The influence of acute elevations in plasma osmolality and serum sodium on sympathetic outflow and blood pressure responses to exercise

Abstract

Elevated plasma osmolality (pOsm) has been shown to increase resting sympathetic nerve activity in animals and humans. The present study tested the hypothesis that increases in pOsm and serum sodium (sNa⁺) concentration would exaggerate muscle sympathetic nerve activity (MSNA) and blood pressure (BP) responses to handgrip (HG) exercise and postexercise ischemia (PEI). BP and MSNA were measured during HG followed by PEI before and after a 23-min hypertonic saline infusion (HSI-3% NaCl). Eighteen participants (age 23 ± 1 yr; BMI 24 ± 1 kg/m²) completed the protocol; pOsm and sNa⁺ increased from pre- to post-HSI (285 ± 1 to 291 ± 1 mosmol/kg H₂O; 138.2 ± 0.3 to 141.3 ± 0.4 mM; P < 0.05 for both). Resting mean BP (90 ± 2 vs. 92 ± 1 mmHg) and MSNA (11 ± 2 vs. 15 ± 2 bursts/min) were increased pre- to post-HSI (P < 0.05 for both). Mean BP responses to HG (106 ± 2 vs. 111 ± 2 mmHg, P < 0.05) and PEI (102 ± 2 vs. 107 ± 2 mmHg, P < 0.05) were higher post-HSI. Similarly, MSNA during HG (20 ± 2 vs. 29 ± 2 bursts/min, P < 0.05) and PEI (19 ± 2 vs. 24 ± 3 bursts/min, P < 0.05) were greater post-HSI. In addition, the change in MSNA was greater post-HSI during HG (Δ9 ± 2 vs. Δ13 ± 3 bursts/min, P < 0.05). A second set of participants (n = 13, age 23 ± 1 yr; BMI 24 ± 1 kg/m²) completed a time control (TC) protocol consisting of quiet rest instead of an infusion. The TC condition yielded no change in resting sNa⁺, pOsm, mean BP, or MSNA (all P > 0.05); responses to HG and PEI were not different pre- to post-quiet rest (P > 0.05). In summary, acutely increasing pOsm and sNa⁺ exaggerates BP and MSNA responses during HG exercise and PEI.

NEW & NOTEWORTHY Elevated plasma osmolality has been shown to increase resting sympathetic activity and blood pressure. This study provides evidence that acute elevations in plasma osmolality and serum sodium exaggerated muscle sympathetic nerve activity and blood pressure responses during exercise pressor reflex activation in healthy young adults.

INTRODUCTION

Elevations in plasma osmolality (pOsm) have been shown to increase resting sympathetic nerve activity in animals (Weiss et al. 1996) and resting muscle sympathetic nerve activity (MSNA) in humans (Greaney et al. 2010). Acutely raising pOsm and serum sodium (sNa⁺) concentration through intravenous infusions of hypertonic saline in rats has been reported to increase lumbar sympathetic nerve activity and decrease splanchnic and renal sympathetic activity (Weiss et al. 1996). Direct intracerebroventricular infusion of hypertonic saline elevates lumbar and adrenal sympathetic activity and causes an increase in blood pressure (BP), a decline in renal sympathetic activity, and no change in splanchnic activity (Stocker et al. 2015). The link between alterations in pOsm and sympathetic outflow has also been demonstrated in other studies examining centrally located sensing mechanisms (Kinsman et al. 2017b; Shi et al. 2007). Acute changes in pOsm and sNa⁺ concentration are detected by specialized neurons within hypothalamic circumventricular organs, which lack a complete blood-brain barrier (Bourque 2008; Kinsman et al. 2017b; Stocker et al. 2015). Subsequent activation of these hypothalamic neurons alters sympathetic nerve activity and elevates BP through mono- or polysynaptic pathways to the rostral ventrolateral medulla, the primary vasomotor site in rats and humans (Stocker et al. 2013).

Despite evidence that acute elevations in pOsm or sNa⁺ level increase resting sympathetic nerve activity, it is unclear if acute elevations in pOsm or sNa⁺ alter sympathetic reactivity. Studies performed in rodents suggest that a high-salt diet elevates plasma Na⁺ concentration to exaggerate sympathetic nerve activity and BP responses to a number of stressors, including direct injection of glutamate into the rostral ventral lateral medulla (Adams et al. 2009; Ito et al. 1999; Pawloski-Dahm and Gordon 1993) and somatosympathetic reflex activation (Simmonds et al. 2014). These exaggerated responses during chronic high-salt diets were prevented by lesioning forebrain regions involved in NaCl sensing. Consistent with this notion, increases in dietary salt intake led to exaggerated exercise pressor reflex responses in rats (Yamauchi et al. 2014). Furthermore, dietary salt has also been shown to acutely raise pOsm and sNa⁺ level in humans (Dickinson et al. 2014).

Stressors such as acute exercise have been shown to cause increases in MSNA, as well as increases in heart rate, cardiac output, and BP (Mark et al. 1985). In a laboratory setting, isometric handgrip (HG) exercise followed by postexercise ischemia (PEI) has been utilized to generate robust increases in MSNA, leading to large increases in BP (Mark et al. 1985). Isometric exercise also allows for the assessment of peripheral feedback mechanisms, such as the exercise pressor reflex. The exercise pressor reflex comprises sensory afferent nerve fibers that respond to skeletal muscle contraction and metabolic by-products generated during exercise (Kaufman and Hayes 2002), which increases MSNA and BP. In healthy individuals, exaggerated BP responses during isometric HG exercise have been shown to predict future risks of cardiovascular disease (Matthews et al. 1993).

The purpose of the current study was to examine whether acute changes in pOsm or sNa⁺ concentration could alter MSNA and BP responses to handgrip exercise. We tested the hypothesis that acute increases in pOsm and sNa⁺ via a hypertonic saline infusion (HSI) would increase BP and sympathetic activity during isometric HG exercise and PEI.

METHODS

Subjects.

All experimental procedures were approved by the Institutional Review Board at the University of Delaware and were in compliance with the guidelines set forth by the Declaration of Helsinki. A total of 18 participants (8 women and 10 men) were recruited to participate in the HSI study, and 13 additional participants (8 women and 5 men) were recruited to serve as time controls (TC). All participants were healthy normotensive young adults. All women were tested during the early follicular phase of their menstrual cycle or during the placebo phase of their oral contraceptive cycle. Two participants completed both HSI and TC trials.

Before their screening visit, participants provided both oral and written consent. All participants were screened and cleared to participate by a board-certified nurse practitioner. During the initial screening visit, participants completed a medical history questionnaire. A 12-lead electrocardiogram (ECG), resting BP, height, weight, and a fasted blood sample were also obtained. Participants were free of any known cardiovascular, renal, metabolic, pulmonary, or neurological diseases. Participants were non-obese (body mass index, BMI < 30 kg/m²) and did not use nicotine products.

Instrumentation.

Participants were instructed to avoid caffeine and exercise 24 h before the testing visit and to arrive in a fasted state (minimum 6 h). Participants were tested in a supine position with their head and one leg supported just above the plane of the body. Participants enrolled in the HSI trial had a catheter placed in the antecubital vein of each arm, one catheter placed in the nondominant arm dedicated to the infusion and the other placed in the dominant arm for blood draws throughout the HSI protocol. Participants in the TC trial had only a catheter placed in the dominant arm for blood draws throughout the protocol. Maximal HG strength was obtained during three maximal voluntary HG contractions (Grip Force Transducer; ADInstruments, Dunedin, New Zealand) of the participant’s dominant hand, with 2 min separating each contraction. This was performed during each study visit.

Beat-to-beat BP, cardiac output, and total peripheral resistance were assessed using a Finometer (finger volume-clamp method; Penáz 1992) and Modelflow estimation (Wesseling et al. 1993), from Finapres Medical Systems (Enschede, The Netherlands), calibrated to brachial BP according to the manufacturer’s recommended calibration procedures. Brachial BP was measured by automated sphygmomanometry (Dinamap Dash 2000; GE Medical Systems, Milwaukee, WI) to verify Finometer BP. Heart rate was obtained using a six-lead ECG (Dinamap Dash 2000; GE Medical Systems). Respiratory measurements were obtained by strain-gauge pneumograph (Pneumotrace; UFI, Morro Bay, CA) placed around the abdomen to ensure participants did not inadvertently perform a Valsalva maneuver during exercise. Femoral artery (FA) blood flow was assessed using duplex Doppler ultrasound (GE Logiq P5 Ultrasound; GE Healthcare, Waukesha, WI) with a linear array transducer. The linear array transducer, operating at a 12-MHz frequency, imaged the FA 2–3 cm proximal to the bifurcation to deep and superficial femoral arteries. Blood velocity data were also collected using the pulse-wave mode at an insonation angle of 60° and linear frequency set to 5 MHz.

Microneurography was utilized to directly assess MSNA throughout the testing protocol. A tungsten recording microelectrode was inserted in the peroneal nerve behind the fibular head, and a reference microelectrode was inserted 2–3 cm from the recording microelectrode, as previously performed in our laboratory (Greaney et al. 2010). The nerve recording was amplified (factor = 90,000), bandpass filtered (700–2,000 Hz), rectified, and integrated (time constant 0.1 s) using a Nerve Traffic Analyzer (model 662c-3; University of Iowa Bioengineering, Iowa City, IA). An adequate nerve recording was confirmed before the experimental protocols, utilizing the following criterion: absence of afferent nerve activity during light skin stroking, increased efferent nerve activity during voluntary end-expiratory apnea, and a visual 3:1 signal-to-noise ratio of spontaneous cardiac cycle gating of efferent nerve bursts.

Protocol.

After instrumentation, a 10-min baseline period was used to assess resting hemodynamics and MSNA. Following baseline, both study groups performed isometric HG exercise with their dominant arm at 40% of their maximal voluntary contraction for 2 min. Immediately before the cessation of HG exercise, an occlusion cuff (Hokanson, Bellevue, WA) wrapped around the upper arm of the exercising limb was rapidly inflated to 250 mmHg and remained inflated for 3 min and 15 s following relaxation of the hand. This period was deemed PEI and was used to isolate the metaboreflex.

A recovery period was performed immediately following HG exercise and PEI to ensure all hemodynamic values were similar to resting baseline values. After recovery, hypertonic saline (513 meq/l) was infused for 23 min at a rate of 0.15 ml·kg⁻¹·min⁻¹ as used previously (Greaney et al. 2010). Instead of receiving an infusion, TC participants rested quietly for 23 min. Following the infusion or TC period, a 5-min baseline was recorded and another HG and PEI trial was performed.

Blood analysis.

Blood draws were performed before the first exercise trial (prebaseline), during the final minute of the first PEI period (pre-PEI), immediately following the 23-min infusion or TC (postbaseline), and again during the final minute of the second PEI period (post-PEI). The blood draws allowed for the assessment of serum electrolyte content (EasyElectrolyte Analyzer; Medica, Bedford, MA) and blood lactate (Lactate Plus Analyzer; Nova Biomedical, Waltham, MA) throughout the protocol. POsm (Advanced 3d3 Osmometer; Advanced Instruments, Norwood, MA), hemoglobin (model Hb 201+; Hemocue, Lake Forest, CA), and hematocrit (pre-calibrated Clay Adams Readacrit centrifuge; Becton Dickinson, Sparks, MD) were measured only at the baseline periods immediately preceding each exercise bout. Blood samples were not obtained in two TC participants.

Data analysis.

FA blood flow analysis was completed using two customized LabVIEW programs (National Instruments, Austin, TX). The first customized LabVIEW program measured FA diameter (mean FA diameter, in cm) and mean blood velocity (V_mean, in cm/s). A second LabVIEW program was then used to synchronized the ultrasound recordings with the ECG, MSNA, and BP data for the precise beat-to-beat measurement of FA blood flow. FA blood flow was calculated using the following equation, as previously described (Fairfax et al. 2013): V_mean (cm/s)·π·[mean FA diameter (cm)/2]·(60 s/min). FA blood flow was assessed during each baseline period and exercise trial.

A total of 20 (n  = 13 HSI; n = 7 TC) MSNA recordings were obtained for analysis, meeting the previously described criterion throughout the experimental protocol. MSNA analysis was performed utilizing a final LabVIEW program (Fairfax et al. 2013), which provided beat-to-beat measures of heart rate, BP, MSNA, and FA blood flow. The program employs R-wave gating to detect MSNA bursts (1.2 ± 0.3 s from initial R wave), and all bursts were visually confirmed by a researcher well trained in microneurography analysis. The highest three detected bursts were averaged and assigned a value of 100 arbitrary units (AU), and all bursts were scaled accordingly within each trial. MSNA was calculated as burst frequency (bursts/min), burst incidence (bursts/100 heartbeats), normalized burst amplitude (mean burst height, AU), and total activity (bursts frequency × mean burst area, AU/min).

The absolute MSNA (burst frequency, burst incidence, and total activity) and hemodynamic values (BP, heart rate, cardiac output, and total peripheral resistance) were compared pre- to postinfusion within each trial type (e.g., pre-PEI vs. post-PEI). To calculate the relative delta (Δ) value, baseline was calculated as the mean value recorded over a 4.5-min period immediately before each HG exercise and PEI trial. To determine BP, MSNA, and hemodynamic responses to HSI and TC, the delta value was calculated from the preceding mean baseline and the mean of the final minute of HG exercise and PEI (e.g., post-PEI MSNA – postbaseline MSNA = post-ΔMSNA value). An index of sympathetic transduction was calculated as the ratio of the percent change in diastolic BP to the percent change in MSNA pre- and postinfusion.

Statistical analysis.

All data are means ± SE. Statistical analyses were performed using GraphPad Prism 6 statistical software (GraphPad Software, La Jolla, CA). Screening data were compared between HSI and TC participants using unpaired two-tailed t-tests. Normality testing was performed using the Shapiro-Wilk normality test on all variables, and variables that did not pass the normality test were log-transformed. To determine the effect of the HSI, a two-way repeated-measures ANOVA was performed comparing the effects of time (baseline vs. HG vs. PEI) and condition (preinfusion vs. postinfusion). Biochemical parameters during the HSI were compared pre- to postinfusion using paired two-tailed t-tests. To assess whether reactivity was enhanced following the infusion during HG and PEI, delta values were calculated relative to the baseline immediately before each HG and PEI trial. The delta values were compared (preinfusion vs. postinfusion) using paired two-tailed t-tests. Separately, the same statistical analyses were performed on the TC data. Post hoc analysis was performed when appropriate using Sidak’s multiple comparisons test. Statistical significance was set to P < 0.05.

RESULTS

Baseline characteristics.

Baseline screening characteristics of all participants are presented in Table 1. All participants were normotensive and non-obese. All participants had liver (aspartate transaminase: 10–40 U/l; alanine transaminase: 7–56 U/l) and kidney function (estimated glomerular filtration rate >60 ml·min⁻¹·1.73 m⁻²) biochemical parameters within normal limits. There were no differences in the screening characteristics between participants completing the HSI and TC trial.

Table 1.

Participant screening characteristics

Screening Characteristics	TC	HSI
No. of subjects	13	18
Sex (F/M)	8F/5M	8F/10M
Age, yr	23 ± 1	23 ± 1
Height, cm	169 ± 4	171 ± 3
Weight, kg	67 ± 4	70 ± 3
BMI, kg/m2	24 ± 1	24 ± 1
SBP, mmHg	118 ± 3	116 ± 4
DBP, mmHg	72 ± 2	70 ± 2
MAP, mmHg	87 ± 2	86 ± 5

Values are means ± SE. TC, time control; HSI, hypertonic saline infusion; F/M, female/male; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; MAP, mean arterial blood pressure.

Biochemical and MSNA during baseline pre- and postinfusion.

Presented in Fig. 1, the 23-min HSI increased pOsm and sNa⁺ (P < 0.05). Hematocrit levels decreased [preinfusion (Pre): 39 ± 1% vs. postinfusion (Post): 38 ± 1%, P < 0.05], suggesting an increase in plasma volume. Hemoglobin was not different following the HSI (Pre: 12.9 ± 0.4 vs. Post: 12.6 ± 0.4 g/dl, P > 0.05). HSI had no influence on baseline blood lactate levels (Pre: 0.9 ± 0.1 vs. Post: 0.9 ± 0.1 mM, P > 0.05).

Fig. 1.

Serum sodium (top) and plasma osmolality (bottom) were both increased in healthy young adults following the osmotic stimulus (3% hypertonic saline infusion). Values are means ± SE. OsM, osmolality. *P < 0.05 vs. preinfusion.

Baseline burst frequency (Pre: 11 ± 2 vs. Post: 15 ± 2 bursts/min, P < 0.05) and burst incidence (Pre: 20 ± 3 vs. Post: 26 ± 4 bursts/100 heartbeats, P < 0.05) increased following the infusion, but the increase in total MSNA activity failed to reach statistical significance (Pre: 526 ± 105 vs. Post: 634 ± 100 AU/min). HSI also did not increase baseline normalized burst amplitude (ANOVA condition, P > 0.05). Linear regression analysis was applied to examine the relation between the change in sNa⁺ pre- to postinfusion to the change in MSNA burst frequency (r = 0.53, P = 0.06) and burst incidence (r = 0.49, P = 0.09). Furthermore, a one-tailed test did reveal that there might indeed be a relation between sNa⁺ and both burst frequency and incidence (P < 0.05). The relations between the change in pOsm and the changes in MSNA burst frequency (r = 0.30, P = 0.32) and burst incidence (r = 0.24, P = 0.44) were weaker.

Cardiovascular and MSNA responses during HG exercise pre- and postinfusion.

Figure 2A is an original recording displaying an increase in BP and MSNA in response to HG exercise preinfusion. Figure 2B displays the same signals postinfusion, demonstrating an exaggerated BP and MSNA response during HG exercise. Absolute BP, hemodynamic, and MSNA values during baseline and HG exercise, pre- and postinfusion, are provided in Table 2 and presented in Figs. 3 and 4. HG exercise resulted in robust increases in systolic BP, diastolic BP, and mean BP, as well as MSNA burst frequency, burst incidence, and total activity (ANOVA time, P < 0.05). Following the infusion, systolic BP, mean BP, and MSNA burst frequency and incidence were elevated compared with the preinfusion values (ANOVA condition, P < 0.05). HSI also increased cardiac output (P < 0.05), whereas total peripheral resistance declined, compared with preinfusion values (P < 0.05). Consistent with the increase in cardiac output, FA blood flow increased during HG exercise both pre- and postinfusion, and the values were greater post-HSI (ANOVA time and condition, P < 0.05 for both). The BP and flow responses noted above were qualitatively and statistically similar in the subset of participants with an adequate nerve recording.

Fig. 2.

Original recording of a handgrip exercise trial pre- (A) and postinfusion of hypertonic saline (B) showing a robust increase in blood pressure (1st trace), muscle sympathetic nerve activity (MSNA; 2nd trace), handgrip force (3rd trace), and postexercise ischemia cuff inflation (4th trace) following an acute increase in pOsm and sNa⁺. Base, baseline recording; HG, 2-min handgrip exercise; PEI, postexercise ischemia. The end of each baseline period occurred ~1 min before the start of handgrip contraction.

Table 2.

Absolute hemodynamic and MSNA responses during HSI and TC trials

	TC		HSI
	Pre-TC	Post-TC	Preinfusion	Postinfusion
Baseline
Heart rate, beats/min	60 ± 4	63 ± 4	58 ± 2	60 ± 2
Cardiac output, l/min	5.2 ± 0.6	5.2 ± 0.5	5.2 ± 0.2	5.8 ± 0.2*
TPR, dyn⋅s⋅cm−5	1,608 ± 234	1,524 ± 154	1,572 ± 81	1,459 ± 58*
Femoral blood flow, ml/min	308 ± 56	321 ± 54	258 ± 44	311 ± 55
Burst amplitude, AU	48 ± 4	41 ± 3	45 ± 2	44 ± 3
40% Handgrip exercise
Heart rate, beats/min	83 ± 4†	89 ± 4†‡	78 ± 2†	83 ± 2*†
Cardiac output, l/min	7.0 ± 0.8†	7.3 ± 0.8†	6.7 ± 0.3†	7.4 ± 0.3*†
TPR, dyn⋅s⋅cm−5	1,380 ± 132	1,324 ± 103	1,468 ± 73†	1,387 ± 59*†
Femoral blood flow, ml/min	473 ± 64†	535 ± 73†	485 ± 100†	543 ± 90†
Burst amplitude, AU	59 ± 9	76 ± 12†	57 ± 4	61 ± 5
Postexercise ischemia
Heart rate, beats/min	63 ± 5	66 ± 5†	60 ± 2†	64 ± 3*†
Cardiac output, l/min	6.0 ± 0.6†	6.0 ± 0.6†	6.0 ± 0.2†	6.6 ± 0.3*†
TPR, dyn⋅s⋅cm−5	1,538 ± 159	1,520 ± 119	1,573 ± 59	1,516 ± 57*†
Femoral blood flow, ml/min	334 ± 52	333 ± 45	358 ± 69†	444 ± 78*†
Burst amplitude, AU	62 ± 11	53 ± 9	57 ± 4	61 ± 5

Values are means ± SE during hypertonic saline infusion (HSI; n = 18) and time control (TC; n = 13) trials. AU, arbitrary units; TPR, total peripheral resistance. Post hoc comparisons:

^*P < 0.05 vs. preinfusion.

^†P < 0.05 vs. baseline.

^‡P < 0.05 vs. pre-TC.

Fig. 3.

Preinfusion (Pre; open circles) and postinfusion (Post; filled squares) blood pressure (BP) in healthy young adults following an osmotic stimulus (hypertonic saline) or time control (TC) protocol. Values are means ± SE. Base, baseline; HG, handgrip; PEI, postexercise ischemia; MAP, mean arterial pressure. *P < 0.05 vs. Pre. †P < 0.05 vs. baseline.

Fig. 4.

Pre (open circles) and Post (filled squares) sympathetic nerve activity in healthy young adults following an osmotic stimulus (hypertonic saline) or time control (TC) protocol. A total of 13 successful nerve recordings were obtained in the hypertonic saline trial, whereas 7 successful nerve recordings were obtained in time controls. Values are means ± SE. AU, arbitrary units; Base, baseline; HG, handgrip; PEI, postexercise ischemia. *P < 0.05 vs. Pre. †P < 0.05 vs. baseline.

Delta responses were also examined and are presented in Table 3. Following HSI, delta diastolic BP and MSNA burst frequency responses were greater during HG exercise compared with preinfusion HG exercise (P < 0.05). The percent increase in MSNA during HG exercise was also examined pre- to postinfusion. Burst frequency (Pre: 113 ± 27% vs. Post: 130 ± 34%), burst incidence (Pre: 55 ± 20% vs. Post: 62 ± 25%), and total activity (Pre: 192 ± 47% vs. Post: 239 ± 57%) were not different pre- to postinfusion (P > 0.05 for all). Sympathetic transduction during HG was similar pre- and postinfusion (0.70 ± 0.36 vs. 0.27 ± 0.17, P = 0.20).

Table 3.

Delta BP, hemodynamic, and MSNA responses during HSI and TC trials

	TC		HSI
	Pre-TC	Post-TC	Preinfusion	Postinfusion
40% Handgrip exercise
ΔSBP, mmHg	18 ± 3	18 ± 3	18 ± 2	19 ± 2
ΔDBP, mmHg	15 ± 2	15 ± 2	16 ± 1	18 ± 1*
ΔMAP, mmHg	16 ± 2	17 ± 2	16 ± 2	18 ± 2
ΔHeart rate, beats/min	23 ± 3	26 ± 3	20 ± 1	23 ± 2
ΔCardiac output, l/min	1.9 ± 0.3	2.1 ± 0.4	1.5 ± 0.1	1.6 ± 0.2
ΔTPR, dyn⋅s⋅cm−5	−228 ± 134	−201 ± 122	−104 ± 34	−72 ± 28
ΔBurst frequency, bursts/min	13 ± 4	11 ± 6	9 ± 2	13 ± 3*
ΔBurst incidence, bursts/100 heartbeats	10 ± 6	5 ± 8	7 ± 3	9 ± 4
ΔTotal activity, AU/min	1,010 ± 332	1,177 ± 478	746 ± 174	1,228 ± 243*
Postexercise ischemia
ΔSBP, mmHg	18 ± 3	20 ± 4	18 ± 3	19 ± 2
ΔDBP, mmHg	11 ± 1	12 ± 2	10 ± 1	12 ± 1*
ΔMAP, mmHg	13 ± 2	15 ± 2	13 ± 2	15 ± 2
ΔHeart rate, beats/min	2 ± 1	4 ± 3	2 ± 1	4 ± 2
ΔCardiac output, l/min	0.8 ± 0.1	0.8 ± 0.2	0.8 ± 0.1	0.8 ± 0.2
ΔTPR, dyn⋅s⋅cm−5	−70 ± 92	−5 ± 82	0 ± 32	57 ± 32
ΔBurst frequency, bursts/min	10 ± 4	12 ± 5	8 ± 2	9 ± 3
ΔBurst incidence, bursts/100 heartbeats	13 ± 7	16 ± 7	13 ± 3	12 ± 4
ΔTotal activity, AU/min	976 ± 406	837 ± 364	260 ± 148	398 ± 153

Values are means ± SE during hypertonic saline infusion (HIS) and time control (TC) trials. SBP, systolic blood pressure; DBP, diastolic blood pressure; TPR, total peripheral resistance.

^*P < 0.05 vs. preinfusion. No differences were found following TC.

Cardiovascular and MSNA responses during PEI pre- and postinfusion.

Figure 2A demonstrates BP and MSNA during PEI preinfusion, and Fig. 2B demonstrates the greater values postinfusion. Presented in Fig. 3 and Fig. 4, absolute BP and MSNA (burst frequency and incidence) responses during PEI were increased under high sNa⁺ (ANOVA time and condition, P < 0.05 for both). Although there was a main effect for condition, total activity was not elevated during PEI compared with preinfusion (post hoc, P > 0.05). During PEI absolute heart rate, cardiac output, FA blood flow, and total peripheral resistance were increased following HSI compared with preinfusion (presented in Table 2; ANOVA time and condition, P < 0.05). The above findings were also confirmed in the subset of participants with an adequate nerve recording.

Delta responses during PEI are presented in Table 3; only diastolic BP response was elevated during PEI following HSI (P < 0.05). The percent increase in MSNA during PEI was also examined pre- to postinfusion. Burst frequency (Pre: 116 ± 36% vs. Post: 84 ± 25%), burst incidence (Pre: 104 ± 36% vs. Post: 70 ± 23%), and total activity (Pre: 109 ± 54% vs. Post: 86 ± 39%) were not different following HSI during PEI (P > 0.05 for all). Sympathetic transduction during PEI was similar pre- and postinfusion (0.49 ± 0.24 vs. 0.96 ± 0.89, P = 0.61). Blood lactate collected during PEI (n = 9; Pre: 3.0 ± 0.2 vs. Post: 3.0 ± 0.3 mM, P > 0.05) was not different following HSI.

Biochemical, cardiovascular, and MSNA responses during time control.

During the TC trial, there were no differences in pOsm (n = 11; Pre: 285 ± 1 vs. Post: 284 ± 1 mosmol/kg H₂O; P > 0.05) and sNa⁺ (Pre: 138.8 ± 0.4 vs. Post: 139.0 ± 0.3 mM, P > 0.05). Quiet rest had no effect on baseline blood lactate (n = 11; Pre: 0.8 ± 0.1 vs. Post: 1.0 ± 0.1 mM, P > 0.05), hematocrit (n = 11; Pre: 39 ± 1% vs. Post: 39 ± 2%, P > 0.05), or hemoglobin (n = 11; Pre: 13.2 ± 0.5 vs. Post: 13.5 ± 0.5 g/dl, P > 0.05). Blood lactate collected during PEI (n = 9; Pre: 2.6 ± 0.2 vs. Post: 2.8 ± 0.3 mM, P > 0.05) was not different following quiet rest.

In TC participants, absolute BP and MSNA increased in response to HG and PEI (Figs. 3 and 4; ANOVA time, P < 0.01), but the responses were not different following the quiet rest period (ANOVA condition, P > 0.05). Absolute hemodynamic data is presented in Table 2. FA blood flow also increased in response to HG and PEI (ANOVA time, P < 0.05), but the response was not different pre- to post-quiet rest (ANOVA condition, P > 0.05). Similarly, delta values were not different following the quiet rest period (see Table 3, P > 0.05 for all). The percent changes in MSNA (burst frequency, burst incidence, and total MSNA) from baseline were not different following quiet rest (P > 0.05 for all).

DISCUSSION

The major finding of the current study is that increases in pOsm and sNa⁺ concentration augment MSNA and BP during acute HG exercise and metaboreflex isolation. Although elevated pOsm or sNa⁺ alters sympathetic outflow at rest in animals (Antunes et al. 2006; Shi et al. 2007; Weiss et al. 1996) and humans (Farquhar et al. 2006; Greaney et al. 2010), the effects of acute elevations in pOsm and sNa⁺ on sympathetic and cardiovascular reactivity during perturbations such as exercise have not been fully explored. These findings are physiologically relevant because the elevations in MSNA and BP were caused by only a small increase in sNa⁺ of ~2%.

Previous studies in animals have documented that acute increases in sNa⁺ increase sympathetic nerve activity and BP. Regardless of method, both intravenous (Antunes et al. 2006; Weiss et al. 1996), intracarotid (Frithiof et al. 2014; Kinsman et al. 2017a), and intracerebroventricular infusions (Kinsman et al. 2017b) increase sympathetic nerve activity and BP. These studies have also established the anteroventral third ventricle region of the brain to be sensitive to acute changes in osmotic concentrations (Antunes et al. 2006; Kinsman et al. 2017b; Stocker et al. 2015). Kinsman et al. (2017b) demonstrated that within this region, neurons respond in a concentration-dependent manner to physiological increases in NaCl. These studies have provided convincing evidence that acute and physiological changes in plasma sodium (5–10 mM) increase sympathetic activity and BP. We report that even small changes in pOsm (~6 mosmol/kg H₂O) and sNa⁺ (~3 mM) increase resting MSNA and BP in humans.

Exaggerated exercise related pressor responses have been associated with adverse cardiovascular and cerebrovascular events (Hoberg et al. 1990; Kokkinos et al. 2002; Mittleman et al. 1993), and individuals with exaggerated BP responses are also more likely to develop hypertension (Dlin et al. 1983; Matthews et al. 1993). During exercise, central command generates feedforward signals to increase sympathetic outflow and BP. Locally in skeletal muscle, muscle contraction stimulates the exercise pressor reflex through the stimulation of mechanosensitive and metabolically sensitive afferent nerve fibers to also increase BP and sympathetic nerve activity (Smith et al. 2006). The findings of the current study demonstrate that acutely elevated sNa⁺ exaggerated MSNA and BP responses of healthy normotensive young adults during exercise pressor reflex activation. Relatedly, rats consuming excess salt for 2–3 wk were found to have an exaggerated exercise pressor reflex response, leading to greater BP (Yamauchi et al. 2014). Therefore, acute or chronic exposure to increased pOsm and sNa⁺ could exaggerate sympathetic reactivity and BP during activities of daily living.

The mechanism by which an elevation in sNa⁺ augments MSNA and pressor responses of humans during exercise pressor reflex activation remains unknown. Possible explanations include 1) exaggerated gain of central autonomic networks or postganglionic neurons or 2) sensitization of skeletal muscle afferents by sNa⁺; however, this explanation would likely not account for the overall increase in resting MSNA. To our knowledge, no prior studies have examined whether acute elevation in sNa⁺ alters muscle afferent sensitivity or postganglionic neuron excitability. Interestingly, normotensive rodents fed a high-salt diet display increases in plasma Na⁺ and exaggerated sympathetic and BP responses to a number of stressors including 1) activation of sciatic afferents (Ito et al. 1999; Simmonds et al. 2014), 2) exercise pressor reflex (Yamauchi et al. 2014), and 3) stimulation of the aortic depressor nerve or vagal afferents (Pawloski-Dahm and Gordon 1993; Simmonds et al. 2014). These effects have been attributed to a sensitization of central autonomic networks such as neurons in the rostral ventrolateral medulla (RVLM; Adams et al. 2007, 2008; Ito et al. 1999). Moreover, water deprivation in rodents increases sNa⁺ and pOsm but also augments glutamate-evoked pressor responses from the RVLM (Brooks et al. 2004a). On the other hand, acute infusion of hypertonic NaCl in rodents increases lumbar sympathetic nerve activity but does not alter pressor responses evoked by RVLM injection of glutamate (Brooks et al. 2004b). Thus the ability of increased sNa⁺ to sensitize central autonomic networks may depend on a number of factors, including the magnitude of the changes in sNa⁺ and duration. Altogether, the mechanism(s) underlying the exaggerated responses during handgrip exercise remains unclear and requires further investigation.

The increase in sNa⁺ that results from a 23-min HSI is greater but parallels the increase in sNa⁺ that has been reported following a high-salt meal (Dickinson et al. 2014; Suckling et al. 2012). Spinelli et al. (1987) reported an increase in sNa⁺ following one high-salt meal, which was accompanied by a reduction in hematocrit levels, suggesting an expansion of the extracellular fluid volume. However, there are also clear differences between the two modes of NaCl delivery (intravenous vs. by mouth), such as the magnitude of the sNa⁺ increase and the time course. Nevertheless, it is likely that subtle increases in sNa⁺ have autonomic effects, which may have relevance for acute and chronic increases in sNa⁺ elicited by diet.

Clinically, HSI is used in the treatment of hemorrhagic and septic shock (Oliveira et al. 2002) and severe traumatic brain injury (Cooper et al. 2004). In patients with hemorrhagic or septic shock, HSI is provided as a treatment to reduce hypoxia and hypotension (Oliveira et al. 2002). In traumatic brain injury patients, HSI treatments have been found to reduce intracranial pressure and increase cerebral perfusion (Mangat et al. 2015). Much of the clinical benefit of HSI is thought to be related to the volume-related effects of infusing 3% NaCl, including fluid shifts. However, these data and other data (Charkoudian et al. 2005; Greaney et al. 2010) suggest that the autonomic effects of acutely raising pOsm may also contribute to the maintenance of BP in conditions such as hemorrhagic shock. For example, in addition to the modest elevation of MSNA post-HSI demonstrated in the present study, two additional studies have demonstrated that HSI increases baroreflex control of MSNA (Charkoudian et al. 2005; Wenner et al. 2007). Thus both volume and autonomic effects of HSI may provide a therapeutic benefit in certain clinical conditions.

There are a few study limitations. First, the control experiments were time controls and not infusion controls (such as isotonic saline infusion). Previous studies utilizing an isotonic saline control have reported that isotonic saline infusions do not increase MSNA or BP (Charkoudian et al. 2003; Farquhar et al. 2006; Greaney et al. 2010). Second, and relatedly, the hypertonic saline infusions and time controls were not randomized, and as stated, the time control experiments were performed in a different set of participants. Although the subjects were well matched, a stronger approach would have been a randomized design where each subject participated in both data collections. Third, HSI-induced increases in urine production and consequently bladder distension could have had an effect on MSNA (Fagius and Karhuvaara 1989). Fagius and Karhuvaara (1989) demonstrated that bladder distension can increase MSNA (i.e., a vesicovascular response). Last, our findings are restricted to acute increases in pOsm and sNa⁺. The study design does not permit any conclusions to be drawn regarding chronic increases in sNa⁺. It is interesting that prolonged exposure to dietary sodium in rodents has been shown to lead to augmented pressor and sympathetic activity during exercise (Yamauchi et al. 2014). Future studies in humans are needed to examine the long-term autonomic effects of prolonged exposure to high sNa⁺.

In summary, acute elevations in pOsm and sNa⁺ increases absolute MSNA and BP during rest and exercise. The elevation in MSNA and BP might have important clinical implications due to the increase in absolute BP during exercise despite only a small increase in pOsm and sNa⁺.

GRANTS

This research was funded by National Heart, Lung, and Blood Institute Grant 1R01HL128388.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

M.S.B., M.M.W., W.C.R., S.D.S., and W.B.F. conceived and designed research; M.S.B., E.L.M., J.C.W., M.C.B., and W.B.F. performed experiments; M.S.B., J.C.W., M.C.B., W.C.R., and W.B.F. analyzed data; M.S.B., E.L.M., M.M.W., S.D.S., and W.B.F. interpreted results of experiments; M.S.B. prepared figures; M.S.B., E.L.M., M.M.W., W.C.R., S.D.S., and W.B.F. drafted manuscript; M.S.B., E.L.M., J.C.W., M.C.B., M.M.W., W.C.R., S.D.S., and W.B.F. edited and revised manuscript; M.S.B., E.L.M., J.C.W., M.C.B., M.M.W., W.C.R., S.D.S., and W.B.F. approved final version of manuscript.