Long-duration bed rest modifies sympathetic neural recruitment strategies in male and female participants

Abstract

To understand the impact of physical deconditioning with head-down tilt bed rest (HDBR) on the malleability of sympathetic discharge patterns, we studied 1) baseline integrated muscle sympathetic nerve activity (MSNA; microneurography) from 13 female participants in the WISE-2005 60-day HDBR study (retrospective analysis), 2) integrated MSNA and multiunit action potential (AP) analysis in 13 male participants performed on data collected at baseline and during physiological stress imposed by end-inspiratory apnea in a new 60-day HDBR study, and 3) a repeatability study (control; n = 6, retrospective analysis, 4 wk between tests). Neither baseline integrated burst frequency nor incidence were altered with HDBR (both P > 0.35). However, baseline integrated burst latency increased in both HDBR studies (male: 1.35 ± 0.02 to 1.39 ± 0.02 s, P < 0.01; female: 1.23 ± 0.02 to 1.29 ± 0.02 s, P < 0.01), whereas controls exhibited no change across two visits (1.25 ± 0.02 to 1.25 ± 0.02 s, group-by-time interaction, P = 0.02). With the exception of increased AP latency (P = 0.03), male baseline AP data did not change with HDBR (all P > 0.19). The change in AP frequency on going from baseline to apnea (∆94 ± 25 to ∆317 ± 55 AP/min, P < 0.01) and the number of active sympathetic clusters per burst (∆0 ± 0.2 to ∆1 ± 0.2 clusters/burst, P = 0.02) were greater post- compared with pre-HDBR. The change in total clusters with apnea was ∆0 ± 0.5 clusters pre- and ∆2 ± 0.7 clusters post-HDBR (P = 0.07). These data indicate that 60-day HDBR modified discharge characteristics in baseline burst latency and sympathetic neural recruitment during apneic stress.

NEW & NOTEWORTHY Long-duration bed rest did not modify baseline sympathetic burst frequency in male and female participants, but examination of additional features of the multiunit signal provided novel evidence to suggest augmented synaptic delays or processing times at baseline for all sympathetic action potentials. Furthermore, long-duration bed rest increased reflex-sympathetic arousal to apneic stress in male participants primarily by mechanisms involving an augmented firing rate of action potential clusters active at baseline.

INTRODUCTION

Both spaceflight and simulated microgravity via head-down tilt bed rest (HDBR) represent chronic homeostatic stressors to multiple systems, including the cardiovascular and nervous systems (23, 34). To date, approaches such as microneurography (most often at the peroneal nerve) (11, 13, 25, 30, 35, 43, 44) or circulating norepinephrine and its spillover (17, 18) yield equivocal findings regarding the impact of spaceflight or HDBR on baseline sympathetic outflow. Varying outcomes between studies [i.e., increases (30), decreases (43), or no change (35) in sympathetic discharge] may be related to small sample sizes. Furthermore, the metrics of sympathetic outflow used to date may limit the ability to observe changes that reflect neurophysiological properties (e.g., axonal recruitment) rather than the average sympathetic outflow over time. For example, the singular outcome of the averaged integrated burst frequency or burst incidence reflects steady-state conditions with measures that, by nature of signal integration, remove information regarding the neural patterns that produce a given burst. Thus, although burst frequency has been considered a standard metric of muscle sympathetic nerve activity (MSNA) (20), this measurement does not provide insight into the broader control of sympathetic outflow, such as burst size and latency, which are a function of neural recruitment strategies (31, 37, 46).

Variations in the size and frequency of integrated MSNA bursts reflect the underlying action potential (AP) content, and new methods to study the discharge patterns of these APs have been developed recently (10, 36, 46, 48). These methods have generated novel information regarding the strategies employed by the sympathetic nervous system during acute physiological stress. For example, stress imposed by severe arterial baroreflex unloading (5, 37, 39), the arterial chemoreflex (3, 5, 6, 46), or the muscle metaboreflex (4) augments the firing rate of sympathetic axons active at rest and elicits the recruitment of larger, latent axons with fast conduction velocities and high recruitment thresholds. To further modulate outflow, the sympathetic nervous system expresses the option to alter AP latency, which reflects the time delay related to the neural arc between the initiation of sympathetic activity by baroreceptor unloading in diastole and AP arrival at the recording site in the sympathetic postganglionic C-fiber (14). Protocols involving effortful stress, such as apneas, are notable for their ability to elicit these changes in AP latency through mechanisms that likely reflect modifiable synaptic delays or central processing times (3–5, 37, 38). Overall, this new approach to MSNA analysis exposes alterations in efferent sympathetic discharge that are not evident in burst frequency analyses used to date.

Maximal end-inspiratory apneas elicit large changes in MSNA in terms of burst frequency, burst size, and burst latency. Concurrently, this maneuver elicits within-burst patterns of AP frequency, newly recruited AP clusters, and reductions in AP latency (3, 46), as outlined above. Intriguingly, the Neurolab Space Shuttle study (11) reported augmented (∼30%) sympathetic burst frequency responses to apnea post-spaceflight compared with pre-spaceflight. Mechanisms underlying this observation are likely multifactorial, as apnea-mediated sympatho-excitation involves chemoreceptor activation, a central command phenomenon related to the drive to breathe, and the removal of lung stretch receptor inhibition (41). Nevertheless, these data suggest that longer-term microgravity exposure modifies AP recruitment. However, to date, no information exists regarding the effect of HDBR or spaceflight on sympathetic neural discharge strategies or reflexive recruitment conditions.

Therefore, this study quantified MSNA discharge patterns in the integrated signal and at the AP scale, during baseline and using a maximal end-inspiratory apnea, to test the hypothesis that 60 days of HDBR modifies baseline sympathetic neural discharge as well as AP recruitment patterns.

METHODS

Sample

This study evaluated data from three groups: a retrospective analysis of baseline data was performed on two previously published studies, including a 60-day HDBR study performed in female participants [WISE-2005; n = 13 (Toulouse France)] (2, 12) and a repeatability study conducted by our laboratory (n = 6) (27). Also, this study includes original data from 13 male participants with successful nerve recordings of sufficient signal-to-noise quality before and after 60 days of HDBR performed at the Institute of Aerospace Medicine, German Aerospace Center (DLR), in Cologne, Germany (Cologne RSL) (29). Data from both the WISE-2005 and Cologne RSL studies were collected by the same microneurographer, using similar protocols and the same equipment. Important differences between studies included the following: 1) the studies were performed in different research settings; 2) the WISE-2005 study tested only female participants, whereas the Cologne RSL study comprised all male participants; 3) Cologne RSL data were collected in the −6° head-down tilt position, whereas the WISE-2005 data were collected in the supine position; and 4) the WISE-2005 data included only the integrated MSNA neurogram collected at 1,000 Hz, precluding multiunit AP analyses, whereas AP analyses could be performed on the Cologne RSL data, as raw data were collected at 10,000 Hz. In the bed rest studies, all participants were healthy as determined by a comprehensive medical questionnaire and history, a complete physical, and electrocardiogram (ECG). The volunteers were normotensive, were nonsmokers, and had normal body mass. Prior to study commencement, all participants gave written, informed consent. All experiments adhered to the ethics principles of the Declaration of Helsinki. The RSL study experimental procedures were approved by the ethics committee of the Northern Rhine Medical Association and the Federal Office for Radiation Protection (Germany). Experimental protocols employed in the WISE-2005 study were approved by the Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale, Midi-Pyrénées (France), Committee for the Protection of Human Subjects at Johnson Space Center, and the Office of Research Ethics at the University of Waterloo. The repeatability study received ethics approval by the University of Western Ontario Review Board for Health Science Research Involving Human Subjects. Participant groups from Cologne RSL, WISE-2005, and the repeatability study will henceforth be referred to as the male, female, and control groups, respectively.

Methodological information for the female (2, 12) and control (27) studies are reported elsewhere. The following outlines methodological information for the male study only, unless specified otherwise.

Head-Down Bed Rest

During the 60-day bed rest period, the participants were strictly confined to a bed in the −6° head-down position. Further information regarding bed rest routines, activities, and diet are reported elsewhere (29).

Data Collection and Experimental Protocol

Data were collected in the −6° head-down position, 4 days before HDBR and on day 58 of the HDBR period. At least 30 min of supine rest occurred before baseline measurements were obtained. Data were collected during two conditions in the pre- and post-HDBR sessions: 1) ≥5 min of baseline rest and 2) end-inspiratory apnea (end-apnea) to volitional break point.

Hemodynamic Data Collection and Analyses

Blood pressure (BP) was measured continuously by finger-cuff plethysmography (Finometer; Finapres Medical Systems, Amsterdam, The Netherlands) on the right middle finger, maintained at heart level. Heart rate (HR) was measured using a three-lead ECG.

All analog signals were sampled in real time at 10,000 Hz with an online acquisition and analysis system (PowerLab; ADInstruments, Castle Hill, New South Wales, Australia) and stored on a computer for subsequent analysis. Research personnel were not blinded during analysis. Baseline mean HR and mean arterial pressure (MAP) were calculated from ≥5 min of data. The last ∼30 s of end-inspiratory apnea were used to calculate mean values. Nadir values were calculated at the onset of end-inspiratory apnea.

MSNA Collection and Analyses

Efferent multiunit sympathetic outflow was measured in the peroneal nerve by microneurography (19). Specifically, a tungsten microelectrode (35 mm long, 200 μm in diameter, and tapered to a 1- to 5-μm uninsulated tip) was inserted percutaneously into the nerve just posterior to the fibular head. A reference electrode was positioned subcutaneously 1–3 cm from the recording site. A suitable recording site was searched for by manually manipulating the microelectrode until a characteristic pulse-synchronous burst pattern was observed. Confirmation of an MSNA site was determined by the absence of skin paresthesia and a signal that increased firing frequency in response to voluntary apnea, but not during arousal to a loud noise (9). The MSNA neurogram was measured with a nerve traffic analysis system (662C-3; Bioengineering University of Iowa, Iowa City, IA). The neural signal was first preamplified with a gain of 1,000 (using preamplifier and isolation amplifier; gains of 100 and 10, respectively) and further amplified with a gain of 75 (using a variable gain amplifier; gain of 0.1–99). The neural activity was then band-pass filtered (bandwidth of 700–2,000 Hz) before being rectified and integrated (using a leaky integrator; 0.1-s time constant) to obtain a mean voltage neurogram. The raw, filtered, and integrated MSNA signals were sampled at a frequency of 10,000 Hz and stored for offline analysis (PowerLab Software; ADInstruments).

Integrated bursts of MSNA were included in the analysis if they exhibited pulse-synchrony, had a signal-to-noise ratio of ≥2:1 with respect to the previous period of neural silence between bursts, and expressed characteristic rising and falling slopes. Occurrence of integrated sympathetic bursts was confirmed by visually inspecting the corresponding raw and filtered neurograms. MSNA bursts were inspected for consistent amplitude to rule out shifts in microelectrode position. Sympathetic activity was quantified as burst frequency (bursts/min) and burst incidence (bursts/100 heartbeats). Burst amplitude was measured in volts and normalized to the mean burst amplitude at baseline, which was given a value of 100. Sympathetic burst latency was measured as the mean time interval(s) from the preceding ECG R-wave to the peak of the corresponding MSNA burst in the integrated neurogram (15). The integrated burst amplitude-latency relationship was calculated for male data only by custom analysis software (36). Indices of integrated sympathetic nerve activity were averaged over ≥5 min of baseline (male and female groups) and the last ∼30 s of end-inspiratory apnea (male group only). Integrated MSNA burst frequency and incidence for control data were derived from those of Kimmerly et al. (27).

Burst Amplitude Probability Distribution Analysis

Burst amplitude probability distribution analysis was performed on male and female baseline data to determine whether the increased latency with HDBR was attributed to sympathetic burst size (49). Unlike measuring the voltage of sympathetic bursts to assess burst amplitude across sessions, electrode position does not impact this technique, and it demonstrates good test-retest reproducibility (27). Sympathetic bursts were normalized to the largest burst (given a value of 100%) that occurred during 5 min of baseline and pre- and post-HDBR. Probability distributions were constructed based on relative burst amplitude. Each burst was placed into one of 10 equally sized bins (i.e., 0–9.9% of maximum burst amplitude, 10–19.9% of max burst amplitude, etc.) based on its relative amplitude, and the probability of each bin was determined as the number of bursts per bin relative to the total number of bursts. Probability distribution graphs were constructed, and the median normalized burst amplitudes (% ± SE) were used for statistical comparisons between pre- and post-HDBR (28, 42, 47).

Sympathetic Baroreflex Analysis

Baroreflex regulation of baseline MSNA in male and female groups was assessed for baroreflex threshold and baroreflex sensitivity (21, 26). The baroreflex threshold analysis method examines the linear portion of the relationship between diastolic BP and the occurrence of sympathetic bursts. For this analysis, all diastolic BPs were grouped into bins of 1 mmHg, and within these bins the percentage of cardiac cycles associated with a sympathetic burst was calculated. The percentage of cardiac cycles associated with a sympathetic burst was plotted as a function of the mean diastolic BP of each bin. The midpoint of the baroreflex threshold analysis curve (T₅₀) was defined as the diastolic BP at which 50% of cardiac cycles were associated with a sympathetic burst. The baroreflex sensitivity analysis method examines the relationship between normalized burst amplitude and the corresponding diastolic BP. The slope of the threshold and sensitivity regression lines, y-intercepts, and Pearson’s correlation coefficients (r) were recorded for each participant pre- and post-HDBR. Baroreflex analysis was performed on 5 min of integrated MSNA baseline data pre- and post-HDBR.

AP Analysis

APs were detected and extracted from the filtered raw MSNA signal using techniques developed in our laboratory (36). Briefly, this technique uses a continuous wavelet transform (CWT) for AP detection. The CWT used a “mother wavelet” that was adapted to an average AP waveform constructed from physiological recordings of postganglionic sympathetic APs (36). The CWT was in turn applied to the filtered MSNA to provide a wavelet coefficient between an individual AP and the mother wavelet such that the wavelet coefficient was largest in the presence of the APs and negligible when applied to noise. Wavelet coefficients related to APs and noise were separated based on thresholding analysis (24). The exact location of the negative peak for each AP was then detected by isolating the largest suprathreshold wavelet coefficient. Using this location information, the AP waveforms were obtained from the original filtered raw signal by putting the estimated location of APs in the center of a predefined window (3.2 ms). In this way, the amplitude and morphology of each extracted action potential remained unaltered. Extracted APs were then ordered based on peak-to-peak amplitude, and histogram analysis was performed to separate APs into amplitude-based clusters. Cluster bin widths were automatically defined based on Scott’s rule, which balanced bin width, data bias, and variance to minimize the integrated mean square error (40). Thus, the number of total clusters varied by subject, and new clusters detected during end-inspiratory apnea represented recruitment of new, larger APs not present during baseline. The signal-to-noise ratio for a period of data was determined as the amplitude of the negative peak of the mean action potential over the standard deviation of the background noise (i.e., during sympathetic silence). According to our simulated analysis, the signal-to-noise ratio of the multiunit AP signal obtained pre-HDBR (baseline, end-inspiratory apnea: 4.1 ± 0.1, 4.1 ± 0.1) and post-HDBR (4.3 ± 0.1, 4.4 ± 0.1) will provide a correct detection rate of >90% and a false positive rate of <3% (36).

AP indices included AP frequency (the no. of action potentials/min), incidence (the no. of action potentials/100 heartbeats), and the mean AP content per integrated burst. Additionally, the number of total clusters detected and the number of active clusters per integrated burst were assessed. The latency of individual APs was determined as the time between the R-wave of the preceding cardiac cycle and the negative peak of the AP waveform (38). AP cluster latency was determined as the mean latency of all APs. AP analysis was performed on similar durations of baseline (112 ± 8 vs. 117 ± 6 s, n = 12, P = 0.33) and end-inspiratory apnea data (33 ± 1 vs. 35 ± 1 s, n = 8, P = 0.26) pre- and post-HDBR. When assessing the AP cluster size-latency relationship at baseline and end-inspiratory apnea, APs were normalized to the largest detected cluster (given a value of 100%) and each cluster was placed into one of 10 equally sized bins (i.e., 0–9.9% of largest detected cluster, 10–19.9% of largest detected cluster, etc.) (3, 4).

Statistical Analysis

All data are presented as means ± SE. Three mixed-model ANOVA designs were conducted to assess 1) the effect of time (2 time points: pre-HDBR/visit 1 vs. post-HDBR/visit 2; time points are referred to as pre/post-HDBR for male and female groups that participated in bed rest and visit 1/visit 2 for the control group, which did not participate in bed rest) and group (3 groups: control vs. male vs. female) on baseline integrated burst frequency, incidence, and latency; 2) the effect of HDBR (2 time points: pre-HDBR vs. post-HDBR) and group (2 groups: male vs. female) on baseline hemodynamic, baroreflex threshold and sensitivity, and MSNA burst probability data; and 3) the effect of HDBR (2 time points: pre-HDBR vs. post-HDBR) and condition (2 conditions: baseline vs. end-apnea) on hemodynamic, integrated MSNA, and AP data in male participants. Interactions were tested, and paired t-tests assessed simple main effects for significant interactions. Paired t-tests tested the impact of bed rest on baseline AP data and the effect of HDBR on changes (∆) in integrated MSNA and AP data from baseline to end-apnea. Statistical analysis was performed using SPSS (version 23, SPSS, Chicago, IL). Tests were two-tailed with α = 0.05.

RESULTS

Baseline

Hemodynamics.

Table 1 provides baseline hemodynamic changes with bed rest. Baseline HR (P = 0.72), MAP (P = 0.45), and breathing frequency (P = 0.60) were similar pre- to post-HDBR in male and female participants. The time (pre-HDBR vs. post-HDBR)-by-group (male vs. female) interactions were all P > 0.07.

Table 1.

Baseline hemodynamic and integrated MSNA over time in controls and male and female participants

	Control		Male		Female
Group (Time)	Visit 1	Visit 2	Pre-HDBR	Post-HDBR	Pre-HDBR	Post-HDBR
Hemodynamics
HR, beats/min	NR	NR	60 (2)	66 (3)	71 (3)	69 (3)
MAP, mmHg	NR	NR	94 (3)	92 (3)	86 (1)	88 (3)
Breathing frequency, breaths/min	NR	NR	15 (1)	15 (1)	15 (1)	15 (1)
Integrated MSNA
Burst frequency, bursts/min	23 (3)	22 (4)	20 (2)	20 (2)	20 (2)	17 (1)
Burst incidence, bursts/100 beats	39 (4)	35 (5)	33 (2)	31 (2)	28 (3)	26 (2)
Burst latency, s*	1.25 (0.02)	1.25 (0.02)	1.35 (0.02)	1.39 (0.02)†	1.23 (0.02)	1.29 (0.02) †

Values are means (SE); n = 6 control, 13 male participants, and 13 female participants. MSNA, muscle sympathetic nerve activity; HDBR, head-down bed rest. HR, heart rate; MAP, mean arterial pressure; NR, not reported. Burst frequency and incidence data for the control group were derived from those of Kimmerly et al. (27).

^*Significant group-by-time interaction (P = 0.02);

^†significantly different from pre-HDBR (P < 0.01).

Integrated MSNA.

Baseline sympathetic burst latency was increased with HDBR in both male participants (P < 0.01) and female participants (P < 0.01), whereas controls exhibited no change across visits [P = 0.67, time (pre-HDBR/visit 1 vs. post-HDBR/visit 2]-by-group (control vs. male vs. female) interaction: P = 0.02; Fig. 1]. Bland-Altman analysis of burst latency in controls demonstrated no fixed (mean difference ± SE, 95% confidence interval: 0.00 ± 0.01 s, −0.02 to 0.02 s, P = 0.67) or proportional biases (r = 0.15, b = −0.05, P = 0.77). Burst amplitude probability distribution analysis of baseline data indicated a time (pre-HDBR vs. post-HDBR)-by-group (male vs. female) interaction (P = 0.05) such that male participants demonstrated a leftward shift in the median normalized amplitude (P = 0.04), suggesting a greater proportion of smaller bursts during baseline post- compared pre-HDBR, whereas female participants demonstrated no change (Fig. 2). However, this interaction had a weak effect (partial η²: 0.16) and was nonsignificant upon removal of the observation, with the largest leftward shift in median normalized amplitude (P = 0.08). The burst latency-amplitude relationship was preserved but shifted upward with HDBR (pre-HDBR: y = −0.39x + 1.46, r² = 0.19, post-HDBR: y = −0.34x + 1.50, r² = 0.19, P = 0.59; Fig. 3) in male participants. In contrast to burst latency, burst frequency and incidence were unchanged with HDBR in either sex (all P > 0.30; Table 1). Time (pre-HDBR/visit 1 vs. post-HDBR/visit 2)-by-group (control vs. male vs. female) interactions for burst frequency and incidence were both P > 0.10.

Fig. 1.

Baseline-integrated sympathetic burst latency over time in controls (n = 6), male participants (n = 13), and female participants (n = 13) (A) and baseline sympathetic action potential (AP) latency pre- to post- head-down bed rest (HDBR) in male participants (B). A mixed-models ANOVA yielded a significant group-by-time interaction for baseline burst latency, P = 0.02. Values are means ± SE. *Significantly different from pre-HDBR within group (P ≤ 0.03, paired t-test).

Fig. 2.

Probability distribution of normalized burst amplitude during baseline in male participants (n = 13; top) and female participants (n = 13; bottom). Values are means ± SE. A mixed-models ANOVA yielded a significant head-down bed rest (HDBR)-by-time interaction, P = 0.05. Numbers and arrows above plots indicate the mean ± SE of the median normalized burst amplitude and direction of shift. *Significantly different from pre-HDBR within group (P = 0.04, paired t-test).

Fig. 3.

Representative upward shift of the integrated burst latency-amplitude relationship at baseline in 1 participant pre- and post-head-down bed rest (HDBR).

AP indices.

Twelve of 13 individuals in the male group had complete baseline AP data with sufficient signal-to-noise ratio before and after bed rest (Table 2). In the pre-HDBR session, AP cluster latency decreased as peak-to-peak AP cluster amplitude increased, as modeled by an exponential decay function, such that larger AP clusters exhibited shorter latencies at baseline (r² = 0.66, P < 0.01). The cluster size-latency relationship was preserved post-HDBR (r² = 0.92, P < 0.01), yet compared with pre-HDBR it was shifted upward (P = 0.03) such that all APs were delayed with HDBR (Fig. 1). Alternatively, baseline AP frequency, incidence, and the number of APs per burst were unchanged with HDBR (all P > 0.19).

Table 2.

Baseline AP indices pre- and post-HDBR in male participants

	Pre-HDBR	Post-HDBR
AP frequency, spikes/min	78 (17)	118 (22)
AP incidence, spikes/100 beats	121 (26)	177 (34)
APs/burst, spikes/burst	5 (0.6)	6 (0.6)
AP latency, s	1.33 (0.03)	1.38 (0.03)*

Values are means (SE); n = 12. AP, action potential; HDBR, head-down bed rest.

^*Significantly different from pre-HDBR (P = 0.03).

Baroreflex threshold and sensitivity analysis.

Baseline-integrated baroreflex threshold (probability gain) and sensitivity (amplitude gain) data were calculated for 13 male participants and 12 female participants with complete data pre- and post-HDBR. The slope of the baroreflex threshold curve was similar pre (male group: y = −8.98x + 702, r² = 0.90; female group: y = −6.50x + 475, r² = 0.80)- and post-HDBR (male group: y = −8.09x + 636, r² = 0.86; women: y = −7.19x + 535, r² = 0.81), indicating no change in baroreflex-mediated regulation of burst occurrence (P = 0.88). The T₅₀ (midpoint of the threshold analysis curve) was unchanged with HDBR (male group: 72 ± 2 to 72 ± 2 mmHg; female group: 65 ± 2 to 68 ± 3 mmHg; P = 0.67). Integrated baroreflex sensitivity, indicating the baroreflex regulation of burst amplitude, was unaltered pre (male group: y = −1.45x + 158, r² = 0.06; female group: pre-HDBR: y = −1.06x + 126, r² = 0.05) to post (male group: y = −1.01x + 126, r² = 0.04; female group: y = −0.92x + 134, r² = 0.04) (P = 0.30). Time (pre-HDBR vs. post-HDBR)-by-group (male vs. female) interactions were all P > 0.22.

Sympathetic Activation with Apnea

Eight of 13 individuals from the male group had complete end-apnea data pre- and post-HDBR. Representative data highlighting the change in BP, integrated MSNA, and AP activity from baseline to end-apnea pre- and post-HDBR are presented in Fig. 4.

Fig. 4.

Representative blood pressure (BP) and sympathetic nerve activity (integrated and raw signal) at baseline and end-apnea in 1 participant pre- and post-head-down bed rest (HDBR).

Hemodynamics.

End-inspiratory apneas were similar in duration across test sessions (pre-HDBR 74 ± 5 s, post-HDBR 85 ± 7 s; P = 0.09). MAP and pulse pressure characteristically drop at the onset of end-inspiratory apnea due to increased intrathoracic pressure, although the drop from baseline to nadir MAP (pre-HDBR 104 ± 5 to 88 ± 6 mmHg, post-HDBR 94 ± 4 to 77 ± 5 mmHg; P = 0.70) and nadir pulse pressure (pre-HDBR 59 ± 3 to 49 ± 5 mmHg, post-HDBR 50 ± 4 to 37 ± 5 mmHg; P = 0.16) were similar pre- and post-HDBR, suggesting similar baroreflex unloading. Also, changes from baseline to end-apnea in HR (pre-HDBR 66 ± 7 to 66 ± 5 beats/min, post-HDBR 68 ± 2 to 70 ± 3 beats/min; P = 0.43) and MAP (pre-HDBR 104 ± 5 to 112 ± 5 mmHg, post-HDBR 94 ± 4 to 103 ± 4 mmHg; P = 0.52) were similar pre- and post-HDBR.

Integrated MSNA.

Integrated MSNA data at baseline and end-apnea, pre- and post-HDBR, along with the P values for the two main-effects (HDBR: pre-HDBR vs. post-HDBR; condition: baseline vs. end-apnea) and interactions (HDBR by condition), are presented in Table 3. Integrated burst frequency (and incidence) increased from baseline to end-apnea pre- and post-HDBR; however, the mixed-model ANOVAs indicated significant HDBR-by-condition interactions, suggesting that the magnitude of increase was greater post-HDBR for both variables. Figure 5 illustrates the interaction for burst frequency. Additional correlation analysis suggested that sympathetic burst frequency responses from pre- to post-HDBR were not related to the change in duration of end-apnea (r = 0.10, P = 0.82). Integrated burst amplitude increased from baseline to end-apnea to a similar extent pre- and post-HDBR (∆normalized burst amplitude pre-HDBR and post-HDBR: 30 ± 6 and 43 ± 10 AU, respectively, P = 0.23). Also, the change in burst latency with end-apnea was not altered with HDBR.

Table 3.

Integrated MSNA and AP indices during baseline and end-apnea pre- and post-HDBR in male participants

	Pre-HDBR		Post-HDBR		P Value
Group (Condition)	BSL	APN	BSL	APN	HDBR	APN	Interaction
Integrated MSNA
Burst frequency, bursts/min	21 (3)	38 (4)	20 (2)	48 (2)*	0.10	<0.01	0.01
Burst incidence, bursts/100 beats	33 (4)	57 (3)	31 (3)	68 (2)*	0.19	<0.01	0.02
Burst latency, s	1.36 (0.04)	1.37 (0.03)	1.39 (0.03)	1.41 (0.03)	0.02	0.10	0.59
AP Indices
AP frequency, spikes/min	62 (22)	156 (44)	106 (27)	423 (74)*	0.04	<0.01	<0.01
AP incidence, spikes/100 beats	92 (31)	226 (54)	159 (40)	602 (107)*	0.16	<0.01	0.02
APs/burst, spikes/burst	4 (0.8)	5 (0.8)	6 (0.9)	9 (1.5)*	0.11	<0.01	0.02
AP latency, s	1.31 (0.04)	1.32 (0.04)	1.36 (0.03)	1.39 (0.04)	0.05	0.23	0.26
Total clusters	7 (0.9)	7 (0.9)	9 (0.6)	11 (0.7)*	0.04	0.04	0.07
Clusters/burst	3 (0.3)	3 (0.3)	3 (0.4)	4 (0.4)*	0.06	<0.01	0.02

Values are means (SE); n = 8. MSNA, muscle sympathetic nerve activity; AP, action potential; HDBR, head-down bed rest; BSL, baseline; APN, end-apnea. Note: for significant interactions, the change from BSL to APN was greater post-HDBR than pre-HDBR, as depicted in Fig. 5.

^*Significantly different from Pre-HDBR APN (P ≤ 0.05).

Fig. 5.

Change (Δ) in integrated sympathetic nerve activity and action potential (AP) indices from baseline to end-apnea pre- and post- head-down bed rest (HDBR) in male participants (n = 8). The significant HDBR-by-condition interactions detailed in Table 3 are highlighted here. Paired sample t-tests were employed. *Significantly different from pre-HDBR, P < 0.05; †trend toward significantly different from pre-HDBR, P = 0.07.

AP indices.

AP indices during baseline and end-apnea are presented in Table 3 along with the P values for the two main effects (HDBR: pre-HDBR vs. post-HDBR; condition: baseline vs. end-apnea) and interactions (HDBR-by-condition). The AP cluster size-latency relationship was maintained with end-apnea both pre- (r² = 0.56, P < 0.01) and post-HDBR (r² = 0.83, P < 0.01). AP analysis demonstrated an exaggerated sympathetic response to end-apnea with HDBR (HDBR-by-condition interaction), which supports integrated MSNA analysis. Compared with baseline, the increase in sympathetic AP frequency (and incidence) with end-apnea was greater in the post-HDBR than in the pre-HDBR session. Although the change in integrated burst size from baseline to end-apnea was similar pre- and post-HDBR, the change in AP content per sympathetic burst with end-inspiratory apnea was greater in the post- than in the pre-HDBR session. When detected APs were binned by peak-to-peak amplitude, there was a trend toward greater recruitment of larger AP clusters from baseline to end-apnea post-HDBR compared with pre-HDBR. This was associated with a greater increase in the number of active AP clusters per burst, from baseline to end-apnea, in the post- than in the pre-HDBR session. Variables that demonstrated greater responses to end-apnea in the post-HDBR session (i.e., significant HDBR-by-condition interactions) are presented in Fig. 5, which shows the change from baseline to end-apnea pre- and post-HDBR.

DISCUSSION

The present study provides novel observations regarding the impact of prolonged HDBR on sympathetic neural discharge patterns at baseline and during reflex-sympathetic activation elicited by end-inspiratory apnea. First, in two separate 60-day HDBR studies, bed rest prolonged the latency of the integrated bursts and the underlying APs within each burst, but not average burst frequency. These neurological changes are attributed to the bed rest condition and not to variability in burst latency between sessions because integrated burst latency was stable in six control individuals across testing sessions separated by 4 wk. Second, reflexive MSNA arousal imposed by end-inspiratory apnea was exaggerated by 60 days of HDBR primarily by mechanisms involving augmented AP firing rate, with modest contributions from the recruitment of latent subpopulations of sympathetic axons. We interpret these findings to suggest that 60 days of HDBR 1) modifies sympathetic neural pathways to prolong synaptic delays or central processing times at one or more sites of sympathetic control and 2) alters central mechanisms regulating the firing rate of active axons and potentially the recruitment of a latent subpopulation of larger sympathetic axons during periods of apnea.

Along with integrated burst frequency, this study examined additional features of the MSNA signal that reflect recruitment and emission patterns. This approach exposed new information such as the upward shift in the latency of baseline integrated bursts in both male and female participants in two 60-day HDBR studies. From these data, we hypothesize that prolonged bed rest modifies synaptic delays or central processing times within the sympathetic neural pathway (49). The stability of integrated burst latency over at least 4 wk in control participants supports this concept. We acknowledge the limitation and offer a caveat here that the control period was shorter than the HDBR period in these retrospective analyses. Nonetheless, these control data argue that baseline burst latency profiles are stable under conditions of normal activity.

Burst latency may be affected by various factors, although a primary determinant appears to be burst size; larger bursts display shorter latencies (49), which is due, in our opinion, to the presence of larger and faster conducting APs (37). However, in this study, the longer burst latency likely did not result from any generalized reduction in burst size with bed rest, as female baseline median normalized burst amplitude was similar pre- to post-HDBR, whereas in male participants, the leftward shift in baseline median normalized burst amplitude was no longer significant when a single observation was removed. Importantly, multiunit AP analysis of the male bed rest data indicated that all sympathetic APs were delayed with HDBR. In accord with previous studies (35, 44), we did not observe differences in baseline burst frequency or incidence with HDBR. Therefore, the present findings indicate that bed rest does not modulate baseline burst frequency but rather alters baseline central processing times or synaptic delays, resulting in prolonged burst latency.

The mechanisms determining sympathetic nerve latency are not known. If this delay reflects a synaptic feature or central processing times, then the possible anatomic sites include any stage of sympathetic integration from the supramedullary modulatory sites, brainstem interneurons, spinal cord integration sites, and ganglia. However, other clues from our earlier studies may provide some insight into the types of factors involved in such a delay. For example, severe baroreflex stress evokes a delay in AP latency (5, 37), whereas several maneuvers with a perceptual aspect, such as maximal end-inspiratory or expiratory apnea, Valsalva’s Maneuver, and fatiguing exercise (but not metaboreflex activation per se), shift the AP cluster size-latency profile downward, reflecting a reduction in the latency of all APs (3–5, 38). These earlier observations suggest that a baroreflex mechanism may affect the overall delay in burst and AP latency observed in the present study. If so, then the baroreflex affects burst frequency differently from burst latency or size, a feature previously suspected in humans (26) and lower mammals (33). Also, our earlier observation that the baroreflex contributes to populations of smaller AP clusters but not the larger and faster conducting axons (39) may represent such a differentiated feature. Brainstem processing of baroreflex inputs likely involves many interneurons (1), and Fagius et al. (14) postulated that more than one neural pathway with varying conduction velocities exists within the human brainstem. Although no anatomic evidence to support this hypothesis in humans exists, multiple neuronal pathways with varying conduction velocities project from the rostral ventrolateral medulla to the intermediolateral cell column of rats (7). Alterations of such conduction delays on the millisecond level will not likely be detected with traditional estimates of baroreflex sensitivity or burst frequency.

Alternatively, prolonged sympathetic AP latencies with HDBR may have resulted from changes in the intrinsic timing of ventricular contraction and the effect that such timing has on cardiac afferent modulation of sympathetic outflow or the timing of the pressure pulse through the aorta and carotid sinus baroreceptor regions. For instance, a previous investigation of the WISE-2005 study demonstrated that women in the control group (i.e., not participating in exercise countermeasures) exhibited an increase in cardiac pre-ejection period with bed rest (22). However, the 0.016-s increase in pre-ejection period observed in this earlier study (the same WISE-2005 study used here) does not account for the 0.04- and 0.06-s increase in burst latency observed in the male and female participants, respectively, studied in the present study.

This study not only examined baseline sympathetic outflow but also examined the details of MSNA recruitment during reflex activation using an end-inspiratory apnea. Previous observations in four astronauts by Eckberg et al. (11) suggest that integrated sympathetic burst frequency responses to end-inspiratory apnea are greater late-mission and on landing day than before spaceflight. The current findings support these earlier flight-based data. Several studies from our laboratory employing multiunit AP analysis and others employing single-unit recordings (32) indicate that end-inspiratory apnea increases the firing frequency of already-active axons and the probability of multiple within-burst axonal firing in young, healthy individuals (4, 46). Our findings from the pre-HDBR session are consistent with these observations. However, following HDBR, we observed larger increases in AP frequency (and incidence) as well as the number of APs per burst in response to end-inspiratory apnea compared with pre-HDBR. Therefore, HDBR enhanced AP recruitment during the apneic stress.

In addition to enhanced axonal firing rate, our previous studies in young, healthy individuals performing apnea in the supine position consistently demonstrate that the sympathetic nervous system recruits a subpopulation of previously silent, larger, and faster conducting sympathetic axons to increase the number of active sympathetic axons firing per burst of activity (3–5, 46). In contrast to these earlier observations, only some individuals demonstrated axonal recruitment in the pre-HDBR session (pre-HDBR: n = 3 of 8; post-HDBR: n = 6 of 8) in the current study, with no statistical difference in the mean group changes in total AP clusters and number of active clusters per burst with end-inspiratory apnea in the pre-HDBR session. We speculate that, at least in the male participants, the −6° head-down position may have attenuated this recruitment phenomenon in the pre-HDBR test through a baroreflex mechanism involving increased venous return, stroke volume, and pulse pressure (50). Nonetheless, compared with pre-HDBR, we observed a modest enhanced recruitment of these larger and faster conducting axons and the number of active sympathetic axons per burst in response to end-inspiratory apnea in the post-HDBR session. Although the exaggerated MSNA response to apnea post-HDBR appears to be driven primarily by the frequency of AP cluster firing with minimal contributions from the appearance of new clusters, these findings suggest that the central neural mechanisms governing sympathetic discharge are enhanced with HDBR.

The mechanisms contributing to enhanced apnea-induced sympathetic axonal firing and recruitment are unknown. Greater sympathetic responses post-HDBR have been attributed to larger reductions in BP and pulse pressure at the onset of end-inspiratory apnea (8, 45). However, because reductions in BP with end-inspiratory apnea onset were similar between sessions in the current study, other factors (e.g., increased chemoreflex sensitivity) are likely responsible for our findings. Furthermore, whether the augmented response to apnea post-HDBR was due to greater chemoreflex stress or a greater perceptual stimulus rather than the exposure to simulated microgravity cannot be explicitly determined in this study. The perceptual stimulus related to the drive to breathe was likely similar pre- and post-HDBR because all participants performed apnea for their maximal voluntary duration. The greater change in integrated burst frequency with apnea pre-HDBR than post-HDBR was not associated with the change in breath-hold time across test sessions (r = 0.10, P = 0.82). Also, the individuals who demonstrated the largest change in burst frequency with apnea across test sessions did not necessarily have the greatest increase in sympathetic outflow. Unfortunately, we were not able to assess whether chemoreflex stress evoked by the apnea was greater post-HDBR. Therefore, the mechanisms mediating the greater post-HDBR sympathetic response to apnea remain unknown.

The physiological significance of the modifications in sympathetic latency and reflex arousal with apnea to neurovascular transduction and cardiovascular control are unclear (41). Our observation of prolonged sympathetic burst latency due to a delay in the underlying APs represents (as discussed above) an intriguing finding from a neurophysiological perspective, as it suggests changes in synaptic delays or central processing. The location, mechanisms, and purpose of these changes remain unknown but do expose a unique aspect available to the sympathetic nervous system for responding to stress. Earlier, the significance of modifiable latency properties in the sympathetic nervous system was hypothesized by Fagius et al. (14) to be reserved for emergency situations to evoke vasomotor adjustments more rapidly. If so, the modest changes in latency observed in this study and in earlier works (4, 5) may affect neurotransmitter release or the postjunctional integration of neural inputs. The recent observations of larger leg vasomotor responses to larger MSNA bursts or integrated area of total bursts (16) support the idea that burst characteristics, and therefore, AP characteristics, contribute to vascular control. Although the mechanisms remain to be understood, the current data add to the growing understanding that complex, variable, and modifiable strategies exist within the sympathetic nervous system.

Methodological Considerations

Prolonged sympathetic burst latency with HDBR may have resulted from a more distal placement of the microelectrode in the post-HDBR session relative to pre-HDBR and/or spinal lengthening with bed rest. Because sympathetic postganglionic nerve APs are conducted at ∼1 m/s (15) and assuming that height increased by ∼1 cm with bed rest (34), the microelectrode would have needed to be inserted 4 cm more distal relative to the fibular head to account for the ∼0.05-s increase in burst latency. This likely did not occur with the fibular head location of the measurement site. In fact, in the male group, the microneurographer (J. K. Shoemaker) documented microelectrode placement and used the same location in the post-HDBR session, reducing the likelihood that electrode placement was responsible for our observations. Also, burst latency increased in the majority of volunteers (n = 21 of 26) participating in two distinct bed rest studies with the same microneurographer (J. K. Shoemaker). Moreover, prolonged latency with bed rest was not a product of analytical bias, as this outcome was observed when manual or automated techniques were employed. Changes in central blood volume due to hypovolemia may have impacted the results. Specifically, HDBR often induces 10–15% reductions in plasma volume (34). However, as central venous pressure was not measured in the current study, the relative contribution of hypovolemia to the present findings cannot be quantified, although previous observations suggest that similar reductions in plasma volume result in a greater proportion of larger bursts at baseline, which may lead to shorter rather than longer burst latencies (28). Finally, our observations may be affected by the implementation of exercise countermeasures in both bed rest studies (male group: n = 8 of 13; female group: n = 6 of 13) as well as the nutrition countermeasure performed in the WISE-2005 study (female group: n = 3 of 13). However, secondary analyses of these issues indicated that our observations regarding baseline burst and AP latency, as well as enhanced burst/AP frequency, AP per burst, and clusters per burst in response to apnea with HDBR, were consistent across countermeasure groups (HDBR by countermeasure interaction, all P > 0.20). Thus, the overall effect of changes in burst and AP latency and AP recruitment during apnea were related to an undisclosed effect of HDBR.

Summary

This study demonstrated chronic malleability of some but not the most commonly studied sympathetic neural discharge properties with 60 days of HDBR. Specifically, 60 days of HBDR did not affect the average burst frequency values observed at baseline but did augment central synaptic delays or processing times at baseline for all APs in two long-term bed rest studies, as inferred from both integrated and multiunit AP analyses. These observations were supported by the stability of this metric in non-HDBR controls across testing sessions separated by 4 wk. Although the apnea maneuver elicited recruitment of a latent group of larger AP clusters, the exaggerated apnea-induced sympathoexcitation following 60 days of HDBR was primarily a function of enhanced firing of AP clusters that were already evident at baseline. This may suggest an enhanced AP cluster firing rate rather than recruitment mechanism. These data suggest that sympathetic discharge strategies are modified with prolonged HDBR, with emphasis on changes in central neural processing. The consequences in terms of the sites of such changes or the impact for end organ control remain unknown.

GRANTS

The WISE-2005 study was supported by contract 9F007-033004/001/ST from the Canadian Space Agency to R. L. Hughson and J. K. Shoemaker, “Centre National d’Etudes Spatiales” (CNES) Grant 4800000367. WISE-2005 was sponsored by the European Space Agency (ESA), the National Aeronautics and Space Administration (NASA), the Canadian Space Agency (CSA), and the French CNES, which was the “promoteur” of the study, according to French law. The study WISE-2005 was performed by MEDES, Institute for Space Physiology and Medicine, Toulouse, France. The 60-day head-down bed rest study in Cologne, Germany, study was partially funded by CNES, the ESA (contract no. 4000113871/15/NL/PG), the Institute of Aerospace Medicine of the German Aerospace Center (DLR), and the Fond Européen de Développement Régional (FEDER). The repeatability study was supported by grants (awarded to J. K. Shoemaker) from the Natural Sciences and Engineering Research Council (NSERC) of Canada, the Heart and Stroke Foundation of Ontario (no. NA 5020), the cooperative activities program (CAP) grant from NSERC, the Canadian Space Agency (216758-98), and the collaborative NSERC-CHRP (Collaborative Health Research Program between NSERC and Canadian Institutes of Health Research) award. Additional funding for this study was provided by NSERC (217916-2013; J. K. Shoemaker). S. A. Klassen was supported by a NSERC Doctoral Scholarship and formerly by an Ontario Graduate Doctoral Scholarship. J. K. Shoemaker is a Tier 1 Canada Research Chair.

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

S.A.K., D.S.K., R.L.H., H.N., and J.K.S. conceived and designed research; S.A.K. and J.K.S. analyzed data; S.A.K. and J.K.S. interpreted results of experiments; S.A.K. and J.K.S. prepared figures; S.A.K. and J.K.S. drafted manuscript; S.A.K., S.D.A., D.K.G., D.S.K., P.A., P.D., R.L.H., H.N., and J.K.S. edited and revised manuscript; S.A.K., S.D.A., D.K.G., D.S.K., P.A., P.D., R.L.H., H.N., and J.K.S. approved final version of manuscript; S.D.A., D.K.G., D.S.K., P.A., P.D., R.L.H., H.N., and J.K.S. performed experiments.