Abstract
Spinal sympathetic neurons are distributed in cord segments from Th1 to L3. High spinal cord injury demonstrates severe orthostatic hypotension, but not lower cord injury. It remains to be clarified as to where is the critical spinal level disturbing neural cardiovascular regulations in response to orthostatic stress. To address this issue, beat-to-beat blood pressure (BP) (measured using a Finapres device) and RR interval (measured electrocardiographically) were recorded at rest and in a 60° head-up position in 26 patients with varying levels of spinal cord injury (C4 to Th12) and in 15 healthy (control) subjects. Sympathetic vascular tone was examined by the Mayer wave power spectrum of systolic blood pressure (SBP) variability. Baroreflex sensitivity was examined by transfer function analysis of SBP and RR interval variabilities. The Mayer wave power spectrum increased in response to postural shift in most patients injured at Th4 or below, whereas this parameter either remained unchanged or decreased in patients with higher-level injury. Baroreflex sensitivity tended to decrease with postural shift in patients injured at Th3 or below, whereas this parameter increased in all patients with higher-level injury. We divided spinal patients into high-level injury (Th3 or above, n = 14) and low-level injury (Th4 or below, n = 12) groups. Systolic blood pressure significantly fell (−10 ± 4 mm Hg, P < .05) with postural shift in high-level injury group but did not change in low-level injury group or in control subjects. The low-level injury group and the control group demonstrated essentially similar autonomic nervous responses to postural shift, ie, a significant increase in Mayer wave power and an insignificant decrease in baroreflex sensitivity. On the contrary, the high-level injury group showed opposite responses, ie, an insignificant decrease in Mayer wave power and a significant increase in baroreflex sensitivity in response to postural shift. We conclude that spinal cord injury at Th3 or above eliminates normal neural cardiovascular responses to mild orthostatic stress in humans. Am J Hypertens 2001;14:141–148 © 2001 American Journal of Hypertension, Ltd.
The sympathetic nervous system lies at the center of neural cardiovascular control and plays a critical role in regulation of blood pressure variability.1 Spinal sympathetic neurons are distributed in cord segments from Th1 to L2 or L3.2 Spinal cord injury causes various degrees of sympathetic dysfunction according to the level of injury, altering blood pressure (BP) variability patterns. We have found that circadian variation of blood pressures is markedly decreased in patients with cervical cord injury, whereas nearly normal circadian BP rhythm has persisted in patients with cord injury below Th6.3
Spinal cord injury could affect short-term BP variability patterns as well. The Mayer wave, representing BP variability having around a 10-sec rhythm, is known to increase during mental or gravitational stresses in humans4–6 and has been proposed as a quantitative marker for sympathetic vasomotor activity receiving central drive.7 However, the reported consequences of complete spinal transection in man are contradictory. Inoue et al have found that resting low-frequency power spectra of systolic blood pressure (SBP) variability, ie, Mayer waves, are lost in tetraplegic patients with cervical cord injury,8 whereas others have reported that similar patients demonstrated a significant level of Mayer waves.9,10 Reasons for these divergent results have not been well explained. Because Mayer waves are very small, even in healthy subjects in the supine position, studies limited to the resting state may be insensitive for demonstrating differences in Mayer wave profiles between spinal patients and normal subjects.
Loss of central regulation of spinal sympathetic neurons could also alter the baroreflex profile. It has been shown that baroreflex sensitivity is reduced by physiologic increase in sympathetic activity such as orthostatism. It has also been shown that reduced baroreflex sensitivity is augmented by the treatment with β-adrenoceptor antagonists in hypertensive humans.11 These data suggest that sympathetic nervous activity physiologically modulates baroreflex sensitivity. Previous studies examined the baroreflex sensitivity of quadriplegic patients at rest.9,10 However, no one has studied dynamic baroreflex response to orthostatic stress in patients with spinal cord injury.
The aim of the present study was to examine the influence of spinal transection on Mayer waves and baroreflex responses to orthostatic stress in humans. Patients with cervical cord injury completely lost central drive of spinal sympathetic neurons, whereas central influences on spinal sympathetic neurons serially increased in cases of lower cord injury. To clarify the critical spinal level for normal neural cardiovascular regulations in humans, we studied spinal patients with varying level lesions ranging from C4 to Th12.
Subjects and methods
Twenty-six patients with spinal cord injury were recruited from among patients admitted to our rehabilitation center between 1995 and 1998, including 18 men and eight women aged 15 to 69 years (mean, 43 years). Inclusion criteria included lack of cardiovascular (coronary artery disease, arrhythmia, or hypertension) or other serious diseases except for spinal cord injury; no use of medications known to affect the cardiovascular system; passage of at least 6 months since injury; a complete spinal cord lesion according to the diagnostic criteria of the American Spinal Injury Association; and no prolonged and serious episodes of autonomic hyperreflexia. Time after injury ranged from 6 months to 10 years. All subjects underwent baseline hematological and chemical laboratory studies, and we obtained chest radiographs and electrocardiograms (ECG). None had diabetes mellitus or renal dysfunction. We also studied 15 healthy volunteers (40 ± 3 years). The study protocol was approved by the ethics committee of Tohoku Rosai Hospital, and all subjects gave their informed consent for participation in the study.
Data recording and analysis
Each subject was positioned supine on a bed with a manually adjustable angle of the upper part of the body. The ECG was monitored using lead II. Blood pressure was monitored in the right middle finger with a digital photoplethysmographic device (Finapres 2300; Ohmeda, Englewood, CO). The procedure for finger BP recording and analysis have been described elsewhere.5,6 After hemodynamic stabilization, analog ECG and BP signals were fed into a signal processor (7T-18; NEC San-ei, Tokyo, Japan) with an R wave detector accurate to within 1 msec. Systolic (SBP) and diastolic blood pressures (DBP) were measured at each R wave. The RR interval was calculated during this period and then digitized to be stored on a floppy disk. Data were collected for 10 min after a 15-min rest in the supine position. Subjects were then raised by 20° every 2 min to a 60° incline. After a 2-min wait for hemodynamic equilibration, data were recorded for another 5 min with the right arm supported at heart level. In all patients with spinal cord injury, the bladder was continuously drained through a urethral catheter during the experiment to minimize the episodes of autonomic hyperreflexia.
On the day of the experiment, blood samples were drawn at 7:30 AM from all subjects who had fasted overnight, while they rested supine. Plasma norepinephrine concentration was measured by high-performance liquid chromatography (TOSOH, Tokyo, Japan).3 Off-line analysis was performed later on a personal computer (PC 9801-RX; NEC, Tokyo, Japan). Computer-generated displays of the RR interval and the systolic and diastolic pressures derived from data obtained over 5 to 10 min in the supine and head-up positions were inspected visually. A 256-sec segment without slow trends, movement artifacts, or premature ventricular contractions was selected for the final analysis. First, descriptive statistics (mean and standard deviation) were computed. Then we examined power spectra of cardiovascular variabilities and the gain of the transfer function using CARSPAN, a fully validated spectral analysis program for cardiovascular time series12 (Fig. 1). Computational techniques and clinical application of this program have been described previously.12,13 In brief, spectral variables of nonequidistant cardiovascular time series (SBP and DBP and RR interval) were calculated by means of discrete Fourier transform. We calculated the Mayer wave (0.07 to 0.14 Hz) and respiratory (0.15 to 0.40 Hz) power spectra for both BP and RR interval variabilities. It remains debatable whether significant levels of Mayer wave powers exist for BP variability in patients with high spinal cord injury.8–10 As a large proportion of the spontaneous changes in BP is due to nonperiodic noise,14 and Mayer waves would be very small even in healthy subjects in the supine position, it would be critically important to define which level of power may still be considered significant. To examine the statistical significance, percent Mayer wave power was obtained by dividing BP power ranges from 0.07 to 0.14 Hz by the total power (0.02 to 0.50 Hz) and multiplying the result by 100. Only components with power > 5% of the total power were considered to be significant.8–10
Systolic blood pressure (SBP) and RR interval power spectra (upper and middle panels), and gain of the transfer function (bottom panels) between systolic blood pressure and RR interval variabilities in a spinal patient with high-level injury (C6, 16-year-old boy) and in a patient with low-level injury (Th10, 19-year-old boy). Dotted and closed lines indicate data in the supine and head-up positions, respectively.
01236-X/2/m_ajh.141.f1.jpeg?Expires=1500713143&Signature=XAMS1t8NXoS0bLr-ULQyp2Uo8BaROusz7aq15LCmIIcMsb7UtcWQLN5PtEneQpZvRIVdPD8xSZ~7K4VmwCtZVtoVTaxTjzesNdOOXEnHIdU-Fpx2S5HNbDrDeoc7WqrFUlv0YwGPHmfn1Uv8PVxaxef2yL8b~dydexoNEhzyylyEfUVgdu7R81nYfYjdR3gx2mGMqdV-b6qj51bgu0N8Qp0qaUYo0m3IJv2cFXtaHbLQBaFu1tLTr5mk1FlnAHA0MNTdEvNeb59YxUeA4AEBAaK-TY3ZrWpV4M556hJo6CKr1CUfEC8UaeRBlfGYAnc-rzuH6-xAioRlF59ydreSvw__&Key-Pair-Id=APKAIUCZBIA4LVPAVW3Q)
To examine the amount of linear coupling between SBP and RR interval variabilities, we calculated the squared coherence between the two signals by dividing the cross-spectral densities by the product of the individual power spectral densities. Sensitivity of the baroreceptor–heart rate reflex was examined by a gain of the transfer function between SBP and RR interval variabilities. In previous studies, a gain in the Mayer wave frequency and that in respiratory frequency had been calculated separately.15,16 However, both gains correlated similarly with baroreflex sensitivity estimated using the vasoactive or sequence method.15,17 We further showed that baroreflex-mediated linkage with BP and RR interval sometimes shifts to a low-frequency (0.02 to 0.06 Hz) band.17 We therefore calculated the mean gain of the transfer function ranging from 0.02 to 0.40 Hz for those frequency points with a coherence > 0.5 as an indicator of baroreflex sensitivity.
Statistical analysis
All data are expressed as mean ± SE. Differences in mean were examined by paired or unpaired t tests. All statistical analyses were performed with a commercially available statistical package (SYSTAT version 5.2, Evanston, IL). A level of P < .05 was accepted as statistically significant.
Results
Fig. 2 and 3 show Mayer wave powers of SBP variability and baroreflex sensitivity at rest and in a 60° head-up position in individual patients with spinal cord injury. The Mayer wave power spectrum increased in response to postural shift in most patients injured at Th4 or below, whereas this parameter remained unchanged or decreased in patients with higher level injury (Fig. 2). Baroreflex sensitivity tended to decrease with postural shift in patients injured at Th3 or below, whereas this parameter increased in most patients with higher level injury (Fig. 3).
Baroreflex sensitivity in the supine (open circles) and head-up position (closed circles) in individual patients with spinal cord injury.
01236-X/2/m_ajh.141.f3.jpeg?Expires=1500713143&Signature=BZb6WDNwzKxPsKzzlyKTa-8wh2JMIH7FHuosbQrHHkmt6U0LIBNDPy8ZhFy4L5ZnqDZ~ZiN2L7Qsg9sqV~mijVfdKl765-hPDBY6XZ2MZxg4cz3TICHCnTLbpdyvN550hFPjqFzvENcm4259cgfsxoF7sfZHtap2hBQbPckw6TtrOvfzeacZq6-KdHTc4VSGm~wLoyIon3KaLy-ZgzQ~66kwJm8Q~sV7ebhAFKSMd1a5DZ1sHIJyEh4bkgzKYw7uzDDRi6CEdoJEmuKREuL4DNwBcZJLsEbkBcriAIPgWp8Zgf2yxNJL~-CDCm6dihksL0xK~YTaGlTV-HZXL6nnhA__&Key-Pair-Id=APKAIUCZBIA4LVPAVW3Q)
Mayer wave power spectra of systolic blood pressure (SBP) variability in the supine position (open circles) and head-up position (closed circles) in individual patients with spinal cord injury.
01236-X/2/m_ajh.141.f2.jpeg?Expires=1500713143&Signature=P5060tyHoAWljpxltCl29aIWvhrhGcEbUlu4MyXJ~Sjm4SRq91Wgvgye5mrKr2GMTmNXp~AXqujDfUlgMCAEQ3G6pM1bX6vXFNYjjrHIFzDvHOem7HKMUwlltBPehlfn3I1VZvGF-kKDqoDYMNbM3GMAmNWA67L0LpyVY12vbqw9Eba7gJ5oIQmFDCoqpDeYRa7no~QC-PuQOqy8b6Ij7QWL4DpQSBHVXadE2yMxR3RfBzToLlUIlf9yKnRbS6ZASR4NKgcJu6Znnujtgd8X54gy1GkYx-YZloCa~VQR9GgO3XOdNMPccG62f2xi5h3LNABKxcurn90cQQWHj8kOnA__&Key-Pair-Id=APKAIUCZBIA4LVPAVW3Q)
We divided spinal patients into high-level injury (Th3 or above, n = 14) and low-level injury (Th4 or below, n = 12) groups. No significant differences were noted with respect to age or gender distribution between spinal patient groups and control subjects (Table 1). Both SBP and DBP were lower and the RR interval was longer in patients with high-level injury than corresponding values for patients with low-level injury or for control subjects (Table 2). This was true in both supine and 60° head-up positions. Blood pressure did not change significantly with postural shift in patients with low-level injury or in control subjects, whereas SBP fell with upward posture (P < .05) in patients with high-level injury. Neither SBP nor DBP variability differed among the three groups in the supine position, but in the head-up position both parameters were significantly lower in patients with high-level injury than in the other two groups.
Clinical and demographic profiles of the two spinal injury groups and control subjects
| Group | High-Level Injury (n = 14) | Low-Level Injury (n = 12) | Controls (n = 15) |
|---|---|---|---|
| Age (years) | 48 ± 4 | 35 ± 5 | 40 ± 3 |
| Sex (M/F) | 10/4 | 8/4 | 9/6 |
| Injury level | C4–Th3 | Th4–Th12 | — |
| Months after injury | 13 ± 7 | 13 ± 8 | — |
| Group | High-Level Injury (n = 14) | Low-Level Injury (n = 12) | Controls (n = 15) |
|---|---|---|---|
| Age (years) | 48 ± 4 | 35 ± 5 | 40 ± 3 |
| Sex (M/F) | 10/4 | 8/4 | 9/6 |
| Injury level | C4–Th3 | Th4–Th12 | — |
| Months after injury | 13 ± 7 | 13 ± 8 | — |
Cardiovascular parameters in supine and 60° head-up positions
| Group | High-Level Injury | Low-Level Injury | Control | |||
|---|---|---|---|---|---|---|
| Position | Supine | Head- Up | Supine | Head-Up | Supine | Head-Up |
| SBP | ||||||
| Mean (mm Hg) | 85 ± 3*† | 75 ± 4‡†§ | 113 ± 5 | 113 ± 5 | 114 ± 5 | 117 ± 6 |
| SD (mm Hg) | 3.9 ± 0.4 | 3.2 ± 0.4‡† | 4.9 ± 0.5 | 5.9 ± 0.9 | 5.0 ± 0.5 | 6.0 ± 0.8 |
| Mayer power (mm Hg2) | 0.9 ± 0.3‡† | 0.6 ± 0.2*∥ | 1.9 ± 0.7 | 4.0 ± 0.9§ | 2.5 ± 0.5 | 4.2 ± 1.1§ |
| Respiratory power (mm Hg2) | 2.1 ± 0.7 | 0.9 ± 0.2 | 1.8 ± 0.7 | 2.2 ± 0.5 | 1.6 ± 0.6 | 2.2 ± 0.5 |
| DBP | ||||||
| Mean (mm Hg) | 44 ± 2‡† | 38 ± 3*∥ | 62 ± 5 | 61 ± 3 | 59 ± 4 | 64 ± 5 |
| SD (mm Hg) | 2.2 ± 0.3 | 1.8 ± 0.1*∥ | 3.1 ± 0.3 | 3.8 ± 0.4 | 3.1 ± 0.3 | 3.6 ± 0.5 |
| Mayer power (mm Hg2) | 0.3 ± 0.1*† | 0.2 ± 0.05‡∥ | 1.1 ± 0.4 | 2.0 ± 0.4§ | 1.1 ± 0.2 | 1.8 ± 0.3§ |
| Respiratory power (mm Hg2) | 0.7 ± 0.3 | 0.5 ± 0.1 | 1.0 ± 0.3 | 1.1 ± 0.3 | 1.4 ± 0.6 | 1.6 ± 0.2 |
| RR interval | ||||||
| Mean (msec) | 1057 ± 44# | 1049 ± 38‡† | 849 ± 49 | 846 ± 49 | 902 ± 45 | 880 ± 47 |
| SD (msec) | 38 ± 5 | 39 ± 5 | 42 ± 7 | 44 ± 11 | 35 ± 5 | 36 ± 4 |
| Mayer power (msec2) | 202 ± 82 | 170 ± 75 | 305 ± 82 | 258 ± 87 | 230 ± 73 | 207 ± 65 |
| Respiratory power (msec2) | 558 ± 251 | 517 ± 234 | 739 ± 353 | 218 ± 88 | 690 ± 235 | 346 ± 81 |
| Baroreflex sensitivity | ||||||
| (msec/mm Hg) | 13.6 ± 2.1 | 17.2 ± 3.1§ | 14.8 ± 3.7 | 11.6 ± 3.5 | 13.6 ± 2.2 | 9.7 ± 1.8 |
| Group | High-Level Injury | Low-Level Injury | Control | |||
|---|---|---|---|---|---|---|
| Position | Supine | Head- Up | Supine | Head-Up | Supine | Head-Up |
| SBP | ||||||
| Mean (mm Hg) | 85 ± 3*† | 75 ± 4‡†§ | 113 ± 5 | 113 ± 5 | 114 ± 5 | 117 ± 6 |
| SD (mm Hg) | 3.9 ± 0.4 | 3.2 ± 0.4‡† | 4.9 ± 0.5 | 5.9 ± 0.9 | 5.0 ± 0.5 | 6.0 ± 0.8 |
| Mayer power (mm Hg2) | 0.9 ± 0.3‡† | 0.6 ± 0.2*∥ | 1.9 ± 0.7 | 4.0 ± 0.9§ | 2.5 ± 0.5 | 4.2 ± 1.1§ |
| Respiratory power (mm Hg2) | 2.1 ± 0.7 | 0.9 ± 0.2 | 1.8 ± 0.7 | 2.2 ± 0.5 | 1.6 ± 0.6 | 2.2 ± 0.5 |
| DBP | ||||||
| Mean (mm Hg) | 44 ± 2‡† | 38 ± 3*∥ | 62 ± 5 | 61 ± 3 | 59 ± 4 | 64 ± 5 |
| SD (mm Hg) | 2.2 ± 0.3 | 1.8 ± 0.1*∥ | 3.1 ± 0.3 | 3.8 ± 0.4 | 3.1 ± 0.3 | 3.6 ± 0.5 |
| Mayer power (mm Hg2) | 0.3 ± 0.1*† | 0.2 ± 0.05‡∥ | 1.1 ± 0.4 | 2.0 ± 0.4§ | 1.1 ± 0.2 | 1.8 ± 0.3§ |
| Respiratory power (mm Hg2) | 0.7 ± 0.3 | 0.5 ± 0.1 | 1.0 ± 0.3 | 1.1 ± 0.3 | 1.4 ± 0.6 | 1.6 ± 0.2 |
| RR interval | ||||||
| Mean (msec) | 1057 ± 44# | 1049 ± 38‡† | 849 ± 49 | 846 ± 49 | 902 ± 45 | 880 ± 47 |
| SD (msec) | 38 ± 5 | 39 ± 5 | 42 ± 7 | 44 ± 11 | 35 ± 5 | 36 ± 4 |
| Mayer power (msec2) | 202 ± 82 | 170 ± 75 | 305 ± 82 | 258 ± 87 | 230 ± 73 | 207 ± 65 |
| Respiratory power (msec2) | 558 ± 251 | 517 ± 234 | 739 ± 353 | 218 ± 88 | 690 ± 235 | 346 ± 81 |
| Baroreflex sensitivity | ||||||
| (msec/mm Hg) | 13.6 ± 2.1 | 17.2 ± 3.1§ | 14.8 ± 3.7 | 11.6 ± 3.5 | 13.6 ± 2.2 | 9.7 ± 1.8 |
SBP = systolic blood pressure; DBP = diastolic blood pressure; SD = standard deviation.
P < .05,
P < .01
P < .001 v control,
P < .05,
P < .01,
P < .01 v low-level injury;
P < .05 v supine.
Systolic and diastolic Mayer wave powers of patients with high-level injury at rest were lower than those of patients with low-level injury or those of control subjects (Fig. 1, Table 2). Both parameters increased with postural changes in patients with low-level injury and in control subjects, but tended to decrease in patients with high-level injury (Fig. 2, Table 2). A significant level of Mayer waves was observed for both SBP and DBP variabilities in spite of position in all but one patient with high-level injury. All patients with low-level injury as well as control subjects demonstrated a significant level of Mayer waves for both SBP and DBP in either position. Respiratory power of SBP tended to increase with postural change in patients with low-level injury and in control subjects, but tended to decrease in patients with high-level injury.
Resting baroreflex sensitivity as assessed by the gain of the transfer function between SBP and RR interval variabilities did not differ among the three groups (Table 2). Five of the 14 patients with high-level injury demonstrated a low coherence (<0.5) with SBP and RR interval variability in the low- and mid-frequency bands and, thus, baroreflex sensitivity was calculated by high-frequency gain only. The remaining nine patients with high-level injury, as well as all patients with low-level injury and control subjects, showed a high coherence either in the low- or mid-frequency band; thus, baroreflex sensitivity was examined by a broad frequency gain including high frequency. Baroreflex sensitivity tended to decrease in the head-up position in patients with low-level injury, but increased significantly in patients with high-level injury (P < .05, Table 2). Consequently, baroreflex sensitivity in the head-up position was higher in patients with high-level injury than in the control subjects (P = .05). In the head-up position, four patients with high-level injury demonstrated a low coherence in the low- and mid-frequency band, whereas all subjects in the other two groups showed a high coherence in those frequency bands.
Plasma norepinephrine concentration was lower in patients with high-level injury (81 ± 11 pg/mL) than in patients with low-level injury (270 ± 60 pg/mL, P < .05) or in control subjects (292 ± 47 pg/mL, P < .01). No parameters differed significantly between patients with low-level injury and control subjects.
Discussion
Mayer wave profiles of patients with high-level injury (C4 to Th3) were very different from those of patients with low-level injury (Th4 to Th12). For both SBP and DBP, the resting Mayer wave powers of patients with high-level injury were much lower than those of patients with low-level injury. Furthermore, the Mayer wave powers in patients with high-level injury tended to decrease in the 60° head-up position from values in the supine position, whereas they increased significantly in patients with low-level injury. The Mayer wave profiles of patients with low-level injury were similar to those of control subjects. Thus, connections between the brain and upper spinal cord play a critical role in the maintenance of normal Mayer wave profiles.
Patients with high-level injury had lower plasma norepinephrine concentration and greater BP fall during mild orthostatic stress compared with findings in control subjects. Both plasma norepinephrine concentration and BP response to the head-up position were similar between patients with low-level injury and control subjects. These data suggest an impaired sympathetic BP regulation in patients with high-level injury. Sympathetic neurons in the spinal cord occupy segments from Th1 to L2 or L3. The present data shows that supraspinal control of the upper one-quarter or more of spinal sympathetic neurons can preserve sympathetic BP regulation.
The significant level of Mayer BP waves in patients with high-level injury, however, suggests that the appearance of Mayer BP waves does not solely depend on higher brain control of the spinal cord. We failed to clarify the mechanisms responsible for residual components of Mayer waves in patients with high spinal cord injury. These may include 1) rhythmic firing of spinal sympathetic motoneurons,10,18,19 2) coordinated vascular smooth muscle contractions,20 or 3) rhythmic firing of vagal cardiac motoneurons.9
The increase in Mayer wave power seen in the 60° head-up position may chiefly reflect a deactivation of cardiopulmonary receptors rather than baroreceptors, because the heart rate did not increase during this maneuver. We did not test responses to standard head-up tilting because some patients with high-level injury complain of dizziness due to hypotension, or of skeletal muscle spasms due to muscle ischemia, which could have interfered with the continuous recording of the stationary data needed to perform spectral analysis. Furthermore, determining the BP response to the head-up position in patients with high-level injury was important because this position is common for meals, reading, and socializing. Maintenance of the 60° head-up position is essential for many daily activities.
Patients with high-level injury demonstrated a significant increase in baroreflex sensitivity with the 60° head-up position compared with the resting condition, in contrast to the other two groups. The mechanisms that cause paradoxically increased baroreflex sensitivity remain unclear. The unusual result was not due to using a near-noise input signal for calculating the gain of the transfer function between SBP and RR interval variabilities, because a significant level of Mayer waves were detected in patients with high-level injury. The increase in baroreflex sensitivity and the lack of tachycardia despite a significant BP fall suggest a baroreceptor resetting, and are very similar to cardiovascular and baroreceptor responses achieved by β-blocker treatment.17,21 It has been shown that the increase in sympathetic nervous activity associated with orthostatic stress exerts an antagonistic effect on the sensitivity of baroreceptor–heart rate reflex, and that treatment with a β-adrenoceptor antagonist could increase the reduced baroreflex sensitivity of standing subjects.11 In our patients with high-level injury, the Mayer wave power of SBP tended to decrease with postural shift in contrast with that in the other two groups. Thus, the paradoxical increase in baroreflex sensitivity seen in patients with high-level injury may be explained by reflex sympathoinhibition. In an absence of supraspinal influences, spinal sympathetic afferent activity is likely to play a critical role in determining a tonic efferent sympathetic discharge.10 We speculate that the reductions in venous return and cardiac dimensions that accompany a head-up maneuver are likely to decrease the discharge of the thoracic sympathetic afferent fibers and thereby the activity of the sympathetic outflow.10
It remains debatable whether baroreflex sensitivity should be evaluated by the gain of the transfer function between SBP and RR interval variabilities around Mayer wave frequency (∼0.1 Hz) only, or should include also the gain in the respiratory frequency (∼0.25 Hz) band. Phase spectrum analysis surely supports the use of the gain around Mayer wave frequency.22 In the high-frequency domain, however, it seems to be difficult to discuss the cause–effect relationship from the phase spectrum because the value is nearly zero.17 We need other studies to clarify the causality in the high-frequency domain.
It has been reported that respiratory variability of RR interval is attributed not only to baroreflex but also to other mechanisms such as direct influence of medullary respiratory neurons on cardiovascular neurons, or to reflex response to lung inflation mediated by thoracic stretch receptors.23,24 However, the gain in the respiratory band correlated well with the baroreflex sensitivity assessed with sequence as well as vasoactive methods,15,17 as did the gain in the Mayer wave frequency. Furthermore, Pitzalis et al have shown that increased respiratory variability in SBP in patients with atrial fibrillation was reduced by about 72% after successful cardioversion.25 They concluded that respiratory sinus arrhythmia is not a prerequisite for SBP variability but may play a role to minimize SBP variability, possibly through the baroreflex mechanism. We support the notion that the gain in the respiratory frequency band could be a measure for baroreflex sensitivity in intact, closed-loop condition. As the frequency of coherence > 0.5 is higher in the respiratory band than in the Mayer wave frequency band,17,26 we considered it more practical to examine the baroreflex sensitivity in combination with both frequency gains than with Mayer wave gain only.
Several reports have described baroreflex sensitivity of patients with complete cervical cord injury,9,27,28 and have also found no significant difference in baroreflex sensitivity between tetraplegic spinal patients and control subjects in the supine position. Examining the response of baroreflex sensitivity to the 60° head-up position, which is very common orthostatic stress in daily life, we found a unique profile of patients with high spinal cord injury. Our maneuver involved very mild orthostatic stress. We therefore cannot exclude the possibility that baroreflex sensitivity could be depressed by more severe orthostatic stress in patients with high spinal cord injury.28
In conclusion, spinal patients injured at Th3 or above demonstrated altered Mayer wave and baroreflex profiles compared with those of healthy control subjects. Both profiles did not differ significantly between spinal patients injured at Th4 or below versus control subjects. Thus, connections between the brain and the upper spinal cord are critical for maintaining normal Mayer wave and baroreflex profiles. Our data suggest that supraspinal control of about the uppermost one-quarter or more of the spinal sympathetic neurons can maintain near-normal autonomic cardiovascular regulations during mild orthostatic stress.
