Blood Pressure Variability and Ocular Vestibular-Evoked Myogenic Potentials Are Independently Associated With Orthostatic Hypotension

Abstract

Background and Purpose

We delineated the association between otolithic dysfunction and blood pressure (BP) variability.

Methods

We prospectively recruited 145 consecutive patients (age=71 [59–79] years, median [interquartile range]; 76 females) with orthostatic intolerance between December 2021 and December 2023 at a tertiary hospital in South Korea. Each patient underwent evaluations of cervical and ocular vestibular-evoked myogenic potentials (oVEMPs), 24-h noninvasive ambulatory BP monitoring (ABPM), and a head-up tilt-table test using the Finometer device. As measures of BP variability, the standard deviations (SDs) of the systolic BP (SBP_SD) and the diastolic BP were calculated based on serial ABPM recordings. Patients were divided into those with orthostatic hypotension (OH, n=68) and those with a normal head-up tilt-table test despite orthostatic intolerance (NOI, n=77) groups.

Results

A multivariable logistic regression analysis showed that OH was associated with bilateral oVEMP abnormalities (p=0.021), SBP_SD (p=0.012), and female sex (p=0.004). SBP_SD was higher in patients with OH than in those with NOI (p<0.001), and was not correlated with n1–p1 amplitude (p=0.491) or normalized p13–n23 amplitude (p=0.193) in patients with OH. The sensitivity and specificity for differentiating OH from NOI were 72.1% and 67.5%, respectively, at a cutoff value of 12.7 mm Hg for SBP_SD, with an area under the receiver operating characteristic curve of 0.73.

Conclusions

Bilaterally deficient oVEMP responses may be associated with OH regardless of 24-h BP variability, reflecting the integrity of the otolith-autonomic reflex during orthostasis. Alternatively, 24-h BP variability is predominantly regulated by the baroreflex, which also participates in securing orthostatic tolerance complementary to the vestibulo-autonomic reflex.

Graphical Abstract

INTRODUCTION

Orthostatic intolerance can occur upon standing because of cerebral hypoperfusion or sympathetic activation, and is relieved by recumbency.¹,² This phenomenon can be ascribed to neurogenic causes affecting the autonomic system, such as peripheral neuropathy, spinal cord injury, or multiple system atrophy.³ Additionally, inadequate release of norepinephrine from sympathetic vasomotor neurons can also lead to vasoconstrictor failure. Alternatively, orthostatic intolerance can be caused by nonneurogenic causes such as hypovolemia, excessive venous pooling, and fever.⁴,⁵

Orthostatic tolerance is ensured by a normal plasma volume, adequate venomotor tone, and intact autonomic reflex.⁵,⁶ The autonomic pathway works against the pull of gravity through two major neural circuits: the baroreflex and the vestibulo-autonomic reflex.⁷,⁸ The vestibulo-autonomic pathway works through a rapid (within 100–200 ms) feedforward mechanism,⁸,⁹,¹⁰ while the slow baroreflex (1–2 s) constantly works to control the cardiovascular tone. These reflexes act in a complementary manner to maintain the cardiovascular tone in response to postural changes.⁷

The blood pressure (BP) constantly fluctuates because of complex interactions between the extrinsic environment and intrinsic autonomic and cardiovascular regulatory mechanisms. Although current antihypertensive therapies largely focus on mean BP values, there is accumulating evidence that BP variability also exerts important effects on cardiovascular and cerebrovascular outcomes.¹¹,¹² BP variations can be both short-term, usually referring to beat-to-beat, minute-to-minute, hour-to-hour, or day-to-night changes, and prolonged (long-term), over days, weeks, months, seasons, or even years. The short-term BP variations mainly reflect the combined influences of central and autonomic reflex modulation (increased central sympathetic drive and reduced baroreflex and vestibulo-autonomic reflex), elastic properties of arteries, and humoral and emotional factors.¹¹,¹³

The association between BP variability and the baroreflex has mainly been investigated in animal and clinical studies,¹⁴,¹⁵,¹⁶,¹⁷ and it is considered that a deficient baroreflex results in increased BP variability.¹⁸ The age-related development of orthostatic hypotension (OH) has also been explained in this context, with the reduction in arterial elasticity associated with aging consequently reducing the carotid baroreceptor sensitivity.¹⁹ Aging is also associated with the occurrence of autonomic dysfunction, which results in large BP variability, thereby giving rise to OH in older individuals.¹⁴

The interactions between the vestibulo-autonomic reflex and BP variability remain unclear. Given the complementarity of the baroreflex and the vestibulo-autonomic reflex, deficient vestibulo-autonomic interactions may also be associated with BP variability. We therefore investigated how the ocular vestibular evoked myogenic potentials (oVEMPs) and cervical vestibular evoked myogenic potentials (cVEMPs) are associated with the within-day BP variability in patients with orthostatic intolerance.

METHODS

Patients

We prospectively recruited 225 consecutive patients with orthostatic dizziness between December 2021 and December 2023 at the Korea University Medical Center. Each patient underwent a detailed examination using Frenzel goggles. Patients with benign paroxysmal positional vertigo or central positional nystagmus were excluded during the recruitment process.

We subsequently excluded patients without adequate measurements for the vestibular evoked myogenic potential (VEMP, n=24), ambulatory BP monitoring (ABPM, n=12), or head-up tilt-table test (HUT, n=10); those whose BP was measured using only a sphygmomanometer cuff (n=4); and those diagnosed with postural orthostatic tachycardia syndrome (n=27) or vasovagal syncope (n=3) according to the consensus criteria proposed in 2011.² The 145 finally enrolled patients were divided into OH (n=68) and NOI (normal HUT results despite orthostatic intolerance, n=77) groups according to the HUT results obtained using the Finometer device.

HUT using the Finometer

The HUT was conducted in a standard electrodiagnostic laboratory environment.²⁰ Any medications that may affect the results of autonomic function tests were discontinued at least 48 h before performing the HUT in all patients.

Patients were kept supine on a standard electrically driven tilt table with a footboard for at least 20 min before it was tilted. The baseline BP was defined as the average BP between 60 s and 30 s before tilting.²¹ Patients were slowly tilted to 70° over a period of approximately 20 s, and then kept in that position for at least 10 min. The test was stopped early if the orthostatic intolerance symptoms were too debilitating. No pharmacological provocation was performed in this study.

Continuous BP changes during tilting were monitored by measuring the beat-to-beat BP using the Finometer device (Finapres Finometer Pro, Finapres Medical Systems B.V., Amsterdam, The Netherlands) placed on the middle finger and a sphygmomanometer cuff placed over the brachial artery. Predefined 5-s time averaging was applied using the Beatscope software.²² We measured the maximal changes in systolic BP (SBP) and diastolic BP (DBP) in 10 min (ΔSBP_{10 min} and ΔDBP_{10 min}) after initiating the tilting.

Patients with OH were further categorized into initial, classic, and delayed types according to the measured decrease in BP.²,²³,²⁴ Classic OH was defined as a sustained decrease in SBP of 20 mm Hg or in DBP of 10 mm Hg within 3 min after tilting the table. Delayed OH was defined as a sustained decrease in BP that occurred >3 min after tilting the table. A decrease in BP of at least 30 mm Hg was required for patients with supine hypertension, since the magnitude of the decrease in orthostatic BP was dependent on the baseline SBP.²³ Finally, initial OH was determined as a transient decrease in SBP of >40 mm Hg or in DBP of >20 mm Hg within 15 s of tilting. The patients did not perform an active standing test.

Vestibular evoked myogenic potentials

The oVEMP and cVEMP responses of patients were obtained as described previously²⁰,²⁵,²⁶,²⁷ using a Nicolet Viking Select unit (Nicolet-Biomedical, Madison, WI, USA). oVEMPs were recorded with the subject sitting while looking at a target >2 m from the eyes and at an angle of >20° upward. An active electrode was placed 1 cm below the center of the lower eyelid, and a reference electrode was attached to the cheek 2 cm below the active electrode. The ground electrode was located on the forehead at Fpz in the international 10–20 system. oVEMPs were elicited by tapping the hairline at AFz using an electric reflex hammer (Tendon hammer, VIASYS Healthcare, Conshohocken, PA, USA). Bilateral responses were simultaneously obtained while the tapping stimuli were applied. The oVEMP responses were obtained at least twice for each ear, and the mean value was calculated. The interaural difference (IAD) of the oVEMP amplitudes was calculated as [(A_right−A_left)/(A_right+A_left)×100%], where A is the n1–p1 amplitude.

cVEMPs were recorded with the subject lying supine on a bed with the head raised at approximately 30° and rotated contralaterally to contract the sternocleidomastoid muscle (SCM). The surface electromyographic activity was measured from an active electrode placed over the belly of the contracted SCM in comparison with the reference electrode placed on the medial clavicle. A ground electrode was attached to the forehead. A short burst of alternating tone (110 dB nHL, 123.5 dB SPL, 500 Hz, rise time=2 ms, plateau=3 ms, and fall time=2 ms) was applied at 2.1 Hz monoaurally via a headphone. The signal was sampled (48 kHz), amplified, and bandpass filtered at 30–1,500 Hz. No rectification or smoothing was performed while recording the cVEMP responses. cVEMP responses to up to 80 stimuli were averaged during data acquisition of each test. Responses were obtained at least twice for each ear, and the mean values were calculated.

While evaluating cVEMPs, the SCM activities were also simultaneously recorded via surface electrodes and digitized at 1 kHz using an analog-to-digital converter (NI PCI-4461, National Instruments, Austin, TX, USA). The LabVIEW program (National Instruments) was used to measure and rectify the peak-to-peak amplitudes of SCM activity and to calculate the mean tonic activation of the SCM during each recording session. The absolute cVEMP amplitudes were then normalized and divided by the mean tonic activation of the SCM. The IAD was calculated to compare the normalized p13–n23 amplitude of the cVEMP on the right side with that on the left side. The peak latencies of p13 and n23 were also calculated.

For the reference ranges, oVEMP and cVEMP responses were obtained from 31 healthy participants (age=55±8 years, mean±standard deviation [SD]; seven males) with no prior history of auditory or vestibular disorders. The oVEMP and cVEMP values were considered to be abnormal when they were absent, delayed, or decreased in amplitude relative to the reference values.²⁵ In the statistical analyses, the mean values for both sides were calculated for n1 and p13 peak latencies and amplitudes.

Ambulatory blood pressure monitoring

All patients underwent a 24-h noninvasive ABPM using a validated oscillometric device (SpaceLabs 90207 Monitor, SpaceLabs Healthcare, Redmond, WA, USA), as described previously.²⁸ BP was measured every 30 min during the daytime (07:00–22:00) and every 1 h during the nighttime (22:00–07:00). SBP >260 mm Hg or <70 mm Hg, DBP >150 mm Hg or <40 mm Hg, and pulse pressure >150 mm Hg or <20 mm Hg were automatically discarded, which resulted in the exclusion of data from three patients. As measures of BP variability, the SDs of the SBP (SBP_SD) and DBP (DBP_SD) were calculated based on serial recordings made over approximately 24 h. Patients with nighttime reductions in SBP and DBP of <10% relative to the mean daytime values were classified as nocturnal nondippers, whereas those with a nighttime reductions of ≥10% were classified as nocturnal dippers.²⁹ To be considered an acceptable recording for inclusion in our study, the 24-h ABPM needed to have produced at least 30 acceptable measurements, which resulted in the exclusion of the data from nine patients.

Statistical analyses

Nominal/independent variables were compared using the χ² or Fisher’s exact test. Continuous/independent variables were compared using Student’s t-test for parametric variables such as the average SBP, average DBP, average heart rate, and ΔSBP_{10 min}, and using Spearman’s correlation or the Mann–Whitney U-test for nonparametric variables such as body weight, the n1 and p13 latencies, n1–p1 amplitude, normalized p13–n23 amplitude, IADs of cVEMP and oVEMP, and ΔDBP_{10 min}. The normality of each variable was assessed using the Shapiro–Wilk test (Supplementary Table 1 in the online-only Data Supplement). In logistic regression analyses, p<0.05 was considered significant in multivariable analyses. Significant variables were selected using the backward variable selection method. Receiver operating characteristic (ROC) analysis was conducted to determine the sensitivity and specificity, the optimal cutoff values of continuous variables that maximized the sensitivity and specificity, and the area under the ROC curve (AUC) for differentiating OH from NOI.

All statistical analyses were conducted using the R software package (version 3.4.0; R Foundation for Statistical Computing, Vienna, Austria; https://www.r-project.org).

Study ethics

This study followed the tenets of the Declaration of Helsinki and was performed according to the guidelines of the Institutional Review Board of Korea University Anam Hospital (2022AN0005). The written informed consent was obtained from all participants.

RESULTS

Clinical characteristics

The detailed clinical characteristics of the 145 patients (age=71 [59–79] years, median [interquartile range]; 76 females) are presented in Table 1. The 68 patients with OH categorized into 54, 13, and 1 with the classic, delayed, and initial types, respectively. The patients had various comorbidities, including hypertension (n=79), dyslipidemia (n=49), diabetes mellitus (n=33), heart failure (n=9), Parkinson’s disease (n=8), prior history of coronary artery occlusive disease (n=27), and transient ischemic attack or cerebrovascular disease (n=14) (Table 1). There were more females in the OH group than in the NOI group (47/68 [69%] vs. 29/77 [38%], respectively, p<0.001). Compared with patients with NOI, patients with OH more frequently exhibited intracranial/extracranial cerebral artery stenosis (8/77 [10%] vs. 17/68 [25%], p=0.020), Parkinson’s disease (1/77 [1%] vs. 7/68 [10%], p=0.026), and benign prostate hyperplasia (5/48 [10%] vs. 10/21 [48%], p=0.001) (Table 1). The presence of bilateral vestibulopathy did not differ between patients with OH and NOI (4/68 [6%] vs. 5/77 [7%], p>0.999).

Table 1. Clinical and demographic characteristics of patients with OH and NOI.

Characteristics	OH (n=68)	NOI (n=77)	p †
Age (yr)	73 [66–78]	67 [54–79]	0.066
Sex, female	47 (69)	29 (38)	<0.001
Body weight (kg)	64 [57–71]	60 [55–70]	0.038
Diabetes mellitus	19 (28)	14 (18)	0.162
Hypertension	37 (54)	42 (55)	0.987
Dyslipidemia	27 (40)	22 (29)	0.157
Bilateral vestibulopathy	4 (6)	5 (7)	>0.999
Intracranial/extracranial cerebral artery stenosis	17 (25)	8 (10)	0.020
Stroke/TIA	8 (12)	6 (8)	0.419
Atrial fibrillation	5 (7)	6 (8)	0.921
Coronary artery occlusive disease	12 (18)	15 (20)	0.777
Congestive heart failure	4 (6)	5 (7)	>0.999
Parkinson’s disease	7 (10)	1 (1)	0.026
Migraine	4 (6)	7 (9)	0.542
Benign prostate hyperplasia*	10 (48)	5 (10)	0.001
Selective serotonin-reuptake inhibitor	5 (8)	5 (7)	0.839
Tricyclic antidepressant	4 (6)	4 (5)	>0.999
Benzodiapezine	5 (7)	9 (12)	0.378

Data are n (%) or median [interquartile range] values.

^*Males only; ^†Variables that did not conform to normality were analyzed using the Mann–Whitney U-test.

NOI, normal head-up tilt-table test despite orthostatic intolerance; OH, orthostatic hypotension; TIA, transient ischemic attack.

oVEMP and cVEMP findings in patients with OH vs. NOI

Abnormal cVEMP responses were seen in 34 of the 145 patients (23%), comprising abnormal unilateral responses in 14 (8 during left ear stimulation) and abnormal bilateral responses in the other 20 patients. The cVEMP abnormalities did not differ significantly between in patients with OH and NOI (p=0.261) (Table 2). We did not detect any difference between patients with OH and NOI in p13 latency (15.5 [14.6–16.1] ms vs. 15.6 [14.6–16.3] ms, p=0.996) or IAD (0% [−4.6% to 7.3%] vs. 0% [−5.4% to 5.7%], p=0.607). However, the normalized p13–n23 amplitude was lower in patients with OH than in those with NOI (1.7 [1.4–2.1] μV vs. 2.1 [1.6–2.9] μV, p=0.019) (Table 2).

Table 2. Findings of ABPM, cVEMP, oVEMP, and HUT in patients with OH and NOI.

		OH (n=68)	NOI (n=77)	p *
Average SBP (mm Hg)		127 [120–138]	119 [113–133]	0.001
Average DBP (mm Hg)		77 [72–82]	72 [67–78]	<0.001
Average HR (beats/min)		70 [65–77]	71 [65–78]	0.923
SBPSD (mm Hg)		14.0 [11.9–17.5]	11.3 [9.4–13.3]	<0.001
DBPSD (mm Hg)		8.6 [7.3–10.7]	8.1 [6.7–9.9]	0.049
HRSD (beats/min)		9.4 [7.3–12.0]	9.4 [7.4–11.2]	0.942
Nocturnal BP dipper		51 (75)	46 (60)	0.051
ΔSBPday-night (%)		-1±9	-4±7	0.064
ΔDBPday-night (%)		-2±9	-6±9	0.013
Average daytime SBP (mm Hg)		128 [120–142]	122 [113–133]	0.005
Average daytime DBP (mm Hg)		76 [73–83]	73 [67–79]	0.004
Daytime SBPSD (mm Hg)		14.3 [11.6–17.9]	10.8 [8.7–13.3]	<0.001
Daytime DBPSD (mm Hg)		8.5 [7.2–10.2]	7.3 [6.0–9.0]	0.013
Average nighttime SBP (mm Hg)		128 [119–143]	115 [107–130]	<0.001
Average nighttime DBP (mm Hg)		77 [68–84]	68 [62–75]	<0.001
Nighttime SBPSD (mm Hg)		10.6 [9.0–12.8]	9.0 [6.8–11.3]	<0.001
Nighttime DBPSD (mm Hg)		8±4	7±3	0.023
cVEMP†				0.261
	Normal	48 (71)	62 (82)
	Unilaterally abnormal	9 (13)	5 (7)
	Bilaterally abnormal	11 (16)	9 (12)
	p13 latency (ms)‡	15.5 [14.6–16.1]	15.6 [14.6–16.3]	0.996
	Normalized p13–n23 amplitude (μV)	1.7 [1.4–2.1]	2.1 [1.6–2.9]	0.019
	IAD (%)	0 [-4.6 to 7.3]	0 [-5.4 to 5.7]	0.607
oVEMP				0.001
	Normal	46 (68)	60 (78)
	Unilaterally abnormal	6 (9)	14 (18)
	Bilaterally abnormal	16 (24)	3 (4)
	n1 latency (ms)‡	7.4 [6.7–7.9]	7.0 [6.5–7.5]	0.018
	n1–p1 amplitude (μV)	7.5 [5.5–11.3]	9.0 [5.8–12.3]	0.188
	IAD (%)	0 [-4.1 to 3.0]	0 [-4.0 to 4.9]	0.868
ΔSBP10 min (mm Hg)		-26.9 [-35.3 to -19.3]	-5.7 [-9.4 to -1.6]	<0.001
ΔDBP10 min (mm Hg)		-3.5 [-8.3 to 0.1]	-0.1 [-1.8 to 1.4]	0.002

Data are n (%), mean±standard deviation, or median [interquartile range] values.

^*Variables that did not confirm to normality were analyzed using the Mann–Whitney U-test; ^†Values of patients with absent responses in either ear were excluded from the analysis; ^‡oVEMP and cVEMP values were considered to be abnormal when they were absent, delayed, and/or decreased in amplitude relative to the reference values.

ABPM, 24-h ambulatory BP monitoring; BP, blood pressure; cVEMP, cervical vestibular-evoked myogenic potential; DBP, diastolic BP; DBP_SD, SD of DBP; ΔDBP_{10 min}, maximal change in DBP at 10 min; HR, heart rate; HR_SD, SD of HR; HUT, head-up tilt-table test; IAD, interaural difference; NOI, normal HUT despite orthostatic intolerance; OH, orthostatic hypotension; oVEMP, ocular vestibular-evoked myogenic potential; SBP, systolic BP; SBP_SD, SD of SBP; ΔSBP_{10 min}, maximal change in SBP at 10 min; SD, standard deviation.

oVEMP responses were abnormal in 39 of the 145 patients (27%), comprising abnormal unilateral responses in 20 (13 during left ear stimulation) and abnormal bilateral responses in the other 19 patients. oVEMP responses were more frequently abnormal in patients with OH than in those with NOI (p=0.001) (Table 2). Additionally, patients with OH showed delayed n1 latency than those with NOI (7.4 [6.7–7.9] ms vs. 7.0 [6.5–7.5] ms, p=0.018). The n1–p1 amplitude (7.5 [5.5–11.3] μV vs. 9.0 [5.8–12.3] μV, p=0.188) and IAD (0% [−4.1% to 3.0%] vs. 0% [−4.0% to 4.9%], p=0.868) did not differ between the two groups.

Findings of ABPM in patients with OH vs. NOI

Compared with patients with NOI, patients with OH had higher average SBP, average DBP, SBP_SD, ΔDBP_day-night, average daytime/nighttime SBP and DBP, daytime SBP_SD, and nighttime SBP_SD and DBP_SD (Table 2).

Correlations of oVEMP and cVEMP responses with ABPM parameters

In patients with OH, SBP_SD was not correlated with the n1–p1 amplitude (r=−0.085, p=0.491) or the normalized p13–n23 amplitude (r=−0.160, p=0.193) (Fig. 1). In patients with NOI, SBP_SD had a weak negative correlation with the normalized p13–n23 amplitude (r=−0.235, p=0.041), but it was not correlated with the n1–p1 amplitude (r=−0.041, p=0.726) (Fig. 1).

Fig. 1. Correlations of the standard deviation of systolic blood pressure on the serial recordings made over approximately 24 h with the normalized p13–n23 amplitude (A) and the n1–p1 amplitude (B) according to the presence of orthostatic hypotension. In patients with OH, SBP_SD was not correlated with the n1–p1 amplitude (r=−0.085, p=0.491) or the normalized p13–n23 amplitude (r=−0.160, p=0.193). In patients with a normal head-up tilt-table test despite orthostatic intolerance, SBP_SD had a weak negative correlation with the normalized p13−n23 amplitude (r=−0.235, p=0.041), but it was not correlated with the n1–p1 amplitude (r=−0.041, p=0.726). NOI, normal head-up tilt-table test despite orthostatic intolerance; OH, orthostatic hypotension; SBP_SD, standard deviation of the systolic blood pressure.

Prediction of OH

Multivariatble logistic regression showed that OH was associated with bilateral oVEMP abnormality (odds ratio [OR]=5.58, 95% confidence interval [CI]=1.29–24.13, p=0.021), SBP_SD (OR=1.17, 95% CI=1.03–1.32, p=0.012), and female sex (OR=3.29, 95% CI=1.46–7.42, p=0.004) (Table 3).

Table 3. Differentiating OH from NOI in logistic regression analyses.

Variable		Univariate analysis		Multivariable analysis*
		OR (95% CI)	p	OR (95% CI)	p
Age		1.03 (1.01–1.05)	0.012
Sex, female		3.70 (1.86–7.39)	<0.001	3.29 (1.46–7.42)	0.004
Body weight		1.01 (0.99–1.04)	0.366
cVEMP abnormality†			0.270
	Unilateral	2.33 (0.73–7.39)	0.153
	Bilateral	1.58 (0.61–4.12)	0.350
oVEMP abnormality†			0.005		0.012
	Unilateral	0.56 (0.20–1.57)	0.269	0.38 (0.11–1.31)	0.124
	Bilateral	6.96 (1.91–25.31)	0.003	5.58 (1.29–24.13)	0.021
Average SBP		1.04 (1.01–1.06)	0.002	1.03 (1.00–1.06)	0.097
SBPSD		1.24 (1.12–1.37)	<0.001	1.17 (1.03–1.32)	0.012
Nocturnal BP dipper		0.56 (0.26–1.17)	0.120
Intracranial/extracranial cerebral artery stenosis		2.88 (1.15–7.18)	0.024
Parkinson’s disease		8.72 (1.05–72.82)	0.045	7.68 (0.71–82.54)	0.093

^*Variables for which p<0.05 in a univariate analysis were included in the multivariable analysis using the backward variable selection method; ^†oVEMP and cVEMP values were considered to be abnormal when they were absent, delayed, and/or decreased in amplitude relative to the reference values.

BP, blood pressure; CI, confidence interval; cVEMP, cervical vestibular-evoked myogenic potential; NOI, normal head-up tilt-table test despite orthostatic intolerance; OH, orthostatic hypotension; OR, odds ratio; oVEMP, ocular vestibular-evoked myogenic potential; SBP, systolic BP; SBP_SD, standard deviation of SBP.

ROC analysis

The sensitivity and specificity for differentiating OH from NOI were 72.1% and 67.5%, respectively, at a cutoff value of 12.7 mm Hg for SBP_SD, with an AUC of 0.73 (95% CI=0.65–0.81) (Fig. 2).

Fig. 2. Receiver operating characteristics analysis of SBP_SD for differentiating OH from NOI. An ROC curve that is closer to the top-left corner indicates that the test is better at predicting OH. The sensitivity was 72.1% and the specificity was 67.5% for a cutoff value of 12.7 mm Hg for SBP_SD, with good statistical power in discriminating between the groups indicated by an area under the ROC curve of 0.73 (95% confidence interval=0.65–0.81). AUC, area under the ROC curve; NOI, normal head-up tilt-table test despite orthostatic intolerance; OH, orthostatic hypotension; ROC, receiver operating characteristic; SBP_SD, standard deviation of the systolic blood pressure.

DISCUSSION

The findings of this study can be summarized as follows: 1) bilateral oVEMP abnormality, SBP_SD, and female sex were independent predictors of OH; 2) SBP_SD was not associated with the n1–p1 amplitude or the normalized p13–n23 amplitude in patients with OH; and 3) the sensitivity and specificity for discriminating OH from NOI were 72.1% and 67.5%, respectively, at a cutoff value of 12.7 mm Hg for SBP_SD, with an AUC of 0.73 (95% CI=0.65–0.81).

Our findings suggest that 24-h BP variability is an independent predictor of OH regardless of the integrity of the otolith-autonomic reflex. OH occurrence and BP variability are commonly attributed to baroreflex impairment.³⁰ It is widely recognized that the baroreflex counteracts the decrease in BP during orthostasis. Baroreceptors provide a servomechanism that triggers appropriate vasomotor and cardiomotor responses to maintain the BP. However, the baroreflex sensitivity decreases with aging;³¹ more specifically, vascular aging decreases arterial elasticity, thereby decreasing the sensitivity of carotid baroreceptors in responding to rapid BP changes.³² Likewise, pharmacologically induced decreases of the baroreflex have been associated with the development of OH³³ and increased BP variability.³⁴ Conversely, administering vagotonic atropine to healthy young adults increased the arterial baroreflex gain and stabilized age-related BP changes.³⁵ There is evidence that BP variability and OH share a common pathomechanism via an impaired baroreflex. Our findings indicate that an SBP_SD higher than 12.7 mm Hg is useful for predicting OH among patients with orthostasis. Our results also suggest that ABPM can aid in the detection of OH. In addition to its low sensitivity or reproducibility,³⁶,³⁷ the utility of the HUT can be further reduced by an extremely high body weight or a low BP.³⁸ In addition, although the HUT is generally safe and complications are rare, caution is necessary for patients having cardiac comorbidities.³⁹ In such cases ABPM can help in the detection of baroreflex dysfunction.

The 24-h BP variability was not associated with oVEMP or cVEMP parameters in this study, implying a negligible contribution of the otolith-autonomic reflex to within-day BP variability. As stated, BP variability is rather related to vascular stiffness and decreased carotid baroreceptor sensitivity owing to vascular aging.³² The galvanic stimulation did not alter the BP variability during the HUT.⁴⁰ Alternatively, the modulation of short-term BP changes via the otolith-sympathetic reflex might not be detectable by measuring the BP variability over 24 h.⁴¹ The otolith inputs for BP modulation are rapid and last for a short time, with the baroreflex taking over the modulation thereafter.⁸,¹⁰ Given that the sensitivity of the HUT is reportedly highly variable (35%–65%),³⁷ our findings suggest that greater BP variability (SBP_SD>12.7 mm Hg) or bilaterally deficient oVEMP responses are a clinical tool for predicting OH.

Otolith dysfunction could independently predict OH. Our findings corroborate that of previous clinical studies showing an association between vestibular dysfunction and OH.²⁰,⁴²,⁴³,⁴⁴,⁴⁵ Similarly, experimental studies have also supported such an association. For example, bilateral transection of the vestibular nerve resulted in a prominent decrease in BP during the HUT in cats.⁴⁶ In addition, pharmacological lesioning of the caudal region of the vestibular nuclei complex has been shown to compromise the sympathetic output and decrease the BP during the HUT in cats.⁴⁷ Provided that this phenomenon is induced by selective severance of the otolith end organs rather than by lesioning the semicircular canals,⁴⁸ the loss of otolithic inputs may be responsible for orthostatic intolerance. Similarly, OH has been found to be more common in patients with abnormal oVEMP responses than in those with normal responses, irrespective of aging.²⁰,⁴⁹ Together these observations indicate that otolith-sympathetic dysfunction may give rise to OH regardless of the integrity of the baroreflex, resulting in a rapid reduction in the feedforward mechanism that underpins orthostatic tolerance. It was particularly interesting that the present study found that within-day BP variability reflects the function of the baroreflex but not the otolith-autonomic reflex.

This study had some limitations. We used patients with NOI as the control group. The orthostatic dizziness in NOI may occur due to mechanisms other than orthostatic intolerance, such as persistent postural perceptual dizziness or benign recurrent vertigo.⁵⁰ A normal result in a single test does not guarantee the absence of OH because HUT results may be subject to intraindividual variability according to the patient’s condition. This limitation of the diagnostic yield of the HUT³⁷ might have resulted in patients with OH being included in NOI group. Thus, the rate of VEMP abnormalities in patients with NOI may have reduced the statistical power in discriminating between the groups. It would be useful to compare utricular dysfunction in patients with OH with that in age- and sex-matched healthy controls. In addition, BP variability over 24 h can be a crude marker since it can fail to detect short-term BP changes at intervals of ≥30 min.⁴¹ Thus, the otolith-autonomic reflex might not be accurately quantified using 24-h ABPM measurements. Spectral analysis of beat-to-beat variations in the R-R interval may accurately reflect the sympathovagal balance in terms of the otolithic control of the cardiovascular system.⁵¹,⁵² Impairments of the baroreflex and the otolith-autonomic reflex may collectively contribute to the development of OH. The mechanism underlying the development of OH could not be clearly characterized by the design of this study, and hence this should also be addressed in future studies. Moreover, further analyses of the ABPM and VEMP results for the three types of OH (initial, classic, and delayed) may improve the understanding of the interactions between the autonomic and vestibular systems.

In conclusion, bilaterally deficient oVEMP responses may be associated with OH regardless of 24-h BP variability, reflecting the integrity of the otolith-autonomic reflex during orthostasis. The 24-h BP variability might instead by predominantly regulated by the baroreflex, which also contributes to orthostatic tolerance complementary to the vestibulo-autonomic reflex.

Availability of Data and Material

The datasets generated or analyzed during the study are available from the corresponding author on reasonable request.

Supplementary Materials

The online-only Data Supplement is available with this article at https://doi.org/10.3988/jcn.2024.0092.

Supplementary Table 1

Shapiro–Wilk test of the variables.