Do vestibular otolith organs participate in human orthostatic blood pressure control?

Autonomic Neuroscience: Basic and Clinical 100 (2002) 77 – 83 www.elsevier.com/locate/autneu

Do vestibular otolith organs participate in human orthostatic blood pressure control? Donald E. Watenpaugh *, Adriena V. Cothron, Stephen L. Wasmund, Wendy L. Wasmund, Robert Carter III, Nicolette K. Muenter, Michael L. Smith University of North Texas Health Science Center, Department of Integrative Physiology, 3500 Camp Bowie Blvd., Fort Worth, TX 76107, USA Received 30 October 2001; received in revised form 17 June 2002; accepted 3 July 2002

Abstract We hypothesized that vestibular otolith organ stimulation contributes to human orthostatic responses. Twelve subjects underwent three 60j upright tilts: (1) with the neck flexed from 0j to 30j relative to the body during 60j tilt, such that the head moved from horizontal to 90j above horizontal (0 to 1 Gz otolith stimulation); (2) with the head and body aligned, such that they tilted together to 60j (0 to 0.87 Gz otolith stimulation); and (3) with the neck flexed 30j relative to the body during supine conditions, and the neck then extended to 30j during 60j body tilting, such that the head remained at 30j above horizontal throughout body tilting (constant 0.5 Gz otolith stimulation). All three tilt procedures increased thoracic impedance, sympathetic nerve activity (N = 8 of 12), arterial pressure, and heart rate relative to supine conditions (all P < 0.04). Within the first 20 s of tilt, arterial pressure increased most obviously in the 0 to 1 Gz otolith condition. Thoracic impedance tended to increase more in otolith-constant conditions, but no dependent variable differed significantly between tilt conditions, and no significant time  tilt interactions emerged. Otolith inputs may contribute to early transient adjustments to orthostasis. However, lack of significant main effects of tilt condition and time  tilt interactions suggests that potential otolith effects on the variables we studied are relatively subtle and ephemeral, or that other mechanisms compensate for a lack of change in otolith input with orthostasis. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Gravity; Head-up tilt; Muscle sympathetic nerve activity; Heart rate; Hemodynamics

1. Introduction Specific neuroanatomy exists which connects the mammalian vestibular system with cardiovascular control centers, and these pathways carry functional significance. Studies in cats demonstrated that vestibular signals impact cardiovascular control centers of the brainstem via pathways through the cerebellum (Doba and Reis, 1972, 1974). Stimulation of the vestibular nerve produces pressor effects in anesthetized cats, and lesions of the fastigial nucleus impair reflex cardiovascular responses to head-up tilting (Doba and Reis, 1972, 1974). Yates (1996), Yates and Miller (1998), and Biaggioni et al. (1998) each reviewed what is known about the neuroanatomy and physiology of vestibular – autonomic interactions. * Corresponding author. Naval Submarine Medical Research Laboratory, Box 900, Groton, CT 06349, USA. Tel.: +1-860-694-5106; fax: +1860-694-2523. E-mail address: [email protected] (D.E. Watenpaugh).

The literature supports a role for both the otolith receptors and semicircular canals in modulating vestibular –autonomic control of cardiovascular function, although Yates (1996) suggested that the otolith plays the more important role. The otolith receptors produce a continuous signal which registers the head’s orientation relative to gravity, and other linear but non-gravitational accelerations of the head. Substantial evidence exists concerning vestibular otolith effects on cardiovascular function in humans. In prone humans, changes in head ‘‘pitch’’ (anterior – posterior neck flexion/extension) alter limb muscle sympathetic nerve activity (MSNA) and vascular resistances (Essandoh et al., 1988; Shortt and Ray, 1997). Essandoh et al. (1988) showed that head-down neck flexion (i.e. hanging the head over the edge of a table so the head and otoliths are upside-down) elicited 35 – 39% reduction in calf and forearm blood flow relative to prone, chin-supported head-up neck extension conditions. They saw no change in arterial pressure or heart rate, indicating that limb vasoconstriction occurred with head-down neck flexion. Subsequently, Shortt and Ray

1566-0702/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S 1 5 6 6 - 0 7 0 2 ( 0 2 ) 0 0 1 4 2 - X

78

D.E. Watenpaugh et al. / Autonomic Neuroscience: Basic and Clinical 100 (2002) 77–83

(1997) measured MSNA during a protocol similar to that used by Essandoh et al. (1988), and found that the limb vasoconstriction was associated with a 63% increase in MSNA during head-down neck flexion. Those works and other studies also reveal that: (1) supine, head-up neck flexion had no significant effect on limb blood flow (N = 3 (Essandoh et al., 1988)); (2) limb SNA and flow responses appear to be specific to the skeletal muscle circulation: the skin is not involved (Essandoh et al., 1988; Ray et al., 1997); (3) sustained sympathoexcitation occurs during 30 min of head-down neck flexion in spite of a progressive rise in blood pressure over that time (Hume and Ray, 1999); (4) neck mechanoreceptors activated by neck flexion and extension may slightly modify some cardiovascular effects attributed to otolithic receptors (Normand et al., 1997); and (5) sympathoexcitation from headdown neck flexion and sympathoexcitation from prone lower body negative pressure (LBNP; simulated orthostasis) are additive in humans (Ray, 2000). The purpose of the present work was not to confirm or further explore the above interesting findings from neck flexion studies in supine and prone subjects. Instead, we sought to determine the role vestibular inputs play in the human cardiovascular adjustments to upright posture (true orthostasis). We tested the hypothesis that supine-to-upright otolith stimulation contributes to cardiovascular responses to orthostasis.

2. Materials and methods 2.1. Subjects Twelve healthy subjects (four women, eight men) participated in this study, which was approved by our institutional review board. Subjects ranged in age from 20 to 45 years, weighed 66.5 F 11.6 kg, and were 174 F 8 cm tall (means F S.D.). No subject reported any medical history of cardiovascular disease or exhibited any illness at the time of their study, and subjects were not taking any medications at the time of their study. 2.2. Protocol Subjects reported to the laboratory and were instrumented for measurement of otolith Gz stimulation, thoracic impedance/electrocardiography, finger arterial blood pressure, and MSNA. We present details of specific measurements below. After instrumentation, subjects underwent a brief trial whole body tilt to 60j to ensure instrumentation and connections remained undisturbed by tilting, and to ensure the MSNA signal remained uncontaminated by artifacts. Thereafter, subjects underwent three different tilting procedures in random order. Each followed the same sequence: 1 min of baseline data collection, 3 min in the

tilted position, and then return to supine baseline conditions. Whole body tilting from horizontal to 60j above horizontal required 10 –15 s. At least 5 min transpired between tilts to ensure full recovery before the next tilt. All three tilts moved the body from horizontal position (0j) to 60j above horizontal, but changes in head orientation differed between the three types of tilt (Fig. 1). In one, the neck was flexed passively from 0j to 30j relative to the body during 60j body tilt, such that the head moved from horizontal position to 90j above horizontal. This increased Gz stimulation of the otolith organ from 0 to 1 Gz (Gz = sin(90j) = 1.0). Therefore, this type of tilt is hereafter designated ‘‘0 to 1 Gz’’. In a second tilt (designated ‘‘0 to 0.87 Gz’’), the head was kept in line with the body, such that the head also moved from horizontal position to 60j above horizontal. This increased Gz stimulation of the otolith organ from 0 to 0.87 Gz (Fig. 1B). These first two tilt conditions were designed to impose normal or near-normal supine to upright Gz stimulation of the otoliths during whole body tilt. In the third type of tilt (designated ‘‘otolithconstant’’), the neck was flexed 30j relative to the body during baseline conditions, and the neck was then passively extended to 30j relative to the body during upright body tilting, such that the head remained at 30j above horizontal throughout the procedure (Fig. 1C). This maintained Gz stimulation of the otolith organ constant at 0.5 Gz during the procedure. Therefore, this tilt condition provided the cardiovascular challenge of whole body tilting without any concurrent increase in otolith Gz stimulation. To accomplish whole body tilting with simultaneous passive head and neck movements, a headboard flexibly attached to the tilt bed supported the subject’s head. An investigator manually controlled headboard angle, and thus head (otolith) Gz stimulation. A plastic bag filled with gel under the head allowed head tilting to be guided smoothly and comfortably such that the subject kept their neck relaxed during tilting. Subjects wore a blindfold to eliminate visual cues that might affect responses. In the 0 to 1 and 0 to 0.87 Gz tilts described above, the head and neck were horizontal during pre-tilt baseline conditions, whereas the neck was flexed 30j (0.5 Gz otolith stimulation) during baseline conditions prior to the otolithconstant tilt (Fig. 1). Therefore, to determine whether these two baseline conditions were comparable (i.e. whether 0.5

Fig. 1. Illustration of head and body tilt angles for each experimental condition. (A) 60j head-up tilt with full orthostatic otolith reorientation (0 to 1 Gz otolith stimulation); (B) 60j head-up tilt with partial orthostatic otolith reorientation (0 to 0.87 Gz otolith stimulation); (C) 60j head-up tilt with no change in otolith orientation (otolith stimulation constant at 0.5 Gz).

D.E. Watenpaugh et al. / Autonomic Neuroscience: Basic and Clinical 100 (2002) 77–83

Gz otolith stimulation by itself elicited significant cardiovascular effects in supine subjects), a subset of eight horizontal supine subjects underwent 30j neck flexion. Essandoh et al. (1988) found no significant effects of this intervention on hemodynamics (N = 3), but we felt a need to confirm those results. Protocol timing was similar to that used for the other conditions described above: 1 min of 0j baseline followed by 3 min of 30j neck flexion. Thoracic impedance, arterial pressure, and heart rate were measured. We also performed two studies to assess the effect of 30j neck flexion and extension per se using the same approach as Normand et al. (1997). Subjects lied horizontal on their left side with their head supported and with the neck neutral, flexed 30j, or extended 30j, as assessed with a goniometer. This strategy allowed neck flexion/extension with no change in otolith orientation to gravity. In one study, nine subjects underwent 1 min of neck flexion followed by 3 min of neck extension. This duplicated neck and head positions in otolith-constant 60j body tilt (Fig. 1C). In another study, seven subjects went from 1 min of head neutral positioning to 3 min of 30j neck flexion. This duplicated neck and head positions in the 0 to 1 Gz tilt (Fig. 1A). Only arterial pressure and heart rate were measured in these two ancillary studies. 2.3. Equipment and measurements A standard tilt bed was used to tilt subjects to a 60j upright position (CircOlectric Bed, Stryker, Kalamazoo, MI). A foot plate was used for support. Subjects stood on their left foot during 60j tilt, because MSNA recording from the right leg required that leg to remain relaxed and unweighted. Accelerometry assessed otolith Gz stimulation. The accelerometer (Model B-1 with NV8A signal amplifier/ filter, Rieker Electronics, Folcroft, PA) was positioned securely on the crown of the head with elastic bands. The signal from the device then provided continuous monitoring of head, and thus otolith, Gz stimulation. A tetrapolar Minnesota impedance cardiograph model 304A (Surcom, Minneapolis, MN) measured changes in thoracic impedance as an index of changes in central blood volume. Decreases in thoracic blood volume are associated with increased thoracic impedance (Ebert et al., 1986). Therefore, we used changes in thoracic impedance as an indirect, inverse measure of changes in central blood volume and, thus, an index of the physiological stress of orthostasis. This measurement involves placement of four Mylar band electrodes circumferentially around the neck and thorax, application of alternating current to the outer two electrodes, and measurement of impedance between the inner two electrodes. To avoid potential artifacts associated with neck movements, we modified the standard electrode placement slightly. The uppermost electrode was placed around the base of the neck, where minimal movement occurred, and the other upper electrode was placed around the upper chest at the level of the axillae.

79

We directly measured leg MSNA with standard microneurographic techniques (Watenpaugh et al., 1999). Briefly, we supported the right leg of the supine subject comfortably on cushions, with the knee elevated and slightly bent. Electrical stimulation through the skin identified the location and course of the peroneal nerve at the popliteal fossa (postero-lateral aspect of the knee). Then, a sterile reference microelectrode was inserted subcutaneously 2 –3 cm from the identified nerve path (tip diameter 5– 10 Am, 35 mm long, Frederick Haer and Co., Bowdoinham, ME). A similar but shielded recording microelectrode was then inserted into the nerve. The nerve was slowly and gently probed with the microelectrode while monitoring the signal for insertion discharges and action potentials. A processing system rectified, integrated, and amplified the nerve signal (University of Iowa Bioengineering, Iowa City, IA). Reproducible activation of pulse-synchronous bursts during a prolonged apnea at residual lung volume, and lack of response to skin stroking or startle stimuli, confirmed recording of muscle, and not skin, sympathetic nerve activity. During baseline control conditions, we assigned the average sympathetic burst amplitude a value of 100 units, and all bursts during subsequent experimental periods were normalized to this value. This measurement was successfully performed in only 8 of our 12 subjects. Arterial pressure was measured non-invasively with a Finapres photoplethysmographic monitor placed around the middle finger (model 2300, Ohmeda, Englewood, CO). This instrument also measured heart rate. Mean arterial pressure was calculated as diastolic pressure plus one third of pulse pressure. 2.4. Statistical analyses We reduced data from the recumbent baseline period, from the first 20 s of tilt to assess the initial phasic responses to tilt, and from the second and third minutes of tilt to assess steady state responses. Two-factor repeated measures analyses of variance (ANOVA) then assessed effects of the two independent variables time and type of tilt (and time  tilt interaction) for each of the dependent variables thoracic impedance, MSNA, heart rate, and mean arterial pressure. For the studies of neck flexion effects, we used single factor ANOVA (factor: time). If significant main effects occurred, post hoc least significant difference (LSD) multiple comparison tests determined specific differences. Significance was assigned when P < 0.05. Results are presented as means F S.E.

3. Results All three types of tilt increased thoracic impedance within 20 s relative to supine conditions (ANOVA F(3,66) = 12.57, P < 0.001; all LSD P < 0.04; Fig. 2). A trend appeared for thoracic impedance to increase more during otolith-constant

80

D.E. Watenpaugh et al. / Autonomic Neuroscience: Basic and Clinical 100 (2002) 77–83

Fig. 2. Thoracic impedance responses to the three types of whole body tilt (means F S.E.; N = 12). See Fig. 1 and Materials and methods for details of the three tilt conditions. All three types of tilt increased thoracic impedance ( P < 0.04), but no significant differences existed in thoracic impedance between tilt conditions ( P > 0.06).

tilting relative to the two types of tilting which altered otolith orientation to gravity (ANOVA F(2,66) = 3.05, P = 0.068; all LSD P > 0.03). However, no interaction occurred between effects of time and tilt condition (ANOVA F(6,66) = 0.90, P = 0.498), indicating effects of time did not depend on tilt condition. Tilting increased mean MSNA 160 –180% within 20 s (ANOVA F(3,42) = 52.67, P < 0.001; all LSD P < 0.001; N = 8), with no significant difference between types of tilt in this increase (ANOVA F(2,42) = 0.315, P = 0.735; N = 8; Fig. 3). MSNA increased further between 20 s and 3 min of tilt for the 0 to 1 Gz (LSD P = 0.009) and otolith-constant tilt conditions (LSD P = 0.031).

Fig. 3. Muscle sympathetic nerve activity (MSNA) responses to the three types of whole body tilt (means F S.E.; N = 8). See Fig. 1 and Materials and methods for details of the three tilt conditions. All three types of tilt increased MSNA ( P < 0.01), but no significant differences existed in MSNA between tilt conditions ( P > 0.7).

Fig. 4. Mean arterial pressure responses to the three types of whole body tilt (means F S.E.; N = 12). See Fig. 1 and Materials and methods for details of the three tilt conditions. All three types of tilt increased mean arterial pressure ( P < 0.01), yet at 20 s, this increase was most apparent for the 0 to 1 Gz condition. However, no significant differences existed in mean arterial pressure between tilt conditions ( P > 0.5).

By 2 min, tilting significantly increased mean arterial pressure f 13% above recumbent values, revealing the highly significant main effect of time (ANOVA F(3,66) = 32.98, P < 0.001; LSD P < 0.001; Fig. 4). However, ANOVA indicated no significant main effect of tilt condition on mean arterial pressure ( F(2,66) = 0.616, P = 0.549). Also, no interaction occurred between effects of time and tilt condition on arterial pressure (ANOVA F(6,66) = 1.15, P = 0.343), which indicates effects of time did not depend on tilt condition. We took the liberty of analyzing LSD results comparing data from times within tilt conditions in spite of lack of time  tilt interaction. In this

Fig. 5. Heart rate responses to the three types of whole body tilt (means F S.E.; N = 12). See Fig. 1 and Materials and methods for details of the three tilt conditions. All three types of tilt increased heart rate ( P < 0.01), but no significant differences existed in heart rate between tilt conditions ( P > 0.13).

D.E. Watenpaugh et al. / Autonomic Neuroscience: Basic and Clinical 100 (2002) 77–83 Table 1 Cardiovascular measurements during 30j neck flexion while supine

Thoracic impedance (V) Mean arterial pressure (mm Hg) Heart rate (beats per min)

Horizontal baseline

Flexion 20 s

2 min

3 min

10.7 F 0.9

10.6 F 0.9

10.0 F 1.0

10.6 F 0.8

85 F 3

85 F 4

87 F 4

87 F 5

54 F 3

70 F 12

63 F 11

55 F 4

81

Table 3 Cardiovascular measurements during 30j neck flexion while lying on left side Baseline

Mean arterial pressure (mm Hg) Heart rate (beats per min)

Flexion 20 s

2 min

3 min

84 F 3

83 F 3

86 F 4

86 F 3

60 F 3

60 F 3

62 F 3

61 F 4

Means F S.E.; N = 7; no significant changes (all ANOVA P > 0.17).

Means F S.E.; N = 8; no significant changes (all ANOVA P > 0.2).

analysis, only the 0 to 1 Gz condition significantly increased arterial pressure between baseline and 20 s of tilt (LSD P = 0.003; LSD P = 0.094 for 0 to 0.87 Gz tilt; LSD P = 0.409 for otolith-constant tilt). Therefore, the lack of significant main effects of tilt and of time  tilt interaction disagreed with the post hoc observation that only the 0 to 1 Gz condition increased arterial pressure within the first 20 s of tilt. We performed a subsequent ANOVA on only the baseline and 20 s mean arterial pressure data with results as follows: time F(1,22) = 7.67, P = 0.018; tilt condition F(2,22) = 0.95, P = 0.401; time  tilt(2,22) = 2.86, P = 0.079. Therefore, this analysis agreed with the ANOVA including all four times. Heart rate increased f 19% within 20 s of tilt for all three conditions (ANOVA F(3,66) = 21.98, P < 0.001; all LSD P < 0.001), with no significant difference between tilt conditions (ANOVA F(2,66) = 0.535, P = 0.593; all LSD P>0.13; Fig. 5). Heart rate increased further between 20 s and 2 min of tilt for the 0 to 0.87 Gz (LSD P = 0.044) and otolith-constant tilt conditions (LSD P = 0.028). In the subset of eight horizontal supine subjects who underwent 30j neck flexion, no significant changes occurred in thoracic impedance, blood pressure, or heart rate (ANOVA, all P>0.2; Table 1). Heart rate tended to increase early in the neck flexion period. No significant cardiovascular responses occurred in the seven to nine subjects who lied on their side while undergoing neck flexion and extension (Tables 2 and 3). ANOVA indicated

Table 2 Cardiovascular measurements during 30j neck flexion followed by 30j neck extension while lying on left side

Mean arterial pressure (mm Hg) Heart rate (beats per min)

Baseline (flexion)

Extension 20 s

2 min

3 min

82 F 2

79 F 3

83 F 3

82 F 3

60 F 3

59 F 2

57 F 3

58 F 3

Means F S.E.; N = 9. Mean arterial pressure: ANOVA P = 0.015; all LSD P > 0.07. Heart rate: ANOVA P > 0.35.

existence of a time effect on arterial pressure during neck extension (Table 2, P = 0.015), but post hoc testing did not confirm this finding (all LSD P > 0.07).

4. Discussion On the surface, the results refuted our hypothesis that supine-to-upright otolith organ stimulation contributes to acute autonomic and cardiovascular responses to orthostasis. We expected that each of the two ‘‘head-up’’ tilt conditions we employed (0 to 1 and 0 to 0.87 Gz) would elicit different MSNA and cardiovascular responses than those seen in the otolith-constant condition if the otolith organs participated importantly in such responses. In spite of our expectations, no statistically definitive differences emerged, so the results suggest that supine-to-upright otolith stimulation may not contribute substantially to cardiovascular responses to orthostasis in humans, at least in the variables we studied. However, two features of the results imply existence of otolith organ effects on orthostatic cardiovascular control. First, a trend exists for thoracic impedance to increase more in the otolith-constant tilt condition (Fig. 2). This suggests greater reduction of central blood volume when whole body tilt occurs without concomitant otolith reorientation. It is possible that this resulted from effects of neck movement on impedance measurement, but we attempted to avoid such effects by placement of the uppermost impedance electrode at the base of the neck (see Materials and methods). Second, during the first 20 s of tilt, only the 0 to 1 Gz otolith condition appeared to increase mean arterial pressure, which suggests the early phasic blood pressure response to tilt requires full upright otolith stimulation (Fig. 4). If the otolith receptors responsible for the vestibulosympathetic reflex are rapidly adapting, then this could explain the transient lack of blood pressure elevation in the otolith-constant and 0 to 0.87 Gz conditions: full headupright (1 Gz) stimulation of such rapidly adapting receptors might be necessary for the early (within 20 s) blood pressure response to tilt. This interpretation agrees with recent work by Kaufmann et al. (2002), suggesting that a short-latency vestibulosympathetic reflex contributes to blood pressure maintenance during forward linear acceleration such as experienced during standing. However, in the

82

D.E. Watenpaugh et al. / Autonomic Neuroscience: Basic and Clinical 100 (2002) 77–83

context of prone head-down neck flexion (a nearly opposite stimulus to those used in the present study), Ray’s group consistently finds prolonged effects of otolith-mediated vestibulosympathetic effects (e.g. 30 min, (Hume and Ray, 1999)). Also, in the present study, no MSNA differences between tilt conditions appeared at 20 s or thereafter. Lack of significant main effects of tilt condition for the studied variables, and lack of significant time  tilt interactions, suggest that potential otolith effects may be relatively subtle and transient. Upright posture elicits lower body vasoconstriction and heart rate elevation, which act to prevent excessive gravitational redistribution of blood volume and to maintain cerebral blood flow (Rowell, 1993). Theoretically, one might expect the vestibular otolith organs to help mediate this response. They sense the head’s orientation to gravity, which often corresponds to the body’s orientation to gravity, and thus the gravitational challenge of maintaining cerebral blood flow. Our results suggest that established mechanisms (arterial and cardiopulmonary baroreflexes, veno-arteriolar reflexes, myogenic vasoconstriction, etc.) may mediate cardiovascular responses to upright posture without vestibular otolith involvement, at least beyond the first minute of orthostasis. The fact that the head’s orientation to gravity does not always correspond to the body’s orientation to gravity (e.g. the otolith-constant condition in the present study), and that otoliths sense other linear but non-gravitational accelerations of the head, may help explain a limited role for otoliths in orthostatic cardiovascular control. Recumbent LBNP simulates many effects of orthostasis and upright activity (Boda et al., 2000; Wolthuis et al., 1974). Ray (2000) recently reported that sympathoexcitation from head-down neck flexion during prone LBNP added to the sympathoexcitation elicited by the LBNP itself. As in earlier related studies, they employed prone, chin-supported head-up neck extension as the control position. The head is upright in this position, as it would be while sitting or standing. Head-down neck flexion in the prone position (hanging the head over the edge of a table) places the head upside-down, which produces otolith stimulation in the opposite direction of that produced by upright posture. Therefore, Ray (2000) results and earlier head-down neck flexion literature (Essandoh et al., 1988; Shortt and Ray, 1997) suggest that upright otolith orientation decreases MSNA and vasoconstriction relative to head-down conditions. However, speculation is complicated concerning implications of the head-down neck flexion literature, and of the head-down neck flexion plus LBNP results (Ray, 2000), for otolith effects in upright posture. Each of our tilt procedures involved different neck flexion/extension conditions. However, we are confident that neck proprioception did not influence our findings. First, other workers have previously examined such possibilities in similar experiments and found either minimal (Normand et al., 1997) or no (Essandoh et al., 1988; Ray

and Hume, 1998) influence of neck positioning on hemodynamics. Second, we found no such effects in the present studies of neck positioning (Tables 1 – 3). We are also confident that visual cues did not influence our results because subjects were blindfolded during all data collection. 4.1. Limitations Our results do not eliminate the possibility that nonotolith mechanisms (e.g. baroreflexes) compensated for disallowance of otolith contributions by the otolith-constant tilt condition (Fig. 1C). Indeed, observations within the first 20 s of tilt suggest this possibility: a subnormal acute blood pressure response to tilt due to lack of full upright otolith stimulation could hypothetically lead to baroreflexive compensation. In addition to baroreflexes, the ‘‘somatic gravity reception’’ described by Mittelstaedt (1992) could also contribute. We did not denervate or block vestibular input. In our ‘‘otolith-constant’’ experimental condition, vestibular otolith input must have continued, even though this experimental condition prevented any change in otolith orientation. Therefore, a constant otolith input remains superimposed on the results from that experimental condition. We stopped at 3 min of 60j body tilt, so it remains possible that vestibular otolith involvement in cardiovascular control becomes measurable at some point beyond 3 min. Our results do not exclude possible involvement of semicircular canal vestibular inputs to orthostatic cardiovascular control. Also, because we studied only healthy subjects, it remains possible that otolith stimulation may influence orthostatic cardiovascular control in certain patient groups and pathologic conditions (Bouvette et al., 1996). For example, orthostatic intolerance in astronauts after space flight may result in part from effects of vestibular re-acclimation to gravity (Mikhailov et al., 1990; Watenpaugh and Hargens, 1996). In at least some cases, post-flight orthostatic intolerance occurs without hypotension (Buckey et al., 1996). Some symptoms of post-flight intolerance resemble those associated with vestibular pathology (e.g. nausea, vomiting, and dizziness). Therefore, vestibular input may influence post-flight orthostatic tolerance by altering autonomic responses to gravitational stress, and by directly eliciting significant syncopallike symptoms. 4.2. Conclusion We compared human acute cardiovascular responses to gravity (whole body tilting) with and without concomitant upright reorientation of the head and otolith organs. Otolith inputs may contribute to early transient blood pressure adjustments to orthostasis. However, lack of significant main effects of tilt condition and time  tilt interactions for the studied variables suggests that potential otolith

D.E. Watenpaugh et al. / Autonomic Neuroscience: Basic and Clinical 100 (2002) 77–83

effects may be relatively subtle and ephemeral, or that other mechanisms compensate for a lack of change in otolith input with orthostasis.

Acknowledgements We thank the subjects for their participation, and Debbie Castillo for administrative assistance. This work was supported in part by the United States National Aeronautics and Space Administration grant NAG 5-3744, and by a National Medical Fellowship from Bristol Myers Squibb, to Adriena Cothron.

References Biaggioni, I., Costa, F., Kaufmann, H., 1998. Vestibular influences on autonomic cardiovascular control in humans. J. Vestib. Res. 8, 35 – 41. Boda, W.L., Watenpaugh, D.E., Ballard, R.E., Hargens, A.R., 2000. Supine lower body negative pressure exercise simulates metabolic and kinetic features of upright exercise [in process citation]. J. Appl. Physiol. 89, 649 – 654. Bouvette, C.M., Mcphee, B.R., Opfer-Gehrking, T.L., Low, P.A., 1996. Role of physical countermaneuvers in the management of orthostatic hypotension: efficacy and biofeedback augmentation. Mayo Clin. Proc. 71, 847 – 853. Buckey Jr., J.C., Lane, L.D., Levine, B.D., Watenpaugh, D.E., Wright, S.J., Moore, W.E., Gaffney, F.A., Blomqvist, C.G., 1996. Orthostatic intolerance after spaceflight. J. Appl. Physiol. 81, 7 – 18. Doba, N., Reis, D.J., 1972. Changes in regional blood flow and cardiodynamics evoked by electrical stimulation of the fastigial nucleus in the cat and their similarity to orthostatic reflexes. J. Physiol. (Lond.) 227, 729 – 747. Doba, N., Reis, D.J., 1974. Role of the cerebellum and the vestibular apparatus in regulation of orthostatic reflexes in the cat. Circ. Res. 40, 9 – 18. Ebert, T.J., Smith, J.J., Barney, J.A., Merrill, D.C., Smith, G.K., 1986. The use of thoracic impedance for determining thoracic blood volume changes in man. Aviat. Space Environ. Med. 57, 49 – 53. Essandoh, L.K., Duprez, D.A., Shepherd, J.T., 1988. Reflex constriction of human limb resistance vessels to head-down neck flexion. J. Appl. Physiol. 64, 767 – 770.

83

Hume, K.M., Ray, C.A., 1999. Sympathetic responses to head-down rotations in humans. J. Appl. Physiol. 86, 1971 – 1976. Kaufmann, H., Biaggioni, I., Voustianiouk, A., Diedrich, A., Costa, F., Clarke, R., Gizzi, M., Raphan, T., Cohen, B., 2002. Vestibular control of sympathetic activity. An otolith-sympathetic reflex in humans. Exp. Brain Res. 143, 463 – 469. Mikhailov, V.M., Kornilova, L.N., Zhernavkov, A.F., Voskresenskii, A.D., Pometov lu, D., Alekseev, V.N., 1990. Correlation of orthostatic tolerance and vestibular function in man after long space flights. Kosm. Biol. Aviakosm. Med. 24, 49 – 50. Mittelstaedt, H., 1992. Somatic versus vestibular gravity reception in man. Ann. N.Y. Acad. Sci. 656, 124 – 139. Normand, H., Etard, O., Denise, P., 1997. Otolithic and tonic neck receptors control of limb blood flow in humans. J. Appl. Physiol. 82, 1734 – 1738. Ray, C.A., 2000. Interaction of the vestibular system and baroreflexes on sympathetic nerve activity in humans. Am. J. Physiol.: Heart Circ. Physiol. 279, H2399 – H2404. Ray, C.A., Hume, K.M., 1998. Neck afferents and muscle sympathetic activity in humans: implications for the vestibulosympathetic reflex. J. Appl. Physiol. 84, 450 – 453. Ray, C.A., Hume, K.M., Shortt, T.L., 1997. Skin sympathetic outflow during head-down neck flexion in humans. Am. J. Physiol. 273, R1142 – R1146. Rowell, L.B., 1993. Human Cardiovascular Control. Oxford Univ. Press, New York, 500 pp. Shortt, T.L., Ray, C.A., 1997. Sympathetic and vascular responses to headdown neck flexion in humans. Am. J. Physiol. 272, H1780 – H1784. Watenpaugh, D.E., Hargens, A.R., 1996. The cardiovascular system in microgravity. In: Fregly, M.J., Blatteis, C.M. (Eds.), Handbook of Physiology: Environmental Physiology. Oxford Univ. Press, New York, pp. 631 – 674. Watenpaugh, D.E., Muenter, N.K., Wasmund, W.L., Wasmund, S.L., Smith, M.L., 1999. Post-apneic inhalation reverses apnea-induced sympathoexcitation before restoration of blood oxygen levels. Sleep 22, 435 – 440. Wolthuis, R.A., Bergman, S.A., Nicogossian, A.E., 1974. Physiological effects of locally applied reduced pressure in man. Physiol. Rev. 54, 566 – 595. Yates, B.J., 1996. Vestibular influences on cardiovascular control. In: Yates, B.J., Miller, A.D. (Eds.), Vestibular Autonomic Regulation. CRC Press, Boca Raton, pp. 97 – 111. Yates, B.J., Miller, A.D., 1998. Physiological evidence that the vestibular system participates in autonomic and respiratory control. J. Vestib. Res. 8, 17 – 25.