,  Guazzi  M,  Boriani  G,  Cicoira  M,  Corrà  U,  Dalla Libera  L, et al. Exercise intolerance in chronic heart failure: Mechanisms and therapies: Part II. Eur J Cardiovasc Prev Rehabil 2010; 17: 643–648.
• 30.    Piepoli  MF,  Guazzi  M,  Boriani  G,  Cicoira  M,  Corrà  U,  Dalla Libera  L, et al. Exercise intolerance in chronic heart failure: Mechanisms and therapies: Part I. Eur J Cardiovasc Prev Rehabil 2010; 17: 637–642.
• 31.    Pina  IL,  Apstein  CS,  Balady  GJ,  Belardinelli  R,  Chaitman  BR,  Duscha  BD, et al. Exercise and heart failure: A statement from the American Heart Association Committee on exercise, rehabilitation, and prevention. Circulation 2003; 107: 1210–1225.
• 32.    Kato  J,  Koike  A,  Hoshimoto-Iwamoto  M,  Nagayama  O,  Sakurada  K,  Sato  A, et al. Relation between oscillatory breathing and cardiopulmonary function during exercise in cardiac patients. Circ J 2013; 77: 661–666.
• 33.    Sobotka  PA,  Mahfoud  F,  Schlaich  MP,  Hoppe  UC,  Bohm  M,  Krum  H. Sympatho-renal axis in chronic disease. Clin Res Cardiol 2011; 100: 1049–1057.
• 34.    DiBona  GF. Physiology in perspective: The wisdom of the body: Neural control of the kidney. Am J Physiol Regul Integr Comp Physiol 2005; 289: R633–R641.
• 35.    Grassi  G,  Quarti-Trevano  F,  Seravalle  G,  Arenare  F,  Volpe  M,  Furiani  S, et al. Early sympathetic activation in the initial clinical stages of chronic renal failure. Hypertension 2011; 57: 846–851.
• 36.    Zucker  IH,  Schultz  HD,  Patel  KP,  Wang  W,  Gao  L. Regulation of central angiotensin type 1 receptors and sympathetic outflow in heart failure. Am J Physiol Heart Circ Physiol 2009; 297: H1557–H1566.
• 37.    Petersson  M,  Friberg  P,  Eisenhofer  G,  Lambert  G,  Rundqvist  B. Long-term outcome in relation to renal sympathetic activity in patients with chronic heart failure. Eur Heart J 2005; 26: 906–913.
• 38.    Despas  F,  Detis  N,  Dumonteil  N,  Labrunee  M,  Bellon  B,  Franchitto  N, et al. Excessive sympathetic activation in heart failure with chronic renal failure: Role of chemoreflex activation. J Hypertens 2009; 27: 1849–1854.
• 39.    Oda  Y,  Joho  S,  Harada  D,  Hirai  T,  Asanoi  H,  Inoue  H. Renal insufficiency coexisting with heart failure is related to elevated sympathetic nerve activity. Auton Neurosci 2010; 155: 104–108.
• 40.    Franchitto  N,  Despas  F,  Labrunee  M,  Vaccaro  A,  Lambert  E,  Lambert  G, et al. Cardiorenal anemia syndrome in chronic heart failure contributes to increased sympathetic nerve activity. Int J Cardiol 2013; 168: 2352–2357.
• 41.    Grassi  G. Sympathetic neural activity in hypertension and related diseases. Am J Hypertens 2010; 23: 1052–1060.
• 42.    Okita  K,  Kinugawa  S,  Tsutsui  H. Exercise intolerance in chronic heart failure: Skeletal muscle dysfunction and potential therapies. Circ J 2013; 77: 293–300.
• 43.    Brink  M,  Price  SR,  Chrast  J,  Bailey  JL,  Anwar  A,  Mitch  WE, et al. Angiotensin II induces skeletal muscle wasting through enhanced protein degradation and down-regulates autocrine insulin-like growth factor I. Endocrinology 2001; 142: 1489–1496.
• 44.    Myers  J,  Arena  R,  Dewey  F,  Bensimhon  D,  Abella  J,  Hsu  L, et al. A cardiopulmonary exercise testing score for predicting outcomes in patients with heart failure. Am Heart J 2008; 156: 1177–1183.
• 45.    McCullough  PA,  Franklin  BA,  Leifer  E,  Fonarow  GC. Impact of reduced kidney function on cardiopulmonary fitness in patients with systolic heart failure. Am J Nephrol 2010; 32: 226–233.
• 46.    Cattadori  G,  Agostoni  P,  Corrà  U,  Di Lenarda  A,  Sinagra  G,  Veglia  F, et al. Severe heart failure prognosis evaluation for transplant selection in the era of beta-blockers: Role of peak oxygen consumption. Int J Cardiol 2013; 168: 5078–5081.
• 47.    Van Laethem  C,  Bartunek  J,  Goethals  M,  Verstreken  S,  Walravens  M,  De Proft  M, et al. Chronic kidney disease is associated with decreased exercise capacity and impaired ventilatory efficiency in heart transplantation patients. J Heart Lung Transplant 2009; 28: 446–452.
• 48.    Iwano  H,  Little  WC. Heart failure: What does ejection fraction have to do with it? J Cardiol 2013; 62: 1–3.
• 49.    Borlaug  BA. Mechanisms of exercise intolerance in heart failure with preserved ejection fraction. Circ J 2014; 78: 20–32.