Exercise pressor response and arterial baroreflex unloading during exercise in chronic kidney disease

Abstract

Patients with chronic kidney disease (CKD) have poor exercise capacity, which contributes to cardiovascular risk. We sought to determine whether patients with stage 2 or stage 3 CKD have an augmented blood pressure (BP) response during exercise, and if so, whether overactivation of the sympathetic nervous system (SNS) during exercise might play a role. In 13 patients with CKD and hypertension and 13 controls with hypertension, we measured hemodynamics and muscle sympathetic nerve activity (MSNA) during the following maneuvers: low-level rhythmic handgrip (RHG 20%), which primarily stimulates mechanoreceptors, and moderate static handgrip exercise (SHG 30%) followed by posthandgrip circulatory arrest (PHGCA), which isolates metaboreceptors. During baseline studies, patients with CKD had significantly greater increases in mean arterial pressure (MAP) during SHG 30% (P = 0.045), RHG 20% (P = 0.031), and PHGCA (P = 0.043); however, the MSNA response was not augmented in patients with CKD compared with controls. We hypothesized that an augmented SNS response during exercise might be revealed in CKD if arterial baroreflex constraint was equalized using nitroprusside (NTP). These exercise maneuvers were repeated in patients with CKD during NTP infusion to equalize the BP response between groups, thereby relieving baroreflex-mediated suppression of SNS activity. With NTP infusion, patients with CKD had significantly increased MSNA responses during SHG 30% (P = 0.0044), and RHG 20% (P = 0.0064), but not during PHGCA (P > 0.05), suggesting increased reflex activation of the SNS during exercise, which may be mediated by mechanoreceptors but not metaboreceptors. Patients with CKD have an exaggerated BP response during rhythmic and static exercise with underlying SNS overactivation that is revealed during arterial baroreflex unloading during exercise.

chronic renal failure, including mild to moderate chronic kidney disease (CKD stages 2 and 3) is characterized by poor physical capacity and exercise intolerance. The pathogenic mechanisms underlying exercise dysfunction in patients with chronic renal failure are multifactorial and not fully understood. The contribution of abnormal hemodynamic and neurocirculatory responses during exercise have not yet been explored. The normal physiologic responses during acute exercise include an increase in blood pressure (BP) and heart rate (HR) that serve to meet the increased metabolic demand of exercising skeletal muscle (12, 36). Conceivably, exaggerated increases in BP could contribute to exercise dysfunction and increase the likelihood of adverse cardiovascular events during physical activity.

One autonomic adjustment that mediates the BP response during exercise is activation of the sympathetic nervous system (SNS). SNS activity directed to nonworking tissues serves to redirect blood flow to exercising skeletal muscle and to prevent excessive metabolically driven vasodilation within exercising muscle itself. In humans, two major systems control the SNS response during exercise: 1) central command, which refers to a signal arising from within the central nervous system that is linked to the perceived effort of exercise and is important in increasing SNS outflow only at maximal or near-maximal effort (39); and 2) the muscle ergoreflex, mediated by groups of sensory nerve fibers within the skeletal muscle that send afferent signals to the central nervous system to increase central SNS outflow when stimulated during exercise (12, 34). These sensory nerve endings include the metaboreceptors, which are activated by ischemic metabolites generated during exercise, and mechanoreceptors, which are largely activated by mechanical stretch. In healthy humans, the muscle metaboreceptors are paramount in generating the reflex increases in SNS activity during static exercise (20). However, in disease states characterized by exercise intolerance such as chronic heart failure (CHF), abnormalities of the exercise pressor reflex are characterized by blunted metaboreceptor control (37) and exaggerated reflex activation of SNS activity mediated by the mechanoreceptors (25, 26).

Our prior work demonstrates that patients with end-stage renal disease (ESRD) have an exaggerated exercise pressor response during both static and rhythmic handgrip exercise compared with nonhypertensive controls (30). We found that muscle metaboreceptor activation of SNS activation was blunted, and the overall SNS response during static and rhythmic exercise was not augmented in ESRD, implicating that the exaggerated pressor response during exercise was not mediated by an exaggerated SNS response. However, because patients with ESRD have intact arterial baroreflex function (4, 13, 16), an exaggerated SNS response during exercise may have been masked by the augmented BP response via baroreflex-mediated dampening of SNS outflow. The purpose of this study was to determine: 1) whether patients with mild to moderate CKD have exaggerated increases in BP during rhythmic and static exercise; 2) whether this exaggerated pressor response is accompanied by and therefore potentially mediated by exaggerated increases in muscle sympathetic nerve activity (MSNA); and 3) which reflex mechanisms (e.g., muscle metaboreceptor, mechanoreceptor, central command) potentially mediate this augmented pressor response, MSNA response, or both.

METHODS

Study Population

The study population consisted of 26 participants (age range 39–65 years): 13 patients with hypertensive CKD and 13 age-matched controls without kidney disease but with hypertension. All study participants were male veterans who were recruited from clinics at the Atlanta Veterans Affairs (VA) Medical Center. All study patients had hypertension and took antihypertensive medications. None of the participants exercised regularly. Patients with CKD exhibited stage 3 as defined as an estimated glomerular filtration rate (eGFR) of 30–59 ml/min as calculated by the modified Modification of Diet in Renal Disease equation (15), or stage 2 CKD defined as an eGFR of 60–89 ml/min with microalbuminuria (defined as a urinary microalbumin-to-creatinine ratio greater than 30 mg per gram of creatinine). All participants had stable renal function defined as no greater than a 5% fluctuation in eGFR within the prior 3 mo. The etiology of renal disease in the patients with CKD included hypertension (5), chronic glomerulonephritis (1), cocaine (1), rhabdomyolysis (1), and unknown (5). Controls had normal renal function (eGFR >60 ml/min) and no microalbuminuria (urine microalbumin:creatinine <0.020). Exclusion criteria for all participants included illicit drug use within the past 12 mo and any major comorbid conditions including diabetes, neuropathy, vascular disease, uncontrolled anemia, or any clinical evidence of heart failure or heart disease determined by electrocardiogram, echocardiogram, stress test, history, or a combination of these. Two of the 13 patients with CKD smoked, and 3 of the 13 controls also smoked. All study participants abstained from smoking for at least 12 h prior to study procedures. This study was approved by the Emory University Institutional Review Board and the Atlanta VA Medical Center Research and Development Committee. Written, informed consent was obtained from each study participant.

Measurements and Procedures

Blood pressure.

Beat-to-beat arterial BP was measured continuously using a noninvasive monitoring device that detects digital blood flow via finger cuffs, and translates blood flow oscillations into continuous pulse pressure waveforms and beat-to-beat values of BP (CNAP, CNSystems). Absolute values of BP were internally calibrated using a concomitant upper arm BP reading, and were calibrated at the start and every 15 min throughout the study. The appropriate sizes of the finger cuffs were selected using a sizing chart, and the bladder of the upper arm cuffs encircled at least 80% of the upper arm. This device has been validated to reflect accurate absolute BP values and accurate beat-to-beat changes in BP as measured via an intra-arterial catheter (9, 11).

Muscle sympathetic nerve activity.

Multiunit postganglionic sympathetic nerve activity directed to muscle (MSNA) was recorded directly from the peroneal nerve by microneurography, as previously described (38, 42). A tungsten microelectrode (tip diameter 5–15 um; Department of Biomedical Engineering, University of Iowa, Iowa City, IA) was inserted into the nerve, and a reference microelectrode was inserted subcutaneously 1–2 cm from the recording electrode. The signals were amplified (total gain 50,000–100,000), filtered (700–2,000 Hz), rectified, and integrated (time constant 0.1 sec) to obtain a mean voltage display of sympathetic nerve activity (Nerve Traffic Analyzer Model 662C-4, University of Iowa) that was recorded by the LabChart 7 Program (PowerLab 16sp, ADInstruments). Lead II of the electrocardiogram (ECG) was recorded simultaneously with the neurogram. All MSNA recordings met previously established criteria (5, 6, 19). Sympathetic bursts were identified by visual inspection of nerve bursts by a single investigator without knowledge of the participant's status as control or patient. MSNA was expressed as burst frequency (bursts per minute) and total activity (arbitrary units per minute). Total activity was determined as the product of the mean burst amplitude and burst frequency per minute. Change in MSNA total activity was quantified as a percent change from baseline total activity, and absolute change in total activity when the largest burst at resting conditions was assigned a value of 1,000 units.

Moderate static handgrip exercise (SHG 30%).

Moderate static handgrip exercise elicits an increase in MSNA (20) by activating metaboreceptors, mechanoreceptors, and central command. Participants were asked to squeeze a hand dynamometer (Stoelting) with maximal force. The highest force attained from three attempts was considered the maximum voluntary contraction (MVC). Moderate static handgrip exercise was performed by squeezing the hand dynamometer at 30% of MVC (SHG 30%) in a sustained manner for 3 min. Participants were instructed to avoid inadvertent valsalva, and to maintain normal breathing patterns.

Post handgrip circulatory arrest (PHGCA).

This maneuver was performed immediately after SHG 30%. It traps ischemic metabolites within the forearm and thereby isolates muscle metaboreceptors from mechanoreceptors and central command (27). Five seconds before the end of the 3-min SHG exercise, an upper-arm blood pressure cuff was inflated to suprasystolic levels (220 mmHg) proximal to the exercising forearm for 2 min to trap the metabolites within the forearm. Participants remained relaxed during this maneuver, eliminating any contribution from mechanoreceptors or central command.

Rhythmic handgrip exercise (RHG 20%).

This low-level rhythmic exercise engages muscle mechanoreceptors and central command without engagement of muscle metaboreceptors (1). Participants squeezed the hand dynamometer intermittently at 20% of MVC for 3 min at a rate of one contraction per 2 s.

Sodium nitroprusside (NTP).

NTP was infused intravenously during handgrip exercises in participants with CKD. The initial dose of 0.3 μg·kg·min was titrated to decrease mean arterial pressure by ∼5–10 mmHg (28), equalizing the BP response between patients with CKD and controls. NTP was infused during static handgrip exercise, posthandgrip circulatory arrest, and rhythmic handgrip exercise to attenuate the exaggerated increase in BP during these maneuvers, and thereby unload baroreflex suppression of MSNA in patients with CKD.

D5W.

D5W was infused intravenously during handgrip exercises in control participants and those with CKD. The purpose of the D5W infusion was to provide a control condition to NTP infusion and ensure that all exercise experiments included placement of an intravenous catheter with infusion, providing uniformity to the studies. D5W was infused at a rate of 1 ml/min during handgrip exercises.

Laboratory measures.

Venous blood samples were drawn and sent to the hospital laboratory for measurement of a basic metabolic panel (BMP: calcium, glucose, sodium, potassium, bicarbonate, chloride, blood urea nitrogen, and creatinine levels). Urine samples were examined for urine albumin and creatinine concentrations to quantify the degree of albuminuria.

Experimental Protocol

All participants were studied in the early morning after abstaining from food, caffeine, and alcohol for at least 12 h and exercise for at least 24 h. The study room was quiet, semidark, and temperate (∼21°C). Blood was drawn for BMP, and a urine sample was collected for to determine urine microalbumin and creatinine concentrations. Participants were placed in a supine position on a comfortable stretcher. A 20-gauge iv catheter was placed into the antecubital vein of the dominant (nonexercising) arm.

Baseline measures.

Finger cuffs were fitted and placed on the fingers of the dominant arm for continuous beat-to-beat arterial BP measurements, and an upper-arm cuff was placed for intermittent automatic calibrations with the finger cuffs. ECG patch electrodes were placed on participants to continuously record HR. Participants' legs were positioned for microneurography, and the tungsten microelectrode was inserted and manipulated to obtain a satisfactory nerve recording. After 10 min of rest, baseline BP, HR, and MSNA were recorded continuously for 10 min.

Exercise measures.

Participants performed the following two maneuvers in random order: 1) RHG 20% for 3 min and 2) SHG 30% for 3 min, followed by PHGCA for 2 min. Each exercise task was performed in the nondominant arm. MSNA, BP, and HR were recorded continuously. Thirty minutes of recovery time was given between the exercise tasks, allowing sufficient time for BP, HR, and MSNA to return to baseline levels before each new intervention. All patients with CKD performed the maneuvers once during the iv infusion of D5W (at 1 ml/min), and once during infusion of NTP (dose 0.3 μg·kg·min). Forty-five minutes of rest was given between the D5W and NTP infusions to ensure sufficient washout of drugs. NTP and D5W were infused only during the exercise maneuvers, and not during the 30-min rest periods. The dose of NTP was titrated to normalize the BP response during each maneuver in patients with CKD with that of controls [i.e., a reduction of mean arterial pressure (MAP) by ∼5–10 mmHg during each maneuver]. Control participants did not receive an NTP infusion and performed the exercise tasks during D5W infusion only.

Data Analysis

Statistical analysis was performed using the SAS 9.2 program (SAS Institutes). Baseline characteristics were compared using independent two-tailed t-tests. Two-way ANOVA with repeated measures was performed using PROC GLM to determine differences between groups (patients with CKD vs. controls) with respect to percent change from baseline and absolute change in MSNA, systolic BP (SBP), diastolic BP (DBP), MAP, and HR with time during each intervention: static handgrip, rhythmic handgrip, and posthandgrip circulatory arrest. When the overall F-test was significant, the contrast option for post hoc analysis was used to compare the groups for change from baseline for each time point. Results were expressed as means ± SE. A value of P < 0.05 was considered statistically significant.

RESULTS

Baseline Characteristics

Patients with CKD and hypertensive controls without kidney disease were well matched for age, body mass index, and baseline hemodynamics (P > 0.05) (see Table 1). All participants from both groups were men, had hypertension, and took antihypertensive medications. Among the 13 controls, 6 were treated with diuretics, 5 with an angiotensin-converting enzyme inhibitor (ACE-I) or angiotensin receptor blocker (ARB), 5 with an dihydropyridine calcium channel blocker (DCCB), 1 with a nondihydropyridine calcium channel blocker (NCCB), 2 with an alpha blocker, and 2 with an aldosterone receptor blocker. Among the 13 patients with CKD, 8 were treated with a diuretic, 11 with an ACE-I or ARB, 5 with a DCCB, 3 with a beta blocker, and 2 with an alpha blocker. There was no significant difference in baseline SBP, DBP, MAP, and HR between the two groups (P > 0.05). Five of the 3 controls and 4 of the 13 patients with CKD had hyperlipidemia treated with HMG-CoA reductase inhibitors.

Table 1.

Baseline demographics, hemodynamics, and sympathetic activity

	Controls	Patients with CKD	P value
N	13	13
Age, years	52.5 ± 2.1	54.5 ± 2.3	0.52
Body mass index, kg/m2	25.6 ± 1.2	25.1 ± 1.0	0.72
Systolic BP, mmHg	134.9 ± 2.9	132.5 ± 4.4	0.65
Diastolic BP, mmHg	77.5 ± 2.0	83.6 ± 2.4	0.06
Mean arterial pressure, mmHg	96.6 ± 2.0	99.9 ± 2.8	0.36
Heart rate, beats/min	63.6 ± 2.9	63.6 ± 3.4	1.0
Serum creatinine, mg/dl	1.0 ± 0.03	1.7 ± 0.10	< 0.0001*
eGFR, mL·min·1.73m2	92.2 ± 3.7	53.6 ± 5.0	< 0.0001*
Urine albumin:creatinine, mg/g	8.8 ± 3.4	253.2 ± 29.2	0.0094*
Maximum voluntary contraction, kg	44.3 ± 1.7	43.5 ± 1.2	0.70
MSNA, bursts/minute	30.6 ± 5.5	38.1 ± 2.2	0.22

Baseline measurements in control participants and patients with chronic kidney disease (CKD). BP, blood pressure; MSNA, muscle sympathetic nerve activity. Estimated glomerular filtration rate (eGFR) was quantified using the four-variable modified Modified Diet in Renal Disease equation. Data are expressed as means ± SE.

^*Significant at P <0.05.

Ten of the 13 patients with CKD had stage 3, and 3 had stage 2. All patients with stage 2 CKD and 6 of the patients with stage 3 CKD had microalbuminuria. As expected, the mean serum creatinine and baseline urine albumin-to-creatinine ratio was significantly higher in patients with CKD. Grip strength quantified as MVC was similar between patients with CKD and controls, and there was no significant difference in baseline MSNA between the two groups.

SHG 30%

The purpose of this study was to determine whether pressor responses during moderate exercise were increased in patients with CKD compared with hypertensive controls, and if so, whether this was accompanied by increased MSNA responses. During SHG 30%, muscle mechanoreceptor, metaboreceptor, and central command are all engaged.

During 3 min of SHG 30% MVC, there was a significantly greater relative percent increase from baseline in SBP in patients with CKD (CKD+D5W) compared with controls (CON+D5W) (ANOVA F-test, P = 0.040) (Fig. 1A) and a trend toward greater absolute (Δ) change in SBP among patients with CKD (ANOVA F-test, P = 0.051) (Fig. 1B). Similarly, there was a significantly greater percent increase (ANOVA F-test, P = 0.045) (Fig. 1E) and Δ increase (ANOVA F-test, P = 0.049) (Fig. 1F) in MAP during D5W infusion in the CKD+D5W group compared with the CON+D5W group. There was no statistically significant difference in change in DBP or HR during D5W infusion between the two groups (Fig. 1, C and D). Infusing NTP in the same patients with CKD (CKD+NTP) during SHG 30% eliminated the difference in MAP and SBP responses between the two groups and reduced BP in patients with CKD to that of controls during exercise. There was no significant difference in MAP and SBP responses during SHG 30% when comparing the CON+D5W group with the CKD+NTP group. Similarly, there was no significant difference in DBP or HR responses between the CON+D5W and CKD+NTP groups.

Fig. 1.

Hemodynamic changes during static handgrip exercise. Percent (%) change and absolute (Δ) change from baseline (BL) levels in systolic blood pressure (SBP), diastolic blood pressure (DBP), mean arterial pressure (MAP), and heart rate (HR) during each minute (M) of static handgrip exercise (SHG 30%) in controls during D5W infusion (CON+D5W, open circles), patients with chronic kidney disease (CKD) during D5W infusion (CKD+D5W, filled circles), and patients with CKD during nitroprusside infusion (CKD+NTP, filled circles with dotted lines). Values are expressed as means ± SE. †Indicates overall ANOVA F-test was significant for a difference between CON+D5W vs. CKD+D5W groups (see RESULTS). *Indicates P < 0.05 in the difference between CON+D5W vs. CKD+D5W groups at that time point.

During D5W infusion, there was no significant difference in percent or Δ change from baseline in MSNA quantified as bursts/minute and total activity during SHG 30% between controls and patients with CKD (ANOVA F-test, P = NS) (Fig. 2, A–D). However, when the BP response was equalized in patients with CKD with NTP infusion to that of controls, patients with CKD had a significantly augmented change in MSNA during SHG 30% when quantified as a percent change in total activity (ANOVA F-test, P = 0.004) (Fig. 2A) and a Δ change in total activity (ANOVA F-test, P = 0.007) (Fig. 2B). Results were similar when MSNA was quantified as a percent change in bursts/minute (ANOVA F-test, P = 0.005) (Fig. 2C) and also a Δ change in MSNA in bursts/minute (ANOVA F-test, P < 0.001) (Fig. 2D). The percent change in MSNA was significantly higher during the first and third minutes of SHG 30% exercise when MSNA was quantified as total activity, and was significantly higher during the second and third minutes of exercise when quantified as bursts/minute (Fig. 2, A and C). The Δ change in MSNA was significantly higher during each minute of exercise when quantified as total activity and bursts/minute (Fig. 2, B and D). The total MSNA in bursts/minute was significantly higher during each minute of SHG 30% in the CKD+NTP group compared with the CON+D5W group (Fig. 3A).

Fig. 2.

Change in MSNA during SHG 30%, RHG 20%, and PHGCA. Change from baseline (BL) in muscle sympathetic nerve activity (MSNA) during each minute (M) of rhythmic (RHG 20%) handgrip exercise, static (SHG 30%) handgrip exercise, and posthandgrip circulatory arrest (PHGCA) in controls during D5W infusion (CON+D5W, open circles), patients with chronic kidney disease (CKD) during D5W infusion (CKD+D5W, filled circles), and patients with CKD during nitroprusside infusion (CKD+NTP, filled circles with dotted lines). MSNA was quantified as bursts/minute and total activity (units/min). Relative (%) change in MSNA, and absolute (Δ) change in MSNA were quantified for both total activity and burst frequency in bursts/minute. Values are expressed as means ± SE. †Indicates ANOVA F-test was significant for a difference between CON+D5W vs. CKD+D5W groups (see RESULTS). *Indicates P < 0.05 in the difference between CON+D5W vs. CKD+D5W groups at that time point. **Indicates P < 0.05 in the difference between CON+D5W vs. CKD+NTP groups at that time point. ‡Indicates ANOVA F-test was significant for a difference between CON+D5W vs. CKD+NTP. In I–L, P values are shown for CON+D5W vs. (VS) both CKD+D5W and CKD+NTP.

Fig. 3.

Absolute levels of MSNA during static and rhythmic handgrip exercise. Absolute levels of muscle sympathetic nerve activity (MSNA) quantified as bursts/minute at baseline (BL) and during each minute (M) of static (SHG 30%) and rhythmic (RHG 20%) handgrip exercise. The overall ANOVA F-test was significant for a difference between controls during D5W infusion (CON+D5W) compared with patients with chronic kidney disease (CKD) during nitroprusside infusion (CKD+NTP) (see RESULTS). *P < 0.05 in the difference between the two groups at that time point.

RHG 20%

The purpose of this study of low-level rhythmic handgrip in which minimal ischemic metabolites are generated, was principally to engage the muscle mechanoreceptors. During 3 min of RHG 20% MVC, there was no statistically significant difference in percent or Δ change in SBP, DBP, or HR during D5W infusion between patients with CKD (CKD+D5W) and controls (CON+D5W) (Fig. 4, A–D, G, and H). There was a significantly greater percent increase (ANOVA F-test, P = 0.031) (Fig. 4E) and Δ increase (ANOVA F-test, P = 0.031) (Fig. 4F) in MAP during D5W infusion in the CKD+D5W group compared with the CON+D5W group. Infusing NTP in the same patients with CKD (CKD+NTP) during RHG 20% eliminated the difference in MAP response between the groups and reduced BP in patients with CKD to that of controls during exercise; there was no significant difference in MAP response during RHG 20% between the CON+D5W and CKD+NTP groups. Similarly, there was no significant difference in SBP, DBP, or HR response between the CON+D5W and CKD+NTP groups.

Fig. 4.

Hemodynamic changes during rhythmic handgrip exercise. Percent (%) change and absolute (Δ) change from baseline (BL) levels in systolic blood pressure (SBP), diastolic blood pressure (DBP), mean arterial pressure (MAP), and heart rate (HR) during each minute (M) of rhythmic handgrip exercise (RHG 20%) in controls during D5W infusion (CON+D5W, open circles), patients with chronic kidney disease (CKD) during D5W infusion (CKD+D5W, filled circles), and patients with CKD during nitroprusside infusion (CKD+NTP, filled circles with dotted lines). Values are expressed as means ± SE. *Indicates P < 0.05 in the difference between CON+D5W vs. CKD+D5W groups at that time point. †Indicates overall ANOVA F-test was significant for a difference between CON+D5W vs. CKD+D5W groups (see RESULTS).

During D5W infusion, there was no significant difference in percent or Δ change from baseline in MSNA quantified as bursts/minute and total activity during SHG 30% between controls and patients with CKD (ANOVA F-test, P = NS) (Fig. 2, E–H). However, when the BP response in patients with CKD was equalized with NTP infusion to that of controls, patients with CKD had a significantly augmented change in MSNA during RHG 20% when quantified as a percent change in total activity (ANOVA F-test, P = 0.006) (Fig. 2E) and a Δ change in total activity (ANOVA F-test, P < 0.001) (Fig. 2F). Results were similar when MSNA was quantified as a percent change in bursts/minute (ANOVA F-test, P = 0.006) (Fig. 2G) and a Δ change in MSNA in bursts/minute (ANOVA F-test, P = 0.002) (Fig. 2H). The percent change and Δ change in MSNA quantified as both bursts/minute and total activity were significantly higher during each minute of RHG 20% in the CKD+NTP group compared with the CON+D5W group (Fig. 2, E–H). The absolute MSNA in bursts/minute was significantly higher during the second and third minutes of RHG 20% in the CKD+NTP group compared with the CON+D5W group (Fig. 3B).

PHGCA

The purpose of this study was to engage only the muscle metaboreceptors because the muscle is no longer contracting, and therefore, the muscle mechanoreceptors and central command are no longer engaged.

During 2 min of PHGCA that immediately followed 3 min of SHG 30%, there was a significantly greater percent and Δ increase from baseline in SBP during D5W infusion in patients with CKD (CKD+D5W) compared with controls (CON+D5W) (ANOVA F-test, P < 0.05) (Fig. 5, A and B). Similarly, there was a significantly greater percent increase in MAP during PHGCA in the CKD+D5W group compared with the CON+D5W group (ANOVA F-test, P < 0.05) (Fig. 5, E and F). There was no statistically significant difference in percent or Δ change in DBP or HR during D5W infusion between the two groups (Fig. 2, C, D, G, and H). Infusing NTP in the same patients with CKD (CKD+NTP) during PHGCA eliminated the difference in MAP and SBP responses between the two groups and reduced BP in patients with CKD to that of controls during PHGCA. There was no significant difference in MAP and SBP responses during PHGCA when comparing the CON+D5W group with the CKD+NTP group. Similarly, there was no significant difference in DBP response between the CON+D5W and CKD+NTP groups. There was a significant difference in percent and Δ change in HR during PHGCA between the CON+D5W and CKD+NTP groups (ANOVA F-test, P < 0.05); HR was significantly higher in the CKD+NTP group during the first minute of PHGCA (Fig. 5, G and H).

Fig. 5.

Hemodynamic changes during posthandgrip circulatory arrest. Percent (%) change and absolute (Δ) change from baseline (BL) levels in systolic blood pressure (SBP), diastolic blood pressure (DBP), mean arterial pressure (MAP), and heart rate (HR) during each minute (M) of posthandgrip circulatory arrest (PHGCA) in controls during D5W infusion (CON+D5W, open circles), patients with chronic kidney disease (CKD) during D5W infusion (CKD+D5W, filled circles), and patients with CKD during nitroprusside infusion (CKD+NTP, filled circles with dotted lines). Values are expressed as means ± SE. *Indicates P < 0.05 in the difference between CON+D5W vs. CKD+D5W groups at that time point. **Indicates P < 0.05 in the difference between the CON+D5W vs. CKD+NTP groups at that time point. †Indicates ANOVA F-test was significant in the difference between CON+D5W vs. CKD+D5W groups (see RESULTS). ‡Indicates ANOVA F-test was significant in the difference between CON+D5W vs. CKD+NTP.

During D5W infusion, the MSNA response to PHGCA was blunted in the CKD+D5W group compared with the CON+D5W group when quantified as a percent change in total activity (ANOVA F-test, P = 0.009) (Fig. 2I) and a Δ change in total activity (ANOVA F-test, P = 0.006) (Fig. 2J). The percent change in MSNA total activity during the first minute of PHGCA was significantly lower in the CKD+D5W group than the CON+D5W group, and was significantly lower during both minutes of PHGCA when quantified as a Δ change in total activity. However, equalizing the BP response in patients with CKD with NTP infusion to that of controls revealed that the MSNA response during PHGCA was not blunted in patients with CKD. NTP infusion revealed that the MSNA response during PHGCA was similar between patients with CKD and controls. MSNA response during the first and second minutes of PHGCA were similar between the CON+D5W and CKD+NTP groups when quantified as both a percent change and a Δ change in MSNA total activity (Fig. 2, I and J). The overall ANOVA F-test was significant for a difference in pattern of MSNA response during PHGCA between the CON+D5W and CKD+NTP groups when quantified as a Δ change in MSNA total activity (Fig. 2J), which is explained by the observation that the MSNA response increased during the second minute of PHGCA compared with the first minute in the CON+D5W group, but decreased during the second minute compared with the first minute in the CKD+NTP group. However, there was no difference in the absolute value of MSNA response during each minute of PHGCA between the CON+D5W and CKD+NTP groups.

Similarly, during D5W infusion, there was no significant difference in percent change in MSNA quantified as bursts/minute during PHGCA between controls and patients with CKD (Fig. 2K); however, there was a trend toward a blunted MSNA response in the CKD+D5W group compared with the CON+D5W group (ANOVA F-test, P = 0.110), which was eliminated during NTP infusion (CON+D5W vs. CKD+NTP, P = 0.899). When MSNA was quantified as an absolute change in bursts/minute, there was no significant difference in Δ change in MSNA between the CON+D5W group vs. the CKD+D5W group (ANOVA F-test, P = 0.866) (Fig. 2L). During NTP infusion, there was a trend toward a greater Δ change in MSNA in the CKD+NTP group vs. the CON+D5W group, which did not reach statistical significance (ANOVA F-test, P = 0.085).

DISCUSSION

The major new findings of this study follow. 1) Patients with stage II and stage III CKD have an exaggerated BP response during low-level rhythmic and moderate static handgrip exercise compared with hypertensive controls without CKD, and 2) despite exaggerated BP responses in patients with CKD during exercise, during the D5W infusion (i.e., without manipulation of the BP response), patients with CKD and controls have a similar MSNA response during low-level rhythmic and moderate static handgrip exercise. Equalizing the BP responses between the two groups by infusing NTP in the patients with CKD during exercise, thereby equalizing baroreflex-mediated restraint of SNS activity to that of controls, revealed an augmented SNS response to exercise in patients with CKD. Specifically, 3) patients with CKD had significantly augmented relative and absolute increases in MSNA during static exercise; and 4) patients with CKD had significantly augmented relative and absolute increases in MSNA during rhythmic exercise, consistent with augmented muscle mechanoreceptor activation. Finally, 5) this exaggerated increase in MSNA during static and rhythmic exercise was not mediated by the muscle metaboreflex because MSNA was not elevated during PHGCA (muscle metaboreceptor isolation).

We observed that patients with stage 2 and stage 3 CKD had exaggerated pressor responses during both low-level rhythmic and moderate static handgrip exercise compared with hypertensive controls without renal disease. Patients with CKD had significantly greater increases in MAP during RHG 20% and significantly greater increases in MAP and SBP during SHG 30%. Conceivably, augmented increases in BP during exercise could contribute to exercise intolerance in CKD by increasing the workload of the heart against an elevated peripheral resistance, and contribute to the pathophysiology of uremic myopathy. Furthermore, augmented increases in BP may contribute to the increased risk of cardiovascular disease and sudden death that characterizes patients with reduced renal function (23). Exaggerated pressor responses during exercise have been shown to correlate with an increased risk for cardiovascular disease (10, 35) and could increase the risk of adverse cardiovascular events during physical activity.

Our previous work demonstrates that patients with ESRD have exaggerated pressor responses during rhythmic and static handgrip exercise (30). Quite strikingly, in the current study, we found that the exaggerated exercise pressor response exists even in patients with far less advanced renal dysfunction not on dialysis. In fact, all patients in this study had stage 2 or early stage 3 CKD, with a mean serum creatinine level of 1.7 mg/dl and an eGFR of 54 ml/min, suggesting that hemodynamic mechanisms of exercise dysfunction begin quite early in the course of renal disease. Similarly, CKD is a strong and independent risk factor for cardiovascular disease, even with minimal reductions in renal function and microalbuminuria (23). Exaggerated pressor and sympathetic neural responses during physical activity could potentially play a role in increasing cardiovascular risk in early CKD. In addition, the previous report of patients with ESRD compared patients undergoing dialysis with nonhypertensive controls. A strength of the current study is the comparison of patients with CKD and hypertension with well-matched hypertensive controls, isolating reduced renal function, rather than comorbid conditions or the dialysis procedure, as the variable associated with the abnormal exercise pressor response.

We next sought to determine whether the augmented pressor response was accompanied by an exaggerated MSNA response. The BP response during exercise is mediated by a balance between vasoconstrictive and vasodilatory forces induced during exercise. The major vasoconstrictive force is reflex activation of the SNS (12, 34). We found that while the BP response was exaggerated during static and rhythmic handgrip exercise, the MSNA response was not different between the two groups (i.e., the same degree of sympathetic output produced a greater pressor response in patients with CKD). This finding suggests that other factors besides SNS overactivation contribute to the overall exaggerated exercise pressor response in CKD. One potential mechanism underlying this difference in relationship between MSNA and BP response between the groups include elaboration of other vasoconstrictors that directly constrict blood vessels or enhance the vasoconstrictive action of norepinephrine released at sympathetic nerve endings during exercise, such as endothelin-1 (18). Endothelin-1 not only acts directly on human blood vessels as a potent vasoconstrictor, it also potentiates the vasoconstrictive action of norepinephrine by sensitizing vascular smooth muscle cells to extracellular calcium (43). Endothelin-1 production increases substantially during exercise both systemically (17) and within nonworking skeletal muscle (18), and is known to be elevated at baseline in patients with CKD. Other potential mechanisms underlying the altered relationship between SNS output and BP during exercise include alterations in the production of concomitant vasodilators generated during exercise such as NO, adenosine (21), differential generation of oxidative stress, or impairment of functional sympatholysis (32) in which SNS-mediated vasoconstrictor responses are blocked locally during exercise by blocking the action of norepinephrine at sympathetic nerve endings. It has been shown that functional sympatholysis is impaired in human hypertension (41), another condition characterized by an exaggerated exercise pressor response. Whether endothelin-1 or other vasoconstrictors, oxidative stress, or impaired functional sympatholysis contribute to the exaggerated exercise pressor response in CKD should be evaluated in future studies.

After establishing that the same intensity of static and rhythmic exercise produced similar MSNA responses but greater pressor responses in patients with CKD, we next sought to determine how the MSNA response might change if the BP responses were equalized between the groups. Although SNS overactivation cannot explain the exaggerated exercise pressor response in CKD given the difference in relationship between MSNA and BP during baseline studies, we could not wholly exclude SNS overactivation during exercise in CKD as a contributing factor that may have been masked by baroreflex-mediated dampening of the SNS response in the setting of an augmented BP response. Patients with chronic renal failure have intact arterial baroreflex sensitivity (4, 13, 16), and arterial baroreflex buffering of MSNA remains intact during exercise (33). Therefore, we hypothesized that ameliorating the augmented BP response, thereby unloading arterial baroreflexes to that of controls, would unmask an exaggerated MSNA response during low-level rhythmic and moderate static handgrip exercise. We further hypothesized that equalizing baroreflex restraint would normalize the blunted MSNA response during metaboreceptor isolation in CKD.

To test the hypothesis that a heightened MSNA response during exercise would be revealed when baroreflex restraint was equalized between the groups, the BP response was normalized to that of controls during exercise by administering NTP to patients with CKD while controls received D5W infusion during exercise. The rationale for these experiments include the following: first, because patients with CKD are known to have intact arterial baroreflex function [with suppression of MSNA when BP is increased with phenylephrine (16)] and baroreflex restraint of SNS activity remains intact during exercise (33), we expected greater suppression of SNS activity during exercise in the setting of an augmented BP response if there was no underlying SNS overactivity (i.e., MSNA should have appeared blunted in patients with CKD). However, our observation that MSNA responses during exercise were not blunted in CKD in the setting of an augmented pressor response and were in fact equivalent to that of controls, suggested that SNS overactivity may play a contributing role in the exaggerated pressor response that was masked by baroreflex-mediated suppression. Second, MSNA during exercise is capable of being suppressed by arterial baroreflexes during exercise in CKD, as evidenced by the profound suppression of MSNA during PHGCA in patients with CKD in the setting of an exaggerated pressor response. During isolation of the muscle metaboreflex activation of MSNA with PHGCA there was an exaggerated BP response but a blunted MSNA response in patients with CKD. When NTP was given to equalize the BP response in patients with CKD there were equivalent increases in MSNA during PHGCA between the groups, suggesting that heightened SNS activity did not contribute to the exaggerated BP response during metaboreflex isolation. Conversely, the MSNA response was not blunted during rhythmic and static handgrip despite the exaggerated BP response, suggesting that the SNS response was not appropriately suppressed and that a heightened MSNA response may be revealed with equalization of baroreflex restraint. Third, in related studies the central sympathomimetic effects of other factors that increase BP both peripherally and centrally, such as cigarette smoking and angiotensin II (ANG II), were not revealed until baroreflex restraint of SNS activity was equalized using NTP. For instance, smoking acutely increases BP, and initial studies reported blunted SNS response to cigarette smoking (7); however, the sympathoexcitatory effects of smoking were revealed when the pressor effect of smoking was attenuated by administration of NTP (28). Similarly, both ANG II and phenylephrine infusion increased BP with concomitant decreases in MSNA (22); however, when NTP was infused simultaneously to maintain BP at baseline levels, MSNA increased significantly during ANG II infusion, but not during phenylephrine (22). Thus, the heightened MSNA response during smoking and ANG II were not revealed until BP responses were normalized using NTP.

In the current study, when the exaggerated BP response during static and rhythmic exercise in patients with CKD was ameliorated with NTP infusion to equal that of controls, thereby equalizing baroreflex-mediated restraint of SNS activity, patients with CKD had significantly augmented relative and absolute increases in MSNA during SHG 30% and RHG 20%. These findings suggest that: a) in patients with CKD, SNS activity is suppressed by arterial baroreflexes during rhythmic and static exercise, as evidenced by the large gradation in MSNA responses between NTP and control (D5W) conditions; and b) this suppression of MSNA is only partial because MSNA levels remain similar to those of controls, despite a significantly augmented BP response during exercise. In summary, SNS activity remains inappropriately high in CKD in the setting of augmented BP responses during exercise and may therefore be a contributory factor in the exaggerated exercise pressor response in these patients. SNS overactivation may contribute to increased risk of adverse cardiovascular events during exercise by increasing arrhythmogenicity (40) and contribute to exercise dysfunction in CKD by increasing vascular oxidative stress (3). New and existing therapeutic strategies alike to reduce exaggerated SNS activity or to blunt its effects in CKD may be an important direction for future clinical studies.

Moderate static handgrip exercise (SHG 30%) evokes reflex activation of SNS activity by engaging both muscle mechanoreceptors and metaboreceptors, as well as central command (20), whereas SNS activation during low-level rhythmic handgrip exercise (RHG 20%) is primarily mediated via muscle mechanoreceptors (1). A previous study using 31P magnetic spectroscopy showed that static exercise produced muscle acidosis and increases in H₂PO₄, whereas prolonged low-level rhythmic handgrip exercise resulted in no muscle metabolite production (1, 8). Furthermore, MSNA increased during rhythmic handgrip exercise but fell to baseline levels during PHGCA, further suggesting that metaboreceptors are not engaged during low-level RHG. Although central command is engaged during RHG 20%, central command contributes minimally to SNS activation during low-intensity rhythmic handgrip exercise and is only important for increases in SNS activity during high-intensity exercise (39). Thus our findings that MSNA responses are exaggerated during SHG 30% and RHG 20% but not during PHGCA, during which metaboreceptors are isolated, suggests that the augmented SNS response during exercise that was revealed during baroreflex unloading is not mediated by enhanced metaboreceptors, but may be potentially mediated by enhanced muscle mechanoreflex control in CKD.

Of note, in the previous study by Batman and colleagues (1) in which young healthy men engaged in prolonged low-level rhythmic handgrip exercise, MSNA did not increase until 6 min into exercise. In contrast, in the current study, we observed that patients with CKD during NTP and D5W infusion, as well as controls, had an immediate increase in MSNA during short-term rhythmic handgrip. Patients with CKD during baroreflex unloading had substantial and significant increases in MSNA of 25–37% during the first 3 min of RHG 20%, and controls also had significantly smaller yet substantial increases in MSNA of 12% within the first minute of RHG 20%. The reasons for the discrepancy between our observations, even among control participants, and the prior study by Batman in which increases in MSNA were not noted until much later in the course of exercise are unclear, but potentially due to differences in participant characteristics. The previous study included young and healthy individuals, whereas the controls in the current study were older and had hypertension. It may be possible that muscle mechanoreflex control of MSNA may be altered in human hypertension and chronic kidney disease. Patients with chronic heart failure have been shown to have significantly increased MSNA responses during short-term rhythmic handgrip exercise, likely due to enhanced muscle mechanoreflex sensitization by ischemic metabolites such as cyclooxygenase products (24, 25). Whether cyclooxygenase products or other ischemic metabolites, and uremic toxins in the case of chronic kidney disease, sensitize muscle mechanoreceptors in patients with hypertension and those with CKD remain to be tested.

Previous investigations of the exercise pressor reflex in other patient populations at increased cardiovascular risk, such as those with hypertension (31), obesity (29), and chronic heart failure (37), have revealed that these groups have blunted metaboreflex control. Importantly, the current study, in which BP responses were normalized, revealed that patients with CKD may have an overall augmented SNS response to static and rhythmic exercise that was unmasked when arterial baroreceptors were unloaded to the same degree as controls, but no evidence of a heightened SNS response during muscle metaboreflex isolation. The mechanisms underlying the altered relationship between the pressor and MSNA response during metaboreceptor isolation in CKD are unclear, but may include a delayed recovery of the heightened BP response following moderate static handgrip exercise. These and other mechanisms underlying the exaggerated pressor response during exercise in CKD should be investigated in future studies.

Limitations

One limitation of the current study is that the potential role of central command in the exaggerated exercise pressor reflex was not studied. Isolating the effect of central command in humans is difficult but can be done by having participants attempt to perform exercise while rendering them incapable of forceful movement through local paralysis with curare (39). In addition, although central command contributes minimally to the exercise pressor reflex during RHG 20%, other methods to isolate mechanoreceptors from central command such as passive hand movement or involuntary muscle contraction elicited by a neuromuscular stimulator (26), were not performed. Second, one patient had CKD attributable to rhabdomyolysis; this occurred more than 5 years ago, and presumably there were no residual muscular effects. Nevertheless, we analyzed data without this individual, and the results were unchanged. Third, although the CKD and control groups were well matched for baseline BP and hypertensive status, more patients with CKD were treated with an ACE-I and ARB. However, ACE-I and ARB are medications are known to decrease central SNS outflow in patients with chronic renal failure (14, 16); despite the sympatholytic effect of these agents, we observed augmented increases in MSNA during exercise and NTP infusion in patients with CKD. When data were analyzed in the subgroup uniformly treated with ACE-I and ARB medications by excluding participants from both groups not treated with these agents, results were similar: MSNA responses during rhythmic and static exercise were significantly higher in patients with CKD during NTP infusion compared with controls. Fourth, experimental studies have shown that dietary salt has a sympathoexcitatory effect (2), but sodium intake was not monitored in this study. However, all participants were studied after a 12-h fast, eliminating variation in sodium intake during the prestudy time frame. Finally, all study participants were male veterans without diabetes, vascular disease, or other comorbidities to isolate the effect of renal dysfunction on the exercise pressor reflex. However, it may not be possible to generalize our findings to a more typical CKD population in which such comorbid conditions are quite common, or to women.

GRANTS

This work was supported in part by the Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development, Clinical Studies Center, Decatur, GA; by the Atlanta Research and Education Foundation; and by ncrr Public Health Service Grant UL1-RR025008 through the Clinical and Translational Science Award program. J. Park was supported by National Institutes of Health Grant K23 HL098744 and the Amgen Nephrology Junior Faculty Award.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: J.P., A.A.Q., and H.R.M. conception and design of research; J.P. performed experiments; J.P. analyzed data; J.P. and H.R.M. interpreted results of experiments; J.P. prepared figures; J.P. drafted manuscript; J.P., A.A.Q., and H.R.M. edited and revised manuscript; J.P., A.A.Q., and H.R.M. approved final version of manuscript.