| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Departments of 1Aerospace Physiology and 2Computer Application, Fourth Military Medical University, Xi'an, China; and 3Institute of Sound and Vibration Research, University of Southampton, Southampton, United Kingdom
Submitted 16 February 2008 ; accepted in final form 30 July 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
postflight cardiovascular deconditioning; hindlimb unweighting; intermittent artificial gravity; cardiovascular variability; spectral analysis; nonlinear analysis
Animal studies have their own advantages over human studies not only in elucidating the underlying mechanisms but also in designing new countermeasures. The tail-suspended, head-down tilt (SUS) rat model (39) and the whole body-suspended (WBS) rat model (42) have been used to investigate the mechanisms underlying postflight cardiovascular dysfunction (for review, see 42, 68, 71). Cardiovascular responses to SUS in rats include transient rise in central venous pressure (35, 50, 53), cephalic fluid shift (29), natriuresis and diuresis (15, 35), hypovolemia (9, 22), impaired ability to distribute cardiac output (38), resting and exercise tachycardia (38), altered baroreflex function (6, 9, 19, 30), reduction in maximum oxygen consumption and exercise performance (18, 46, 65), and decreased tolerance to LBNP (45, 67).
In the past decade, depression in myocardial contractility (67) and region-specific adaptation of vessels in different anatomic regions (16, 17, 51, 68, 71) due to SUS have been well documented. It has been speculated that microgravity-induced adaptation in structure and function of myocardium and vessels might be among the most important mechanisms responsible for postflight cardiovascular dysfunction (16, 17, 51, 62, 68, 71). Our work has further shown that daily short-duration exposure to –Gx (dorsoventral) gravitation by standing (STD) to restore the rat's orthostatic posture, or –Gx with +Gz component by +45° head-up tilt, which mimics the IAG countermeasure, is surprisingly effective in preventing myocardial contractility depression (70) and vascular changes (55, 69). For example, it has been demonstrated that daily 1-h –Gx by STD is sufficient to prevent the impairment in myocardial contractility and differential changes in cerebral and hindquarter arteries that might occur due to a 28-day simulated microgravity alone (55, 69, 70). Therefore, it is of great interest to test the hypothesis that the countermeasure of 1-h/day STD can also be effective in preventing postsuspension cardiovascular alterations in conscious rats.
However, postsuspension cardiovascular responses of conscious rats are seldom investigated, and only a few studies have been reported on various results. For example, no significant changes in both blood pressure (BP) and heart rate (HR) (9, 24, 45, 50), a significant tachycardia with a trend of fall in BP (35, 56), and a significant fall in BP with nonsignificant changes in HR (4–6) have all been reported. These discrepancies were apparently related to the lack of consistency in animal model, control group, experimental protocol, recording procedure, etc. In addition, to the best of our knowledge, the duration of suspension was limited to 1–7 days in most of these studies (5, 6, 35–37, 43), 14 days in two studies (24, 56), and 21 days in one (4). For our research goal, it is appropriate to extend the duration to 28 days with more attention to the standardization of the animal model (39) and experimental procedure, because our previous studies were mostly conducted within such a time frame to allow the cardiovascular deconditioning to attain a new steady state (55, 68, 70). According to the findings reported in Refs. 10, 24, 40, 42, 52, and 56, we further hypothesized that the postsuspension cardiovascular manifestation in conscious rats might be hypertensive and tachycardiac.
Moreover, the spectral analysis of HR and BP variability (HRV and BPV) has usually been used in space and bed-rest human studies to probe the overall changes in the dynamics of cardiovascular autonomic regulation (2, 20, 57), but it has rarely been conducted in animal studies. A few studies adopted the nonparametric periodogram method to analyze the postsuspension cardiovascular signals of conscious rats, but no significant changes have been reported (24, 37, 56). Hence, we have adopted a parametric autoregressive (AR) method to analyze the auto- and cross-spectrum of HRV and BPV signals to get smoother spectral components and a more accurate estimation of power spectral density (PSD) (33, 57). Cross-spectral analysis and follow-up transfer function analysis may provide information regarding the sensitivity of the baroreceptor-HR reflex responsiveness (BRS) (11, 66). However, in previous studies on postsuspension BRS, the pharmacological method was used exclusively, but the results reported were inconsistent (6, 19, 24). In addition, the cardiovascular signals were also analyzed by the recurrence quantification analysis (RQA) (14, 63) and the approximate entropy (ApEn) measures (49). It was speculated that these nonlinear methods might provide valuable information on the dynamics of cardiovascular regulation in conscious rats after SUS with and without countermeasure.
Therefore, the aims of the present study were 1) to assess the BP and HR responses of conscious rats having been subjected to a 28-day SUS before and immediately after being released from suspension and during the first 2-h period of postsuspension, 2) to estimate the overall changes in the dynamics of cardiovascular autonomic regulation by auto- and cross-spectral analysis of the HRV and BPV signals using the AR method and by nonlinear ApEn and RQA measures, and 3) to test the hypothesis that the postsuspension cardiovascular changes in conscious rats after a 28-day simulated microgravity can be prevented by a gravitation-based countermeasure of 1-h/day –Gx by STD.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Tail-suspended, hindlimb-unloaded rat model. The technique of tail suspension (39) with modification from our laboratory has been described in detail previously (55, 70). The rats were attached via the plastic bar in the tape to a swivel mounted at the top of the cage allowing free 360° rotation. The rats were maintained in about –30° head-down tilt position with their hindlimbs unloaded. The controls were housed in identical Plexiglas cages, except that the tail suspension device was removed. All animals received standard lab chow and water ad libitum and were caged individually in a room maintained at 23°C on a 12:12-h light-dark cycle.
Model of daily short-duration –Gx gravitation. Daily stationary ground support in normal orthostatic posture, or standing (STD), for 1 h was adopted to simulate the countermeasure effect of IAG as previously described (55, 69, 70). For short-duration STD, the suspended rat was released from suspension and then placed into a 50-cm-long, tubelike metallic mesh cage maintained in horizontal position for 1 h. The rat could move forward and backward, but it could not turn around. Food and water were provided ad libitum at the front end of the cage. The gravity vector was –Gx.
Experimental design. All protocols and procedures were reviewed and approved by the Animal Care and Use Committee of the Fourth Military Medical University. Twenty-four male Sprague-Dawley rats obtained from the Animal Center of the Fourth Military Medical University were randomly assigned to three experimental groups (n = 8 rats/group): control (CON), tail suspension (SUS), and suspension for 23 h/day plus STD for 1 h/day (SUS + STD1). During the 28-day period, daily 1-h –Gx gravitation treatment was conducted between 0800 and 1200.
Hemodynamic parameters of conscious rats from the three groups were recorded on suspension day 29. After the completion of physiological recordings, the animals were anesthetized with pentobarbital sodium (50 mg/kg ip) and killed by exsanguination via the abdominal aorta. The left soleus and tibia were removed, and muscle wet weight and bone length were measured to confirm the deconditioning and the efficacy of countermeasure and to monitor any effect on growth.
Direct Arterial BP Recording in Conscious, Free-Moving Rats
Surgical procedures. Forty-eight hours before recording, the rats were anesthetized with pentobarbital sodium and were chronically cannulated. A polyethylene catheter (PE-10 connected to PE-50) filled with heparinized saline (320 U/ml) was inserted via the right femoral artery into the lower abdominal aorta. Then the catheter was tunneled subcutaneously, exteriorized, and fixed onto the saddle mounted on the posterior cervical region and was plugged. After surgery, all rats were placed in their individual cages for recovery, and the SUS and SUS + STD1 rats maintained their head-down tilt posture without and with the countermeasure.
Hemodynamic recording. After 48 h of recovery, the rats were placed in their individual recording cages in a quiet room, and the arterial catheter was connected to a Statham pressure transducer (P23 ID, Gould Statham) 40 cm above the rat, via a rotating swivel, which allowed the rats to move freely. The pressure transducer was calibrated and adjusted to 0 mmHg at the heart level. To avoid clotting, the arterial catheter was continuously flushed with a heparin-containing saline (320 U/ml) at a rate of 0.3 ml/h by a microsyringe (WZS-50F2, Zhejiang University, Hangzhou, China) while recording was not made, and the total amount of infused saline was <1.5 ml. The transducer output was amplified by a carrier amplifier (AP-621G; Nihon-Kohden, Tokyo, Japan), digitized by an analog-to-digital converter (PCI-6220, National Instrument), and stored in a computer (Dell 5100). BP recording was started 4 h after the rats had been connected to the pressure transducer at the following four sessions each for 20 min. They were: just before the release from suspension (R-0–), and immediately (R-0+), 1 h (R-1h), and 2 h (R-2h) after release. During the R-0– session, the CON rats remained in their orthostatic posture.
Signal Analysis of Systolic BP and HR Time Series
Preprocessing of digitized BP data. Off-line data preprocessing was performed as previously described (61). Briefly, the peaks and valleys of the arterial pressure were detected according to a strategy incorporating thresholding, template-matching, and wavelet transform that minimizes the influence of baseline changing. The time series of systolic and diastolic BPs (SBP and DBP) were obtained as the amplitude of arterial pressure at the peaks and valleys, and HR was derived as the inverse of the peak-to-peak interval. Stationary SBP and HR signal segments of 2 min were selected from each condition. The segments were resampled at 12 Hz using spline interpolation and linearly detrended for further analysis.
Autospectral analysis of SBP and HR signals.
The autospectra of these data sets were first estimated by the classical periodogram method based on the fast Fourier transform. Then autospectral analysis of the same data sets was performed by the autoregressive (AR) method. The model parameters were estimated by the Marple algorithm (33), which can be expressed as
 | (1) |
) is the autospectrum obtained and p is the model order. To reduce the estimation variation, running window of 30 s with 10 s overlapping was applied. The autospectra were estimated from each 30-s data segment of 2 min and then averaged. The optimal number of order is estimated according to the Akaike information criterion (AIC). After comparing order numbers from 10 to 25, a fixed order number 20 was selected as it was close to the average optimal order number of all data and it minimized the influence of varying order. The definitions of frequency domain spectral indexes of the short-term HRV and SBP variability (SBPV) are according to the reference (57), and the range of the frequency band of each spectral component specifically for rats are according to Yang et al. (66).
Cross-spectral analysis of SBP and HR signals.
Cross-spectrum of the same data sets was carried out with the AR method from which two follow-up functions were estimated. First, a magnitude-squared coherence function was estimated
 | (2) |
0.5 was considered to be statistically significant. Second, a transfer function was estimated
 | (3) |
 | (4) |
index, may be used to estimate the baroreflex sensitivity (11, 66).
RQA of SBP and HR time series.
The method of RQA was first introduced to visualize the time-dependent behavior of the dynamics of systems (63). RQA was performed as previously described (14, 63). Briefly, the recurrence plot was constructed by embedding the SBP or HR data in a p-dimensional Euclidean space, and the three important features of the signals were calculated. The main step of this visualization is the calculation of the N x N matrix:
 | (5) |
is a small threshold distance, ||·|| is a norm, and 
(·) is the Heaviside function. Yt = {x(t), x(t + 
), ..., x[t + (m – 1)]}, where is the time delay and m is the dimension of the system. Some measures are extracted from the recurrence plot (RP) based on the recurrence point density and the diagonal structures in the RP. We calculated the recurrence percentage (RC%), determinism percentage (DT%), and the longest diagonal line (Lmax) in the plot. RC% and DT% are a measure of the stationarity and regularity of the signal over time. Lmax is inversely related to the Lyapunov exponent. The presence of diagonal lines in recurrence plots also indicates that deterministic rules are present in the dynamic (63).
ApEn measurement.
ApEn is a statistic quantifying the regularity and complexity of a time series (49). ApEn is defined by ApEn(m, r, N) = 
m(r) – m+1(r), where m(r) is defined as the average value of lnC
(r), i.e., m(r) = (N – m + 1)–1 
lnC (r).
For this study, ApEn was calculated over n = 1,000 consecutive data points in a data set. We chose m = 2 and r = 15% of the SD of the data set.
Statistical Analysis
Values are expressed as means ± SE (except for body weight data, which are means ± SD). Two-way ANOVA with repeated measures was used to determine the overall differences among different groups and different recording sessions of the same group, and then the Student-Newman-Keuls post hoc test was used to determine the group differences. The 0.05 level of probability was chosen as significant for all analyses.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
The data are summarized in Table 1. There were no significant differences in the final body weight and tibia length among the three groups. The wet weight of soleus of SUS rats was 46% less than that of the CON rats (P < 0.01). However, the soleus wet weight of SUS + STD1 rats was 27% less than the CON rats (P < 0.01), indicating the countermeasure effectiveness of 1-h/day STD in attenuating muscle atrophy (69, 70).
|
Representative tracings of BP and HR of one CON, one SUS, and one SUS+ STD1 rat over a 2-min period from each of the four recording sessions (R-0–, R-0+, R-1h, and R-2h) are shown in Fig. 1. In both the SUS rat and SUS + STD1 rat, no noticeable step changes were observed in both the BP and HR between the R-0– and R-0+ session, except that in the SUS rat both the BP and HR showed enhanced and less regular oscillations during R-0–. The BP and HR level of the SUS rat was apparently higher than the CON rat, whereas the BP level of the SUS + STD1 rat was comparable with that of the CON rat. The group data are summarized in Table 2. ANOVA confirmed that the SUS group showed a significantly higher SBP, DBP, and HR during all the four recording sessions compared with the CON rats (P < 0.05, or <0.01). However, for the SUS + STD1 group, the SBP value was significantly lower than that of the SUS group (P < 0.05), but the HR was significantly higher than the CON group (P < 0.05).
|
|
Representative autospectra of the SBP and HR data sets computed by the AR method from one CON, one SUS, and one SUS+ STD1 rat are shown in Fig. 2A. Two oscillatory rhythms at distinct frequencies with peaks are evident for both the SBP and HR time series. The LF oscillation in the rats ranges from 0.25 to 0.8 Hz, and the HF oscillation is 0.8–2.4 Hz. Table 3 summarizes the group data computed by the AR method. For comparison, data computed by the periodogram method for the same data sets are listed in Table 4. The results obtained by these two methods are comparable. Except that the total power (TP) and the power of LF of HRV in the SUS and SUS + STD1 groups were significantly higher than that of CON group during the R-0– session, no significant differences in the spectral indexes of SBP and HR were detected either among the three groups or different sessions of the same group.
|
|
|
Cross-spectral analysis further revealed a close link between the SBP and HR oscillations in the LF and HF bands. Representative plots of coherence and transfer function between SBP and HR computed by the AR method are shown in Fig. 2B. Coherence 0.5 is considered to be statistically significant, and corresponding frequency ranges in the magnitude of transfer functions are denoted by a heavy line. The group data of the transfer function magnitude are summarized in Table 5. No significant differences in the magnitude data were detected by the ANOVA either among the three groups or within the same group in different sessions.
|
The results of RQA analysis are depicted in Fig. 3. SUS resulted in a significant decrease in all the three nonlinear indexes (RC%, DT%, and Lmax) of SBP (P < 0.01) compared with those of the CON group, whereas 1-h/day STD partially alleviated the decrement of these indexes. Although all the values of the three indexes of the SUS + STD1 rats were still significantly smaller (P < 0.01) than those of CON rats, they were greater than SUS rats with the differences in DT% and Lmax being significant (P < 0.05 or <0.01) (Fig. 3A). In contrast, there were no significant differences in RC% and DT% of HR among the three groups. However, SUS significantly decreased Lmax of HR and daily 1-h STD did not show any alleviating effect (Fig. 3B). For all the three RQA indexes of SBP and HR, no significant differences were found among the four recording sessions of each group (Fig. 3). Figure 4 shows the differential effect of SUS on the ApEn measure of HR and SBP. In all the four sessions, there were no significant differences in the ApEn value of HR between the SUS and SUS + STD1 groups, but the ApEn value of both the two groups was significantly greater (P < 0.05) than that of the CON group. In contrast, the ApEn value of SBP did not show significant change either among the three groups or within the same group in different sessions.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Postsuspension Cardiovascular Manifestation of Conscious Rats
Previously reported postsuspension cardiovascular response has been variable and inconsistent (4–6, 9, 24, 35–37, 45, 50, 56). This discrepancy should be further elucidated, because it is related to two major concerns of whether the SUS rat is a valid model for studying cardiovascular deconditioning due to microgravity exposure (24, 42) and whether the countermeasure efficacy evaluated solely according to postsuspension cardiovascular response of conscious rats is reliable. In our view, the differences in rat model, control group, suspension duration, and BP measurement might be among important factors that lead to these discrepancies.
1) Rat model. Compared with the whole body suspension (41), the tail-suspension appears to be less stressful (39). However, important cautions and technical considerations listed in Ref. 39 have not been sufficiently mentioned in some work. This makes it difficult to clarify the technical reasons for these discrepancies.
2) Control groups. More than one control group may be needed in some cases (39). Caution should be taken to interpret the results while the tail-attached (or -tethered) orthostatic rats was used as the sole controls as in some studies (4–6, 24, 35), since attachment itself has the restraint effect and may impose a stress on the animal. However, it remains of great interest to further elucidate the underlying mechanisms involved (8).
3) Suspension duration. It is well recognized that the first 2 days of suspension are a time of particular stress evidenced by loss of body weight and reduction of water and food intake (42). Popovic (50) reported that the mean BP was decreased during the first 2 days of suspension, followed by an elevation to normal level. It should also be noted that the myocardial and vascular adaptations to simulated microgravity occur over the first 4 wk of SUS and approach a steady state thereafter (68, 71).
4) BP measurement. The following factors might bias the true result. For example, the position of catheter tip varied from the femoral artery (5, 35, 37, 40), the common carotid artery (4), to the lower abdominal aorta (24, 56) and the aortic arch (43), and the pressure transducer was positioned at heart level (4, 5, 35, 37, 40, 43) or a definite distance above the free-moving rat (24, 56). The postsurgical recovery period was 24 h (56) or 48 h (24, 35, 37, 40). Recently, it has been shown that the BP level of the mice was persistently elevated during the first 4 days after implanting the probe and gradually returned to its normal level with physiological diurnal rhythms only 10 days after surgical operation (64).
In the present study, it seems that the potential influence of the possible confounding factors, such as the rat model, body weight, control group, and short duration, should all be excluded except for the factor of short recovery after surgery. Nevertheless, we believe that the hypertensive and tachycardiac response just before and on postsuspension is a "true" response to a 28-day SUS, not simply a general stress response, because the postsurgical recovery period was identical for all the three experimental groups, but only the SUS rats did show significant changes in BP and HR. Our findings are consistent with that reported by Mueller et al. (40). They showed that on postsuspension both the mean BP and HR were significantly higher in conscious rats after a 14-day SUS. Furthermore, their experimental results have further suggested that a 14-day SUS may reduce peripheral NO and cardiac parasympathetic tone and increase the sympathetic tone to the heart and possibly the vasculature (40). Actually, curves depicted in Fig. 2 of Ref. 24 also showed the trend of elevated SBP, DBP, and mean BP in suspended rats compared with the controls before and after releasing from suspension. Thus the postsuspension hypertension is not due to the change in gravity vector on releasing but reflects changes in cardiovascular function of intact animals during a medium-term SUS.
Given that there is a greater difference in gravitationally dependent distribution of blood volume between the small quadruped rats and the bipedal humans (52), it is reasonable to assume that the postsuspension cardiovascular response is not necessarily to be fully comparable with that of humans postflight or post-bed rest. It is also reasonable to consider that the lack of postsuspension hypotension in conscious rats does not mean that the SUS rat model is invalid for studying the mechanism of cardiovascular deconditioning due to microgravity as questioned in Refs. 24 and 42. Moreover, integrative studies using dynamic tests to challenge the overall circulatory function, such as LBNP (45, 67), tilting (35, 67), exercise (18, 38, 46, 65), autonomic blockade, sympathectomy, sympathomimetics infusion, or NO-nitric oxide synthase inhibition (40, 45, 56), and hemorrhage (30) have provided data supporting postsuspension cardiovascular dysfunction.
Then how to integrate the findings from the present intact animal study with those from studies on vascular function at the organ level (16, 17, 68)? One possible explanation is that the sympathetic hyperactivity is a compensatory mechanism mobilized in the intact animal to counteract the SUS-induced hyporesponsiveness of resistance vessels. Two relevant studies have provided evidence supporting our reasoning. Tarasova et al. (56) reported that although intact postsuspension rats showed diminished pressor responses to sympathomimetics, their mean BP level was comparable to that of intact controls, and postsuspension hypotension became apparent only after the elimination of sympathetic influences by chemical sympathectomy. This is also consistent with the findings that a 21-day SUS results in a decreased responsiveness of hindlimb vessels to sympathetic nerve stimulation (51). Mueller et al. (40) showed that hexamethonium-induced depressor response was enhanced after a 14-day SUS. In addition, postsuspension sympathetic hyperactivity is also evidenced by an increased norepinephrine turnover in medulla oblongata and cardiac and kidney tissues examined 6 h after release (25). In human studies, reduced responsiveness of the resistance vessels to the increased sympathetic nerve traffic has also been observed in postflight astronauts (10). Given that the postsuspension hypertension is a main manifestation of cardiovascular deconditioning in rats, we try to explain why 1-h/day –Gx countermeasure is also effective in intact animals. It seems that rats having been subjected to a daily 1-h standing no longer need a sympathetic hyperactivity to maintain their circulatory homeostasis while being released from SUS and reassuming their normal horizontal posture, since the decrement in myocardial contractility and vascular responsiveness of resistance vessels has been fully prevented (55, 70). Therefore, present integrative animal study has also provided a mechanistic view on the efficacy of IAG in preventing cardiovascular deconditioning in humans during prolonged microgravity exposure.
Signal Analysis of BPV and HRV of Conscious Rats
Few studies using spectral analysis of cardiovascular signals have been performed in rats (1, 12, 31, 34, 48, 66). However, in the previous work, the periodogram method was used exclusively. Although the AR method demonstrates an excellent performance in estimating high-resolution spectra out of short data segments, the difficulty and complexity in choosing an appropriate order of autoregression (p) in computation have prevented its use in studies with rats (12, 34). In the present study, the AR method was used and compared with the results obtained by the periodogram method. The number 20 was chosen as the model order (p) according to the information criteria and repetitive trials. Although smoother spectral curves were obtained by using the AR method (Fig. 2), the spectral components and the powers of spectral indexes computed by the two methods are comparable (see Tables 3 and 4). The result of cross-spectral analysis is also similar to that computed by the Fourier transform (11, 66). While this study was in preparation, we learned that the AR method has been used in autospectral analysis of mice cardiovascular signals; however, the model order issue has not been mentioned in this reference (26).
As in humans and dogs, the spectral distribution of BPV and HRV in conscious rats is also characterized by the three principal spectral components, but with different frequency ranges due to a faster HR in rats. Except that during the suspension period, the powers of TP and LF of HRV in SUS and SUS + STD1 groups were significantly greater than that of the CON group, and no significant differences in the spectral indexes of the SBPV and HRV were found among the three groups. Despite that the AR method was adopted, our results are also consistent with those obtained by the three previous studies using periodogram (24, 37, 56). With regard to the statement by Tarasova et al. (56) that "transition from rest to movement strongly affected fluctuations of MAP (mean arterial pressure)," we suggest that it should be further verified by the time-frequency methods, since traditional spectral analysis requires stationarity of the signals along a considered time window. Finally, the results of cross-spectral and coherence analyses confirmed that SBP and HR signals in our experiment are linearly related at both the LF and HF bands as previously reported (11, 66). However, no significant changes were found in the results of transfer function analysis among CON, SUS, and SUS + STD1 groups. It has been established that the magnitude of the transfer function can also be used as an estimate of the baroreceptor-HR reflex sensitivity (BRS) in rats (11, 66). To the best of our knowledge, in the previous studies, the BRS of postsuspension rats were evaluated by the pharmacological method exclusively, but with variable results. Both impairment (6, 19, 36, 67) and no modifications (24) in BRS have been reported. Our result by spectral analysis is consistent with that reported by Fagette et al. (24).
Why were there no noticeable changes in spectral indices of cardiovascular signals among the three groups? One possible explanation is that the conventional, short-term spectral analysis of HRV and BPV has many limitations, making it difficult to reveal subtle changes in spontaneous HR and BP oscillations. Another possible explanation is that the HRV spectral indexes measure fluctuations in autonomic inputs rather than the mean level of autonomic inputs, i.e., it can be used as markers of cardiac autonomic influences on the modulation of HR, but cannot be used as measures of the tone of cardiac autonomic influences under various conditions (57). Similar discrepancies have also been reported in studies on BPV of genetically hypertensive rats (12, 48).
The method of nonlinear dynamics opens up a new and essentially different approach to detect the complexity of cardiovascular autonomic modulation in different physiological states (7, 14, 28, 49, 63). RQA can easily quantify the subtle patterns by its quantitative descriptors in much richer dynamics, which are not apparent in the original time series (63). The present study has provided promising data. SUS for 28 days resulted in a significant decrease in the three RQA indexes of SBP (RC%, DT%, and Lmax) (P < 0.01), which was partially alleviated by daily 1-h STD (see Fig. 3). These changes do agree with the trend that SUS results in postsuspension hypertension that can be prevented by 1-h/day –Gx. Dabire et al. (14) reported that modifications of RQA indexes are negatively correlated to changes of BP level. Sympathetic blockade by 1-adrenoreceptor blockade or ganglionic blockade by hexamethonium decreased BP level but increased the three nonlinear indexes of BP (14). Another observation is that in spontaneously hypertensive rats (SHR), the nonlinear index Lmax is significantly lower than that of the normotensive Wistar-Kyoto (WKY) rats; treatment with clonidine (a drug known to inhibit sympathetic tone) significantly decreased BP level but increased Lmax of BPV (14). This is also in accordance with one key point of RQA that Lmax is the reverse of the Lyapunov exponent. It has been shown that, in conscious dogs, baroreceptor denervation increased BP level but reduced the highest Lyapunov exponent (cited from Ref. 14). SUS, but not SUS + STD1, also results in a significant decrease in Lmax of HRV, which suggests a modulation on cardiac parasympathetic tone during SUS without the countermeasure. Relevant studies have shown that cholinergic blockade increases the Lmax (14) and decreases the Lyapunov exponent (7) of the HRV in rats. With regard to ApEn measures, only the results of HRV showed significant changes. ApEn values of HRV in both SUS and SUS + STD1 groups were significantly higher (P < 0.05) than the CON group (see Fig. 4). Higher values of ApEn indicate a more complex structure in the time series (49). A plausible explanation is that the vagally mediated HRV is enhanced to oppose the rise in BPV in SUS and SUS + STD1 groups during the four recording sessions, as it has been shown that the HRV may play an antioscillatory role in stabilizing the BPV (27, 58). The present study also supports the idea that RQA and ApEn represent a robust approach to space cardiovascular research, and they can be applied to rather short and even nonstationary data (49, 63).
Lack of clear and distinct changes in HRV and BRS has also been reported in some recent space and ground-based human studies (2, 20). Di Rienzo et al. (20) have suggested that the reported reduction of BRS at reentry is not caused by a baroreflex deconditioning during flight but rather is due to the influence of other factors. Thus it is also speculated that methods of nonlinear dynamics might provide valuable information in future ground-based and on-board human studies. For example, cross recurrence quantification (CRQ), a bivariate tool to quantify the RP, has been used to investigate the decoupling between HR and SBP of bed-rest subjects with promising data (3).
Study Limitations
First, a longer postoperative recovery period should be considered to exclude the possible effects of operation on cardiovascular responses of conscious rats (64). Second, a three-dimensional spectral analysis and time-frequency analysis of a longer data segment is worth trying to avoid the bias due to the many limitations brought about by the batch analysis of short data sets. Third, a battery of tools derived from nonlinear dynamics would be better suited for characterizing a system in its entirety (7, 49, 57, 63). Fourth, a possible systemic error due to the use of pulse interval instead of RR interval from ECG recording should be further assessed.
In conclusion, the present study has demonstrated that conscious rats released from a medium-term (28 days) tail-suspension manifest hypertensive and tachycardiac response, which can be prevented by a countermeasure of daily 1-h –Gx exposure. Although auto- and cross-spectral analysis does not reveal clear and distinct changes, the nonlinear methods, RQA and ApEn measure, have provided promising data in detecting subtle changes in cardiovascular autonomic modulation due to SUS with and without countermeasure.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
O2max and regional blood flow responses to exercise. J Appl Physiol 66: 653–659, 1989.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
