The impetus for the present study was an observation made during a clinical consultation in a patient with VHL who had disabling OH after neurosurgery for brainstem hemangioblastomas [12]. As expected, in this patient the OH was neurogenic, based on both physiological [11] and neurochemical [18] data. Unexpectedly, head-up tilt table testing evoked large BP oscillations at the relatively high frequency of breathing (Traube-Hering waves) [1]. Such oscillations were not present before or after the tilting. It seemed that tilt-induced HF BP oscillations might occur even in the setting of baroreflex-sympathoneural failure. This would be in contrast with relatively low frequency Mayer waves, which complexly reflect baroreflex-mediated modulation of sympathetic cardiovascular outflows [3, 5, 24–27]. The purpose of the present study therefore was to explore mechanisms of tilt-evoked, breathing-driven BP oscillations, by conducting power spectral analyses of systolic BP variability in patient groups with nOH (PAF, PD, MSA-P) and a comparison cohort without OH (HVs, PD No OH).
The main new findings were that (1) across both the nOH and No OH cohorts HF power increased during tilting, (2) the cohorts did not differ in the magnitude of HF power during Tilt, and (3) the cohorts had a similar increase in mean HF power from Pre-Tilt to Tilt. From these results we infer that tilt-evoked BP oscillations at the periodicity of breathing occur independently of baroreflex-sympathoneural modulation.
If Traube-Hering waves evoked by head-up tilting were independent of baroreflex-sympathoneural modulation, what would be the mechanism of the increase in HF power? One might consider that the HF BP oscillations involve mechanical changes in thoracic pressure as individuals breathe in and out and that these changes are augmented by decreased venous return to the heart. Compared to the amplitude of BP oscillations, changes in intrathoracic pressure are minimal during inspiration and expiration, meaning that the changes in respiration are not the sole determinant of Traube-Hering waves [4].
Slow, deep, metronomic breathing increases the magnitude of respiratory sinus arrhythmia (a measure of baroreflex-cardiovagal function) and elicits Traube-Hering waves in healthy individuals [1]. Although respiratory sinus arrhythmia participates in the generation of Traube-Hering waves, additional mechanisms other than parasympathetic input to the heart modulate BP oscillation on the timescale of respiration. Thus, lung transplant patients lack respiratory sinus arrhythmia but have respiratory modulation of BP [35]. It has been proposed that breathing-driven increased BP variability occurs by two independent mechanisms—respiratory sinus arrhythmia and respiratory modulation of pulse pressure [1]. Until the present study, however, whether head-up tilting increases Traube-Hering waves in a manner independent of baroreflex-sympathoneural outflow was not explored.
Baroreflexes contribute importantly to BP oscillations. Indeed, arterial baroreflex failure is always associated with decreased ability to buffer BP changes evoked by virtually any internal or external stimulus. For example, patients with afferent baroreflex dysfunction as a late consequence of neck irradiation have highly variable BP during 24-hour ambulatory monitoring [34], and ablation of the nucleus of the solitary tract, the site of initial synapses for baroreflexes, evokes chronic, labile hypertension in rats [30]. In the present study, the mean baroslope, a measure of baroreflex-cardiovagal function [11], was lower in the nOH than the No OH cohort; however, HF power of BP variability during Tilt was unrelated to the baroslope across individuals in both cohorts. As for HF power of BP in the frequency domain, BP variability in the time domain (measured by SD BP) increased during Tilt in both the nOH and No OH cohorts. It was evident that some patients had slower breathing, and thus part of their breathing-driven oscillations fell into the LF range. Effects of orthostatic changes on respiratory frequency could complicate the results based on power calculations within pre-defined frequency bins.
Another possible complicating factor is the chemoreflex [33]. It might be informative to correlate exhaled CO2 by capnography, baroreflex-cardiovagal gain by the sequence method [31], and HF power of BP variability during orthostasis.
LF power of BP variability (corresponding to Mayer waves) was lower in the nOH than No OH cohort, as one would expect given that all the patients with nOH had baroreflex-sympathoneural dysfunction [7]. It has been proposed that they are generated by brainstem and spinal cord interneuronal microcircuit oscillators that likely are modulated by baroreflexes [8, 9], although the exact mechanism in humans remains poorly understood.
A further test of the hypothesis that tilt-evoked, breathing-driven BP oscillations can occur independently of baroreflex-sympathoneural modulation would be to evaluate individuals with chronic high spinal cord injury (SCI). Such patients typically have nOH due to disruption of neurotransmission in the spinal cord. Baroreflex-sympathoneural failure in SCI would be expected to occur without baroreflex-cardiovagal failure, since the vagus nerve exits the central nervous system from above the level of the SCI. A study is under way on HF BP oscillations during head-up tilt in SCI. Preliminarily, individuals with chronic SCI can have tilt-evoked Traube-Hering waves that resemble those in the nOH cohort in the present study.
Limitations
In all the patient groups and in the HVs there was substantial individual variability in both LF and HF BP oscillations, despite application of a standardized autonomic function testing schema in the same dedicated patient testing room. Carrying out statistical tests necessitated log transformation of the data. The duration of Tilt was relatively brief, and we did not take into account potential tilt-evoked shifts in the frequency of respiration. There may also have been large individual differences in the effects of orthostasis on venous return to the heart and consequently on cardiac stroke volume and pulse pressure. Although the automated finger cuff systems we used for tracking BP continuously came with software applications that reported estimated values for beat-to-beat stroke volume, to our knowledge the algorithms have not been validated in nOH.