The six-item visual vertigo analogue scale: A modified questionnaire for assessment of visual vertigo

Abstract

Background

Individuals with vestibular disorders often describe symptoms of dizziness, disorientation, and impaired balance induced by visual stimuli, commonly referred to as visual vertigo.

Objective

The purpose of this study was to improve assessment of visually induced symptoms by modifying the original, nine-item Visual Vertigo Analogue Scale (VVAS) to produce a more robust, interpretable, and clinically useful measure of visual vertigo. We also aimed to establish a cut-off score for identifying patients with vestibular disorders and low- versus high-severity visual vertigo.

Methods

Patients with vestibular disorders (n = 199) and controls (n = 107) completed the modified VVAS (mVVAS). Exploratory factor analysis was conducted to develop a shorter scale and internal consistency analyses were performed. A K-means cluster analysis was used to determine a cut-off score for identifying low- and high-severity groups.

Results

A reduced six-item scale (VVAS-6), which included three items from the VVAS and three items from the mVVAS, had excellent internal consistency (Cronbach’s α = 0.93). We identified four concepts within the VVAS-6 (walking with visual stimulation, visual stimulation only, escalator, car). A cut-off score of 37 was able to identify patients with vestibular disorders and high severity of visual vertigo.

Conclusions

The use of a cut-off score of 37 for the VVAS-6 scale resulted in an improvement in correctly identifying patients with vestibular disorders and high severity of visual vertigo from controls.

Keywords

visual vertigo dizziness vestibular assessment

Introduction

Visual vertigo describes symptoms of dizziness, disorientation, and impaired balance induced by environments with conflicting visual and vestibular information or complex visual stimuli.¹ Visual vertigo has shared characteristics with space and motion discomfort,² phobic postural vertigo,³ and chronic subjective dizziness.^4,5 Although these conditions share overlapping symptom profiles, they represent distinct clinical constructs; visual vertigo refers to visually induced vertiginous symptoms,⁶ whereas persistent postural-perceptual dizziness is a formal diagnosis characterized by persistent symptoms exacerbated by moving or complex visual stimuli.⁷ While clinicians have long recognized visually-induced complaints in patients presenting with central and peripheral causes of dizziness, a formal means of assessment under the moniker of visual vertigo was not presented until 2011.⁸ Dannenbaum and colleagues proposed a nine-item Visual Vertigo Analogue Scale (VVAS)⁸ based on the qualitative examination questions of Longridge and Mallinson.⁹

The VVAS is used to rate the intensity of dizziness induced by dynamic visual input during daily situations, and was designed to be faster to administer in the clinic than the 41-item Situational Characteristics Questionnaire¹⁰ for quantifying space and motion discomfort. Patients are asked to make a vertical mark on each 10 cm long analogue scale of the VVAS, indicating how much dizziness is provoked by a particular situation. The scales are anchored from zero to ten, with zero representing no dizziness and ten representing extreme dizziness and/or activity avoidance due to dizziness.⁸ The distance from the zero anchor to the patient’s marking is measured to the nearest half-centimeter, and two final scores are calculated: VVAS positive and VVAS severity.⁸ VVAS positive indicates that the patient rated two or more items on the scale above zero.⁸ VVAS severity is the mean score of the completed analogue scales multiplied by 10, yielding a score of 0 (patient does not experience visual vertigo) to 100 (patient has severe visual vertigo).⁸ The VVAS is valid and responsive to change, making it a useful tool for identifying and evaluating the progression of symptoms of visual vertigo.¹¹

While the VVAS is internally consistent and reliable (Cronbach’s α = 0.94),⁸ there are aspects of the scale that present challenges for use. First, patients are classified as VVAS positive if they rate two or more items on the scale above zero.⁸ This makes identifying and grouping patients challenging, as a patient who rates two items at 0.5 cm and a patient who rates nine items at 9.5 cm would both be classified as VVAS positive. These patients at the two extremes in scoring would likely present with very different symptoms in the clinic. While VVAS severity can also be calculated, Dannenbaum et al.¹² did not provide a cut-off score for identifying patients with vestibular disorders and high severity of visual vertigo. Second, the scales are anchored from zero to ten, which can cause confusion for interpreting scales that are not marked. If a patient does not make a marking on a scale, it may be because the situation is not applicable (“I don’t do that”), is not provocative (“I do that but it does not bother me”), or is not completed (“I did not respond to that item”). This makes the VVAS difficult to score consistently and interpret. Third, measuring to the nearest half-centimeter makes calculating VVAS severity imprecise.

The purpose of this study was to improve assessment of visually induced symptoms by modifying the original, nine-item VVAS to produce a more robust, interpretable, and clinically useful measure of visual vertigo. We aimed to (1) construct a measure that was internally consistent, reliable, and clinically interpretable; and (2) establish a cut-off score for identifying patients with vestibular disorders and low- versus high-severity visual vertigo. We also aimed to explore relationships between visual vertigo scale scores and self-perceived dizziness handicap scores. We hypothesized that patients with vestibular disorders and high severity of visual vertigo would have greater self-perceived dizziness handicap scores, depression, and anxiety than patients with vestibular disorders and low severity of visual vertigo.

Materials and methods

Participants

One hundred and ninety-nine patients with vestibular disorders (Table 1) and 107 controls who were healthy or seeking physical therapy for orthopedic disorders participated from February 2018 to August 2019. Participants in the vestibular group were recruited from a balance disorders clinic, were over the age of 18, and had complaints of dizziness. All patients were evaluated and diagnosed by a board-certified neurologist based on clinical history and examination, with diagnostic determinations supported as appropriate by vestibular function testing, neuroimaging, and laboratory studies (diagnoses included vestibular hypofunction, benign paroxysmal positional vertigo, vestibular migraine, concussion, and unspecified dizziness). The designation of “unspecified dizziness” was used when patients met clinical criteria for vestibular symptoms, but a specific etiologic diagnosis could not be established at the time of evaluation. Control participants were volunteers who were either referred for physical therapy for lower back or upper/lower extremity orthopedic conditions or were healthy adults from the community. Control participants with a self-reported history of dizziness, migraine, cervical pain, or prior vestibular trauma were excluded. Participants were not excluded based on psychiatric or other neurological conditions. Informed consent was obtained from all participants, and the study was approved by the University of Pittsburgh Institutional Review Board. Participants were sequentially enrolled and completed the patient self-report measures using pen and paper during a single visit. As there was no exclusion criteria based on duration of dizziness, some participants had undergone vestibular rehabilitation prior to the baseline visit and some had not.

Table 1.

The most frequent diagnoses for 199 patients with vestibular disorders.

Diagnoses	n	%
Unspecified dizziness	83	41.7
Unilateral peripheral vestibulopathy	50	25.1
Vestibular migraine	21	10.6
Bilateral peripheral vestibulopathy	10	5.0
Concussion	9	4.5
Benign paroxysmal positional vertigo	8	4.0
Other ^a	18	9.0

^aOther diagnoses included Meniere’s disease, abnormality of gait, gait instability, functional gait disorder, central vertigo, and downbeat nystagmus; each represented a small number of participants.

Modifications to the VVAS

We proposed several modifications to the original, nine-item VVAS to address limitations related to interpretability and clinical application. We changed the instructions to simply and clearly state: “Indicate the amount of dizziness you experience in the following situations by marking off the scales below. Check the box to the left if you do not do the activity (it is not applicable) or don’t know if it would bother you.” As stated in these instructions, a box has been added to the left of each item that the patient can mark if an item could not be rated; however, it did not distinguish between reasons such as non-applicability, uncertainty, or activity avoidance.

Additionally, we added six items to the scale, making 15 items total in the modified VVAS (mVVAS). The selection of new items for the mVVAS was accomplished through discussion and consensus among a group of physical therapists and a board-certified neurologist with fellowship training in nonsurgical neurotology. A formal item-generation process (such as Delphi methods) was not used. Instead, items were identified based on clinical experience treating patients with vestibular disorders and visual vertigo, with the goal of capturing visually provocative situations commonly reported by these patients. Each newly added item was selected because it represented a distinct and clinically meaningful visual–vestibular stimulus not fully addressed by the original VVAS. No additional items beyond these six were considered for inclusion.

The first two items were added to determine if the location of being a passenger in the front seat or back seat of a vehicle is more provocative. Sitting in the front seat allows for fixation on the road ahead, while sitting in the back seat exposes the passenger to visual-vestibular mismatch and constant motion in the peripheral field of view, which can provoke motion sickness.¹³ The next item, “being under flickering lights,” was added because many of our patients with vestibular disorders and visual vertigo report that it is not the type of light (fluorescent) but the flickering of the lights that provokes their symptoms. Two additional items relating to ambulation, anxiety, and space and motion discomfort (“walking in crowds” and “walking in open spaces”) were added to the scale. Lastly, “going down an escalator” was added to complement the original item “going up an escalator.”

Finally, we modified the scoring to make the measure of dizziness more precise. Patients were asked to make a vertical mark on each 10 cm long analogue scale of the mVVAS, indicating how much dizziness is provoked by a particular situation. The scales are anchored from zero to ten, with zero representing no dizziness and ten representing extreme dizziness and/or activity avoidance due to dizziness.⁸ The distance from the zero anchor to the patient’s marking were measured to the nearest tenth of a centimeter. To remain consistent with the scoring proposed in the original measure, two final scores were calculated for the mVVAS: VVAS positive and VVAS severity.

Patient self-report measures

In addition to completing the VVAS and the mVVAS, patients with vestibular disorders completed the Dizziness Handicap Inventory (DHI)¹⁰ and the Hospital Anxiety and Depression Scale (HADS).¹⁴ The DHI was developed to assess the self-perceived handicapping impact of dizziness on daily life. The DHI consists of 25 items with a total possible score of 100.¹⁵ The DHI is internally consistent and reliable (Cronbach’s α = 0.89).¹⁵ The HADS is used to identify and measure self-reported severity of anxiety and depression. There are seven items related to anxiety and seven items related to depression. Each subscale has a total score range from 0 to 21 (each item scored 0–3), and is reliable and valid for detecting and quantifying the severity of anxiety and/or depression.¹⁶ A cut-off score of 8 or more on each section of the HADS was used to identify probable cases of anxiety and/or depression.¹⁴ The control group did not complete the DHI or HADS.

Statistical analyses

Data were analyzed using IBM SPSS Statistics 26 (IBM Corporation, Armonk, NY). Descriptive statistics were calculated for demographic variables and patient self-report measures. Data were checked for normality using Shapiro-Wilk tests. Spearman rank correlations were used to explore relationships between visual vertigo scales and the DHI. Independent-samples t-tests were used to compare DHI and HADS scores between patients with vestibular disorders who were VVAS positive and those who were not VVAS positive, and between patients classified as having low versus high visual vertigo severity. The internal consistency of the mVVAS was examined using Cronbach’s alpha.

To achieve the goal of constructing a measure that was internally consistent and concise, an exploratory factor analysis (EFA) was conducted using all 15 items of the mVVAS. To determine if there were a sufficient number of items that were significantly related, Bartlett’s test of sphericity and the Kaiser–Meyer–Olkin test were conducted. Then, the EFA was completed, and the scree plot was visualized to determine how many factors to retain. Also, the number of factors with Eigenvalues >1 were considered.

After we determined the number of factors to retain, we organized the items into concepts. We then included items with the highest factor loading from each concept and assessed Cronbach’s alpha. We repeated this analysis to determine the point at which the addition of more items did not result in a significant improvement in Cronbach’s alpha with the goal of constructing a short scale meeting the requirements of internal consistency and reliability. It was determined that a scale of six items satisfied these criteria (VVAS-6), including: walking over a patterned floor; walking through a supermarket aisle; going up an escalator; watching traffic at a busy intersection; being under flickering lights; and being a passenger in a car in the front seat. These six items, their associated conceptual domains, and factor loadings are presented in Table 2.

Table 2.

Proposed six-item visual vertigo analogue scale.

Item	Original/Newly added	Concept	Factor 1 loading
Walking over a patterned floor	Original	Walking with visual stimulation	0.899
Walking through a supermarket aisle	Original	Walking with visual stimulation	0.888
Going up an escalator	Newly added	Escalator	0.874
Watching traffic at a busy intersection	Original	Visual stimulation only	0.858
Being under flickering lights	Newly added	Visual stimulation only	0.848
Being a passenger in a car in the front seat	Newly added	Car	0.732

Once the EFA was completed, K-means clustering was used to divide the patients with vestibular disorders into two non-overlapping clusters utilizing the VVAS-6. The goal was to produce low-severity and high-severity groups with high similarity within groups and low similarity between groups.^17–19 We established a cut-off score between the maximum low-severity group score and minimum high-severity group score. To determine the stability of the clustering, the sample was randomly divided into two halves and the cluster analysis was repeated on each half.¹²

Results

The control participants (71 female, 34 male, 2 unreported gender) had a mean age of 42 (SD = 18) years. The patients with vestibular disorders (125 female, 72 male, 2 unreported gender) had a mean age of 54 (SD = 16) years. The most frequent diagnoses based on the primary 10th revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-10) code were abstracted from the electronic medical record and are listed in Table 1.

Patients with vestibular disorders reported an average DHI score of 38 (SD = 23). The patients with vestibular disorders who were VVAS positive (two or more items scored greater than zero; n = 179) reported significantly higher DHI scores (M = 40, SD = 22) than the patients with vestibular disorders who were not VVAS positive (n = 20; M = 13, SD = 8), (t (60) = 6.97, p < 0.001). Patients with vestibular disorders reported an average HADS anxiety score of 6 (SD = 4) and HADS depression score of 7 (SD = 4). Using a cut-off of 8 on the HADS anxiety and depression subscales for probable anxiety and depression,²⁰ 93 patients with vestibular disorders (47%) scored 8 or above on the depression subscale, 70 (35%) scored 8 or above on the anxiety subscale, and 53 (27%) scored 8 and above on both subscales in our study. Patients with vestibular disorders who were VVAS positive had more anxiety symptoms evidenced by HADS anxiety scores (M = 6.2, SD = 4.3) than patients with vestibular disorders without visual vertigo (M = 3.0, SD = 2.5), t (28) = 4.93, p < 0.001.

Amongst all participants, 179 (90%) patients with vestibular disorders and 60 (56%) controls were VVAS positive (two or more items scored greater than zero). Patients with vestibular disorders who were VVAS positive reported a mean severity score of 34 (SD = 25) on the VVAS and 37 (SD = 26) on the mVVAS. Controls who were VVAS positive reported a mean severity score of 5 (SD = 6) on the VVAS and 8 (SD = 10) on the mVVAS. Among participants classified as VVAS positive, the percentage endorsing each item and corresponding mean ratings for the VVAS and mVVAS are presented in Table 3. Because Table 3 includes only participants who met VVAS-positive criteria, it does not reflect the proportion of all patients or controls classified as VVAS positive, which is reported separately in the Results. The VVAS and the mVVAS were moderately correlated with the DHI (ρ = 0.63, p < 0.001 and ρ = 0.62, p < 0.001, respectively).

Table 3.

Item endorsement (%) and mean ratings among participants classified as Visual Vertigo Analogue Scale (VVAS) positive.

Item	Patients with vestibular disorders (n = 179)		Controls (n = 60)
Item	%	M (SD)	%	M (SD)
Walking through a supermarket aisle	94.4	3.5 (3.2)	100	0.2 (0.4)
Being a passenger in a car	94.4	3.8 (3.3)	98.3	1.5 (2.1)
Being under fluorescent lights	91.1	2.7 (3.2)	100	0.5 (1.23)
Watching traffic at a busy intersection	91.1	3.2 (3.2)	100	0.2 (0.7)
Walking through a shopping mall	92.7	3.6 (3.3)	100	0.4 (1.4)
Going down an escalator	83.2	3.9 (3.5)	98.3	0.6 (1.0)
Watching a movie at the movie theater	72.6	2.7 (3.3)	98.3	0.3 (0.5)
Walking over a patterned floor	91.1	3.7 (3.4)	98.3	0.7 (1.2)
Watching action television	91.6	2.8 (3.2)	100	0.2 (0.6)
Being a passenger in a car in the front seat^a	93.9	3.2 (3.1)	98.3	0.9 (1.7)
Being a passenger in a car in the back seat^a	80.4	4.8 (3.5)	96.7	1.8 (2.3)
Being under flickering lights^a	88.3	4.4 (3.6)	98.3	0.9 (1.7)
Walking in crowds^a	91.1	3.7 (3.2)	100	0.5 (1.3)
Walking in open spaces^a	93.3	2.6 (2.9)	98.3	0.2 (0.8)
Going up an escalator^a	82.1	3.6 (3.5)	100	1.4 (1.1)

^aNewly added items.

The Cronbach’s alpha coefficient was estimated at 0.94 for the VVAS and 0.96 for the mVVAS. Item-to-total correlation coefficients for both scales are included in Table 4. According to the inter-item correlation matrix of the original, nine-item VVAS, several items were highly correlated (r > 0.80).

Table 4.

Item-to-total correlation coefficients and Cronbach’s alpha index for the nine-item Visual Vertigo Analogue Scale (original) and the fifteen-item modified Visual Vertigo Analogue Scale.

Item	VVAS		mVVAS
Item	Correlation with total	Cronbach’s alpha if item deleted	Correlation with total	Cronbach’s alpha if item deleted
Walking through a supermarket aisle	0.84	0.93	0.83	0.95
Being a passenger in a car	0.57	0.94	0.62	0.96
Being under fluorescent lights	0.73	0.93	0.74	0.95
Watching traffic at a busy intersection	0.79	0.93	0.80	0.95
Walking through a shopping mall	0.85	0.92	0.84	0.95
Going down an escalator	0.78	0.93	0.80	0.95
Watching a movie at the movie theater	0.72	0.93	0.70	0.95
Walking over a patterned floor	0.88	0.92	0.87	0.95
Watching action television	0.71	0.93	0.69	0.95
Being a passenger in a car in the front seat	(Items are not part of the original VVAS.)		0.66	0.96
Being a passenger in a car in the back seat			0.67	0.96
Being under flickering lights			0.78	0.95
Walking in crowds			0.81	0.95
Walking in open spaces			0.68	0.96
Going up an escalator			0.81	0.95

VVAS, Visual Vertigo Analogue Scale; mVVAS, modified Visual Vertigo Analogue Scale.

We included all 15 items from the mVVAS in the EFA. Bartlett’s test of sphericity was significant (X² (105) = 3546.4, p < 0.001) and the KMO statistic approached 1 (KMO = 0.919), suggesting that there were a sufficient number of related items to conduct the EFA. There were 2 factors that had Eigenvalues >1. The three items that loaded on the second factor were related to riding in a car. Although these three items had strong factor loadings on factor 2, they loaded more strongly to factor 1. Therefore, only one factor was retained and the three items with strong loadings on both factors were included in factor 1. We then organized the items by strength of factor loadings and identified four concepts within the items (walking with visual stimulation, visual stimulation only, escalator, and car) (Table 5).

Table 5.

Items in the Visual Vertigo Analogue Scale (original items) and modified Visual Vertigo Analogue Scale (newly added items) organized by strength of factor loadings and concept (walking with visual stimulation, visual stimulation only, escalator, car).

Item number	Item	Original/Newly added	Concept	Factor 1 loading
8	Walking over a patterned floor	Original	Walking with visual stimulation	0.899
1	Walking through a supermarket aisle	Original	Walking with visual stimulation	0.888
5	Walking through a shopping mall	Original	Walking with visual stimulation	0.885
13	Walking in crowds	Newly added	Walking with visual stimulation	0.880
15	Going up an escalator	Newly added	Escalator	0.874
6	Going down an escalator	Original	Escalator	0.864
4	Watching traffic at a busy intersection	Original	Visual stimulation only	0.858
12	Being under flickering lights	Newly added	Visual stimulation only	0.848
3	Being under fluorescent lights	Original	Visual stimulation only	0.802
9	Watching action television	Original	Visual stimulation only	0.787
7	Watching a movie at the movie theater	Original	Visual stimulation only	0.784
14	Walking in open spaces	Newly added	Walking with visual stimulation	0.778
11	Being a passenger in a car in the back seat	Newly added	Car	0.753
10	Being a passenger in a car in the front seat	Newly added	Car	0.732
2	Being a passenger in a car	Original	Car	0.718

To construct a concise scale with good reliability while still maintaining the measurement of the four identified concepts, we assessed Cronbach’s alpha for the items with the highest loadings within each concept (items 8, 15, 4, 11). The Cronbach’s alpha was 0.90 for these four items. We then included the items with the second-highest factor loadings from each concept (items 1, 6, 10, 12) and Cronbach’s alpha was 0.94. This indicates excellent internal consistency and reliability. The addition of three more items resulted in a Cronbach’s alpha of 0.96 indicating possible redundancy. Also, there were several items among those 11 items that were highly correlated (r > 0.80). Therefore, only the first eight items were retained. Upon further review of the Cronbach’s alpha analysis, items 15 and 6 were highly correlated (r > 0.90) and item 11 would improve Cronbach’s alpha if it was deleted. The research team made the decision to remove item 6 because it had a lower factor loading than item 15 and to remove item 11 because it did not contribute to the internal consistency of the scale and because it was marked “not applicable” by 39 participants (patients with vestibular disorders and controls). The remaining six items (8, 15, 4, 1, 10, 12) had excellent internal consistency with Cronbach’s alpha of 0.93 and did not have any highly correlated items (r > 0.80). The proposed six-item scale (VVAS-6; Appendix) includes three items from the VVAS and three items from the mVVAS (Table 2).

K-means clustering was used to divide the patients with vestibular disorders into two, non-overlapping clusters based on VVAS-6 score. The first high-severity cluster (n = 83) had an average VVAS-6 score of 61 (SD = 16; range 37–100). The second low-severity cluster (n = 114) had an average VVAS-6 score of 12 (SD = 11; range 0–37). Using a cut-off score of 37, only one of the controls (with a VVAS-6 score of 49) had high VVAS severity.

The sample was randomly divided into two, and the cluster analysis was repeated to determine the stability of the clustering. For the first half, the first cluster (n = 47) had an average VVAS-6 score of 60 (SD = 24; range 37–100) and the second cluster (n = 53) had an average VVAS-6 score of 13 (SD = 16; range 0–36). For the second half, the first cluster (n = 36) had an average VVAS-6 score of 62 (SD = 26; range 39–96) and the second cluster (n = 61) had an average VVAS-6 score of 12 (SD = 14; range 0-37).

Based on a cut-off score of 37 on the VVAS-6, 83 (42%) patients with vestibular disorders had high severity of visual vertigo. The patients with vestibular disorders and high severity of visual vertigo (n = 83) reported an average DHI score of 52 (SD = 22), while the patients with vestibular disorders and low severity of visual vertigo (n = 114) reported an average DHI score of 27 (SD = 16). The difference in DHI scores between the high- and low-severity VVAS-6 groups was statistically significant (t (167) = 8.80, p < 0.001). The patients with vestibular disorders and high severity of visual vertigo (M = 5.8, SD = 4.3) had more anxiety symptoms than the patients with vestibular disorders and low severity of visual vertigo (M = 4.3, SD = 3.3). Patients with vestibular disorders in the high-severity VVAS-6 group had significantly higher HADS anxiety scores than those in the low-severity group (t (147) = 2.81, p = 0.006). Patients with vestibular disorders and high severity of visual vertigo (M = 8.6, SD = 4.1) also had more depressive symptoms than the patients with vestibular disorders and low severity of visual vertigo (M = 6.2, SD = 4.0). Patients with vestibular disorders in the high-severity VVAS-6 group also had significantly higher HADS depression scores than those in the low-severity group (t (174) = 4.08, p < 0.001). The VVAS-6 was moderately correlated with the DHI (ρ = 0.65, p < 0.001).

Applying each of the scales to describe the patients with vestibular disorders resulted in the following: VVAS (M = 30.2, Mdn = 24.0, range 0.0–98.7), mVVAS (M = 34.1, Mdn = 31.6, range 0.0–100.00), and VVAS-6 (M = 32.9, Mdn = 27.3, range 0.0–99.5). The VVAS was highly correlated with the mVVAS and the VVAS-6 (ρ = 0.90, p < 0.001, and ρ = 0.97, p < 0.001, respectively). Similarly, the mVVAS and the VVAS-6 were highly correlated (ρ = 0.92, p < 0.001).

Discussion

The purpose of this study was to improve the assessment of visually-induced symptoms by modifying the original, nine-item VVAS to produce a clinically more informative measure with an empirically derived severity cut-off. We propose a six-item scale (VVAS-6) with a cut-off score of 37 that allows clinicians to identify patients with vestibular disorders who have high severity of visual vertigo, addressing a key limitation of the original “VVAS positive” classification. By using the cluster analysis to separate patients with vestibular disorders into groups with low- and high-severity of visual vertigo, we were able to overcome a limitation of the original VVAS, which classifies most individuals with vestibular dysfunction as VVAS positive, calling into question the clinical utility of this designation. It is important to note that the goal of the present study was not to formally compare “positive” classifications between the original VVAS and the mVVAS, but rather to highlight limitations of binary classification and emphasize severity-based interpretation. The binary VVAS-positive designation has limited clinical utility, as it identifies the majority of patients with vestibular disorders as “positive” and does not differentiate severity. For this reason, our focus was on developing a clinically meaningful severity-based subgrouping approach rather than on comparing positive/negative classifications across versions of the scale. The VVAS-6 cut-off score provides a clearer distinction between low- and high-severity visual vertigo than a positive/negative classification and therefore offers greater value for clinical decision-making. While the original VVAS also provided a measure of severity, Dannenbaum et al.¹² did not provide a cut-off score for identifying patients with high severity of visual vertigo.

Additionally, we were able to establish a cut-off score of 37 for the VVAS-6. We performed split-sample validation and found that the cluster profiles of the sub-samples maintained the clustering pattern of the original, full-sample solution. This allows us to reliably conclude the existence of two meaningful subgroups within the sample (low-severity and high-severity groups). The use of a cut-off score of 37 for the VVAS-6 scale resulted in an improvement in correctly identifying patients with vestibular disorders and high severity of visual vertigo from controls who were healthy or seeking physical therapy for orthopedic disorders. Amongst all participants, 90% of patients with vestibular disorders and 56% of controls were VVAS positive using the criteria established by Dannenbaum et al. (two or more items scored greater than zero). Less than 1% of controls classified in the high-severity group using the cut-off score of 37. This severity-based cut-off score may help clinicians to accurately identify patients with vestibular disorders and visual vertigo so that they can prescribe more appropriate vestibular rehabilitation interventions including optokinetic and/or habituation-type exercises for individuals who score 37 and above.

Using the cut-off score of 37 on the VVAS-6, 42% of patients with vestibular disorders were identified as having high severity of visual vertigo. They reported a mean severity score of 61. Using the criteria for VVAS positive (two or more items scored greater than zero), 90% of patients with vestibular disorders and 56% controls were identified as having some degree of visually-induced dizziness; however, severity scores revealed marked differences between groups. Among patients with vestibular disorders who met VVAS-positive criteria, mean severity scores were 34 on the VVAS and 37 on the mVVAS, values consistent with the mean severity score of 30.3 reported in the vestibular group described by Dannenbaum et al.⁸ In contrast, patients with vestibular disorders classified in the high-severity group using the VVAS-6 reported substantially higher severity scores, indicating that reliance on the VVAS-positive classification alone might underestimate the severity of visual vertigo.

This apparent underestimation is likely due to differences in the psychometric structure of the scales. The original VVAS and the mVVAS include items that are less provocative, less frequently endorsed, or represent weaker contributors to visually-induced dizziness. When these lower-severity items are included in the mean score, they dilute the overall severity rating and compress differences between individuals with mild and more substantial visually-induced symptoms. In contrast, the VVAS-6 is composed exclusively of the highest-loading items across the constructs identified in our factor analysis, each of which strongly discriminates between individuals with low versus high symptom burden. By focusing the scale on the most sensitive and clinically meaningful visual–vestibular situations, the VVAS-6 amplifies the severity signal rather than averaging it with less provocative contexts. As a result of this refinement, the VVAS-6 produced a clearer separation between severity groups and demonstrated strong internal consistency without redundancy. This supports the use of the VVAS-6 as a concise tool that preserves the conceptual breadth of visually-induced dizziness while providing a more accurate reflection of symptom severity.

Patients with vestibular disorders and high severity of visual vertigo in our sample experienced substantial functional impact from their dizziness. Their mean DHI score was 52, indicating that they perceived a moderate level of handicap due to dizziness.²¹ Patients with visual vertigo are known to have higher DHI scores than those with vestibular dysfunction without visual vertigo and healthy controls.²² In our study, the patients with vestibular disorders in the high-severity group did report a higher average DHI score than those who were in the low-severity group, but we were unable to make comparisons to our controls as they did not complete the DHI.

Patients with vestibular disorders and high severity of visual vertigo had higher anxiety symptom scores than patients with vestibular disorders and low severity of visual vertigo. Patients with vestibular dysfunction and visual vertigo,²³ vestibular disorders,²⁴ and Meniere’s disease²⁵ are known to have increased levels of anxiety and depression. Depression scores have been correlated with increased reports of symptoms provoked or aggravated by complex visual stimuli.²³ The increased levels of depression reported by the patients with vestibular disorders in the high-severity group in this study may be due to the vestibular disorder, activity limitations, and participation restrictions.²⁶ Patients with vestibular disorders are known to have increased prevalence of anxiety.^23,27 As observed in this study, increased anxiety has been reported in patients with visual vertigo,²² chronic subjective dizziness,⁴ and persistent postural-perceptual dizziness.²⁸ Anxiety-related gaze instability has been proposed as one possible pathophysiologic mechanism for visual vertigo.²⁹

The VVAS was moderately correlated with the DHI in this study. Previous studies have noted a moderate correlation between the VVAS and the DHI, ranging from 0.54³⁰ to 0.67.⁸ The mVVAS and the VVAS-6 were also moderately correlated with the DHI. This may indicate that the DHI does not completely reflect the presence or severity of visually-induced dizziness and may provide preliminary evidence for the importance of using a separate tool (the VVAS-6) to measure these aspects of the patient’s vestibular dysfunction. Yip and Strupp found that the DHI did not correlate with tests of vestibular function or postural sway on posturography.³¹ They also concluded that the DHI and vestibular function tests plus posturography measure different aspects of the patient’s vestibular disorder.³¹ Accordingly, the DHI and the VVAS-6 should be viewed as complementary tools within a comprehensive vestibular assessment, each providing unique and clinically relevant information.

Both the VVAS and the mVVAS had a high level of internal consistency. The alpha coefficients estimated in this sample were slightly higher than that reported for the vestibular group (Cronbach’s α = 0.92) by Dannenbaum et al.⁸ Correlation coefficients showed high associations between individual items for both scales, indicating that some items may be redundant. Using EFA, we were able to reduce the mVVAS to six items. These items included three items from the original VVAS and three items from the mVVAS. Item-level endorsement patterns among VVAS-positive participants demonstrated that visually complex walking environments and dynamic visual–vestibular contexts were commonly provocative, supporting the construct validity of the retained VVAS-6 items. The item, “going down an escalator,” was added to the mVVAS to complement the original item, “going up an escalator.” Although both ascending and descending an escalator can provoke visually-induced dizziness, only the “going up an escalator” item demonstrated a sufficiently strong factor loading to retain in the final scale. The item, “being under flickering lights,” affirms the report that it is not the type of light (i.e., fluorescent), but the flickering of the lights that provokes the symptoms of patients with vestibular disorders. The factor analysis supported inclusion of a car-related item in the final scale, reflecting the clinical relevance of visually induced symptoms during vehicular motion. This refined scale maintains excellent internal consistency while providing clinicians with a practical, severity-based tool for identifying and stratifying visually induced dizziness.

A box was added to the left of each item that the patient can mark if an item could not be rated. Some individuals may not participate in certain activities based on personal preferences or may be unsure if an activity is bothersome (they have not had an opportunity to complete a certain activity since the onset of their dizziness). Other patients may be unable to participate in certain activities due to their dizziness and/or desire to avoid certain activities known to aggravate their dizziness. The addition of this box may help clinicians better understand the patient’s activity limitations and participation restrictions by cueing the clinician to further explore why a patient does not perform certain activities. If left blank, however, there is still uncertainty as to why a particular scale was not marked off. Although the checkbox may help identify items that were intentionally left unrated, it provides limited information regarding the reason an item was not scored. Specifically, it does not differentiate between non-applicability, uncertainty, or avoidance due to symptom provocation, which remains a limitation of the VVAS, mVVAS, and VVAS-6.

Our sample differed from that originally described by Dannenbaum and colleagues because we included individuals diagnosed with benign paroxysmal positional vertigo, bilateral vestibular hypofunction, and central vestibular disorders. This may have affected the variability of VVAS scores in this sample but likely represents the population of patients treated for dizziness in tertiary care settings. Future studies should explore the test-retest reliability, convergent validity, and responsiveness of the newly proposed VVAS-6.

Conclusions

The VVAS-6 is a short and internally consistent measure of visually-induced dizziness that can be used in clinical settings. A cut-off score of 37 is recommended for identifying patients with vestibular disorders and high severity of visual vertigo. Future studies should explore its reliability, validity, and responsiveness when administered as a stand-alone test.

Footnotes

Acknowledgements

The authors give special thanks to Jehan Alomar, MS, PT, Suchita Rajan, MS, PT, and Anita Lieb, MS for their help on this project.

ORCID iDs

Carrie W. Hoppes

Pamela M. Dunlap

Patrick J. Sparto

Kefah M. Alshebber

Brooke N. Klatt

Susan L. Whitney

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Partial support for this project was funded by a grant (P30 AG024827) from the Pittsburgh Claude D. Pepper Older Americans Independence Center, and the Eye and Ear Foundation of Pittsburgh. Participant recruitment was facilitated by the Clinical and Translational Science Institute, supported by the National Institutes of Health through Grant Number UL1TR001857.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Disclaimer

The views and information expressed are those of the authors and do not represent the official position of the 59th Medical Wing, Defense Health Agency, Department of the Air Force, Department of War, or U.S. Government.

Appendix

References

Bronstein

. Visual vertigo syndrome: clinical and posturography findings. J Neurol Neurosurg Psychiatry 1995; 59: 472–476.

Jacob

Woody

Clark

, et al. Discomfort with space and motion: a possible marker of vestibular dysfunction assessed by the situational characteristics questionnaire. J Psychopathol Behav Assess 1993; 15: 299–324.

Brandt

. Phobic postural vertigo. Neurology 1996; 46: 1515–1519.

Staab

. Chronic subjective dizziness. Continuum 2012; 18: 1118–1141.

Staab

Ruckenstein

. Expanding the differential diagnosis of chronic dizziness. Arch Otolaryngol Head Neck Surg 2007; 133: 170–176.

Bronstein

. The visual vertigo syndrome. Acta Otolaryngol Suppl 1995; 115: 45–48.

Staab

Eckhardt-Henn

Horii

, et al. Diagnostic criteria for persistent postural-perceptual dizziness (PPPD): consensus document of the committee for the classification of vestibular disorders of the Bárány Society. J Vestib Res 2017; 27: 191–208.

Dannenbaum

Chilingaryan

Fung

. Visual vertigo analogue scale: an assessment questionnaire for visual vertigo. J Vestib Res 2011; 21: 153–159.

Longridge

Mallinson

. Visual vestibular mismatch in work-related vestibular injury. Otol Neurotol 2005; 26: 691–694.

10.

Jacob

Lilienfeld

Furman

, et al. Panic disorder with vestibular dysfunction: further clinical observations and description of space and motion phobic stimuli. J Anxiety Disord 1989; 3: 117–130.

11.

Dannenbaum

Chilingarian

Fung

. Validity and responsiveness of the visual vertigo analogue scale. J Neurol Phys Ther 2019; 43: 117–121.

12.

Clatworthy

Buick

Hankins

, et al. The use and reporting of cluster analysis in health psychology: a review. Br J Health Psychol 2005; 10: 329–358.

13.

Hellerstein

Scheiman

Fishman

, et al. Visual rehabilitation for patients with brain injury. In: Scheiman

(ed). Understanding and Managing Vision Deficits: A Guide for Occupational Therapists. Slack Incorporated, 2011, p. 210.

14.

Bjelland

Dahl

Haug

, et al. The validity of the hospital anxiety and depression scale. J Psychosom Res 2002; 52: 69–77.

15.

Jacobson

Newman

. The development of the dizziness handicap inventory. Arch Otolaryngol Head Neck Surg 1990; 116: 424–427.

16.

Zigmond

Snaith

. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983; 67: 361–370.

17.

Morissette

Chartier

. The k-means clustering technique: general considerations and implementation in Mathematica. Tutor Quant Methods Psychol 2013; 9: 15–24.

18.

Pérez-Ortega

Almanza-Ortega

Romero

. Balancing effort and benefit of K-means clustering algorithms in Big Data realms. PLoS One 2018; 13: e0201874.

19.

Peterson

Ghosh

Maitra

. Merging K‐means with hierarchical clustering for identifying general‐shaped groups. Stat 2018; 7: e172.

20.

Olssøn

Mykletun

Dahl

. The hospital anxiety and depression rating scale: a cross-sectional study of psychometrics and case finding abilities in general practice. BMC Psychiatry 2005; 5: 46.

21.

Whitney

Wrisley

Brown

, et al.

Is perception of handicap related to functional performance in persons with vestibular dysfunction?

Otol Neurotol 2004; 25: 139–143.

22.

Zur

Schoen

Dickstein

, et al. Anxiety among individuals with visual vertigo and vestibulopathy. Disabil Rehabil 2015; 37: 2197–2202.

23.

Pavlou

Davies

Bronstein

. The assessment of increased sensitivity to visual stimuli in patients with chronic dizziness. J Vestib Res 2006; 16: 223–231.

24.

Grunfeld

Gresty

Bronstein

, et al. Screening for depression among neuro-otology patients with and without identifiable vestibular lesions. Int J Audiol 2003; 42: 161–165.

25.

Kirby

Yardley

. Understanding psychological distress in Ménière's disease: a systematic review. Psychol Health Med 2008; 13: 257–273.

26.

Staab

Ruckenstein

. Which comes first? Psychogenic dizziness versus otogenic anxiety. Laryngoscope 2003; 113: 1714–1718.

27.

Guerraz

Yardley

Bertholon

, et al. Visual vertigo: symptom assessment, spatial orientation and postural control. Brain 2001; 124: 1646–1656.

28.

Yan

Cui

, et al. Analysis of the characteristics of persistent postural-perceptual dizziness: a clinical-based study in China. Int J Audiol 2017; 56: 33–37.

29.

Staab

. The influence of anxiety on ocular motor control and gaze. Curr Opin Neurol 2014; 27: 118–124.

30.

Grigol

TAd.Ae.S

Silva

Ferreira

, et al. Dizziness handicap inventory and visual vertigo analog scale in vestibular dysfunction. International Archives of Otorhinolaryngology, 2015.

31.

Yip

Strupp

. The dizziness handicap inventory does not correlate with vestibular function tests: a prospective study. J Neurol 2018; 265: 1210–1218.