Enhanced open-loop but not closed-loop cardiac baroreflex sensitivity ...

Am J Physiol Regul Integr Comp Physiol 301: R1591–R1598, 2011. First published September 7, 2011; doi:10.1152/ajpregu.00347.2011.

Enhanced open-loop but not closed-loop cardiac baroreflex sensitivity during orthostatic stress in humans Toshinari Akimoto,1 Jun Sugawara,2 Daisuke Ichikawa,1 Nobuyuki Terada,1 Paul J. Fadel,3 and Shigehiko Ogoh1 1

Department of Biomedical Engineering, Toyo University, Kawagoe-Shi, Saitama; 2Human Technology Research Institute, National Institute of Advanced Industrial Science and Technology, Tsukuba, Ibaraki, Japan; and 3Department of Medical Pharmacology and Physiology, Dalton Cardiovascular Research Center, University of Missouri, Columbia, Missouri Submitted 28 June 2011; accepted in final form 5 September 2011

Akimoto T, Sugawara J, Ichikawa D, Terada N, Fadel PJ, Ogoh S. Enhanced open-loop but not closed-loop cardiac baroreflex sensitivity during orthostatic stress in humans. Am J Physiol Regul Integr Comp Physiol 301: R1591–R1598, 2011. First published September 7, 2011; doi:10.1152/ajpregu.00347.2011.—The neural interaction between the cardiopulmonary and arterial baroreflex may be critical for the regulation of blood pressure during orthostatic stress. However, studies have reported conflicting results: some indicate increases and others decreases in cardiac baroreflex sensitivity (i.e., gain) with cardiopulmonary unloading. Thus the effect of orthostatic stress-induced central hypovolemia on regulation of heart rate via the arterial baroreflex remains unclear. We sought to comprehensively assess baroreflex function during orthostatic stress by identifying and comparing open- and closed-loop dynamic cardiac baroreflex gains at supine rest and during 60° head-up tilt (HUT) in 10 healthy men. Closed-loop dynamic “spontaneous” cardiac baroreflex sensitivities were calculated by the sequence technique and transfer function and compared with two open-loop carotid-cardiac baroreflex measures using the neck chamber system: 1) a binary white-noise method and 2) a rapid-pulse neck pressure-neck suction technique. The gain from the sequence technique was decreased from ⫺1.19 ⫾ 0.14 beats·min⫺1·mmHg⫺1 at rest to ⫺0.78 ⫾ 0.10 beats·min⫺1·mmHg⫺1 during HUT (P ⫽ 0.005). Similarly, closed-loop low-frequency baroreflex transfer function gain was reduced during HUT (P ⫽ 0.033). In contrast, open-loop low-frequency transfer function gain between estimated carotid sinus pressure and heart rate during white-noise stimulation was augmented during HUT (P ⫽ 0.01). This result was consistent with the maximal gain of the carotid-cardiac baroreflex stimulus-response curve (from 0.47 ⫾ 0.15 beats·min⫺1·mmHg⫺1 at rest to 0.60 ⫾ 0.20 beats·min⫺1·mmHg⫺1 at HUT, P ⫽ 0.037). These findings suggest that open-loop cardiac baroreflex gain was enhanced during HUT. Moreover, under closed-loop conditions, spontaneous baroreflex analyses without external stimulation may not represent open-loop cardiac baroreflex characteristics during orthostatic stress. arterial blood pressure; arterial baroreceptors; neck pressure; neck suction; head-up tilt

(ABP) is well maintained during orthostatic stress, despite a decrease in stroke volume, as unloading of the arterial and cardiopulmonary baroreceptors leads to increases in heart rate (HR), sympathetic nerve activity, and peripheral vascular resistance (22, 33, 34, 42). Notably, carotid baroreflex responses to orthostatic stress are not only due to changes in carotid sinus pressure but also may be influenced by an interaction with cardiopulmonary baroreceptors responding to decreases in central blood volume or pres-

ARTERIAL BLOOD PRESSURE

Address for reprint requests and other correspondence: S. Ogoh, Dept. of Biomedical Engineering, Toyo Univ., 2100 Kujirai, Kawagoe-shi, Saitama 350-8585, Japan (e-mail: [email protected]). http://www.ajpregu.org

sure (6, 26, 29, 43). Indeed, in human studies, reductions in cardiac filling with low-level lower body negative pressure to unload cardiopulmonary baroreceptors augment carotid baroreflex control of HR (4, 29) and forearm vascular resistance (6, 43). Although the selectivity of low-level lower body negative pressure for stimulation of cardiopulmonary receptors has continually been challenged (9), these data clearly indicate the potential influence of unloading cardiopulmonary baroreceptors on carotid baroreflex function. Moreover, it has been suggested that the interaction between carotid and cardiopulmonary baroreflexes is critical for the maintenance of blood pressure during orthostatic stress (5, 26). Therefore, this neural reflex interaction may be of particular importance during postural changes. Numerous techniques assessing the dynamic relationship between “spontaneous” fluctuations in ABP and HR, including the sequence technique (16), the autoregressive multivariate technique (1), transfer function analysis (2, 11, 13, 36, 44), and the autoregressive exogenous input model (24), have been used to provide an estimate of cardiac baroreflex sensitivity. These techniques are widely used, because baroreflex sensitivity can be derived from continuous recordings of cardiovascular data without physiological perturbation to the arterial baroreceptors. Overall, there is considerable debate on the utility of these dynamic baroreflex measures to estimate cardiac baroreflex sensitivity (3, 8, 21, 30). This is likely due, in part, to the inability of dynamic measures to isolate arterial baroreflex responses from other closed-loop influences. Indeed, dynamic closed-loop baroreflex analyses have been shown to underestimate the open-loop characteristics of the arterial baroreflex (18). Previous human studies employing these spontaneous dynamic analyses reported conflicting results regarding the interaction between the arterial and cardiopulmonary baroreceptors during orthostatic stress. In contrast to previous reports using the neck chamber technique to directly stimulate carotid baroreceptors (4 – 6, 26, 29, 43), dynamic analyses have suggested that orthostatic stress-induced central hypovolemia caused a reduction in cardiac baroreflex sensitivity (1, 2, 10, 11, 13, 36, 41, 44). The exact reason for these discrepancies is unclear but may be due to the different approaches used to assess baroreflex control of HR, in that the results from spontaneous dynamic analyses may not represent open-loop baroreflex characteristics. With this background in mind, the purpose of the present study was to comprehensively assess cardiac baroreflex function during orthostatic stress in humans. This was accomplished by identifying open-loop cardiac baroreflex gains at rest and during 60° head-up tilt (HUT) and comparing them

0363-6119/11 Copyright © 2011 the American Physiological Society

R1591

R1592

CARDIOPULMONARY AND CAROTID BAROREFLEX INTERACTIONS

with closed-loop spontaneous dynamic cardiac baroreflex measures. Kawada et al. (18) demonstrated that a binary whitenoise perturbation technique could identify the carotid sinus open-loop baroreflex transfer characteristics under closed-loop conditions. We employed this technique using the variablepressure neck chamber system to investigate the effect of orthostatic stress on the carotid sinus open-loop dynamic baroreflex transfer characteristics. Carotid-cardiac baroreflex measures were also derived from the rapid-pulse neck pressure (NP)-neck suction (NS) technique, a more traditional openloop carotid-cardiac baroreflex function measure. Both measures were compared with closed-loop dynamic cardiac baroreflex sensitivity derived from spontaneous oscillations in ABP and HR calculated using the sequence technique and the transfer function. Importantly, the assessment of open- and closed-loop methods during orthostatic stress in the same set of individuals would allow for direct comparisons of these methods, data that are not currently available in the literature. Indeed, it is difficult to compare data from previous reports using closed-loop baroreflex measures in one set of subjects with data from another study in a completely different group of subjects using open-loop baroreflex measures. In addition to different subject populations, direct comparisons are further confounded by the potential influences of different protocols and data analyses, as well as the degree of orthostatic stress induced. We tested the hypothesis that orthostatic stress would enhance carotid-cardiac open-loop baroreflex gain measurements, whereas in the same subjects a closed-loop spontaneous measure would be unable to detect this increase. METHODS

Subjects Ten healthy men (mean ⫾ SD: 22 ⫾ 4 yr old, 170 ⫾ 6 cm height, 65 ⫾ 9 kg body wt) volunteered to participate in the study. Subjects were moderately active and typically engaged in low (e.g., walking)to-moderate (e.g., jogging, stationary bike)-intensity aerobic activities (3– 4 days/wk), but importantly none competed in endurance events. Each subject provided written, informed consent after all potential risks and procedures were explained. All experimental procedures and protocols conformed to the Declaration of Helsinki and were approved by the Institutional Review Board (IRB) of the Faculty of Biomedical Engineering, Toyo University (IRB no. 2009-R-01). All subjects were free of known cardiovascular or neurological disease, and no subjects were taking prescription or over-the-counter medications. Each subject was familiarized with the equipment and procedures before the experimental session. Physiological Measurements HR was continuously monitored using a lead II ECG. Beat-to-beat ABP was measured using noninvasive tonometry monitoring (model BP-608, Omron Colin, Kyoto, Japan) on the wrist of the left hand (the radial artery), which was supported at the level of the right atrium on an adjustable padded bedside support. The respiratory rate was counted from the respiratory waveform by a nose-tip thermistor (model TR-511G, Nihon Kohden, Tokyo, Japan). All measurements were connected to an acquisition system (PowerLab 8/30, ADInstruments, Bella Vista, Australia) interfaced with a personal computer equipped with data acquisition software (LabChart 6, ADInstruments) for beat-to-beat recording of variables. Waveforms were sampled at 1 kHz, and real-time beat-to-beat values of HR and systolic, mean, and diastolic blood pressures (SBP, MBP, and DBP) were stored for off-line analysis. AJP-Regul Integr Comp Physiol • VOL

Experimental Protocols On experimental days, the subjects arrived at the laboratory ⱖ2 h following a light meal. The subjects were placed supine on a tilt table and then instrumented for HR, ABP, and respiratory rate. In addition, the subjects were fitted with a malleable lead neck collar that encircled the anterior two-thirds of the neck for application of NP and NS. The experimental session was composed of three 60° HUTs separated by ⱖ20 min. During HUT, the subjects were supported by a bicycle saddle without footboard to reduce leg muscle contraction and were requested to abstain from leg movement to reduce muscle pump activity. Each HUT was preceded by a period of quiet supine rest in which one of the baroreflex protocols described below was performed. The subject was then tilted to 60°, and after 3 min the same baroreflex protocol was implemented. An automatically controlled bed tilted the subject from supine to 60° at a rate of 2–3°/s. Thus, in ⬃30 s, the subject was tilted from supine to 60°. The order of the baroreflex protocols was randomized among the subjects. Protocol 1: closed-loop baroreflex identification. Baseline hemodynamic data were collected over a period of 10 min during supine rest for calculation of cardiac baroreflex sensitivity using the sequence technique and transfer function analysis between HR and ABP without baroreceptor stimulation. Then the subject was tilted for ⬃13 min, with data collected for HUT sequences and transfer function analysis in the last 10 min. During data collection, the subject’s respiration was controlled to maintain the respiratory rate relatively constant by using a metronome (⬃0.26 Hz). Protocols 2 and 3: open-loop identification protocols. Carotid cardiac baroreflex function during supine rest and HUT was determined using the following two variable-pressure neck chamber methods. PROTOCOL 2: WHITE-NOISE STIMULATION. The carotid sinus pressure was perturbed by a neck chamber system according to a binary white-noise signal (14) with a minimum switching interval of 0.5 s. We first generated a Gaussian white noise with zero mean based on a Fourier type of expansion (23) and treated the negative values as high values (⫺40 mmHg) and all others as low values (0 mmHg) of the binary signal. We used different sequences of a binary white noise for different subjects to ensure random stimulation. The white-noise stimulation was applied for 10 min at rest and during HUT. Upon assumption of the 60° HUT, we waited 3 min before applying white noise. During data collection, the subject’s respiration was also controlled to maintain the respiratory rate constant by using a metronome (⬃0.26 Hz). PROTOCOL 3: RAPID-PULSE NP AND NS STIMULATION. Rapid trains of NP and NS were performed based on the protocol outlined by Pawelczyk and Raven (29). Twelve consecutive pulses of NP and NS ranging from ⫹40 to ⫺80 mmHg, each 500 ms in duration, were delivered to the carotid sinus precisely 50 ms after the R wave of the ECG to elicit maximum carotid baroreflex responses. These pulses were applied during a 10- to 15-s breath hold at end expiration to minimize respiratory-related modulation of HR and ABP. A minimum of three trains of NP and NS were executed at rest and during HUT, with ⬃45 s between successive trials to allow full recovery of HR and MAP (39). As with the other baroreflex protocols, there was a 3-min delay prior to application of the first NP and NS train during HUT. Transfer Function Analysis Beat-to-beat SBP, MAP, estimated carotid sinus pressure (ECSP ⫽ MAP ⫺ neck chamber pressure), HR, and R-R interval (RRI) were obtained by integration of analog signals within each cardiac cycle and then linearly interpolated and resampled at 2 Hz for spectral analysis (25). At rest and during HUT, the transfer function gain between SBP, MAP (protocol 1), or ECSP (protocol 2) and HR (or RRI) fluctuations was calculated to estimate dynamic cardiac baroreflex sensitivity. The transfer function H(f) between SBP, MAP, or 301 • NOVEMBER 2011 •

www.ajpregu.org


ECSP and HR (or RRI) was computed from the cross-spectrum between SBP, MAP, or ECSP and HR (or RRI) variability and the auto-spectrum SBP, MAP, or ECSP variability using the Welch method: H(f) ⫽ Sxy(f)/Sxx(f), where Sxx(f) is the auto-spectrum SBP, MAP, or ECSP variability and Sxy(f) is the cross-spectrum between SBP, MAP, or ECSP and HR (or RRI) variability. The resampled data were segmented into 50% overlapping bins of 256 data points each. Linear trend removal followed by Hamming window treatment was applied to each segment before calculation of the transfer functions. The input power spectra of the binary command signals were fairly constant up to 0.5 Hz and diminished to noise level at 1 Hz. We estimated the transfer function data up to 0.5 Hz, and not beyond, since the estimation became greatly dispersed due to the lack of input power in the higher-frequency range. Spectral power was calculated in the low-frequency (LF, 0.05– 0.20 Hz) and high-frequency (HF, 0.20 – 0.30 Hz) ranges. The ABP fluctuations in the HF range, such as induced by the respiratory frequency, are transferred to HR; however, ABP fluctuations in the LF range are independent of the respiratory frequency and primarily reflect baroreflex mechanisms (25, 45). Sequence Technique Analysis The beat-to-beat time series of SBP or MAP and HR or RRI were analyzed off-line using a customized computer program (Nevrokard, Izola, Slovenia). Briefly, sequences of three or more consecutive beats where ABP and HR changed in the opposite direction or ABP and RRI changed in the same direction were identified as arterial baroreflex sequences (15, 25). A liner regression was applied to each individual sequence, and only those sequences in which r2 ⬎ 0.85 were accepted. The slope of these sequences was calculated as a measure of spontaneous cardiac baroreflex sensitivity. Carotid-Cardiac Baroreflex Function Analysis by Rapid NP and NS Stimulation Carotid baroreflex HR responses were evaluated by plotting the peak changes in HR against the ECSP. The carotid baroreflex stimulus-response data were fitted to the logistic model described by Kent et al. (19). This function incorporates the following equation: HR ⫽ A1{1 ⫹ exp[A2(ECSP ⫺ A3)]}⫺1 ⫹ A4, where HR is the dependent variable, A1 is the range of response of the dependent variable (maximum ⫺ minimum), A2 is the gain coefficient, A3 is the carotid sinus pressure required to elicit equal pressor and depressor responses (centering point), and A4 is the minimum response of HR. The data were fitted to this model by nonlinear least-squares regression (using a Marquardt-Levenberg algorithm), which minimized the sum-ofsquares error term to predict a curve of “best fit” for each set of raw data. The gain was calculated from the first derivative of the logistic function, and the maximal gain (Gmax) was applied as the index of carotid-cardiac baroreflex responsiveness. Gmax was calculated as follows: Gmax ⫽ ⫺A1A2/4. Statistics Values are means ⫾ SE. A two-way ANOVA with repeated measures was employed to determine significant differences in cardiovascular variables and respiration during supine rest and HUT under control conditions and during binary white-noise NS. A Student-Newman-Keuls test was employed post hoc when main effects were significant. In addition, paired t-tests were used for comparison of the different cardiac baroreflex measures between supine rest and HUT. P ⬍ 0.05 was considered to represent statistically significant differences. RESULTS

During HUT, HR was increased significantly, while SBP, DBP, and MAP were unchanged from supine rest with and AJP-Regul Integr Comp Physiol • VOL

R1593

without NS binary white noise (Fig. 1, Table 1). The closedloop cardiac baroreflex sensitivity derived from spontaneous oscillation in HR and BP, without NS stimulation, was decreased during HUT (Fig. 2). These results were consistent for the sequence technique (Fig. 2A) and the transfer function gain (Fig. 2B), and there were no differences when RRI or HR was used as the dependent variable or when SBP or MAP was used as the independent variable. In contrast to these closed-loop analyses, the LF transfer function gain between ECSP and HR during the application of white-noise stimulation for open-loop identification of cardiac baroreflex sensitivity was increased from supine rest to HUT (from 0.12 ⫾ 0.03 to 0.17 ⫾ 0.03 beats·min⫺1·mmHg⫺1, P ⫽ 0.02; Fig. 3A). Conversely, the HF transfer function gain between ECSP and HR was unchanged during HUT (from 0.09 ⫾ 0.01 to 0.08 ⫾ 0.02 beats·min⫺1·mmHg⫺1, P ⫽ 0.62). Similar to the open-loop LF transfer function gain between ECSP and HR, HUT increased Gmax of the carotidcardiac baroreflex function curve determined by the rapid-pulse NP-NS technique (from 0.47 ⫾ 0.15 beats·min⫺1·mmHg⫺1 at rest to 0.60 ⫾ 0.20 beats·min⫺1·mmHg⫺1 at HUT, P ⫽ 0.037; Fig. 3B). DISCUSSION

In the present study, dynamic open-loop cardiac baroreflex LF transfer function gain was enhanced during orthostatic stress using a binary white-noise perturbation technique with the neck chamber system. An increase in Gmax of the carotidcardiac baroreflex function curve was also observed during HUT. These data were in contrast to the reductions in closedloop cardiac baroreflex sensitivity derived from spontaneous oscillations in ABP and HR calculated using the sequence technique and transfer function analysis. These findings suggest that orthostatic stress-induced central hypovolemia enhanced open-loop dynamic cardiac baroreflex gain. Moreover, under closed-loop conditions, spontaneous baroreflex analyses without external stimulation may not represent open-loop arterial baroreflex characteristics during orthostatic stress. Collectively, these data add importantly to the ongoing debate of use of open- vs. closed-loop baroreflex measures to assess cardiac baroreflex characteristics. Indeed, the reduction in cardiac baroreflex sensitivity using closed-loop dynamic baroreflex measures without external stimulation during orthostatic stress may be due to the methodological approach used, rather than the physiological response. Thus the reliability of the concept that orthostatic stress causes dysfunction of the systemic circulation that is related to an attenuated cardiac baroreflex function seems debatable. Frequency domain analysis for identifying closed-loop dynamic cardiac baroreflex sensitivity is based on the concept that HR changes in response to spontaneous beat-to-beat oscillations of ABP via arterial baroreceptors (27). Therefore, spontaneous dynamic cardiac baroreflex sensitivity should reflect arterial-cardiac baroreflex gain around the operating point of the full baroreflex function curve (17, 25). Indeed, Ogoh et al. (25) demonstrated that the gain at the operating point of the carotid-cardiac baroreflex function curve was associated with dynamic baroreflex gain measures calculated by the sequence technique and transfer function analysis at rest and during increasing exercise workloads. In contrast, cardiac baroreflex sensitivity measures derived from spontaneous fluctuations in 301 • NOVEMBER 2011 •

www.ajpregu.org

R1594


Fig. 1. Original recordings of neck chamber pressure (CP), heart rate [HR, beats/min (BPM)], and arterial blood pressure (ABP) from 1 subject during supine rest and 60° head-up tilt (60 HUT) without (A) and with (B) binary white-noise neck suction. Solid line in ABP trace represents mean arterial pressure.

ABP and HR during orthostatic stress (1, 2, 10, 11, 13, 36, 41, 44) were not consistent with the gains calculated from the carotid-cardiac baroreflex function curve evaluated with the NP-NS technique (4 – 6, 26, 29, 43). Indeed, studies using dynamic analyses to assess cardiac baroreflex sensitivity have consistently reported a reduced gain with cardiopulmonary unloading (1, 2, 10, 11, 13, 36, 41, 44), whereas an increased carotid-cardiac baroreflex gain has been observed during orTable 1. Cardiorespiratory data with and without a binary white noise of neck suction during supine rest and 60° HUT Control Variables

Supine

Neck Suction 60° HUT

Supine

60° HUT

HR, beats/min 75.7 ⫾ 4.3 85.2 ⫾ 5.5† 70.3 ⫾ 5.5* 86.2 ⫾ 4.8† RRI, ms 816.5 ⫾ 48.1 729.1 ⫾ 44.0† 874.9 ⫾ 45.9* 715.3 ⫾ 39.3† MAP, mmHg 87.1 ⫾ 3.2 86.1 ⫾ 3.7 89.7 ⫾ 3.6 84.1 ⫾ 3.1 SBP, mmHg 122.8 ⫾ 3.6 119.2 ⫾ 4.4 125.0 ⫾ 3.9 115.7 ⫾ 4.1 DBP, mmHg 62.3 ⫾ 7.6 64.2 ⫾ 8.0 66.0 ⫾ 8.1 60.8 ⫾ 7.3 Respiratory rate, Hz 0.26 ⫾ 0.0 0.26 ⫾ 0.0 0.26 ⫾ 0.0 0.27 ⫾ 0.0 Values are means ⫾ SE. HUT, head-up tilt; HR, heart rate; RRI, R-R interval; MAP, mean arterial blood pressure; SBP, systolic blood pressure; DBP, diastolic blood pressure. *Significantly different from control (P ⬍ 0.05). †Significantly different from supine (P ⬍ 0.05). AJP-Regul Integr Comp Physiol • VOL

thostatic stress using the neck chamber technique to stimulate carotid baroreceptors (4 – 6, 26, 29, 43). One possibility for this discrepancy is that the dynamic spontaneous analyses in humans include closed-loop influences and, therefore, cannot accurately represent open-loop baroreflex characteristics during HUT. However, to date, no study has identified dynamic open-loop cardiac baroreflex function during orthostatic stress. The findings of the present study provide insight into the discrepancy between these previous studies by directly comparing closed- and open-loop baroreflex measures during orthostatic stress in the same subjects. As reported previously (11, 13, 41), cardiac baroreflex sensitivity derived from the sequence technique and transfer function gain between continuous recordings of ABP and HR were significantly decreased during HUT. Thus, without exogenous perturbation of the baroreceptors, a conclusion of diminished arterial baroreflex function would have been derived. However, in the present study, the carotid sinus open-loop dynamic baroreflex LF transfer function gain determined by the binary white-noise technique was significantly increased during HUT (P ⫽ 0.02; Fig. 3A), while the HF gain was unchanged during HUT (P ⫽ 0.62). Because the HF range is significantly influenced by respiration, we focused our interpretation on the LF range, in which the ABP fluctuations are independent of the respiratory 301 • NOVEMBER 2011 •

www.ajpregu.org


R1595

Fig. 2. Group summary data for closed-loop measures of cardiac baroreflex sensitivity during supine rest and 60° HUT: data from the sequence technique (A) and from transfer function analyses (B) at low frequency (0.05– 0.20 Hz) and high frequency (0.20 – 0.30 Hz). To be inclusive, these spontaneous cardiac baroreflex measures are presented using R-R interval (RRI) and HR as the dependent variable and systolic blood pressure (SBP) and mean arterial pressure (MAP) as the independent variable. *Significantly different from supine (P ⬍ 0.05).

frequency and, therefore, provide an index that primarily reflects baroreflex mechanisms (25, 45). Results similar to the open-loop LF gain were found with Gmax of the carotid-cardiac baroreflex function curve identified using the rapid-pulse NP-NS technique. Previous reports indicate that the mechanism of augmentation of cardiac baroreflex gain is due to an interaction with cardiopulmonary baroreceptors (6, 26, 29, 43). Collectively, these data strongly suggest that open-loop dynamic cardiac baroreflex gain is augmented and that cardiac baroreflex measures derived without external stimulation do not represent open-loop dynamic baroreflex characteristics during orthostatic stress. AJP-Regul Integr Comp Physiol • VOL

Estimates of the transfer function characteristics among cardiovascular variables are useful in providing insight into the mechanisms of circulatory regulation, particularly at rest (18, 25, 37). Since the arterial baroreflex is activated dynamically on a beat-to-beat basis, analysis should include consideration of these transient, dynamic changes (18). However, it is difficult to identify the dynamic transfer function characteristic of the arterial baroreflex because of the closed-loop system. Under such closed-loop conditions, changes in ABP lead to alterations in HR and sympathetic nerve activity via the arterial baroreflex, whereas the resulting changes in HR and sympathetic nerve activity, in turn, affect ABP. Therefore, this 301 • NOVEMBER 2011 •

www.ajpregu.org

R1596


Fig. 3. Group summary data for open-loop measures of cardiac baroreflex sensitivity during supine rest and 60° HUT: transfer function gain between estimated carotid sinus pressure (ECSP) and HR with binary white-noise neck suction at low (0.05– 0.20 Hz) and high (0.20 – 0.30 Hz) frequency (A) and carotid-cardiac baroreflex maximal gain (Gmax) calculated from the modeled carotid baroreflex function curve (B). *Significantly different from supine (P ⬍ 0.05).

closed-feedback loop makes it difficult to identify the openloop transfer function characteristics by means of conventional frequency domain methods based on the Fourier transformation, which is mathematically noncausal (18). Orthostatic stress significantly increases HR and sympathetic nerve activity, indicating that the net hemodynamic data collected include more feedback signals (i.e., counteracting signals), which may not be related to arterial baroreflex responses. Indeed, Kawada et al. (18) demonstrated that, even at rest, under the closed-loop condition, dynamic analysis calculated from continuously recorded hemodynamic data underestimates open-loop dynamic baroreflex function. In addition, it is plausible that the increases in cardiovascular parameters with orthostatic stress contribute to decreases in arterial baroreflex sensitivity identified by these closed-loop dynamic analyses. In other words, the increases in HR and cardiac output, along with other circulatory changes during orthostatis, would influence ABP and, potentially, under closed-loop conditions without external stimulation, enhance the possibility that baroreflex function will be underestimated. In addition, some studies have used the parametric estimation technique to assess closed-loop dynamic baroreflex function, such as the autoregressive moving average model of dynamic analysis (1, 44). Of note, parametric estimation methods depend on the structure and order of the model selected. Thus, baroreflex function may not only be changed by the actual system response but also by the model used (18). In general, there is considerable debate on the utility of dynamic baroreflex measures to assess cardiac baroreflex sensitivity, with some studies demonstrating correlations with traditional measures of cardiac baroreflex function (28, 35, 37), while others have reported weak relationships (3, 18, 21, 30). Perhaps some consideration should be given to the inherent differences in the methodological approaches used to assess baroreflex function, rather than expecting them to provide identical information (12). In the present study, we have compared cardiac baroreflex sensitivity measurements estimated from closed-loop dynamic transfer function analysis and the sequence technique with two open-loop carotid-cardiac baroreflex function measures derived with the application of NP-NS at rest and during HUT. In human studies, the variable-pressure neck chamber technique has several advantages to identify baroreflex function. Importantly, this technique can isolate carotid baroreflex-mediated AJP-Regul Integr Comp Physiol • VOL

responses without counteraction from aortic and cardiopulmonary baroreceptors under closed-loop conditions. Indeed, the brief period of stimulation typically used obviates any counterresponses emanating from extracarotid baroreceptors (7, 31–33). In addition, the peak reflex responses of HR and ABP to NP-NS are unaffected by baseline hemodynamic changes (33). Therefore, the baroreflex function curve can be identified and compared under a variety of conditions, including orthostatic stress. While pharmacological approaches could be used to assess arterial baroreflex control of HR (e.g., “modified” Oxford technique), this would be more challenging during HUT, as the systemic administration of pharmacological agents, particularly depressor drugs, can be problematic during the imposed orthostatic stress. In addition, the bolus infusions of phenylephrine and nitroprusside transiently load and unload cardiopulmonary baroreceptors, respectively, which would directly influence the cardiopulmonary baroreceptors. To be comprehensive, the present study identified dynamic carotidcardiac baroreflex transfer function with a binary white-noise technique, as well as with the rapid-pulse NP-NS technique. Importantly, the results were consistent and in contrast to those derived from closed-loop transfer function measures. Perspectives and Significance Taken together, the data from the present study indicate that previous studies using dynamic spontaneous cardiac baroreflex measures calculated from continuous recordings of hemodynamic data without exogenous perturbation (1, 2, 10, 11, 13, 36, 41, 44) appear to have not identified open-loop cardiac dynamic baroreflex function during orthostatic stress. Thus the reliability of the concept that orthostatic stress causes dysfunction of the systemic circulation, which was related to an attenuated arterial baroreflex function, seems questionable. Indeed, it has been suggested that the increase in the sensitivity of the carotid baroreceptor during orthostatic stress was related to orthostatic tolerance (5, 26). Patients with syncope fail to increase carotid baroreflex gain during HUT (5). Importantly, these previous studies suggest that an increase in the sensitivity of the carotid baroreceptor reflex may be important in the maintenance of ABP during orthostatic stress, meaning that the failure of this mechanism contributes to posturally mediated syncope. Interestingly, the interaction between the carotid and 301 • NOVEMBER 2011 •

www.ajpregu.org


cardiopulmonary baroreceptors has been shown to be attenuated in high-fit subjects (26), possibly contributing to greater orthostatic intolerance in this group. However, reductions in cardiopulmonary (38) and aortic baroreflex (39) sensitivity have also been reported in endurance-trained athletes. An important caveat for the present study is that the dynamic closed-loop baroreflex measures presumably encompass aortic and carotid baroreflexes, whereas the dynamic open-loop measures isolate the carotid baroreflex. The neck chamber technique has been used extensively to estimate arterial baroreflex control in human investigations. Importantly, with the use of the variable-pressure neck chamber to selectively describe carotid baroreflex control, the assumption is made that the aortic baroreflex operates in parallel with the carotid baroreflex and, therefore, will respond similarly (7, 32, 33). Physiologically, aortic baroreceptors are integrated within the central nervous system in a manner similar to the carotid sinus baroreceptors, making this supposition very reasonable. Thus we believe that the response of the carotid baroreflex to orthostatic stress would be representative of the arterial baroreflex. However, since open-loop testing did not directly assess aortic baroreflex function, we cannot completely rule out the possibility that the differences between the closed- and open-loop techniques result from the inability to manipulate the full arterial baroreflex with the latter methodology. In summary, we have provided insight into the discrepancy between previous studies reporting that cardiac baroreflex sensitivity was reduced or increased during orthostatic stress. The present studies clearly demonstrate that this discrepancy depends on the methodology used to assess dynamic cardiac baroreflex function. Both of the open-loop measures of dynamic carotid-cardiac baroreflex gain were augmented during orthostatic stress, whereas the sequence technique and the closed-loop transfer function gain derived from spontaneous oscillations in ABP and HR were reduced. Therefore, in previous studies using transfer function analysis without baroreceptor stimulation, the finding of the reduction of cardiac baroreflex sensitivity during orthostatic stress may not represent the open-loop dynamic baroreflex characteristics. Overall, care should be taken when closed-loop dynamic analyses are used as a primary measure of cardiac baroreflex sensitivity during orthostatic stress. While these measurements are noninvasive and may have value under resting conditions, the current results demonstrate that, during HUT, further measures that drive the baroreflex system and attempt to “open” the closed-loop system by evoking clear baroreflex-mediated responses are needed for a full evaluation of baroreflex function. ACKNOWLEDGMENTS We sincerely thank the subjects for their interest and cooperation. GRANTS This study was supported by Center for Academic Research Promotion (Toyo University Research Institution of Industrial Technology) Grant 7. DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the authors. AUTHOR CONTRIBUTIONS T.A. and S.O. analyzed the data; T.A. and S.O. interpreted the results of the experiments; J.S., P.J.F., and S.O. edited and revised the manuscript; D.I., N.T., and S.O. performed the experiments; P.J.F. and S.O. drafted the manuAJP-Regul Integr Comp Physiol • VOL

R1597

script; P.J.F. and S.O. approved the final version of the manuscript; S.O. was responsible for conception and design of the research; S.O. prepared the figures. REFERENCES 1. Barbieri R, Triedman JK, Saul JP. Heart rate control and mechanical cardiopulmonary coupling to assess central volume: a systems analysis. Am J Physiol Regul Integr Comp Physiol 283: R1210 –R1220, 2002. 2. Blaber AP, Yamamoto Y, Hughson RL. Change in phase relationship between SBP and R-R interval during lower body negative pressure. Am J Physiol Heart Circ Physiol 268: H1688 –H1693, 1995. 3. Colombo R, Mazzuero G, Spinatonda G, Lanfranchi P, Giannuzzi P, Ponikowski P, Coats AJ, Minuco G. Comparison between spectral analysis and the phenylephrine method for the assessment of baroreflex sensitivity in chronic heart failure. Clin Sci (Lond) 97: 503–513, 1999. 4. Cooper VL, Hainsworth R. Carotid baroreceptor reflexes in humans during orthostatic stress. Exp Physiol 86: 677–681, 2001. 5. Cooper VL, Hainsworth R. Effects of head-up tilting on baroreceptor control in subjects with different tolerances to orthostatic stress. Clin Sci (Lond) 103: 221–226, 2002. 6. Ebert TJ. Carotid baroreceptor reflex regulation of forearm vascular resistance in man. J Physiol 337: 655–664, 1983. 7. Fadel PJ, Ogoh S, Keller DM, Raven PB. Recent insights into carotid baroreflex function in humans using the variable pressure neck chamber. Exp Physiol 88: 671–680, 2003. 8. Fisher JP, Ogoh S, Junor C, Khaja A, Northrup M, Fadel PJ. Spontaneous baroreflex measures are unable to detect age-related impairments in cardiac baroreflex function during dynamic exercise in humans. Exp Physiol 94: 447–458, 2009. 9. Fu Q, Shibata S, Hastings JL, Prasad A, Palmer MD, Levine BD. Evidence for unloading arterial baroreceptors during low levels of lower body negative pressure in humans. Am J Physiol Heart Circ Physiol 296: H480 –H488, 2009. 10. Furlan R, Jacob G, Palazzolo L, Rimoldi A, Diedrich A, Harris PA, Porta A, Malliani A, Mosqueda-Garcia R, Robertson D. Sequential modulation of cardiac autonomic control induced by cardiopulmonary and arterial baroreflex mechanisms. Circulation 104: 2932–2937, 2001. 11. Galbreath MM, Shibata S, Vangundy TB, Okazaki K, Fu Q, Levine BD. Effects of exercise training on arterial-cardiac baroreflex function in POTS. Clin Auton Res 21: 73–80, 2011. 12. Gasch J, Reimann M, Reichmann H, Rudiger H, Ziemssen T. Determination of baroreflex sensitivity during the modified Oxford maneuver by trigonometric regressive spectral analysis. PLos One 6: e18061, 2011. 13. Gulli G, Claydon VE, Slessarev M, Zenebe G, Gebremedhin A, Rivera-Ch M, Appenzeller O, Hainsworth R. Autonomic regulation during orthostatic stress in highlanders: comparison with sea-level residents. Exp Physiol 92: 427–435, 2007. 14. Harada S, Ando S, Imaizumi T, Hirooka Y, Sunagawa K, Takeshita A. Arterial baroreflex control of cardiac and renal sympathetic nerve activities is uniform in frequency domain. Am J Physiol Regul Integr Comp Physiol 261: R296 –R300, 1991. 15. Iellamo F, Hughson RL, Castrucci F, Legramante JM, Raimondi G, Peruzzi G, Tallarida G. Evaluation of spontaneous baroreflex modulation of sinus node during isometric exercise in healthy humans. Am J Physiol Heart Circ Physiol 267: H994 –H1001, 1994. 16. Iellamo F, Legramante JM, Raimondi G, Peruzzi G. Baroreflex control of sinus node during dynamic exercise in humans: effects of central command and muscle reflexes. Am J Physiol Heart Circ Physiol 272: H1157–H1164, 1997. 17. Iwasaki K, Zhang R, Perhonen MA, Zuckerman JH, Levine BD. Reduced baroreflex control of heart period after bed rest is normalized by acute plasma volume restoration. Am J Physiol Regul Integr Comp Physiol 287: R1256 –R1262, 2004. 18. Kawada T, Sugimachi M, Sato T, Miyano H, Shishido T, Miyashita H, Yoshimura R, Takaki H, Alexander J Jr, Sunagawa K. Closed-loop identification of carotid sinus baroreflex open-loop transfer characteristics in rabbits. Am J Physiol Heart Circ Physiol 273: H1024 –H1031, 1997. 19. Kent BB, Drane JW, Blumenstein B, Manning JW. A mathematical model to assess changes in the baroreceptor reflex. Cardiology 57: 295– 310, 1972. 20. Lacolley PJ, Pannier BM, Slama MA, Cuche JL, Hoeks AP, Laurent S, London GM, Safar ME. Carotid arterial haemodynamics after mild degrees of lower-body negative pressure in man. Clin Sci (Lond) 83: 535–540, 1992. 301 • NOVEMBER 2011 •

www.ajpregu.org

R1598


21. Lipman RD, Salisbury JK, Taylor JA. Spontaneous indices are inconsistent with arterial baroreflex gain. Hypertension 42: 481–487, 2003. 22. Mano T. Muscle sympathetic nerve activity in blood pressure control against gravitational stress. J Cardiovasc Pharmacol 38 Suppl 1: S7–S11, 2001. 23. Marmarelis ZP, Marmarelis ZV. The white noise method in system identification. In: Analysis of Physiological Systems. New York: Plenum, 1978, p. 131–221. 24. Mukkamala R, Kim JK, Li Y, Sala-Mercado J, Hammond RL, Scislo TJ, O’Leary DS. Estimation of arterial and cardiopulmonary total peripheral resistance baroreflex gain values: validation by chronic arterial baroreceptor denervation. Am J Physiol Heart Circ Physiol 290: H1830 – H1836, 2006. 25. Ogoh S, Fisher JP, Dawson EA, White MJ, Secher NH, Raven PB. Autonomic nervous system influence on arterial baroreflex control of heart rate during exercise in humans. J Physiol 566: 599 –611, 2005. 26. Ogoh S, Volianitis S, Nissen P, Wray DW, Secher NH, Raven PB. Carotid baroreflex responsiveness to head-up tilt-induced central hypovolaemia: effect of aerobic fitness. J Physiol 551: 601–608, 2003. 27. Parati G, Mancia G, Di Rienzo M, Castiglioni P. Point: cardiovascular variability is/is not an index of autonomic control of circulation. J Appl Physiol 101: 676 –678; discussion 681– 672, 2006. 28. Parlow J, Viale JP, Annat G, Hughson R, Quintin L. Spontaneous cardiac baroreflex in humans. Comparison with drug-induced responses. Hypertension 25: 1058 –1068, 1995. 29. Pawelczyk JA, Raven PB. Reductions in central venous pressure improve carotid baroreflex responses in conscious men. Am J Physiol Heart Circ Physiol 257: H1389 –H1395, 1989. 30. Pitzalis MV, Mastropasqua F, Passantino A, Massari F, Ligurgo L, Forleo C, Balducci C, Lombardi F, Rizzon P. Comparison between noninvasive indices of baroreceptor sensitivity and the phenylephrine method in post-myocardial infarction patients. Circulation 97: 1362–1367, 1998. 31. Potts JT, Shi XR, Raven PB. Carotid baroreflex responsiveness during dynamic exercise in humans. Am J Physiol Heart Circ Physiol 265: H1928 –H1938, 1993. 32. Raven PB, Fadel PJ, Ogoh S. Arterial baroreflex resetting during exercise: a current perspective. Exp Physiol 91: 37–49, 2006. 33. Raven PB, Potts JT, Shi X. Baroreflex regulation of blood pressure during dynamic exercise. Exerc Sport Sci Rev 25: 365–389, 1997.

AJP-Regul Integr Comp Physiol • VOL

34. Rea RF, Wallin BG. Sympathetic nerve activity in arm and leg muscles during lower body negative pressure in humans. J Appl Physiol 66: 2778 –2781, 1989. 35. Robbe HW, Mulder LJ, Ruddel H, Langewitz WA, Veldman JB, Mulder G. Assessment of baroreceptor reflex sensitivity by means of spectral analysis. Hypertension 10: 538 –543, 1987. 36. Saitoh T, Ogawa Y, Aoki K, Shibata S, Otsubo A, Kato J, Iwasaki K. Bell-shaped relationship between central blood volume and spontaneous baroreflex function. Auton Neurosci 143: 46 –52, 2008. 37. Saul JP, Berger RD, Albrecht P, Stein SP, Chen MH, Cohen RJ. Transfer function analysis of the circulation: unique insights into cardiovascular regulation. Am J Physiol Heart Circ Physiol 261: H1231–H1245, 1991. 38. Shi X, Gallagher KM, Smith SA, Bryant KH, Raven PB. Diminished forearm vasomotor response to central hypervolemic loading in aerobically fit individuals. Med Sci Sports Exerc 28: 1388 –1395, 1996. 39. Smith SA, Querry RG, Fadel PJ, Welch-O’Connor RM, OlivenciaYurvati A, Shi X, Raven PB. Differential baroreflex control of heart rate in sedentary and aerobically fit individuals. Med Sci Sports Exerc 32: 1419 –1430, 2000. 40. Taylor JA, Halliwill JR, Brown TE, Hayano J, Eckberg DL. “Nonhypotensive” hypovolaemia reduces ascending aortic dimensions in humans. J Physiol 483: 289 –298, 1995. 41. Triedman JK, Cohen RJ, Saul JP. Mild hypovolemic stress alters autonomic modulation of heart rate. Hypertension 21: 236 –247, 1993. 42. Victor RG, Leimbach WN Jr. Effects of lower body negative pressure on sympathetic discharge to leg muscles in humans. J Appl Physiol 63: 2558 –2562, 1987. 43. Victor RG, Mark AL. Interaction of cardiopulmonary and carotid baroreflex control of vascular resistance in humans. J Clin Invest 76: 1592– 1598, 1985. 44. Zhang R, Behbehani K, Crandall CG, Zuckerman JH, Levine BD. Dynamic regulation of heart rate during acute hypotension: new insight into baroreflex function. Am J Physiol Heart Circ Physiol 280: H407– H419, 2001. 45. Zhang R, Zuckerman JH, Giller CA, Levine BD. Transfer function analysis of dynamic cerebral autoregulation in humans. Am J Physiol Heart Circ Physiol 274: H233–H241, 1998.

301 • NOVEMBER 2011 •

www.ajpregu.org