The Loop Gain of Autonomic Reflex Function. Li Hui CHOW and Hsing I. CHEN*

Transcription

1 Japanese Journal of Physiology, 39, , 1989 The Loop Gain of Autonomic Reflex Function in Orthostatic Hypotension Li Hui CHOW and Hsing I. CHEN* Clinical Research Center, Triservice General Hospital, Taipei, Taiwan, Republic of China *Department of Physiology and Biophysics, National Defense Medical Center, Taipei, Taiwan, Republic of China Abstract The loop gain (G) of the autonomic reflex function in orthostatic stress was assessed in anesthetized dogs subjected to 45 and 90 head-up tilt. We observed the magnitude of orthostatic hypotension before and after 1) sinus denervation and vagotomy (SDVT), or 2) ganglionic blockade (GB) with hexamethonium. The decreases in arterial pressure during the orthostatic stress before and after interruption of the autonomic reflex from either the afferent or efferent limb were defined as E and D, respectively. The loop G of the compensatory system was calculated using closed-loop analysis: G = (D/E) -1. In the SDVT experiments, the average values of E, D, and G were 18.6 mmhg, 62.6 mmhg, and 2.36, respectively for 45 tilt; and 31.2 mmhg, 82.7 mmhg, and 1.63, respectively for 90 tilt. In the GB experiment, the E, D, and G values were 14.6 mmhg, 51.6 mmhg and 2.53 for 45 tilt; and 28.9 mmhg, 72.6 mmhg, and 1.51 for 90 tilt. The data provide a quantitative measure of the autonomic reflex function in orthostatic hypotension. Furthermore, we found that the corresponding G values in the SDVT and GB experiments were not significantly different. In each experiment, the G value in 90 tilt was lower than that in 45 tilt. The findings suggest that reflexes from the arterial baroreceptors and cardiopulmonary receptors account for a large part of the autonomic compensation to the orthostatic stress. The whole control system operates in a nonlinear fashion, because the gain value tends to decrease as the degree of tilt is increased. Key words : closed-loop analysis, attenuation factor, arterial baroreceptors, cardiopulmonary receptors. In physiological system, the loop gain of a regulatory system is defined to be the ratio of output response to input stimulus (MILHORN, 1966; SAGAWA, 1978; CHEN and BISHOP, 1983). The value of gain is used for a quantitative measure of Received for publication September 5, 1988 * To whom all correspondence should be addressed. 673

2 674 L. H. CHOW and H. I. CHEN the compensatory capability in minimizing the disturbance. In a linear control system, the disturbance D added to the controlled variable is attenuated to an error E by a negative feedback mechanism in a relationship: E=D/ (1 +gain) (MILHORN, 1966; MCRITCHIE et al., 1976; HOsoMI, 1978; CHEN and BISHOP, 1983). Consequently, gain value can be calculated from the magnitude of E and D determined before and after elimination of the compensatory mechanism. CHEN and BISHOP (1983) provided data to correlate the open-loop gain (output/input) of the baroreflex function with the arterial pressure compensation during hemorrhagic hypotension in a closed-loop condition. Because of the nonlinearity of the reflex control system, the gain value was not a constant function, varying with the degree of circularoty perturbation. The authors thus suggested that different range of arterial pressure disturbance should be produced for the determination of circulatory compensation in terms of gain. Most of the studies dealing with the closed-loop analysis of loop gain for circulatory compensation used arterial hemorrhage as the disturbance (KUMADA et al., 1970; EDIS,1971; PELLETIER et al., 1971; HOSOMI,1978; CHEN and BISHOP, 1983). Systemic hypotension induced by vasodilators was employed in a few experiments (MCRITCHIE et al., 1976; CHEN, 1979). Although postural change or lower body negative pressure has been known to cause circulatory changes (ABEL et al., 1963; GAUER and THRON, 1965; JOHNSON et al., 1974; ABBOUD et al., 1979; MARK and MANCIA,1983), a quatitative analysis of the autonomic reflex function in orthostatic stress is not found in the literature. In the present study, we determined the values of loop gain from the degree of orthostatic hypotension during head-up tilt before and after 1) deafferentation of the baroreceptors and cardiopulmonary receptors, and 2) pharmacological blockade of the autonomic efferents. The gain values were compared between these two experimental conditions. In addition, two degrees of tilt were produced to observe whether the gain value is altered with the magnitude of orthostatic stress. METHODS Preparation. Mongrel dogs of either sex, weighing 8-14 kg, were anesthetized with an intravenous dose (30 mg/kg) of pentobarbital sodium. The trachea was intubated and connected to a ventilator (Harvard Apparatus model 613). The animal was artificially ventilated with ambient air supplemented with 95%02-5%C02. Gallamine triethiodide (2 mg/kg, Lederle-Parentals) was given to avoid muscular activity. An arterial catheter was inserted through the femoral artery and advanced to a level near the aortic arch. The systemic arterial pressure (SAP) was monitored with a Statham P23dB transducer. The signal was damped with a low-pass filter (1/2 amplitude high frequency 0.5 Hz) to record the mean arterial pressure (MAP). Heart rate (HR) was monitored with a cardiotachometer (Grass 7P4F) triggered by the arterial pulses. A catheter was placed in the vena cava via the external jugular vein and connected with the pressure transducer (Statham P23dB) to measure the Japanese Journal of Physiology

3 REFLEX GAIN IN ORTHOSTATIC STRESS 675 central venous pressure (CVP). The peripheral venous pressure (PVP) was recorded by a transducer connected with saphenous vein catheter. All recordings were simultaneously displayed on a Grass 7P polygraph recorder. Drug was administered via a femoral vein catheter. Rectal temperature was maintained at C by an infrared lamp and a thermostatically regulated heating pad. Arterial blood gases and ph were measured every min with ph/blood gas analyzer (Radiometer PHM 73, Copenhagen). The tidal volume and respiratory rate were adjusted to maintain ph at , P~o2 at 34x-40 mmhg, and Po t in excess of 100 mmhg. Tilt. The dog was positioned on a tilt table which was operated by an oil-pressure-driven motor. The animal was fixed with a special cradle and fastener to prevent body movement during tilting. The right atrial level (approximated one-third thorax in the anteroposterior plane) was taken as the zero reference point for SAP and CVP (ABEL et al., 1963; GAUER and THRON, 1965). This level was fixed at the same plane as the tilt axis. To avoid gravitational forces during tilting, the pressure transducers for SAP and CVP were fixed along the tilt axis. The PVP transducer was located near the ankle region, which was approximately 65 cm from the tilt axis. Since this transducer was moved along with the table during tilting, it detected the hydrostatic pressure changes in the saphenous vein. Deafferentation and autonomic blockade. In one group of animals, the common carotid arteries and the vagus nerves were isolated in the neck region and looped with threads for subsequent occlusion or section. The carotid sinus regions were also exposed. At an appropriate time, bilateral cervical vagotomy was performed. This procedure increased the SAP and HR and enhanced the pressure response to common carotid occlusion. Subsequent denervation of the carotid sinuses was achieved by stripping the adventitia of the carotid sinuses and painting the area with 5% phenol-95% alcohol. The abolition of pressor response to common carotid occlusion was taken as the evidence of sinus denervation. Pharmacological blockade of the autonomic ganglia was done in other group of dogs by hexamethonium (10 mg/kg, iv.). Tilt trials were conducted 5-20 min after hexamethonium. Subsequently, the decreased SAP and HR after hexamethonium were elevated to levels close to the control by epinephrine infusion ( j g/(kg min), iv.), and tilt trials were repeated. Experimental protocols. The dogs were subjected to head-up tilt of 45 and 90. Each degree of tilt was accomplished within 5 s, repeated 2 times, and lasted 2-4 min. A resting period of more than 5 min was allowed to elapse between tilt trials. After the control runs, the same procedures were repeated after sinoaortic denervation + vagotomy (SDVT) in one group, or hexamethonium and hexamethonium + epinephrine in the other group. Calculation of loop gain and data analysis. After head-up tilt of 45 and 90, there were changes in SAP, CVP, PVP, and HR. These parameters reached a steady state within 2 min after postural alterations. The steady-state values of 2 trials for each degree of tilt were averaged for individual data. The decreases in SAP before and after baroreflex and autonomic blockade were defined as E and D, respectively. Vol. 39, No. 5, 1989

4 676 L. H. CHOW and H. I. CHEN The loop gain (G) was calculated by closed-loop analysis (MILHORN,1966; HosoMi, 1978; CHEN and BISHOP, 1983) as follows: G = (D/E) -1. All data were expressed as means + S.E. For comparison between groups of data, the significance in difference was evaluated by unpaired t-test. RESULTS Table 1 shows the control values of MAP, HR, CVP, and PVP before orthostatic stress. In this group of animals (n =8) after SDVT, the steady-state MAP and HR were elevated, the CVP was slightly but not significantly increased (p >0.1), and the PVP remained unaltered (p >0.1). Figure 1 illustrates an example of the cardiovascular responses to 45 tilt before and after SDVT. Head-up tilt caused an abrupt increase in PVP. Concomitantly, the SAP and CVP were decreased. Reflex responses such as tachycardia and partial recovery of hypotension occurred shortly after these changes (Fig, 1A). Following SDVT, the resting SAP and HR were elevated. The increase in PVP during head-up tilt remained essentially the same as the magnitude observed before SDVT. The reflex tachycardia was abolished. The fall in CVP became slightly greater. The orthostatic effect on the decrease in SAP was greatly accentuated after SDVT (Fig. 1B). Table 2 summarizes the steady-state responses to 45 and 90 tilt in 8 dogs. The orthostatic effects on the changes in arterial and venous pressures were apparently greater in 90 tilt than 45 (p<0.001). In either 45 or 90 head-up tilt, the increase in PVP was not affected by SDVT. The fall in SAP due to the same orthostatic effect was increased by an average of 44.0 ( ) mmhg in 45 tilt and 51.5 ( ) mmhg in 90 tilt after SDVT. In contrast to SDVT, autonomic blockade with hexamethonium (Hex) Table 1. The control values before head-up tilt in the group subjected to sham operation and baroreceptor deafferentation. Japanese Journal of Physiology

5 v REFLEX GAIN IN ORTHOSTATIC STRESS 677 vas -H~a.b ) E a Q, v a), c ) v IH c a U c,,-.. E 0 v, I 0 o p Q ~ 0 c v t.~. 'b c.- 0 Q)._ ) -- 4) c td.~.d ~ ) ' v cd - o ~~~~ 3 U b w c Vol. 39, No. 5, 1989

6 678 L. H. CHOW and H. I. CHEN Table 2. Effects of baroreceptor deafferentation on the cardiovascular responses to 45 and 90 tilt. decreased the resting SAP and HR (Fig. 2 and Table 3). Epinephrine (Ep) was infused intravenously to raise the SAP and HR towards control levels (Fig. 2 and Table 3). After Hex with or without Ep infusion, the fall in SAP during the orthostatic stress was much greater than the magnitude observed in control condition (Fig. 2). Table 4 summarizes the changes in MAP, HR, CVP, and PVP to 45 and 90 head-up tilt in three conditions: A) control, B) after Hex, and C) after Hex + Ep. The data showed that Hex did not affect the increase in PVP, but abolished the reflex tachycardia and augmented the orthostatic hypotension. Although the degree of hypotension after autonomic blockade was slightly smaller with Ep infusion, the difference was not statistically significant (p >0.1). The calculation of gain values (Table 5) was based on the magnitude of orthostatic hypotension observed before (a magnitude designated as E) and after (designated as D) cardiovascular deafferentation and autonomic blockade. In the experiments of autonomic blockade, the gain values with Ep infusion were smaller than those with Hex alone (2.19 ± 0.48 vs in 45 tilt, and vs in 90 tilt). However, the differences were not statistically significant (p >0.02). In either 45 or 90 tilt, the gain values obtained from the SDVT experiment Japanese Journal of Physiology

7 REFLEX GAIN IN ORTHOSTATIC STRESS E 'c3. ba U2.~ a U a~~u V a ~ 0 E i E 0 - a~ a d),.,. C.) U 0 j C) E o ~ S1 ~" 0 4) C" o.0-0h<.~ " 0-0 UC~ a GQ Vol. 39, No. 5, 1989

8 680 L. H. CHOW and H. I. CHEN a a 0 U a U U b a U a-4). N - by r..+ U 3 U n 'd 0 ~a U 0 _0 ' n a 0 ba a ba 0 U W N cc H Japanese Journal of Physiology

9 REFLEX GAIN IN ORTHOSTATIC STRESS 681 Table 5. Gain values calculated from the changes in mean arterial pressure before and after baroreceptor denervation or autonomic blockade. ( for 45 tilt and for 90 tilt) also did not differ from the corresponding values (2.53 ± 0.44 or 2.19 ± 0.48 for 45 tilt, and 1.51 ± 0.23 or for 90 tilt) in experiment with autonomic blockade (p >0.1). Thus the reflex responses from the arterial baroreceptors and cardiopulmonary receptors may account for a large, if not the entire, part of rapid autonomic compensation to orthostatic hypotension. On the other hand, it is noteworthy that difference exists in the gain values between 45 and 90 tilt. The significant decrease in the gain value with 90 tilt indicates that the compensatory capability of the autonomic reflex functions became relatively small in the face of a greater perturbation. DISCUSSION The nonlinearity of the open-loop characteristics of the baroreflex function has been demonstrated by many investigators (LEVISON et al., 1966; CHEN, 1979; CHEN et al., 1979; CHEN and BISHOP, 1983). The sigmoid relationship between the input pressure (carotid sinus pressure) and the output pressure (SAP) reflects that the loop G is not a constant function, but decreases toward the saturation pressure (CHEN and BISHOP, 1983). CHEN and BISHOP (1983) also found that the values of the carotid baroreflex G obtained from the closed-loop experiment after hemorrhagic hypotension were decreased significantly as the volume of arterial bleeding was increased. The data of the present experiment revealed that the G value was Vol. 39, No. 5, 1989

10 682 L. H. CHOW and H. I. CHEN significantly reduced when the degree of head-up tilt was increased from 45 to 90 (Table 5). Accordingly, the compensatory capacity in terms of the G value became less in the face of a greater perturbation. Previous studies dealing with the circulatory changes during the orthostatic stress usually used 1 degree (30-60 ) of tilt (EDHOLM, 1940; ABEL et al., 1963; GAUER and THRON, 1965). Although we had only two degrees of head-up tilt, our findings appeared to agree with the nonlinearity of the baroreflex input-output relationship. Substantial evidence supports the functional importance of the sympathetic nervous system in maintaining the arterial pressure during orthostatic stress. In man and animals, vasoconstrictor responses to central and peripheral sympathetic stimulation result in increased vascular resistance and decreased vascular capacity (MELLANDER, 1960; MELLANDER and JOHANSON, 1968; SHOUKAS and SAGAWA, 1973; CHEN and WANG, 1984). The changes in vascular resistance and capacity due to sympathetic stimulation minimize the systemic hypotension due to postural changes (GAUER and THRON, 1965, JOHNSON et al., 1974; SHEPHERD and VANHOUTTE, 1978; MARK and MANCIA, 1983). In the present experiment, autonomic blockade with hexamethonium greatly increased the magnitude of systemic hypotension. A steady infusion of epinephrine sufficient to raise the arterial pressure to the control level had only a small beneficial effect on the orthostatic hypotension. The overall loop gain of the autonomic efferent compensation was calculated to be 2.53 and 1.51 in 45 and 90 tilt, respectively. The values were derived on the basis of the autonomic compensation to orthostatic hypotension by 70.3% (45 tilt) and 62.3% (90 tilt) of the arterial pressure disturbance. Although many compensatory mechanisms may contribute to the attenuation of orthostatic hypotension in upright position, the immediate responses are autonomic reflex functions, muscle contraction, and respiratory changes, etc. (MURRAY et al., 1968; BANNISTER et al., 1977; ABBOUD et al., 1979; DAMPNEY et al., 1979; SCHATZ, 1984). In the present study, experiments were performed in animals paralyzed with neuromuscular blockade and ventilated with a respirator to avoid muscle movements and respiratory changes. We thus confined the analysis of compensatory mechanism in the scope of autonomic reflex function. In addition, we obtained the steady-state values in 2 min after the postural changes. Other delayed compensatory mechanisms were not likely involved. However, it should be noted that the experiments were done in anesthetized animals. The reflex automatic compensation could be greater in a conscious state. With respect to the afferent limbs of the reflex circulatory control, the relative importance of arterial baroreceptors and cardiopulmonary receptors in orthostatic stress may vary with animal species (EDHOLM, 1940; MAYERSON, 1942; ABEL et al., 1963; MARK and MANCIA, 1983) and range of operation (JOHNSON et al., 1974; ABBOUD et a1.,1979; DAMPNEY et at., 1979; MARK and MANCIA, 1983). Nevertheless, our experiments with SDVT and hexamethonium revealed that the gain values were virtually the same regardless of whether the reflex loop was interrupted from the afferent or efferent limbs. Accordingly, the reflex function from the arterial Japanese Journal of Physiology

11 REFLEX GAIN IN ORTHOSTATIC STRESS 683 baroreceptors and cardiopulmonary receptors accounts for a large, if not the entire, part of the autonomic compensation in the orthostatic stress. In this experiment, SDVT and hexamethonium increased the postural changes in SAP and CVP, but not the saphenous venous pressure (PVP). Thus the elevation of PVP might only reflect the height of hydrostatic column below the heart. In this connection, several reports (RosE et al., 1961; BROWSE et al., 1966; EPSTEIN et al., 1968; SHEPHERD and VANHOUTTE, 1978) suggested that reflex sympathetic activation did not significantly alter the venous tone or volume in the limb veins. Another possibility for the unaltered PVP in the face of reflex sympathetic activation was a reduction in venous capacity without significant change in the pressure. The baroreflex G in arterial pressure compensation have been assessed using open- and closed-loop methods (LEVISON et al., 1966; ALLISON et al., 1969; DONALD and EDIS, 1971; MCRITCHIE et al., 1976; CHEN, 1979; CHEN et al., 1979; CHEN and BISHOP, 1983). CHEN and BISHOP (1983) were the first to correlate the G values from the open-loop experiment with the data from the closed-loop approach. The study suggested that the G values obtained from the open- and closed-loop methods were essentially identical within the same input magnitude. Since an open-loop analysis requires vascular isolation of the baroreceptor area, the surgical procedure may cause interference with the innervation and the compliant property of the vescular wall in the reflexogenic area (EDIS, 1971; CHEN and BISHOP, 1983). In this and the other experiments (MCRITCHIE et al., 1976; HOSOMI, 1978; CHEN and BISHOP, 1983), the loop gain was obtained simply by determining the fall in arterial pressure during the orthostatic stress or hemorrhage before and after sinoaortic denervation and cervical vagotomy. The procedure inevitably elevated the baseline SAP before tilt. However, CHEN and BISHOP (1983) showed that the loop G for hemorrhagic hypotension was not determined by the baseline SAP. In addition, hexamethonium decreased while epinephrine elevated the baseline SAP. The values of G were not much different in various levels of baseline SAP after SDVT, Hex, and Hex + Ep. Our data revealed that the total baroreflex compensation in terms of the loop gain during the orthostatic stress averaged 2.36 and 1.63 for 45 and 90 head-up tilt, respectively. For a comparison, the gain values obtained from open-loop analysis of the carotid and aortic baroreflexes in anesthetized animals varied from (LEVISON et al., 1966; KUMADA et al., 1970; CHEN, 1979; CHEN et al., 1979; CHEN and BISHOP, 1983). In the dog, aortic baroreceptors were shown to play a minor role in the antihypotensive function (ALLISON et al., 1969; DONALD and EDIS, 1971; EDIS,1971). The difference between the values of in this study and in others may be attributed to: 1) the simple technique without vascular isolation in our closed-loop experiment, 2) the additional reflex compensation from the cardiopulmonary receptors, and 3) the nonlinearity of the baroreflex function and hence the dependence of G on the input magnitude. This study was supported in part by the National Science Council, ROC (Grant No. NSC B016-33) and a CRC Grant from the Academia Sinica. Vol. 39, No. 5, 1989

12 684 L. H. CHOW and H. I. CHEN The authors wish to thank Mr. H. C. Liu and Miss L. W. Huang for the assistance and Miss F. H. Chang for the preparation of this paper. technical REFERENCES ABBOUD, F. M., ECKBERG, D. L., JOHANNSEN, U. J., and MARK, A. L. (1979) Carotid and cardiopulmonary baroreceptor control of splanchnic and forearm vascular resistance during venous pooling in man. J. Physiol. (Lond.), 286: ABEL, F. L., PIERCE, J. H., and GUNTHEROTH, W. G. (1963) Baroreceptor influence on postural changes in blood pressure and carotid blood flow. Am. J. Physiol., 205: ALLISON, J. L., SAGAWA, K., and KUMADA, M. (1969) Open-loop analysis of the aortic arch barostatic reflex. Am. J. Physiol., 217: BANNISTER, R., SEVER, P., and and GROSS, M. (1977) Cardiovascular reflexes and biochemical responses in progressive autonomic failure. Brain, 100: BROWSE, N. L., DoNAL, D. E., and SHEPHERD, J. T. (1966) Role of the veins in the carotid sinus reflex. Am. J. Physiol., 210: CHEN, H. I. (1979) Interaction between the baroreceptor and Bezold-Jarisch reflexes. Am. J. Physiol., 237 (Heart Circ. Physiol., 6): H655-H661. CHEN, H. I. and BISHOP, V. S. (1983) Baroreflex open-loop gain and arterial pressure compensation in hemorrhagic hypotension. Am. J. Physiol., 245 (Heart Circ. Physiol., 14): H54-H59. CHEN, H. I., CHAI, C. Y., TUNG, C. S., and CHEN, H. C. (1979) Modulation of the carotid baroreflex function during volume expansion. Am. J. Physiol., 237 (Heart Circ. Physiol., 6): H153-H158. CHEN, H. I. and WANG, D. J. (1984) Systemic and pulmonary hemodynamic responses to intracranial hypertension. Am. J. Physiol., 247 (Heart Circ. Physiol.,16): H715-H721. DAMPNEY, R. A., STELLA, L. A., Golln, R., and ZANCHETTI, A. (1979) Vagal and sinoaortic reflexes in postural control of circulation and renin release. Am. J. Physiol., 237 (Heart Circ. Physiol., 6): H 146-H 152. DONALD, D. E. and EDIS, A. J. (1971) Comparison of aortic and carotid baroreflexes in the dog. J. Physiol. (Loud.), 215: EDIS, A. J. (1971) Aortic baroreflex function in the dog. Am. J. Physiol., 221: EDHOLM, 0. G. (1940) Effect of gravity on the blood pressure of the cat. J. Physiol. (Lond.), 98: EPSTEIN, S. E., BEISTER, G. D., STAMPFER, M., and BRAUNWALD, E. (1968) Role of the venous system in baroreceptor-mediated reflex in man. J. Clin. Invest., 47: GAUER, 0. H. and THRON, H. L. (1965) Postural changes in circulation. In: Handbook of Physiology, ed. by HAMILTON, W. F. and DOW, P., American Physiological Society, Bethesda, Maryland, Circulation 3, Sect. 2, Chap. 67, pp HOSOMI, H. (1978) Overall characteristics of arterial pressure control systems studied by mild hemorrhage. Am. J. Physiol., 234 (Regul. Intergr. Comp. Physiol., 3): R104-R109. JOHNSON, J. M., ROWELL, L. B., MANFRED, N., and EISMAN, M. M. (1974) Human splanchnic and forearm vasoconstrictor responses to reductions of right atrial and aortic pressures. Circ. Res., 34: KUMADA, M., SCHMIDT, R. M., SAGAWA, K., and TAN, K. S. (1970) Carotid sinus reflex in response to hemorrhage. Am. J. Physiol., 219: LEVISON, W. H., BARNETT, G. 0., and JACKSON, W. D. (1966) Nonlinear analysis of the Japanese Journal of Physiology

13 REFLEX GAIN IN ORTHOSTATIC STRESS 685 baroreceptor reflex system. Circ. Res., 18: MARK, A. L. and MANCIA, G. (1983) Cardiopulmonary baroreflexes in humans. In: Handbook of Physiology, the Cardiovascular System, ed. by HAMILTON, W. F. and Dow, P., American Physiological Society, Bethesda, Maryland, III, Chap. 21, pp MAYERSON, H. S. (1942) Effect of gravity on the blood pressure of the dog. Am. J. Physiol., 135: MCRITCHIE, R. J., VATNER, S. F., HEYNDRICKX, G. R., and BRAUNWALD, E. (1976) The role of arterial baroreceptors in the regulation of arterial pressure in conscious dogs. Circ. Res., 39: MELLANDER S. (1960) Comparative studies on the adrenergic neurohormonal control of resistance and capacitance blood vessels in the cat. Acta Physiol. Scand. (Suppl. 176), 50: MELLANDER, S. and JOHANSON, B. (1968) Control of resistance, exchange and capacitance functions in the peripheral circulation. Pharmacol. Rev., 20, MILHORN, H. T., Jr. (1966) The Application of Control Theory to Physiological Systems, Saunders, Philadelphia, PA. MURRAY, R. H., THOMPSON, L. J., BOWERS, J. A., and ALBRIGHT, C. D. (1968) Hemodynamic effects of graded hypovolemia and vasodepressor syncope induced by lower body negative pressure. Am. Heart J., 76: PELLETIER, C. L., EDIS, A. J., and SHEPHERD, J. T. (1971) Circulatory reflex from vagal afferents in response to hemorrhage in the dog. Circ. Res., 29: Ross, J., Jr., FRAHM, C. J., and BRAUNWALD, E. (1961) Influence of carotid baroreceptor and vasoactive drugs on systemic vascular volume and venous distensibility. Circ Res., 9: SAGAWA, K. (1978) Concerning "gain." Am. J. Physiol., 235 (Heart Circ. Physiol., 4): H117. SCHATZ I. J. (1984) Orthostatic hypotension, I. Functional and neurogenic causes. Arch. Intern. Med., 144: SHEPHERD, J. T. and VANHOUTTE, P. M. (1978) Role of the venous system in circulatory control. Mayo Clin. Proc., 53: SHOUKAS, A. A. and SAGAWA, K. (1973) Control of total systemic vascular capacity by the carotid sinus baroreceptor reflex. Circ. Res., 33: Vol. 39, No. 5, 1989