Increased Prevalence of Vestibular Loss in Mild Cognitive Impairment and Alzheimer’s Disease

Abstract

Background/Aims:

Recent evidence has shown that Alzheimer’s Disease (AD) patients have reduced vestibular function relative to healthy controls. In this study, we evaluated whether patients with Mild Cognitive Impairment (MCI) also have reduced vestibular function relative to controls, and compared the level of vestibular impairment between MCI and AD patients.

Methods:

Vestibular physiologic function was assessed in 77 patients (26 MCI, 51 AD) and 295 matched controls using 3 clinical vestibular tests. The association between vestibular loss and cognitive impairment was evaluated using conditional logistic regression models.

Results:

Individuals with vestibular impairment had a 3 to 4-fold increased odds of being in the MCI vs. control group (p-values < 0.05). MCI patients had a level of vestibular impairment that was intermediate between controls and AD.

Conclusion:

These findings suggest a dose-response relationship between vestibular loss and cognitive status, and support the hypothesis that vestibular loss contributes to cognitive decline.

1. INTRODUCTION

The vestibular (inner ear balance) system is integral for postural control, gait, and spatial orientation. Recent evidence has shown a link between vestibular and cognitive function, particularly spatial cognitive function, in older adults [1–7]. The vestibular system provides critical input to the hippocampal networks involved in spatial orientation and navigation, and reduced vestibular function has been suggested to cause impairment in these spatial abilities [8–11]. Moreover, in prior work, we demonstrated that vestibular loss is substantially more prevalent in patients with Alzheimer’s Disease (AD) relative to age-matched controls [12]. Many AD patients suffer from impairments in spatial cognition, leading to impairments in gait, spatial memory, spatial navigation, and to falls [13–15]. Reduction in vestibular function may underlie these spatial deficits in AD patients; indeed, we observed in a subsequent study that AD patients with vestibular loss have poorer spatial cognitive skills relative to AD patients with normal vestibular function [16]. As such, vestibular loss may contribute to a “spatial” subtype of AD.

However, whether patients with mild cognitive impairment (MCI) also exhibit vestibular impairment has not been established. MCI affects an estimated 15–20% of adults ages 65 and older [17–19], and a recent systematic review of studies tracking the progression of MCI over a period of 5 or more years found that at least one third of MCI patients progress to AD during follow-up [20]. Given the intermediate phenotype of MCI and the high conversion rate from MCI to AD, MCI is regarded as a clinical stage on the continuum of cognitive decline between normal aging and dementia [21]. Therefore, we sought to evaluate whether patients with MCI also demonstrate vestibular impairment. We further hypothesized that there would be a dose-response relationship between vestibular loss and level of cognitive impairment (i.e. a greater level of vestibular dysfunction in individuals with AD compared to MCI).

In this study, we characterized vestibular function in 77 patients (26 MCI and 51 AD). MCI and AD cases were matched to controls evaluated in the Baltimore Longitudinal Study of Aging (BLSA). We compared the proportion of vestibular impairment in MCI and AD patients compared to matched controls in conditional logistic regression models. Furthermore, we modeled cognitive decline as an ordinal outcome variable with MCI as an intermediate stage between healthy controls and AD using the conditional adjacent-category logistic regression model [22]. Our study explores at what stage along the continuum from healthy aging to AD does vestibular loss become evident. These data could assist in making efforts at early detection and vestibular intervention in patients with cognitive impairment to mitigate subsequent risks of spatial disorientation and falls.

2. METHODS

2.1. Study Participants

2.1.1. Cases

Patients with MCI and AD were recruited from December 2014 to August 2017 from the Johns Hopkins Memory and Alzheimer’s Treatment Centre, an interdisciplinary memory disorder clinic, and the Johns Hopkins Alzheimer’s Disease Research Centre. Inclusion criteria for the study were 1) Age ≥ 55 years; 2) Diagnosis of MCI or AD; 3) Fluency in English; 4) Ability to obtain informed consent from the participant or a legally authorized representative. A diagnosis of MCI or AD was made using the National Institute on Aging-Alzheimer’s Association diagnostic criteria [23,24]. Prior to participation, a Memory Center physician determined the participant’s ability to follow examination procedures.

Patients were excluded from the study if they were unable to understand and participate in the examination procedures. Demographic characteristics (age, sex, and education) and score on the Mini Mental Status Exam (MMSE) were obtained from the patients’ charts. Education was classified as less than college, college, or greater than college. All the participants provided written informed consent. The study was approved by the Johns Hopkins Institutional Review Board (Protocol # NA_00035749).

2.1.2. Controls

For each case, up to 4 healthy controls were matched without replacement from the Baltimore Longitudinal Study of Aging (BLSA) based on age (within 5 years), sex, and education. The BLSA is a prospective cohort study of participant volunteers aged 21 to 103 in the National Institute on Aging Clinical Research Unit at Harbor Hospital in Baltimore, Maryland, conducted by the National Institute on Aging Intramural Research Program. This study evaluated a cross-sectional sample of BLSA participants who underwent vestibular testing from February 2013 to December 2016. All participants in the BLSA provided written informed consent, and the study protocol was approved by the institutional review board of the National Institute of Environmental Health Science, National Institutes of Health.

2.2. Vestibular Function Tests

The vestibular end-organ consists of two otolith organs (the saccule and utricle) and three semicircular canals (horizontal, anterior and posterior). We employed a standard array of vestibular physiologic tests that measure otolith and semicircular function. Specifically, the sound-evoked cervical vestibular evoked myogenic potential (cVEMP) was used to assess saccular function, the vibration-evoked ocular vestibular evoked myogenic potential (oVEMP) was used to assess utricular function, and video head impulse testing (vHIT) was used to assess horizontal semicircular canal function. Details of these three vestibular physiologic tests have been previously published in detail [25–29], and are briefly described below.

2.2.1. Vestibular-evoked Myogenic Potentials

Vestibular-evoked myogenic potential recordings in patients with cognitive impairment were made using a commercial electromyographic system (software version 21.1; Natus Neurology, Middletown, WI), with self-adhesive electrodes from GN Otometrics (Schaumburg, IL). A similar system was employed at the BLSA (software version 14.1; CareFusion Synergy, Dublin, OH).

In sound-evoked cVEMP testing, trained examiners placed recording electromyographic electrodes on the sternocleidomastoid (SCM) muscle and the sternoclavicular junction bilaterally. A ground electrode was placed on the manubrium sterni. Patients were instructed to turn their heads against resistance to activate the SCM muscle. A sweep of auditory tone bursts (500 Hz, 125 dB SPL) was delivered monaurally through headphones (VIASYS Healthcare, Madison, WI), and inhibitory potentials were recorded from the ipsilateral SCM muscle. The presence or absence of a cVEMP response was recorded for each ear, indicating normal or impaired saccular function, respectively, as characterized by published guidelines [25–27].

In vibration-evoked oVEMP testing, a non-inverting electrode was placed on the cheek directly beneath the pupil approximately 3 mm below the eye. A second inverting electrode was placed 2 cm below the non-inverting electrode, and a ground electrode was placed on the manubrium sterni. An absent response was determined if the characteristic waveforms did not occur per published guidelines [25–27], and was confirmed with a repeat assessment. For both cVEMP and oVEMP tests, normal vestibular function was defined as the presence of a vestibular evoked myogenic potential in both ears. Abnormal vestibular function was defined as either unilaterally or bilaterally absent function.

2.2.2. Video Head Impulse Testing (vHIT)

The vHIT was performed in the plane of the right and left horizontal semicircular canals using the EyeSeeCam system (Interacoustics, Eden Prairie, MN), which has been shown to accurately assess the Vestibulo-ocular Reflex (VOR) [28–31]. The EyeSeeCam video-oculography system consists of a light-weight goggle frame with a built-in high-speed digital camera that tracks the pupil and an accelerometer to record head movements. The head was moved a small amplitude (approximately 5° – 15°) with high velocity (typically 150° – 250° per second) horizontally toward the right and left sides randomly, at least 10 times in each direction. The horizontal VOR gain was defined as the ratio of the area under the eye velocity curve to the head velocity curve from HIT onset until the head velocity returned to 0. Individuals with normal semicircular canal function have a VOR gain of 1.0, i.e. the eye movement velocity is equal to and fully compensates for the head movement velocity. Individuals with semicircular canal impairment make insufficient eye movements, resulting in VOR gains that decrease from 1.0. The VOR gain of the worse ear was used for the analysis.

Three patients (3 AD) had missing cVEMP data, 11 patients (1 MCI, 10 AD) had missing oVEMP data, and 16 patients (3 MCI, 13 AD) had missing vHIT data due to the inability of the patient to follow testing instructions on one or more tests or because the equipment was unavailable for testing. There were no significant differences in age, sex, or education between patients with missing data for one or more vestibular physiologic tests and patients without any missing data according to bivariate statistics (t-tests for continuous variables and χ2 for categorical variables). Patients with missing data for one or more vestibular physiologic tests had a significantly lower MMSE score than patients without any missing data (p = 0.02), according to two-sample t-test.

2.3. Statistical Analysis

Cases (MCI and AD) were matched 1:4 to controls from the BLSA without replacement based on age (within 5 years), sex, and education, in order to increase the statistical power of the study. Among cases, 24 MCI and 40 AD matched to a maximum of 4 controls, while the remaining cases (2 MCI, 11 AD) matched to 3 controls.

The main outcome of interest was diagnosis (AD versus MCI versus controls). The main variables of interest were abnormal VEMPs and VOR gain. Bivariate statistics (t-tests for continuous variables and χ2 for categorical variables) were utilized to assess differences in demographic and vestibular variables between cases (MCI and AD) and their matched controls. Conditional logistic regression analyses adjusted for age, sex, and education were conducted to compare vestibular function (abnormal cVEMPs, abnormal oVEMPs, and VOR gain) between MCI patients and their matched controls, as well as AD patients and their matched controls. Moreover, the conditional adjacent-category logistic regression model was adapted to explore the association between vestibular function and diagnosis (i.e. level of cognitive impairment). The conditional adjacent-category logistic model was developed to analyze matched case-control data where there is one control group, and multiple disease states with a natural ordering among them [22]. This statistical model is advantageous because its framework allows for more than one level of disease. The model assumes a constant OR across all pairs, i.e. the OR estimate comparing MCI and healthy controls is equivalent to the OR estimate comparing AD and MCI. Additionally, a random-effects Tobit model was used to assess the relationship between vestibular function and MMSE score, while accounting for the case-control structure of the data. Neither the conditional logistic regression, conditional adjacent logistic regression, nor the tobit models assume a normal distribution of its independent variables. All the analyses were performed using STATA version 14 (College Station, TX, USA).

3. RESULTS

Demographic information for the sample of 77 cases (26 MCI and 51 AD) and 295 matched controls and the results of bivariate statistics (t-tests for continuous variables and χ2 for categorical variables) are provided in Table 1. With successively greater levels of cognitive impairment, there was a decrease in normal cVEMP responses (51.1% in controls, 34.6% in MCI, and 14.6% in AD) and a decrease in normal oVEMP responses (80.1% in controls, 52.0% in MCI, and 51.2% in AD), while VOR gain was lowest (poorest) in MCI patients (mean = 0.84, SD = 0.16), followed by AD patients (mean = 0.93, SD = 0.17), then controls (mean = 0.94, SD = 0.15). “Mean MMSE scores were 26.5 (SD 2.1; range 22, 30) in MCI patients, and 20.2 (SD 4.8; range 9, 29) in AD patients, and 28.5 (SD 1.4; range 24, 30) in healthy controls. Four AD patients and 34 controls had missing MMSE scores.”

Table 1.

Characteristics of sample by level of cognitive impairment.

-	Total	MCI			AD
	N=372	Cases (N=26)	Controls (N=102)	p-value *	Cases (N=51)	Controls (N=193)	p-value†
Sex, n (%)	-	-	-	0.94	-	-	0.82
Male	130 (34.9%)	11 (42.3%)	44 (43.1%)	-	15 (29.4%)	60 (31.1%)	-
Female	242 (65.1%)	15 (57.7%)	58 (56.9%)	-	36 (70.6%)	133 (68.9%)	-
Mean age (SD)	74.5 (8.1)	72.7 (9.3)	72.5 (9.5)	0.95	76.0 (6.6)	75.4 (7.3)	0.61
Education, n (%)	-	-	-	0.99	-	-	0.98
Less than college	117 (31.5%)	7 (26.9%)	26 (25.5%)	-	18 (35.3%)	66 (34.2%)	-
College	150 (40.3%)	9 (34.6%)	36 (35.3%)	-	22 (43.1%)	83 (43.0%)	-
Greater than college	105 (28.2%)	10 (38.5%)	40 (39.2%)	-	11 (21.6%)	44 (22.8%)	-
Mean MMSE (SD)^	27.1 (3.6)	26.5 (2.1)	28.3 (1.4)	<0.001	20.2 (4.8)	28.5 (1.4)	<0.001
cVEMP, n (%)‡	-	-	-	0.03	-	-	<0.001
Normal Function	136 (44.0%)	9 (34.6%)	49 (58.3%)	-	7 (14.6%)	71 (47.0%)	-
Abnormal Function	173 (56.0%)	17 (65.4%)	35 (41.7%)	-	41 (85.4%)	80 (53.0%)	-
oVEMP, n (%)§	-	-	-	0.008	-	-	<0.001
Normal Function	227 (73.9%)	13 (52.0%)	70 (78.7%)	-	21 (51.2%)	123 (80.9%)	-
Abnormal Function	80 (26.1%)	12 (48.0%)	19 (21.3%)	-	20 (48.8%)	29 (19.1%)	-
Mean VOR gain (SD)¶	0.93 (0.16)	0.84 (0.16)	0.94 (0.14)	0.008	0.93 (0.17)	0.95 (0.16)	0.53

^*p-values are from bivariate statistics comparing MCI cases and matched controls.

^†p-values are from bivariate statistics comparing AD cases and matched controls.

^{^}334 participants completed MMSE (261 controls, 26 MCI, and 47 AD).

^‡309 participants completed cVEMP testing (235 controls, 26 MCI, and 48 AD).

^§307 participants completed oVEMP testing (241 controls, 25 MCI, and 41 AD).

^¶282 participants completed VOR testing (221 controls, 23 MCI, and 38 AD).

Conditional logistic regression analyses were conducted to assess the association between vestibular loss and cognitive impairment categories, controlling for demographic factors (Table 2). In comparing MCI to controls, abnormal cVEMPs were associated with a 3-fold increased odds of MCI (OR 3.0, 95% CI 1.1, 8.5) and abnormal oVEMPs were associated with a nearly 4-fold increased odds of MCI (OR 3.9, 95% CI 1.4, 11.3). Additionally, for every standard deviation increase in VOR gain, there was a significantly decreased odds of MCI (OR 0.36, 95% CI 0.17, 0.74). In comparing AD to controls, abnormal cVEMPs were associated with a 5-fold increased odds of AD (OR 5.0, 95% CI 2.0, 12.3) while abnormal oVEMPs were associated with an over 4-fold increased odds of AD (OR 4.2, 95% CI 1.9, 9.1). VOR gain was not significantly associated with increased odds of AD (OR 0.87, 0.59, 1.30).

Table 2.

Conditional logistic regression model for odds of cognitive impairment by vestibular physiologic function.

	MCI		AD
Vestibular Function	OR (95% CI)	p-value	OR (95% CI)	p-value
Abnormal cVEMPs *	3.0 (1.1, 8.5)	0.04	5.0 (2.0, 12.3)	0.001
Abnormal oVEMPs †	4.1 (1.4, 12.4)	0.01	4.2 (1.9, 9.1)	<0.001
VOR gain ‡ §	0.36 (0.17, 0.74)	0.006	0.87 (0.59, 1.30)	0.51

^*cVEMP data includes 26 MCI and 48 AD cases and their matched controls.

^†oVEMP data includes 25 MCI and 41 AD cases and their matched controls.

^‡VOR gain data includes 23 MCI and 38 AD cases and their matched controls.

^§Odds ratios for VOR gain are standardized based on the standard deviation of VOR gain.

Additionally, we used the conditional adjacent-category logistic regression model to assess whether there was a dose-response relationship between vestibular loss and cognitive impairment. We applied this model only to cVEMP and oVEMP data, given that for those 2 vestibular measures, we observed an increasing level of vestibular impairment by severity of cognitive impairment in descriptive analyses (Table 1). Controlling for demographic factors, individuals with abnormal cVEMPs had a 4-fold increased proportional odds (OR 4.0, 95% CI 2.0, 7.9) of a successively greater level of cognitive impairment. Similarly, individuals with abnormal oVEMP responses also had an over 4-fold increased proportional odds (OR 4.2, 95% CI 2.2, 7.9) of a successively greater level of cognitive impairment. Moreover, we used the random effects Tobit model to evaluate whether there was a relationship between vestibular loss and MMSE score. Controlling for demographic factors, abnormal cVEMPs (β −1.5, 95% CI −2.6, −0.5, p = 0.005) and oVEMPs (β −1.7, 95% CI - 2.9, −0.5, p = 0.007) were significantly associated with lower MMSE scores.

4. DISCUSSION

AD is a slowly progressive dementia affecting 55 million Americans in 2019 characterized by neuropathological features of neuritic plaques and neurofibrillary tangles [32]. The clinical progression of AD is thought to be preceded by a transitional state involving mild cognitive complaints, most commonly mild memory complaints [33]. Individuals with MCI, who have mild cognitive complaints with minimal or no function decline, have been shown to have an increased risk of progressing to clinically probably AD [34]. In this case-control study of MCI and AD patients, we observed a significantly higher proportion of vestibular impairment in patients with MCI relative to matched controls. Further, we found evidence that vestibular loss increases the odds of successive levels of cognitive impairment (normal cognition to MCI, MCI to AD) ~4-fold. The dose-response relationship we observed between abnormal cVEMPs and cognitive impairment (i.e. healthy controls vs. MCI vs. AD) provides support for our hypothesis that vestibular loss, namely saccular dysfunction, may be causally related to cognitive decline. Our study adds to prior work by demonstrating that vestibular loss is evident even at earlier stages of cognitive impairment than AD, specifically at the MCI stage. We observed statistically greater rates of saccular, utricular and semicircular canal impairment among patients with MCI relative to age-matched controls. Our finding is consistent with previous work showing that patients with MCI and AD have de creased performance on clinical tests of equilibrium, limb coordination, and balance compared to controls [35, 36]. The dose-response relationship we observed between vestibular loss and cognitive impairment provides support for the hypothesis that vestibular loss may be causally related to cognitive decline.

This study builds on a growing body of literature demonstrating an association between vestibular loss and cognitive impairment, notably spatial cognitive impairment. In cognitively normal older adults, reduced vestibular function has been associated with poorer performance on neurocognitive tests that measure visuospatial skills such as the digit symbol substitution test [1, 37], the Benton Visual Retention Test, and the Trail Making Test Part B. The link between vestibular and spatial cognitive function may be explained by the critical role played by vestibular inputs in hippocampal spatial orientation and navigation networks. Animal studies have found that vestibular ablation leads to decreased muscarinic receptor expression in the hippocampus, as well as disruption of location-specific firing of hippocampal place cells [8, 9]. In humans, functional neuroimaging studies have shown activation of the hippocampus following peripheral vestibular stimulation [38, 39], and structural neuroimaging studies have shown that patients with vestibular disorders have reduced hippocampal volume [40, 41]. The role of neurofibrillary tau deposition in brainstem nuclei with diffuse projections across the cerebral cortex has also been theorized as a potential contributor to the association between vestibular impairment and AD [42]. Recent evidence has also shown that vestibular loss is 2-fold more prevalent in patients with Alzheimer’s disease and dementia relative to age-matched controls [12]. More broadly, it has been suggested that microgravity, which causes reduced stimulation of the peripheral vestibular system, is associated with accelerated development of neurodegenerative diseases [43]. Moreover, AD patients with vestibular loss were significantly more likely to have impairment in spatial cognitive skills (as measured by performance on the Money Road Map test and the Trail Making Test Part B) and are more likely to experience difficulty driving relative to AD patients without vestibular loss [16, 44]. Vestibular loss has been hypothesized to be causally related to the onset of AD and dementia [45]. The clinical phenotype of AD is heterogeneous, with some patients having more of a classic amnestic type and others having disproportionately greater impairment in motoric and spatial function [13, 14, 46]. Vestibular loss may contribute specifically to a spatial subtype of AD, possibly via a spatial subtype of MCI. Future longitudinal studies will be needed to further test this hypothesis. Additionally, given the well-studied association between APOE genotype and AD, future studies are needed to evaluate whether the APOE genotype might be a confounder between vestibular loss and AD.

We note several limitations in our study. This was a cross-sectional case-control study and thus cannot support causal inferences between vestibular loss and cognitive impairment. Although patients and controls were matched by age, sex and education level, it is possible that other differences between the groups (i.e. confounding factors) may explain the association between vestibular loss and cognitive impairment. Additionally, the control sample was drawn from the BLSA which is known to include very healthy older adults and may not be fully comparable to the MCI/AD patient sample. Another source of selection bias may arise from the fact that patients are more likely to present for clinical assessment when they have both vestibular loss and cognitive impairment vs. either condition alone, leading to a potential overestimate of the proportion of vestibular impairment in MCI and AD patients (Berksonian bias). Additionally, given that we excluded patients with severe dementia who may have been unable to complete the vestibular testing procedures, our findings may only be applicable to patients with mild-to-moderate cognitive impairment. Several MCI and AD patients were unable to complete one or more vestibular physiologic function tests. These individuals had significantly lower MMSE scores compared to patients who were able to complete all 3 tests. This may contribute to an underestimation of the association between vestibular loss and cognitive impairment, given that those with the most severe levels of cognitive impairment would be expected to have the highest rates of vestibular loss, but are more likely to not be able to complete vestibular physiologic function tests. Further, it is also possible that patients with cognitive impairment and dementia may exhibit poor vestibular function because of their inability to perform test procedures. However, all individuals who were tested in this study had to be able to successfully follow the instructions to calibrate each individual test (i.e., follow the calibration laser point for vHIT, generate sufficient background neck contraction for cVEMP testing, and generate vertical saccades for oVEMP testing) prior to participating in vestibular testing. Another potential limitation of this study was our use of the conditional adjacent-category logistic regression model, which assumes a constant dose-response relationship between vestibular loss and levels of cognitive impairment, i.e. a constant OR across all pairs. We selected this model for its simplicity and the resulting increased statistical efficiency. However, it is possible that the dose-response relationship between vestibular loss and cognitive impairment is more complex, perhaps according to the brain’s compensatory abilities which vary with stage of cognitive impairment. Future studies in larger samples will be needed to evaluate for potentially more complex, non-linear relationships between vestibular loss and cognitive impairment. Lastly, in this study we did not differentiate different subtypes of MCI, which may or may not be of likely AD cause. Future studies assessing vestibular function in MCI patients with likely AD etiology (prodromal AD) will be needed to determine whether vestibular loss is specifically associated with the onset of AD-related neuropathology.

CONCLUSION

In summary, in this case-control study of MCI and AD patients, we observed that MCI patients also have significant vestibular impairment, at a level that appears to be intermediate between normal cognition and AD. This finding builds on existing literature suggesting an association between vestibular loss and AD by demonstrating for the first time that vestibular impairment begins to occur in MCI, a clinical stage that is often a precursor to AD and other dementias. This study provides support for our hypothesis that vestibular loss increases the risk of progression to dementia, via a preclinical dementia stage. Future longitudinal studies will be needed to confirm this hypothesis by assessing the temporal relationship between vestibular loss, cognitive decline and incident dementia. Additionally, further studies will be needed to explore whether vestibular loss specifically contributes to the onset of AD vs. other or mixed dementias, by evaluating the relationship between vestibular loss and AD-related pathology such as amyloid-beta and tau accumulation and medial temporal lobe atrophy. Lastly, although in this study we selected a battery of tests consistent with the latest literature that measures across the entire vestibular end-organ [47–52], future studies evaluating whether specific vestibular disorders are associated with cognitive impairment and Alzheimer’s disease may be useful, as well as evaluating the impact of unilateral versus bilateral vestibular loss. Vestibular impairments are modifiable through vestibular rehabilitation, which consists of a program of specific exercises that improve gait and gaze instability. At present, vestibular rehabilitation is rarely offered to MCI or AD patients, who may benefit both to reduce their risk of typical consequences of vestibular impairment such as falls and also their risk of further cognitive decline. Future studies will need to evaluate the potential diagnostic and/or prognostic value of vestibular loss with respect to MCI and AD, and studies will also need to examine the potential benefit of vestibular therapy in cognitively-normal individuals to reduce the progression of cognitive impairment. Vestibular loss is a treatable condition, therefore identifying it as a potential risk factor or co-factor in the progression to AD could be of substantial public health benefit.

LIST OF ABBREVIATIONS

AD

Alzheimer’s disease

BLSA

Baltimore Longitudinal Study of Aging

cVEMP

Cervical Vestibular Evoked Myogenic Potential

CI

Confidence Interval

MCI

Mild Cognitive Impairment

oVEMP

Ocular Vestibular Evoked Myogenic Potential

SCM

Sternocleidomastoid

vHIT

Video Heat Impulse Test

VOR

Vestibulo-ocular Reflex

AVAILABILITY OF DATA AND MATERIALS

The data supporting the findings of the article will be shared by the corresponding author [Eric X Wei] to other researchers upon request.