| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the Departments of *Hygiene and Space Medicine and Anesthesiology, Nihon University School of Medicine, Tokyo, Japan.
Address correspondence and reprint requests to Ken-ichi Iwasaki, MD, PhD, Associate Professor, Department of Hygiene and Space Medicine, Nihon University School of Medicine, 30-1 Oyaguchi-Kamimachi, Itabashi-Ku, Tokyo 173-8610, Japan. Address e-mail to kiwasaki{at}med.nihon-u.ac.jp.
Abstract
BACKGROUND: Frequent changes in the perioperative central blood volume could affect cerebral autoregulation through alterations in sympathetic nerve activity, cardiac output, blood viscosity, and cerebral vasomotor tone. However, the effect of dynamic cerebral autoregulation has not been studied during acute wide-ranging changes in central blood volume, especially with respect to central hypervolemia with hemodilution.
METHODS: We evaluated dynamic cerebral autoregulation during central hypovolemia and central hypervolemia with hemodilution using spectral and transfer function analysis between mean arterial blood pressure (MBP) and cerebral blood flow (CBF) velocity variability in 12 individuals. Rapid changes in central blood volume were achieved using two levels of lower body negative pressure (–15 and –30 mm Hg) and two discrete infusions of normal saline (15 mL/kg and total 30 mL/kg). We then estimated changes in central blood volume as central venous pressure (CVP) and/or cardiac output using impedance cardiography.
RESULTS: Steady-state CBF velocity and cardiac output decreased at –30 mm Hg lower body negative pressure (changes of CVP approximately –4 mm Hg) or were increased by each saline infusion (changes of CVP 4–6 mm Hg), without a significant change in MBP. However, transfer function gain (magnitude of transfer) between MBP and CBF velocity variability significantly increased only after saline infusion, suggesting an increased magnitude of transfer from MBP oscillations to CBF fluctuations during central hypervolemia with hemodilution.
CONCLUSION: Our results suggest that, although steady-state CBF velocity changes under both central hypervolemia and hypovolemia, only hypervolemic hemodilution impairs dynamic cerebral autoregulation.
Various perioperative stimuli, such as infusions, dehydration associated with fasting, postural changes, vasodilatation due to anesthetics, hemorrhage, and transfusions, can rapidly and extensively alter central blood volume. Changes in central blood volume affect sympathetic nerve activity (1,2), cardiac output (3), and blood viscosity (4,5). Central hypervolemia (3,6) or hemodilution (4,7) increases cerebral blood flow (CBF) velocity without significantly changing arterial blood pressure. Such changes in CBF velocity imply alterations in cerebral autoregulation.
Several studies of cerebral autoregulation have enabled evaluation of "dynamic" cerebral autoregulation in addition to assessment of steady-state CBF. The ability of the cerebral vascular bed to buffer changes in CBF velocity induced by transient changes in arterial blood pressure is referred to as "dynamic cerebral autoregulation," and it has been estimated by spectral and transfer function analysis between arterial blood pressure and CBF velocity variability (8–10). This method has been clinically used and it has provided insight into dynamic cerebral autoregulation (11–15).
However, the effect of dynamic cerebral autoregulation has not been studied during acute wide-ranging changes in central blood volume, especially with respect to central hypervolemia with hemodilution.
We therefore evaluated dynamic cerebral autoregulation during central hypervolemia with hemodilution and central hypovolemia to test our hypothesis that changes in central blood volume affect cerebral autoregulation. Dynamic cerebral autoregulation was estimated using spectral and transfer function analysis between arterial blood pressure variability and CBF velocity variability.
METHODS
The Institutional Ethical Committee of Nihon University School of Medicine approved this study. All study participants provided written informed consent as well as a medical history, and underwent a physical examination including electrocardiography (ECG), arterial blood pressure measurements and echocardiography. One volunteer was excluded because CBF velocity signals in the middle cerebral artery (MCA) could not be obtained by transcranial Doppler (TCD) ultrasonography. We investigated 12 healthy, normotensive, males (aged 21.4 ± 1.2 yr; height 171.9 ± 4.1 cm; weight 63.1 ± 5.9 kg; means ± sd).
All participants fasted for at least 2 h before the study, and refrained from heavy exercise and from consuming caffeinated or alcoholic beverages for at least 24 h before the study. All participants were familiarized with the measurement techniques and experimental conditions before starting the study.
Participants lay supine in the lower body chamber in an environmentally controlled experimental room at an ambient temperature of 23°C–25°C. An ECG, arterial oxygen saturation, and end-tidal carbon dioxide pressure (ETco2) monitor (Life scope BSM-5132; NIHON KOHDEN, Tokyo, Japan) were applied. Continuous arterial blood pressure was measured in the radial artery using tonometry of a noninvasive arterial blood pressure monitor at the heart level on a beat-to-beat basis, and calibrated by intermittent arterial blood pressure measured using an oscillometric method with a cuff sphygmomanometer placed over the brachial artery (JENTOW 7700; COLIN, Komaki, Japan). The CBF velocity in the MCA was continuously measured by TCD ultrasonography (WAKI; Atys Medical, St. Genislaval, France). A 2-MHz probe was placed over the temporal window and fixed at a constant angle with a probe holder customized to fit individual facial bone structures (16). Each waveform of continuous arterial blood pressure, CBF velocity, and ECG were recorded at a sampling rate of 1 kHz using commercial software (Notocord-hem 3.3; Notocord, Paris, France) throughout the study.
Cardiac output and stroke volume were measured using an impedance cardiograph (CIC1000; Sobra Medical Systems, Milwaukee, WI), which provided a noninvasive index of central blood volume (17,18). An 18-gauge catheter was inserted into a forearm vein for saline infusion. In addition, central venous pressure (CVP) was measured and recorded in five individuals by advancing a central catheter (First PICC catheter 18 GA (4F) 1.35 mm x 65 cm, Becton Dickinson, NJ) that was inserted into an antecubital vein to the level of the superior vena cava. Positioning of the central catheter was estimated by measuring the external distance from the antecubital fossa to the manubrium. This catheter was connected to a pressure transducer (DX-312, Becton Dickinson, NJ) placed at the level of the heart.
Baseline data were measured after at least 30 min of quiet rest, and then lower body negative pressure (LBNP) that provided steady suction was used to decrease central blood volume without dehydration in a graded fashion. The negative pressure of the chamber was monitored with a manometer. The magnitude of the suction was increased in a stepwise fashion according to the following protocol: –15 mm Hg for 7 min (LBNP-15), and –30 mm Hg for 7 min (LBNP-30). The LBNP was terminated if signs and/or symptoms of presyncope such as nausea, sweating, light-headedness, bradycardia, or hypotension (sustained systolic blood pressure <80 mm Hg) developed.
After completion of the protocols with LBNP, the subject rested for 15–20 min to enable hemodynamic recovery from the LBNP protocols. Recovery was confirmed by means of stabilized hemodynamics determined as values of heart rate (HR), intermittent arterial blood pressure, CBF velocity, and CVP that were similar to those at baseline. Two discrete normal saline (NS) infusions were rapidly administered to increase central blood volume with hemodilution. NS (15 mL/kg) was infused at a rate of 100 mL/min (first infusion protocol; NS15), followed by an additional 15 mL/kg (total 30 mL/kg) at the same rate (second infusion protocol; NS30). Total infusion duration was approximately 10 min per step.
Cardiac output and stroke volume were measured using an impedance cardiograph at the initiation of each step. A clinical laboratory (SRL CO, Tokyo, Japan) also performed a complete blood count including hematocrit (Hct) calculated as a ratio (%) in five subjects after each saline infusion. Respiratory rate and ETco2 were recorded every minute during each step. Mean values for steady-state mean arterial blood pressure (MBP), CBF velocity, HR, CVP, respiratory rate, and ETco2 obtained by averaging the 6-min segments for each subject were group-averaged for statistical analysis. Six minutes of continuous MBP, CBF velocity, ECG and CVP were used during LBNP and immediately after saline infusion for spectral and transfer function analysis with spontaneous respiration.
The beat-to-beat values of MBP and CBF velocity were obtained by integrating signals within each cardiac cycle using PC-based Notocord-hem 3.3 software. Beat-to-beat data for MBP and CBF velocity were then linearly interpolated and resampled at 2 Hz for spectral and transfer function analysis. Fast Fourier Transform and transfer function analysis were performed using a Hanning window on 256-point segments with 50% overlap, and this process resulted in five segments of data recordings over 6 min with which to assess the dynamic pressure-flow relationship (8,9). We analyzed these data using DADiSP software (DSP Development, Cambridge, MA).
The spectral power of MBP and CBF velocity, mean value of transfer function gain, phase and coherence function were calculated in the very-low- (0.02–0.07 Hz), low- (0.07–0.20 Hz), high- (0.20–0.30 Hz) frequency ranges, and over the entire frequency range (0.02–0.30 Hz) (Fig. 1). These ranges were specifically selected to reflect different patterns of the dynamic pressure-flow relationship, as identified by transfer function analysis (8,9). Coherence function reflects the linear relationship between these two variables and the reliability of the transfer function gain and phase. Thus, low coherence indicates nonlinear relationship between pressure and velocity, and/or an effective autoregulation with low reliability of the transfer function gain. Conversely, higher coherence means that pressure and velocity vary together very closely and indicate greater dependence of CBF velocity on arterial blood pressure. Temporal relationships between these two variables were estimated using phase shifts. Transfer function gain reflects the ability of the cerebrovascular bed to buffer changes in CBF velocity induced by transient changes in arterial blood pressure at different frequencies. A larger gain implies that any given change in pressure leads to a larger change in flow, implying impaired autoregulation (8–10). Previous studies have confirmed that these estimates are stable over 1–2 h under supine resting conditions (9,12).
|
In addition, the arterial oxygen content (CaO2) was calculated as:
|
|
where Pao2 was substituted with 100 mm Hg.
The equivalent of cerebral oxygen transport was calculated as:
|
|
Cerebral vascular resistance index (CVRi) was expressed as:
|
|
where cerebral perfusion pressure = MBP minus CVP. Normalized steady-state CBF velocity to Hct was expressed as steady-state CBF velocity divided by the reciprocal number of the Hct in each saline infusion protocol.
Variables were compared using one-way repeated-measures ANOVA with stage (baseline, LBNP-15, LBNP-30, NS15, and NS30), in conjunction with the Bonferroni post hoc test for comparisons with a baseline value. A P value of <0.05 was considered statistically significant. The analysis was performed using the PC-based software (SigmaStat, Systat Software, Inc., CA). Data are presented as means ± sd.
RESULTS
The LBNP of –30 mm Hg was terminated at 4 min with <80 mm Hg of systolic blood pressure, despite the absence of presyncope symptoms in one participant. We excluded these values from the group-averaged data for statistical analysis.
Table 1 shows the average values of steady-state hemodynamic and respiratory data under each condition. Compared with baseline, cardiac output significantly decreased at LBNP-30 (P < 0.001) as a result of a large reduction in stroke volume (P < 0.001) despite the increased HR (P < 0.001), and significantly increased at NS15 (P = 0.038) and NS30 (P < 0.001) as a slight augmentation in stroke volume and a result of increase in HR (NS30, P < 0.001). Although MBP did not change under either condition, steady-state CBF velocity significantly decreased at LBNP-30 (P = 0.046) and significantly increased at NS15 (P = 0.001) and NS30 (P < 0.001). The tendency of %changes in CBF velocity [(baseline – each step)/baseline x 100] was similar in terms of absolute values of steady-state CBF velocity. The amplitude of MBP fluctuations was about 15–20 mm Hg under all conditions. Respiratory rate and ETco2 remained unchanged under both conditions. In five subjects, CVP significantly decreased with LBNP-15 (P = 0.002) and LBNP-30 (P < 0.001), and significantly increased with NS15 (P < 0.001) and NS30 (P < 0.001). Although CVRi did not change during LBNP, it significantly decreased at NS30 (P = 0.002). Furthermore, Hct significantly decreased with NS15 (P < 0.001) and NS30 (P < 0.001). Cerebral oxygen transport significantly decreased with NS15 (P < 0.001) and NS30 (P = 0.005). Normalized steady-state CBF velocity to Hct also decreased significantly with NS15 (P = 0.005) and NS30 (P = 0.022).
|
Table 2 shows the average frequency domain data under each condition. Figure 1 shows the group-averaged transfer function analysis of beat-to-beat changes in MBP and CBF velocity at baseline, LBNP-30, and NS30. Compared with baseline, the spectral power of MBP variability did not change at all frequency ranges under both conditions, except for increases in the high-frequency range at LBNP-30 (P = 0.005) and decreases in the very-low-frequency range at NS30 (P = 0.037). The spectral power of CBF velocity variability did not significantly change at all frequency ranges under both conditions. Although it did not change at all frequency ranges during LBNP, transfer function gain increased significantly at the very-low- and low-frequency ranges at NS15 (very low frequency, P = 0.001; low frequency, P = 0.001) and at NS30 (very low frequency, P = 0.011; low frequency, P = 0.001). Moreover, transfer function gain in the high-frequency range increased significantly at NS15 (P = 0.041) with a trend toward an increase at NS30 (P = 0.060). Coherence in the very-low-frequency range was below 0.5 under all conditions. Coherence in the low- and high-frequency range was above 0.5 under both conditions with a significant increase at the high-frequency range at LBNP-30 (P = 0.032). Phase did not change over the all-frequency range under all conditions.
|
Figure 2 shows group-averaged power spectral density and transfer function gain in the entire frequency range between MBP and CBF velocity under each condition. Associated with changes in each frequency range, transfer function gain calculated over the entire frequency range increased significantly with NS15 (P < 0.001) and NS30 (P < 0.001) despite unchanged MBP and CBF velocity variabilities.
|
The correlation between steady-state CBF velocity and MBP (R2 = 0.005) or cardiac output (R2 = 0.24) was weak. The correlation between transfer function gain over the entire frequency range and cardiac output (R2 = 0.31) or MBP variability in this frequency range (R2 = 0.10) was also weak. However, transfer function gain over the entire frequency range correlated with CVRi (R2 = 0.52) (Fig. 3). The correlation between transfer function gain in the each frequency range and CVRi were R2 = 0.11, 0.44, and 0.33 at very low, low, and high frequency, respectively.
|
DISCUSSION
The primary findings of the present study were that, although steady-state CBF velocity was changed by both acute central hypervolemia and hypovolemia, transfer function gain between MBP variability and CBF velocity variability significantly increased only after an infusion of NS. In other words, this increase of transfer function gain induced by central hypervolemia with hemodilution implied that, for any given change in arterial blood pressure, the fluctuation in CBF was substantially greater than normovolemia. These findings indicated that dynamic cerebral autoregulation was impaired during hypervolemia with hemodilution in addition to increases in steady-state CBF velocity.
The traditional concept of cerebral autoregulation is that steady-state CBF remains relatively constant despite large changes in perfusion pressure (19). Changes in perfusion pressure produce obvious changes in cerebrovascular resistance, thereby maintaining CBF relatively constant. However, continuous measurements of CBF by methods that provide high temporal resolution such as the TCD technique used in large cerebral vessels have revealed prominent beat-to-beat fluctuations, similar to the oscillation observed in arterial blood pressure (8,9). Beat-to-beat CBF velocity, therefore, spontaneously fluctuates in response to dynamic changes in arterial blood pressure (8–10). The amplitude of spontaneous blood pressure fluctuation is comparable to changes in blood pressure during traditional autoregulation testing, such as the thigh cuff method and phenylephrine infusion. For example, the amplitude of MBP fluctuations was about 15–20 mm Hg in the present study. Here, we used cross-spectral methods for transfer function analysis to quantify the ability of the cerebral vascular bed to buffer this relatively large fluctuation of arterial blood pressure (8,9).
During central hypervolemia (change of CVP approximately 4–6 mm Hg) in the present study, saline infusion increased transfer function gain over the entire frequency range with increases in steady-state CBF velocity. This result suggests that central hypervolemia with hemodilution impairs dynamic cerebral autoregulation. The degree of such impairment would be similar to that expected from 5% carbon dioxide inhalation (8), sevoflurane anesthesia (11), and exposure to hypoxia (12). On the other hand, the present findings, that transfer function gain did not change during central hypovolemia induced by LBNP despite decrease in CBF velocity, suggest that dynamic cerebral autoregulation is not altered by mild central hypovolemia (CVP changes of approximately –3 to –4 mm Hg). These findings are consistent with the results of mild hypovolemia in a previous study (20).
The change in central blood volume with hemodilution alters several factors that influence regulation of CBF, such as sympathetic activity (21), cardiac output (3), cerebral vascular resistance (22), blood viscosity, and oxygen delivery (4,7). Some mechanisms might explain the alteration of dynamic cerebral autoregulation observed in the present study, although we did not attempt to elucidate the factors or mechanisms behind changing cerebral autoregulation and CBF.
First, changes in transfer function gain might be related to cerebrovascular state (22) without alterations in other mechanisms, such as neural and endothelial processes. One study found inverse-correlations between transfer function gain and cerebral vascular resistance (22). The present study also found a similar inverse-correlation between transfer function gain and CVRi, especially with gain in the low-frequency range. Therefore, reductions in cerebral vascular resistance during hypervolemia with hemodilution might induce an increase in transfer function gain. Although interpretation of this inverse-correlation is controversial, hemodilution might be a key mechanism in the present study. Reduced blood viscosity decreases peripheral vascular resistance (5) and reduced oxygen delivery induced by hemodilution dilates cerebral arterioles, thus decreasing cerebral vascular resistance (4). Constriction of smooth muscle in the dilated cerebral arteriolar would be weakened. Therefore, myogenic mechanisms for dynamic cerebral regulation might be attenuated. Insights from the frequency domain indicate that dynamic cerebral autoregulation at the low-frequency range might be partly modulated by myogenic mechanisms (8,19). We found here that transfer function gain at this frequency range significantly increased during hypervolemic hemodilution, implying that dynamic cerebral autoregulation associated with myogenic mechanisms might be impaired by cerebral arteriolar vasodilation. However, we could not reach a definitive conclusion without estimating cerebral arteriolar vasodilation indicated by increases in cerebral blood volume. We measured only CBF velocity, which might not always be identical to cerebral blood volume (23).
Another possibility is that cerebral oxygen transport in the present study significantly decreased despite the increased steady-state CBF velocity induced by saline infusion. Hypervolemia with hemodilution would not perfectly compensate for reduced oxygen delivery with increased CBF velocity, although a significant portion of the response of the CBF velocity might be related to oxygen delivery. Thus, hypoxia due to decreased cerebral oxygen transport might be related to altered dynamic cerebral autoregulation, since acute exposure to normobaric mild hypoxia alters dynamic cerebral autoregulation in the very-low-frequency range (12). Changes in dynamic cerebral autoregulation in this range under these conditions might be induced by similar mechanisms via reduced oxygen delivery.
The autonomic nervous system partly modulates dynamic cerebral autoregulation in the very-low-frequency range (21). Others have reported that LBNP induces activation of the sympathetic system (1). However, central hypervolemia induces a reduction of peripheral vasomotor sympathetic activity (2), weakening sympathetically mediated cerebral dynamic regulation. Thus, differences in sympathetic activity between hypovolemia and hypervolemia might preserve the transfer function gain in the very-low-frequency range during LBNP, or might increase it during saline infusion. Nevertheless, arterial blood pressure variability in the low-frequency range, including some vasomotor sympathetic nerve activity, was mildly changed in the present study. Therefore, the influence of sympathetic nerve activity on dynamic cerebral autoregulation might be minimal, although we cannot exclude the possibility of an influence of sympathetic nerve activity on dynamic cerebral autoregulation. The Bainbridge reflex might also be associated with the present results, as showed by increases in HR during hypervolemia (24).
Others have reported that cardiac output does not alter dynamic cerebral autoregulation (3,22), although the relationship between changes in CBF velocity and cardiac output are linear and highly significant (5). We also found a weak correlation between cardiac output and transfer function gain. The previous results together with our findings suggest that the effects of cardiac output on dynamic cerebral autoregulation are minimal.
The primary limitation of the present study is the difference between central hypovolemia induced by LBNP and hypovolemia induced by dehydration. Acute central hypovolemia induced by LBNP does not alter either blood concentration or blood viscosity. Therefore, the effects of hypovolemia on dynamic cerebral autoregulation might differ between LBNP and dehydration. A previous study that determined dynamic cerebral autoregulation during central hypovolemia without hemoconcentration induced by tilting the head up also found a decrease of steady-state CBF velocity and unchanged dynamic cerebral autoregulation (25). Further studies should compare dehydration with central hypovolemia induced by LBNP or by tilting the head to understand the effects of hemoconcentration on dynamic cerebral autoregulation. Also, changes in CVP were not actually comparable between hypovolemia (change in CVP, –4 mm Hg) and hypervolemia (change in CVP, 6 mm Hg) in the present study.
Another limitation of the present study is the use of TCD ultrasonography for the MCA. This approach is based on the assumption that the diameter of the MCA remains relatively constant, as shown by others (26–28). Moreover, previous studies have suggested that the diameter of large arteries minimally changes during hemodilution (7,29). Therefore, changes in blood flow velocity in the MCA would reflect a change in CBF. However, the MCA diameter might change during hypervolemic hemodilution.
There were two limitations with our study protocol. The possibility of potential changes in arterial CO2 and acidosis due to saline infusion cannot be excluded because we did not measure arterial CO2 and pH. However, ETco2 or respiratory rate in the present study remained unchanged, suggesting no remarkable changes in arterial CO2 and pH. The other limitation is the effect of LBNP on the results of saline infusion. The magnitude of lower body suction was mild compared with that in previous studies, which determined dynamic cerebral autoregulation during orthostatic stress using LBNP (20,30), and only one of our 12 subjects developed slight asymptomatic hypotension. Thus, presyncope symptoms might not have affected the present results. However, a second baseline should be established after LBNP or for any control group, to exclude the effects of LBNP and/or time factors.
Central blood volume frequently changes under clinical conditions and/or during perioperative maneuvers. Hypervolemic hemodilution is remarkable, especially in the fluid volume expansion required for kidney transplant surgery (31), in preoperative hypervolemic hemodilution to avoid blood transfusion during major surgery or for those whose religious beliefs proscribe such manipulations (32,33). Here, we investigated the effects of hypervolemic hemodilution and central hypovolemia on dynamic cerebral autoregulation by using spectral and transfer function analysis between arterial blood pressure and CBF velocity variability. We discovered that although steady-state CBF velocity changes under both central hypervolemia and hypovolemia, only hypervolemic hemodilution impairs dynamic cerebral autoregulation. The fluctuation in CBF would be substantial during hypervolemic hemodilution when the cerebral circulation is exposed to large and rapid fluctuations in arterial blood pressure induced by postural changes, injected vasoactive drugs and/or operative stimuli.
Footnotes
Accepted for publication July 9, 2007.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  E. A. M. Frost Neurosurgical Anesthesia and Critical Care, Vol. 25, No. 3 in Anesthesiology Clinics Anesth. Analg., May 1, 2008; 106(5): 1595 - 1596. [Full Text] [PDF]
|
|
|
|
|
|
J. K. Shoemaker Hemodilution Impairs Cerebral Autoregulation, Demonstrating the Complexity of Integrative Physiology Anesth. Analg., November 1, 2007; 105(5): 1179 - 1181. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
