Peak Oxygen Consumption is Positively Associated with Estimates of Oxygen Extraction and Microvascular Blood Volume in Veterans with Chronic Kidney Disease

Abstract

Peak oxygen consumption ($\dot{V}{O}_{2\text{peak}}$) is reduced in patients with chronic kidney disease (CKD). Although cardiovascular and skeletal muscle factors are implicated in the declines of VO_2peak, few studies have evaluated muscle oxygenation responses during exercise. We hypothesized that lower VO_2peak in CKD would be associated with attenuated responses in muscle oxygenation compared to those without CKD. Forty-six male Veterans (CKD stages 3 & 4, n=23; referent controls (REF), n=23) completed the study. Cardiopulmonary exercise testing (CPET) was performed on a treadmill using the Modified Bruce protocol. Peak change in dominant medial gastrocnemius deoxygenated hemoglobin/myoglobin (Δ[deoxy(Hb-Mb)]_peak), total hemoglobin/myoglobin (Δ[total(Hb-Mb)]_peak), tissue saturation index (ΔTSI), and ΔTSI reoxygenation half-time recovery (ΔTSI_reoxy1/2time), were assessed via near-infrared spectroscopy (NIRS). $\dot{V}{O}_{2\text{peak}}$, exercise time, HR_peak, $\dot{V}{O}_2$ at GET, and exercise time after GET were lower in the CKD group versus the REF group (p=0.002, p<0.001, p=0.020, p=0.044, and p=0.005, respectively). For NIRS outcomes, Δ[total(Hb-Mb)]_peak was lower, and ΔTSI_reoxy1/2time prolonged, in the CKD group compared to REF group (p=0.032 and p=0.031, respectively). $\dot{V}{O}_{2\text{peak}}$ was positively associated with HR_peak (CKD, r=0.57, p=0.005; REF, r=0.63, p=0.001) and Δ[total(Hb-Mb)]_peak (CKD, r=0.63, p=0.001; REF, r=0.52, p=0.012) in both groups. Conversely, $\dot{V}{O}_{2\text{peak}}$ was positively associated with Δ[deoxy(Hb-Mb)]_peak in the CKD group only (r=0.64, p<0.001). These findings suggest that skeletal muscle impairments, in addition to cardiovascular impairments, contribute to reduced $\dot{V}{O}_{2\text{peak}}$ in patients with CKD.

New and Noteworthy

Peak oxygen consumption is associated with peak heart rate, oxygen extraction, and microvascular blood volume in patients with chronic kidney disease (CKD), highlighting the importance of cardiovascular and skeletal muscle health in this patient population. Future studies are necessary to determine which exercise approaches are most efficacious at enhancing cardiorespiratory fitness and whether, and to what extent, improvements in cardiorespiratory fitness result from changes in cardiovascular factors, skeletal muscle factors, or a combination of both.

Graphical Abstract

INTRODUCTION

Chronic kidney disease (CKD) is among the leading causes of death worldwide (1). In 2022, the financial burden of CKD in the United Sates (U.S.) alone was $95.7 billion, with an additional $45.3 billion spent on treatment of end-stage kidney disease (ESKD) (2). Cardiovascular disease (CVD), rather than ESKD, is the primary cause of death in patients with CKD (3–5). Increased risk for CVD and cardiovascular mortality in the CKD population is, in part, due to the bidirectional interactions between the heart and kidney’s (6). Risk factors for the development of CVD in CKD include hypertension, dyslipidemia, smoking, hyperglycemia, vascular calcifications, chronic inflammation, and increased proteinuria (4).

Cardiorespiratory fitness, defined as the capacity of the cardiorespiratory system to deliver and utilize oxygen to meet the energetic requirements for maximal physical effort, is a strong predictor of adverse cardiovascular events, and disability and mortality in patients with CKD (7–10). Maximal or peak oxygen consumption ($\dot{V}{O}_{2\text{peak}}$) is the gold standard for assessing cardiorespiratory fitness and is routinely found to be compromised in patients with CKD when compared to that in their non-CKD counterparts (8, 11, 12). The increased risk for CVD and cardiovascular impairments in patients with CKD implicates reduced oxygen delivery as a contributor to decreased $\dot{V}{O}_{2\text{peak}}$ in this patient population (8, 11, 13–15).

Patients with CKD also experience skeletal muscle alterations which may attenuate oxygen utilization at the tissue level. Decreased mitochondrial number and function, reductions in oxidative enzymes, lower capillarization, and compromised diffusion capacity have been found to be present in patients with CKD versus adults with preserved kidney function (16–20). While compromised oxygen utilization has been suggested to play a role in the reduced cardiorespiratory fitness of patients with CKD (21, 22), few studies have evaluated skeletal muscle oxygenation responses of patients with CKD during exercise (25–27).

Near-infrared spectroscopy (NIRS) provides a non-invasive tool to evaluate factors controlling oxygen delivery and utilization at the tissue level (23, 24). Studies assessing skeletal muscle oxygenation using NIRS in patients with type 2 diabetes mellitus (T2DM) and patients with ESKD have reported impaired responses compared to those in healthy controls (25–27). Findings from studies using NIRS in patients with CKD demonstrate lower skeletal muscle oxidative capacity compared to healthy controls (25, 28–32). However, the associations of cardiovascular and skeletal muscle oxygenation factors with $\dot{V}{O}_{2\text{peak}}$ in patients with CKD are currently unclear.

The present study aimed to determine the associations of cardiovascular and skeletal muscle oxygenation factors with $\dot{V}{O}_{2\text{peak}}$ in older male U.S. Veterans with and without CKD stages 3 or 4. We hypothesized that, in addition to cardiovascular factors, lower $\dot{V}{O}_{2\text{peak}}$ in Veterans with CKD would be associated with attenuated responses in skeletal muscle oxygenation compared to those in older Veterans without evidence of CKD.

MATERIALS AND METHODS

Study Design and Protocol

This study used a cross-sectional design comparing older male U.S. Veterans with CKD stages 3 & 4 (n=23) to a similar group of male U.S. Veterans without CKD (n=23). Community-dwelling U.S. Veterans were screened for potential enrollment at the Veterans Affairs Medical Center in Washington D.C. (DC VAMC) from October 2022-September 2025. The study was approved by the DC VAMC Institutional Review Board (IRB) and Research and Development (R&D) Committee prior to the commencement of research activities. Inclusion criteria for study enrollment required participants to be ambulatory with or without the use of assistive devices (e.g., cane, walker), aged 50-84 years, and diagnosed with CKD stage 3 or 4 defined as estimated glomerular filtration rate (eGFR) 59-15 ml/min/1.73 m² according to Kidney Disease Improving Global Outcomes (KDIGO) guidelines, and not requiring kidney replacement therapy (33). Exclusion criteria were a history of acute kidney injury, inability to follow study instructions, peripheral artery disease, and any uncontrolled cardiovascular or musculoskeletal problems that would make study participation unsafe. Male Veterans without clinical evidence of CKD stages 3-5 (i.e., eGFR >60 ml/min/1.73 m²), ambulatory with or without the use of assistive devices, aged 50-84 years, and free of any uncontrolled cardiovascular or musculoskeletal problems that would make study participation unsafe formed a referent group (REF) for comparison. Veterans were recruited via mailed flyers and clinician referrals. All CKD and REF study participants voluntarily provided written informed consent using a DC VAMC IRB and R&D approved form before study participation. The data presented herein are part of a larger randomized clinical trial (ClinicalTrials.gov NCT04397159).

Procedures

After informed consent and formal enrollment into the study, all participants completed a standard clinical blood draw and urinalysis, followed by a symptom-limited cardiopulmonary exercise test (CPET) on a treadmill using the Modified Bruce protocol (34). Breath-by-breath gas exchange, electrocardiography (ECG), and NIRS data were collected simultaneously during CPET to determine $\dot{V}{O}_{2\text{peak}}$, heart rate (HR), and changes in skeletal muscle oxygenation of the dominant medial gastrocnemius. Body mass (BM), fat mass, and fat-free mass were determined prior to CPET using multi-frequency segmental bioelectrical impedance analysis (Tanita MC-980U Plus).

Gas exchange was measured continuously during the CPET using breath-by-breath open circuit spirometry (COSMED USA Inc, Concord, CA, USA). Standard calibration was performed before participant preparation, in accordance with the manufacturer’s specifications. Each participant completed a 3-minute standing rest period prior to exercise initiation to obtain resting baseline values. Following the standing rest, participants completed successive 3-minute stages that increased in speed and/or grade until volitional exhaustion. Rating of perceived fatigue (RPF) was recorded immediately following the termination of exercise using a 0-10 Likert scale, where 0 represents “not fatigued at all” and 10 represents “total fatigue and exhaustion-nothing left” (35). Participants were familiarized with the RPF scale and accompanying diagrams prior to starting the exercise test and were provided standardized instructions for rating their perceived fatigue levels (35). Given the clinical nature of the study, obtaining a “true” maximal test was not a requirement. However, a respiratory quotient (RQ) of ≥1.00 and 85% of age-predicted maximal HR were used as criteria for establishing if a maximal test was performed for interpretation purposes (36).

Skeletal muscle oxygenation indices were obtained using a portable continuous-wave NIRS device with spatial resolution (Portamon, Artinis Medical Systems, The Netherlands). The NIRS device emits continuous-wave light at 2 wavelengths (760 nm and 850 nm) at a sampling rate of 10 Hz per second to measure micromolar (μM) changes of oxygenated (Δ[oxy(Hb-Mb)]), deoxygenated (Δ[deoxy(Hb-Mb)]), and total (Δ[total(Hb-Mb)]) hemoglobin-myoglobin concentrations. Because the NIRS device used in the current study is only able to provide relative changes, an initial value was arbitrarily set to zero using the manufactures software at the start of each CPET. The device also calculates tissue saturation index (TSI) as an estimate of oxygen saturation of the tissue being interrogated. Readers are referred to previously published reviews for detailed descriptions on NIRS technology (23, 24, 37, 38). For the purposes of this study, Δ[deoxy(Hb-Mb)] (i.e., surrogate of fractional oxygen extraction), Δ[total(Hb-Mb)] (i.e., surrogate of microvascular blood volume), and TSI were examined. The NIRS optode was placed vertically on the belly of the medial gastrocnemius muscle of the dominant limb. The site of NIRS optode placement was shaved and cleaned with 70% isopropyl alcohol prior to attachment. The optode was secured to the skin using double-sided adhesive and wrapped with self-adherent elastic band to block environmental light penetration. Subcutaneous adipose tissue thickness (ATT) was quantified via longitudinal B-mode ultrasound scans of the medial gastrocnemius at the site of the optode placement by a single experienced technician to account for potential differences between groups (Nobulus, Hitachi Aloka Medical, Parsippany, NJ, USA).

Data Analysis

Prior to completing the CPET protocol, resting baseline measures were obtained as the participants stood quietly on the treadmill for 3 minutes. Baseline $\dot{V}{O}_2$ ($\dot{V}{O}_{\text{2BSL}}$) and HR (HR_BSL) were calculated as the average of the last 30-seconds during the standing rest period whereas $\dot{V}{O}_{2\text{peak}}$ and peak HR (HR_peak) were measured as the average of the final 30-seconds completed during the CPET. Gas-exchange threshold (GET) was determined by 2 independent reviewers using the V-slope and dual criterion methods (39). The V-slope method identifies the GET as the point at which the rate of CO2 production ($\dot{V}{\mathrm{CO}}_2$) accelerates beyond the rate of $\dot{V}{O}_2$, without hyperventilation. The Dual Criterion method involves simultaneous evaluation of the ventilatory equivalent for oxygen ($\dot{V}\text{E}/\dot{V}{O}_2$) and ventilatory equivalent for CO2 ($\dot{V}\text{E}/\dot{V}{\mathrm{CO}}_2$); GET is identified as the point at which there is a systematic increase in $\dot{V}\text{E}/\dot{V}{O}_2$ without a simultaneous increase in $\dot{V}\text{E}/\dot{V}{\mathrm{CO}}_2$. Two independent raters first identified $\dot{V}{O}_2$ at GET using the V-slope method with data averaged in 10-sec epochs. This value was cross-referenced with the dual criterion method, making manual adjustments if necessary. Interrater acceptability criteria were defined a priori as ≤ 10% or a ≤ 0.15-L/min difference between the two $\dot{V}{O}_2$ values. If interrater criteria were not met, differences were resolved via consensus during an in-person meeting and review of relevant CPET data. GET was considered indeterminant when inter-rater criteria were not met and consensus could not be reached or when both reviewers agreed that GET could not reliably be identified. Estimated $\dot{V}{O}_{2\text{peak}}$ was determined using Eq 1 (40). Percent predicted $\dot{V}{O}_{2\text{peak}}$ ($\dot{V}{O}_{2\text{peak}}$PP) and percent predicted $\dot{V}{O}_2$ at the GET ($\dot{V}{O}_2$PP at GET) were then calculated as a percentage of the actual $\dot{V}{O}_2$ achieved relative to the estimated $\dot{V}{O}_{2\text{peak}}$ using Eq. 1.

$$\dot{V}{O}_{2\text{peak}}\left({\text{ml·kg}}^{-1}·{\text{min}}^{-1}\right)=45.2–0.35\text{x age}–10.9\text{x gender}\left(\text{male}=1;\text{female}=2\right)–0.15\text{x weight}\left(\text{lb}\right)+0.68\text{x height}\left(\text{in}\right)–0.46\text{x exercise mode}\left(\text{treadmill}=1;\text{bike}=2\right)$$ Equation 1:

NIRS variables were converted to second-by-second data points offline by averaging 10 data points per second. After the onset of exercise there is an initial time delay (~10 seconds) before an observed rise in Δ[deoxy(Hb-Mb)] and Δ[total(Hb-Mb)] variables (Fig. 1). The initial time delay is suggested to reflect a complex balance between Δ[deoxy(Hb-Mb)], oxygen delivery, and the effect of muscle contraction on microvascular volume (41). Thus, to isolate the metabolic response of the medial gastrocnemius, peak Δ[deoxy(Hb-Mb)] (Δ[deoxy(Hb-Mb)]_peak) and peak Δ[total(Hb-Mb)] (Δ[total(Hb-Mb)]_peak) were determined by subtracting the nadir achieved after the initial time delay from the end of exercise. The nadir was identified by visual inspection of the second-by-second data and quantified as the average of the lowest 3 data points following the onset of exercise while the end of exercise was quantified as the average of the last 3 data points achieved during the CPET. The Δ[deoxy(Hb-Mb)] response was adjusted for Δ[total(Hb-Mb)] using methods by Ryan et al. (42). Baseline TSI (TSI_BSL) was determined by averaging the last 30-seconds of the standing rest period and peak ΔTSI (ΔTSI_peak) was calculated by subtracting the TSI_BSL from the last 3-seconds achieved during the CPET. The reoxygenation rate of TSI during the recovery of exercise was quantified as the time to achieve 50% of the baseline ΔTSI value (ΔTSI_reoxy1/2time) immediately following the termination of exercise (29, 43). In a few instances, the reoxygenation of TSI did not increase following exercise termination and therefore data from these participants were excluded from the analysis.

Figure 1.

(a) Example of change in the dominant medial gastrocnemius ∆[deoxy(Hb-Mb)] and Δ[total(Hb-Mb)] and (b) ΔTSI in a Veteran with CKD stage 3 in response to CPET using the Modified Bruce protocol performed on a treadmill. Black circles and solid lines represent amplitude change from nadir in Δ[deoxy(Hb-Mb)] and open circles and dashed lines represent amplitude change from nadir in Δ[total(Hb-Mb)].

Abbreviations. Δ[deoxy(Hb-Mb)], deoxygenated hemoglobin-myoglobin; Δ[total(Hb-Mb)], total hemoglobin-myoglobin; CKD, chronic kidney disease; CPET, cardiopulmonary exercise testing; ΔTSI, tissue saturation index; ΔTSI_reoxy1/2time, time to reach 50% of resting ΔTSI.

Statistical Analysis

Sample size was estimated a-priori to be 17 participants per group based on differences between two independent means for $\dot{V}{O}_{2\text{peak}}$ and ΔTSI_reoxy1/2time for two tailed hypothesis, large effect size (1.0), alpha of p<0.05, and at a power of 0.8, in accordance with previously published data in patients with CKD (11, 29). All data were normally distributed with equal variances according to the Kolmogorov-Smirnov test and the Levene’s test. Differences in participant characteristics and peak responses were analyzed using independent samples t-test. Chi-square (χ²) test of independence was used to test differences between groups for categorical data. General linear model for repeated measures was used to assess within-group and between-group differences at 0%, 25%, 50%, 75%, and 100% of total exercise time. Pearson product moment correlation coefficients (r) were used to assess bivariate associations of $\dot{V}{O}_{2\text{peak}}$ with HR_peak, Δ[deoxy(Hb-Mb)]_peak, Δ[total(Hb-Mb)]_peak, ΔTSI_peak, and ΔTSI_reoxy1/2time. Statistical significance was set at p<0.05. All values are expressed as means ± SD. Statistical analyses were performed using SPSS version 24 (IBM, Inc., Armonk, NY, USA).

RESULTS

Baseline eGFR was significantly lower in CKD than REF in participants (p<0.001). The CKD group was comprised of 8 participants with stage 3a (eGFR 59-45 mL/min per 1.73 m²), 10 participants with stage 3b (eGFR 44-30 mL/min per 1.73 m²), and 5 participants with stage 4 (eGFR 29-15 mL/min per 1.73 m²). CKD and REF groups were well matched for age, body mass, body mass index (BMI), hemoglobin, fat mass, fat free mass, and medial gastrocnemius ATT (p>0.05) (Table 1). One CKD participant presented with severe anemia (hemoglobin = 7.6 g/dl). No significant differences were present in $\dot{V}{O}_{\text{2BSL}}$, HR_BSL, TSI_BSL, resting systolic blood pressure, or resting diastolic blood pressure (p>0.05). The CKD group had a higher prevalence of T2DM compared to the REF group, though the latter difference was not statistically significant (p=0.116 and p=0.005, respectively). Racial/ethnic differences were observed between groups with the CKD group being predominantly African American/Black (p=0.028), whereas the REF group was demographically diverse. The CKD group also had a greater percentage of individuals on diuretics and ACE-inhibitors compared to the REF group (p=0.009 and p=0.007, respectively), whereas there were no significant group differences in the use of beta blockers, angiotensin II receptor blockers or calcium channel blockers.

Table 1.

Participant characteristics.

	CKD (n=23)	REF (n=23)	p-value
Age (years)	74.9 ± 6.2	72.0 ± 7.0	0.140
Body mass (kg)	86.6 ± 19.7	87.0 ± 12.1	0.839
BMI (kg∙m−2)	28.1 ± 5.6	27.0 ± 3.3	0.440
eGFR (mL/min per 1.73 m2)	39.3 ± 11.0	78.0 ± 11.9	<0.001†
Fat mass (kg)	21.4 ± 11.6	20.4 ± 6.3	0.721
Fat free mass (kg)	64.6 ± 8.5	67.1 ± 7.5	0.304
Hemoglobin (g/dl)	13.8 ± 2.2	14.6 ± 0.9	0.089
ATT (mm)	3.7 ± 2.4	3.8 ± 2.2	0.844
Diabetes mellitus (%)	43	22	0.116
Hypertension (%)	87	48	0.005†
African American/Black (%)	83	52	0.028*
Caucasian/White (%)	17	35	0.179
Latino/Hispanic (%)	0	9	0.148
Asian (%)	0	4	0.312
Diuretics (%)	43	9	0.009†
Beta-blockers (%)	30	9	0.063
ACE inhibitors (%)	43	9	0.007†
Angiotensin II receptor blockers (%)	9	14	0.636
Calcium channel blockers (%)	30	26	0.734
$\dot{\mathit{V}}{O}_{\text{2BSL}}$ (ml·kg−1·min−1)	3.9 ± 1.1	4.6 ± 2.1	0.180
HRBSL (bpm)	76 ± 14	77 ± 16	0.863
TSIBSL (%)	58.4 ± 5.0	57.3 ± 4.5	0.401
Resting systolic blood pressure (mmHg)	137 ± 17	127 ± 19	0.128
Resting diastolic blood pressure (mmHg)	77 ± 11	80 ± 11	0.272

Abbreviations. CKD, participants with stages 3 and 4 chronic kidney disease; REF, reference participants without CKD; eGFR, estimated glomerular filtration rate; BMI, body mass index; $\dot{V}{O}_{\text{2BSL}}$, baseline oxygen consumption; HR_BSL, baseline heart rate; TSI_BSL, baseline tissue saturation index; kg, kilogram; g/dl, grams per deciliter; m, meter; mm, millimeter; ATT, adipose tissue thickness; ml·kg^-1·min^-1, milliliter per kilogram per minute; bpm, beats per minute; mmHg, millimeters of mercury; %, percent.

^*Significant difference between CKD and REF groups at p<0.05.

^†Significant difference between CKD and REF groups at p<0.01.

Peak CPET and Muscle Oxygenation Responses to Acute Exercise

$\dot{V}{O}_{2\text{peak}}$ (p=0.002), $\dot{V}{O}_{2\text{peak}}$PP (p=0.020), HR_peak (p=0.009), exercise time (p<0.001), $\dot{V}{O}_2$ at GET (p=0.044), time after GET (p=0.006), RQ (p=0.007), and Δ[total(Hb-Mb)]_peak (p=0.032) were all lower, and ΔTSI_reoxy1/2time was prolonged (p=0.031), in the CKD versus REF group (Table 2). $\dot{V}{O}_2$ at GET was indeterminant in 1 CKD participant and 3 REF participants. Similarly, ΔTSI_reoxy1/2time was indeterminant in 3 participants from each group. There were no significant group differences in $\dot{V}{O}_2$PP at GET (p=0.717), time to GET (p=0.095), peak systolic blood pressure (p=0.669), peak diastolic blood pressure (p=0.542), Δ[deoxy(Hb-Mb)]_peak (p=0.162), ΔTSI_peak (p=0.503), or RPF (p=0.736).

Table 2.

Peak cardiorespiratory and muscle oxygenation responses to acute exercise.

	CKD	REF	p-value
$\dot{\mathit{V}}{O}_{\text{2peak}}$ (ml·kg−1·min−1)	18.1 ± 5.1	23.1 ± 5.2	0.002†
$\dot{\mathit{V}}{O}_{\text{2peak}}$PP (%)	71.7 ± 16.8	82.9 ± 14.2	0.020*
HRpeak (bpm)	127 ± 21	143 ± 18	0.009†
Exercise time (min)	10.6 ± 3.7	14.6 ± 2.7	<0.001†
$\dot{\mathit{V}}{O}_2$ at GET (ml·kg−1·min−1)	12.9 ± 2.9	14.6 ± 2.1	0.044*
$\dot{\mathit{V}}{O}_{\text{2PP}}$ at GET (%)	53.0 ± 14.9	51.6 ± 7.3	0.717
Time to GET (min)	4.9 ± 3.0	6.6 ± 3.3	0.095
Time after GET (min)	5.8 ± 2.6	8.1 ± 2.5	0.006†
Peak systolic blood pressure (mmHg)	187 ± 27	190 ± 26	0.669
Peak diastolic blood pressure (mmHg)	79 ± 20	82 ± 16	0.542
RQ	1.0 ± 0.1	1.1 ± 0.1	0.007†
∆[deoxy(Hb-Mb)]peak (µM)	11.2 ± 6.1	14.4 ± 8.6	0.162
∆[total(Hb-Mb)]peak (µM)	9.2 ± 6.7	13.9 ± 7.8	0.032*
∆TSIpeak (%)	-9.4 ± 6.7	-11.0 ± 9.2	0.503
∆TSIreoxy1/2time (s)	40.8 ± 19.4	29.7 ± 10.5	0.031*
RPF (a.u.)	4.7 ± 3.0	5.0 ± 2.0	0.736

Abbreviations. $\dot{V}{O}_{\text{2peak}}$, peak oxygen consumption; $\dot{V}{O}_{\text{2peakPP}}$, actual $\dot{\mathrm{V}}{O}_2$ achieved as a percentage of estimated $\dot{\mathrm{V}}{O}_{\text{2peak}}$ at peak exercise; HR_peak, peak heart rate; GET, gas-exchange threshold; $\dot{V}{O}_{\text{2PP}}$ at GET, $\dot{\mathrm{V}}{O}_{\text{2}}$ achieved at GET as a percentage of estimated $\dot{\mathrm{V}}{O}_{\text{2peak}}$; RQ, respiratory quotient; ∆[deoxy(Hb-Mb)]_peak, change in deoxygenated hemoglobin; ∆[total(Hb-Mb)]_peak, change in total hemoglobin; ∆TSI, change in tissue saturation index; ∆TSI_reoxy1/2time, half time to reach 50% of resting ∆TSI immediately following exercise cessation; ml·kg^-1·min^-1, milliliters per kilogram per minute; RPF, rating of perceived fatigue; bpm, beats per minute; s, seconds; a.u., arbitrary unit; ml/beat, milliliters per beat; %, percent; mmHg, millimeters of mercury.

^*Significant difference between CKD and REF groups at p<0.05.

^†Significant difference between CKD and REF groups at p<0.01.

Relative Changes in CPET and Muscle Oxygenation Responses to Acute Exercise

Significant increases in $\dot{V}{O}_2$ and HR were observed throughout the duration of the CPET in both CKD and REF groups (Fig. 2). Time by group interactions revealed that the CKD group had significantly lower $\dot{V}{O}_2$ compared to the REF group at 25, 50, 75, and 100% of total exercise time (p=0.024, p=0.006, p<0.001, and p=0.002, respectively) whereas HR was only significantly lower in the CKD group versus REF group at peak exercise (p=0.009).

Figure 2.

Changes in (a) oxygen consumption ($\dot{V}{O}_2$), (b) heart rate (HR), (c) deoxygenated hemoglobin-myoglobin (Δ[deoxy(Hb-Mb)]), (d) total hemoglobin-myoglobin (Δ[total(Hb-Mb)]), and (e) tissue saturation index (ΔTSI) at 0, 25, 50, 75, and 100% total exercise time achieved during cardiopulmonary exercise testing (CPET). Veterans with chronic kidney disease (CKD) are represented by the black circles and solid lines while Veterans without CKD are represented by the gray circles and dashed lines.

†Indicates within-group significant difference from prior time point for CKD at p<0.05

‡Indicates within-group significant difference from prior time point for REF at p<0.05

*Indicates between-group significant difference for a given time point at p<0.05

Significant increases in Δ[deoxy(Hb-Mb)] were observed at 25% of total exercise time in the CKD group after which Δ[deoxy(Hb-Mb)] plateaued. Conversely, Δ[deoxy(Hb-Mb)] in the REF group increased significantly until exercise termination. However, no significant group differences were noted in Δ[deoxy(Hb-Mb)] at any time point. Both CKD and REF groups significantly increased Δ[total(Hb-Mb)] to 75% of total exercise time with between group differences observed at 25%, 50%, 75%, and 100% total exercise time (p=0.024, p=0.007, p=0.008, and p=0.032). ΔTSI decreased to 50% of total exercise time in the CKD group and 75% of total exercise in the REF group with no between-group differences observed at any time point.

Bivariate Correlates of $\dot{\mathit{V}}$O_2peak with CPET Outcomes

In the CKD group, positive associations were observed between $\dot{V}{O}_{2\text{peak}}$ and both HR_peak and RQ (p=0.005 and p=0.008, respectively) (Fig. 3). In the REF group, $\dot{V}{O}_{2\text{peak}}$ was positively associated with HR_peak (p=0.001).

Figure 3.

Scatter plots depicting the relationship of peak oxygen consumption ($\dot{V}{O}_{2\text{peak}}$) with (a) peak heart rate (HR_peak) and (b) respiratory quotient (RQ) in Veterans with and without chronic kidney disease (CKD) stages 3 and 4. Veterans with CKD are represented by black circles and solid lines of best fit while referent (REF) Veterans are represented by gray circles and dashed lines of best fit.

*Correlation significant at p<0.05

Bivariate Correlates of $\dot{\mathit{V}}$O_2peak with Muscle Oxygenation Outcomes

In the CKD group, positive associations were noted between $\dot{V}{O}_{2\text{peak}}$ and both Δ[deoxy(Hb-Mb)]_peak and Δ[total(Hb-Mb)]_peak (p<0.001 and p=0.001, respectively) (Fig. 4). In the REF group, $\dot{V}{O}_{2\text{peak}}$ was positively associated only with Δ[total(Hb-Mb)]_peak only (p=0.012). There were no significant associations of $\dot{V}{O}_{2\text{peak}}$ with ΔTSI_reoxy1/2time or ΔTSI_peak in either the CKD or REF group.

Figure 4.

Scatter plots depicting the relationship of peak oxygen consumption ($\dot{V}{O}_{2\text{peak}}$) with (a) peak deoxygenated hemoglobin-myoglobin (Δ[deoxy(Hb-Mb)]_peak), (b) peak total hemoglobin-myoglobin (Δ[total(Hb-Mb)]_peak), (c) tissue saturation index reoxygenation half-time (ΔTSI_reoxy1/2time), and peak tissue saturation index (ΔTSI_peak) in Veterans with and without chronic kidney disease (CKD) stages 3 and 4. Veterans with CKD are represented by black circles and solid lines of best fit while referent (REF) Veterans are represented by gray circles and dashed lines of best fit.

*Correlation significant at p<0.05

DISCUSSION

The present study aimed to advance the understanding of selected cardiovascular and skeletal muscle factors underlying compromised $\dot{V}{O}_{2\text{peak}}$ in older male U.S. Veterans with CKD stages 3 or 4 not on dialysis. The primary findings of this study are that male Veterans with CKD undergoing CPET demonstrate lower $\dot{V}{O}_2$ and Δ[total(Hb-Mb)] responses compared to male Veterans without CKD similar in age and body composition (Fig. 2). Additionally, those with CKD exhibited 1) modestly reduced $\dot{V}{O}_2$ at the GET (p=0.044), 2) reduced exercise time after GET (p=0.006), and 3) prolonged ΔTSI_reoxy1/2time immediately following exercise cessation (p=0.031), when compared to Veterans without CKD. The positive association between $\dot{V}{O}_{2\text{peak}}$ and Δ[deoxy(Hb-Mb)]_peak (r=0.64, p<0.001) in the CKD group, but not the REF group, suggests that Veterans with CKD achieving a higher $\dot{V}{O}_{2\text{peak}}$ and are able to extract a greater amount of oxygen at the medial gastrocnemius during exercise. Collectively, these findings further highlight the diminished aerobic capacity in older male Veterans with CKD and provide novel insights into the relationship between $\dot{V}{O}_{2\text{peak}}$ and changes in skeletal muscle oxygenation.

Notably, the $\dot{V}{O}_{2\text{peak}}$ of 18.1 ml·kg⁻¹·min⁻¹ achieved by the CKD group represents the ~35^th percentile for healthy men aged 70-79 years (36), underscoring the extent to which aerobic capacity is diminished in CKD patients (8, 11, 44–46). In contrast, the $\dot{V}{O}_{2\text{peak}}$ of 23.1 ml·kg^-1·min⁻¹ achieved by the REF group represents the ~70^th percentile for healthy men of the same age category. The risk for loss of functional independence is increased once $\dot{V}{O}_{2\text{peak}}$ reaches a level of 18 ml·kg⁻¹·min⁻¹ or below (47). Of the CKD participants enrolled in the present study, 50% had a $\dot{V}{O}_{2\text{peak}}$ below 18 ml·kg⁻¹·min⁻¹ compared to only 13% in the REF group. Similarly, a $\dot{V}{O}_2$ of <11 ml·kg⁻¹·min⁻¹ at the GET is commonly used as a threshold to identify at-risk patients across multiple conditions (48). In the current study, 26% of CKD participants and 22% of REF participants had a $\dot{V}{O}_2$ at the GET of <11 ml·kg⁻¹·min⁻1. However, an additional 37% of participants in the CKD had a $\dot{V}{O}_2$ at the GET of <12 ml·kg⁻¹·min⁻¹ compared to only 1% in the REF group. Together, data from $\dot{V}{O}_{2\text{peak}}$ and _$\dot{V}$O₂ at the GET highlight the extent to which aerobic capacity is compromised in Veterans with CKD and the potential for increased risk of future complications in these patients.

The Δtotal[Hb-Mb] variable from NIRS is thought to reflect changes in total blood volume in the microvasculature (23, 24). In older adults, total blood volume has been identified as an independent determinant of $\dot{V}{O}_{2\text{peak}}$ and peak exercise stroke volume (49–51). The ability to increase Δtotal[Hb-Mb] with progression of exercise intensity is attenuated in individuals with lower fitness levels and in clinical populations in which microvascular structure and function are impaired (26, 27, 52). For example, Δ[total(Hb-Mb)] response was found to be lower in diabetic patients with and without peripheral arterial disease when compared to healthy control subjects (27). More recently, an impaired Δ[total(Hb-Mb)] response was reported in patients on dialysis compared to healthy control subjects, with Δ[total(Hb-Mb)]_peak positively associated with $\dot{V}{O}_{2\text{peak}}$ in both groups (25). Our findings extend those previously reported in diabetic and dialysis patients, and demonstrate that patients with CKD have significantly lower Δ[total(Hb-Mb)] responses at all time points of the exercise test compared to REF participants. Furthermore, we observed a positive association was observed between $\dot{V}{O}_{2\text{peak}}$ and Δ[total(Hb-Mb)]_peak in both the CKD and REF groups. Thus, the inability to increase microvascular blood volume as exercise intensity progresses may serve as a limiting factor in the $\dot{V}{O}_{2\text{peak}}$ achieved by those with CKD.

Several factors may impede the ability to adequately increase Δ[total(Hb-Mb)] in the microvasculature in response to acute exercise in patients with CKD. Endothelial dysfunction is a primary contributor to atherosclerosis and CVD in CKD patients, and may limit the vasodilatory response to acute exercise (19). Downey et al. (53) found that lower flow mediated dilation was associated with reduced $\dot{V}{O}_{2\text{peak}}$ in patients with stage 3 CKD. Additionally, reduced capillary density, thickening of the capillary endothelium and capillary basement membranes, and enlargement of the interstitial space have been detected in hemodialysis patients (16, 18, 54). Importantly, capillary density and post-occlusive hyperemia are progressively reduced with advancing stages of CKD prior to kidney failure, indicating alterations in capillary structure and function are present in early stages of the disease and not unique to kidney failure patients (55). The mechanisms underlying the reduced Δ[total(Hb-Mb)] responses to exercise in CKD participants in the current study are unclear and require further investigation.

The inability to extract sufficient oxygen from the circulation to meet the energetic demands of muscle contraction during exercise is implicated as a potential factor for reduced $\dot{V}{O}_{2\text{peak}}$ in adult patients with versus without CKD (18, 21, 22). For example, Chinnappa et al. (21) reported that $\dot{V}{O}_{2\text{peak}}$ was more strongly associated with arterio-venous O₂ difference than with cardiac output in patients with CKD. This finding contrasts with those observed in healthy control subjects and age-matched patients with heart failure, in whom $\dot{V}{O}_{2\text{peak}}$ was more strongly associated with cardiac output versus arterio-venous O₂ difference (21). In hemodialysis patients, the Δ[deoxy(Hb-Mb)]_peak response during exercise was shown to be lower than that observed in healthy control subjects, although it was not significantly associated with $\dot{V}{O}_{2\text{peak}}$ (25). Our finding in the present study of a positive association between $\dot{V}{O}_{2\text{peak}}$ and Δ[deoxy(Hb-Mb)]_peak in patients with CKD indicates that patients with CKD with a higher $\dot{V}{O}_{2\text{peak}}$ had the greatest change in Δ[deoxy(Hb-Mb)]_peak. However, it is not possible to establish causality from these data, and further investigations are required to determine if the increased $\dot{V}{O}_{2\text{peak}}$ was driven by greater change in Δ[deoxy(Hb-Mb)]_peak or vice versa, in patients with CKD.

The rate of skeletal muscle reoxygenation (i.e., ΔTSI_reoxy1/2time) following the termination of acute exercise has been used as a measure of skeletal muscle oxidative capacity (37). Factors influencing ΔTSI_reoxy1/2time include mechanisms governing adenosine triphosphate (ATP) synthesis, age, physical activity status, fitness level, and peripheral circulation (56–59). Several studies have documented delayed ΔTSI_reoxy1/2time following exercise in various disease populations versus healthy controls, including patients with CKD (29, 43, 60). For example, Wilkinson et al. (29) found that ΔTSI_reoxy1/2time was slower in patients with CKD compared to that in healthy controls following an overground incremental shuttle walk test, and that CKD patients with higher fitness levels had faster ΔTSI_reoxy1/2time responses. Our ΔTSI_reoxy1/2time results are in accordance with those reported by Wilkinson et al. (29) and demonstrated that ΔTSI_reoxy1/2time response in the CKD group was ~31% slower versus that in the REF group after graded treadmill exercise, suggesting that skeletal muscle oxidative capacity is compromised in Veterans with CKD.

The lower prevalence of CKD participants with a RQ of ≥1.00 (~61%), compared to that in the REF group (~83%), suggests that some CKD participants may have ended their exercise test for reasons other than reaching the maximal cardiorespiratory limit. Patients with CKD are reported to be more fatigable than their non-CKD counterparts which may partly explain the reduced exercise tolerance and RQ in the CKD versus REF group (28, 61–63). Although the CKD group had a lower RQ at peak exercise, $\dot{V}{O}_2$ relative to total exercise time was reduced in the CKD group compared to the REF group across the entire duration of the CPET (Fig. 2). Furthermore, the reduced $\dot{V}{O}_2$ at the GET in the CKD group represents a fitness marker independent of effort (64). Finally, the delayed recovery of ΔTSI following exercise termination (i.e., ΔTSI_reoxy1/2time) in the CKD group compared to REF group is considered to reflect reduced skeletal muscle oxidative capacity in those with CKD (37). Taken together, these data suggest that CKD participants exhaust their aerobic energy system more rapidly than do REF participants, and become reliant on anaerobic metabolism, even at low exercise intensities. Thus, the decreased exercise capacity in the CKD group may have resulted from accelerated accumulation of anaerobic metabolic by-products. Specifically, hydrogen ions (H⁺) and inorganic phosphate (P_i) are implicated in compromised cross-bridge formation and, in turn, prevent sufficient force production to maintain activity (65–67). Further studies are required to determine how, and the extent to which, fatigability and altered skeletal muscle bioenergetics contribute to $\dot{V}{O}_{2\text{peak}}$ in the CKD population.

Limitations.

Several study limitations must be acknowledged. First, the CKD group had a greater percentage of African American/Black participants, and those with type 2 diabetes and hypertension, compared with the REF group. Given the relatively small sample size, we are unable to discern if the findings of this study are the direct result of the pathogenesis of CKD, racial/ethnic differences, medication use, comorbidities, or a combination of these factors. Second, our sample included only older men and therefore findings are not generalizable to younger men or adult women of any age. Third, a symptom-limited CPET was performed due to the clinical nature of the investigation. Fourth, a physiologic calibration is recommended when using continuous-wave NIRS technology. However, this was not possible in the current study and thus caution is warranted when comparing the muscle oxygenation data from the present study to other studies. Finally, although the data provided herein offer new insights into the potential contributions of skeletal muscle oxygenation to $\dot{V}{O}_{2\text{peak}}$ in patients with CKD, the present study only explored a few of the many factors known to influence cardiorespiratory fitness. Therefore, further research is needed to elucidate the interactions of multiple other potential mechanisms that influence $\dot{V}{O}_{2\text{peak}}$, and the effects of various interventions designed to improve $\dot{V}{O}_{2\text{peak}}$.

CONCLUSIONS

The present findings suggest that skeletal muscle impairments, in addition to cardiovascular impairments, contribute to reduced $\dot{V}{O}_{2\text{peak}}$ in patients with CKD. Larger scale studies in more diverse populations of patients with moderate to severe CKD are needed to confirm the present findings. Further interventional research is also necessary to determine which exercise approaches are most efficacious at enhancing $\dot{V}{O}_{2\text{peak}}$ in this population and whether, and to what extent, improvements in $\dot{V}{O}_{2\text{peak}}$ result from changes in cardiovascular factors, skeletal muscle factors, or a combination of both.