As mentioned, fall risk in older adults may comprise both physical and psychological factors

Introduction

Trauma and mortality due to fall accidents in older adults are increasingly serious medical and socioeconomic issues in developed countries with rapidly aging populations. In Japan, over 8,800 older adults died from fall accidents in 2018. Among older adults (≥75 years old), mortality doubles as the population age increases by 5 years, as reported by the Consumer Affairs Agency of the Japanese Government. Multiple factors affect fall risk, including static and dynamic postural balance, visual inputs, sensory inputs, hearing loss, cognitive decline and other psychological distress, medications for conditions such as cardiovascular diseases, and living environment ^{1, 2}. For dizzy older adults presenting to otolaryngology clinics, the ability to predict fall risk is important. As mentioned, fall risk in older adults may comprise both physical and psychological factors.

Among dizzy older adults visiting our otolaryngology clinic, static postural balance is assessed by posturography and the one-leg standing test (OLS). Static postural balance in older adults is generally impaired due to reductions in vestibular, visual, and sensory inputs and muscle weakness ³. Postural balance as measured by posturography reportedly provides a significant predictor of fall risk in healthy, non-institutionalized older individuals ⁴.

Subjective dizziness-associated handicap can be quantitatively evaluated using the Dizziness Handicap Inventory (DHI) questionnaire. The DHI is the most widely used self- reported measure for dizzy patients and consistently evaluates patient handicaps in daily life

using total scores ranging from 0 to 100 ^{5, 6}. As similar quantitative measures, general anxiety and depression can be assessed by scores ranging from 0 to 21 using the Hospital Anxiety and Depression Scale (HADS) questionnaire. HADS is widely used as a measurement of the severity of emotional disorder in outpatient medical settings and is applicable for dizzy patients ^7-9. Older adults may be vulnerable to mental health problems such as depression, anxiety, and dementia ¹⁰. Subjective handicaps associated with dizziness/vertigo are thus thought to be affected by complex psychological factors in older individuals.

Different opinions on the concordance of objective data and subjective symptoms associated with dizziness and vertigo have been reported. For example, Menant et al. found that postural sway with eyes closed and a positive Dix Hallpike test were significant predictors of moderate/severe dizziness handicap as assessed by DHI in adults (≥50 years old) ¹¹. In contrast, Yip et al. and Chiarovano reported that postural sway in posturography did not correlate with DHI score ^{12, 13}. In a study of 34 older adults, practically no correlation was identified between DHI score and static postural balance as assessed by computerized dynamic posturography¹⁴.

Given these conceptions, there is a paucity of published data analyzing how both objective and subjective assessments of balance can predict fall risk in older adults visiting otolaryngology clinics. This study reviewed the medical charts of dizzy patients who visited our otolaryngology clinic, and examined how objective data (postural balance by

posturography and OLS, spontaneous nystagmus and hearing level), as well as subjective symptoms (self-perceived handicap as assessed by DHI and HADS) affect fall risk.

Materials and methods

Study participants

This is a cross-sectional study. The medical charts in our Otolaryngology Department between November 2018 and December 2019 were retrospectively reviewed. Eligibility criteria were outpatients (female or male) with dizziness or vertigo and ≥65 years old. The study protocol was approved by the institutional ethics committee at Okayama University Hospital (protocol number, 2001-041; Principal investigator, Y. M.) and was conducted in compliance with the Declaration of Helsinki. Informed consent was obtained as the form of opt-out on the web-site.

Collection of patient data in the clinic

All the subjects were outpatients complaining of dizziness/vertigo with or without hearing loss. They represented the older adult patients who routinely visited our otolaryngology clinic with chief complaints of dizziness/vertigo or hearing loss. During a session of consultation in the clinic, the attending physician asked the patient if they feel dizzy or vertiginous in daily

life. It was verified that all the subjects included in this study were experiencing dizziness/vertigo. If the answer was “yes”, physical examinations comprising posturography, OLS, and spontaneous nystagmus, as well as assessments of self-perceived symptoms (DHI and HADS) were performed to provide appropriate medical care for dizziness/vertigo symptoms. Other information from patients was collected using the in-house interview sheet at our clinic. This interview sheet included the following questions: “Have you experienced falls during the last year?” (answer selected from: no; once; twice; three times; four times;

five times or more); “Do you exercise in a standing position for more than 40 min/day?”; “Are you taking medications for any of the following: diabetes mellitus, hypertension, arrhythmia, or hyperlipidemia?”; “Have you ever experienced any of the following conditions: cerebral infarction; myocardial infarction; arrhythmia; or malignant neoplasm?”; and “Do you regularly see a doctor for any other diseases?”.

Posturography

Static postural balance was measured by posturography (GS-7; Anima, Tokyo, Japan) which consisted of a force plate and an operation board. Subjects were asked to stand on a solid plate of the posturography on the floor with the feet close together without shoes. The subjects were requested to align their feet along footprint outlines (the shapes of the right and left footprints) drawn on the plate put on the floor. The subjects stood upright while looking

straight ahead at a target 2 m away, with arms to the sides. Postural sway was measured for 1 min with eyes open, followed by another 1 min with eyes closed. Areas of center-of-gravity sway with eyes open and closed were used to quantitatively evaluate static postural balance as postural sway area with eyes open (PSAEO) and postural sway area with eyes closed (PSAEC).

OLS

This test measures the duration (in s) for which the subject can stand on one leg without support. The patient was asked to stand on the right or left leg, whichever they considered easier, with eyes open. The OLS is a widely used clinical test to quantitatively evaluate postural balance in a static position for older individuals ¹⁵.

Nystagmus examination

Spontaneous nystagmus was observed in both sitting and supine positions in the dark using an infrared charge-coupled device (CCD) camera. Examination was also performed with Dix-Hallpike maneuver. We assessed whether any nystagmus suggestive of vestibular dysfunction was observed with the patient blindfolded using the CCD camera. Gaze nystagmus was examined without CCD camera, but no gaze nystagmus was observed in any of our patients.

Pure tone audiometry

As vestibulocochlear diseases are primary differential diagnoses in patients encountered at our otolaryngology clinic, pure tone audiograms were obtained from all patients. Hearing level was calculated as the average of air-conduction thresholds at 6 frequencies (250, 500, 1000, 2000, 4000, and 8000 Hz). Mean values in the right and left ears were designated as mean hearing level (MHL) for each patient in this paper.

Evaluation of subjective symptoms by DHI and HADS

On the same day as the physical examinations, self-perceived symptoms of the patient were evaluated using questionnaires of DHI and HADS. DHI is a 25-item questionnaire to evaluate dizziness-associated handicaps in daily life and is widely used to quantitatively assess self-perceived outcomes of dizziness/vertigo ^{5, 6}. DHI comprised 3 domains of functional, emotional, and physical aspects of dizziness and unsteadiness, with total DHI score ranging from 0 (no handicap) to 100 (most severe handicap). HADS is a 14-item self- reported questionnaire used to screen for anxiety and depression in outpatient clinics ^{7, 8}. This tool comprises 7 items eachfor anxiety and depression, with scores ranging from 0 (no anxiety or depression) to 21 (most severe anxiety or depression). For both anxiety and depression scores, a cut-off value of 8 was used to differentiate subjects with anxiety or

depression (HADS scores 8–21) from subjects without either condition (HADS scores 0–7).

Data analysis

First, to examine whether objective data on balance and hearing correlate with subjective symptoms, correlations of the continuous variables of PSAEO, PSAEC, OLS, MHL with total DHI score, HADS anxiety score, and HADS depression score were calculated using Pearson’s correlation coefficient.

Second, outcome for fall risk (experience of falls during the preceding year) was estimated by multiple logistic regression modelling with predictive variables of objective data (PSAEC, OLS, nystagmus, and MHL) and subjective symptoms (self-perceived total DHI score, HADS anxiety score, and HADS depression score) at the same time. The model was also adjusted for sex and presence of history of cardiovascular diseases (myocardial infarction and arrhythmia), as these conditions reportedly affect fall risk in older adults ¹⁶. In addition, stratified analysis was conducted, and the odds ratio for falls was also estimated to evaluate effect modification by the presence of anxiety or depression.

Third, as regression analyses in the second step suggested that nystagmus and PSAEC may predict fall risk, the cut-off value for PSAEC was determined by receiver operating characteristic (ROC) analyses to identify patients who experienced falls among those with nystagmus.

All statistical analyses were performed using SPSS version 24 (SPSS, Chicago, IL) and STATA/SE version 16 (Stata Corp, College Station, TX) software.

Results

Seventy-six older adults (mean age, 74.2±6.3 years; 67.1% female) with medical records of physical examinations, self-perceived symptoms, and in-house interview sheets were identified during the study period (Table 1). Slightly over one-third (34.2%) of these older adults presented with complaints of nonspecific dizziness/vertigo with bilateral sensorineural hearing loss (Table 2). Twenty subjects were fallers during the preceding year, and 49 were non-fallers. Data on fall incidence were missing for 7 patients, because these patients omitted answers regarding falls in the interview sheet. No sex bias was seen in age, PSAEO, PSAEC, OLS, MHL, DHI functional, DHI emotional, DHI physical, DHI total, HADS anxiety, or HADS depression scores (P>0.05 by t-test), or in incidence of falls or history of cardiovascular diseases (P>0.05 by Pearson’s chi-squared test and Fisher’s exact test). Nystagmus was more frequent in males (P<0.01 by Pearson’s chi-squared test and Fisher’s exact test).

PSAEO and PSAEC were within the range of 0–50 cm²and severity of PSAEC and PSAEO was expressed as the value divided by 10 (range, 0–5), and analyzed by regression modelling.

Correlations between objective data and subjective symptoms

Within variables of static postural balance data, PSAEC correlated significantly with PSAEO (r=0.63; P<0.01) (Table 3). PSAEC showed a weak but significant correlation with OLS (r=-0.24; P<0.05). Similarly, within variables of subjective symptoms, total DHI score correlated significantly with HADS anxiety score (r=0.38; P<0.01). However, no objective data on postural balance (PSAEO, PSAEC, or OLS) showed significant correlations with any subjective symptoms (total DHI score, HADS anxiety score, and HADS depression score;

P>0.05 each). Interestingly, worse MHL correlated significantly with poor OLS (r=-0.34;

P<0.01), whereas MHL showed no correlation with other metrics.

Model for predicting fall risk

Logistic regression analysis showed that objective data for PSAEC significantly predicted fall risk (P<0.05; 4.9 [1.4–16.8] per 10-cm² increment) (Table 4; chi-squared test for model:

P=0.015). Nystagmus tended to be associated with fall risk with a high odds ratio (P=0.088;

4.1 [0.8–20.4]), whereas MHL showed no association with falls (P=0.97; 1.0 [0.97–1.03]). In contrast, subjective symptoms of total DHI score provided no predictive value for fall risk (P=0.43; 1.0 [0.9–1.03]). Higher HADS depression score was negatively associated with lower fall risk (P<0.05; 0.8 [0.6–0.99]).

In stratified analysis, even though confidence intervals (CIs) were too wide and not significant in patients with anxiety or depression (HADS anxiety or depression score ≥8), the

odds ratio associating PSAEC with fall risk was 133.1 [0.4–50,043.9]. In contrast, the corresponding odds ratio was 2.8 [0.6–12.8] in those without anxiety or depression (HADS anxiety and depression scores both 7).

Finally by ROC analysis, a cut-off value for postural balance (PSAEC score) of >6.1 cm² was proposed to predict fall in our patients with nystagmus, with 78% sensitivity and 67%

specificity (area under the curve 0.74; 95% CI 0.48–0.997). Subjective symptoms of total DHI score did not differ significantly between the 9 fallers and 12 non-fallers with nystagmus (P=0.59 by t-test; mean ± standard deviation, 27.8±19.3 vs 33.0±22.8).

Discussion

In general, mean self-perceived dizziness handicap for DHI score in healthy older adults (70–95 years old) is 5-6 ¹⁷. In our older adults who visited the otolaryngology clinic, mean DHI score was 20–21, and patients included in this study typically experienced mild dizziness/vertigo.

According to our data, objective data for postural balance, as represented by postural sway with eyes closed, did not correlate with subjective symptoms associated with dizziness in older adults. On the other hand, logistic regression modelling showed that poor postural balance (postural sway with eyes closed) significantly predicted fall risk. Nystagmus also showed a strong association with fall risk, even if the association was not significant. Self-

perceived symptoms of total DHI score, however, were not associated with higher fall risk.

Previously, scarcely any analyses have been published regarding the relationship between DHI score and fall risk, although one paper by Criter et al. reported that low DHI score may be effective for identifying older adults without a history of recent falls ¹⁸. Based on our prediction model, self-perceived DHI score is at most an inadequate tool to predict fall risk when we considered objective data at the same time in older adults at otolaryngology clinics.

We assume that this is because the subjective DHI score in older adults is affected by complex psychological factors such as anxiety and depression, and does not effectively reflect the objective data of postural balance. Poor postural balance is the primary factor to predict increased fall risk. In our dataset of older adults, DHI score correlated with the anxiety and depression scores assessed by the HADS questionnaire.

Regarding the more general mental statuses of anxiety and depression, high HADS depression score was significantly associated with reduced fall risk. We assumed that patients with high HADS depression score were prone to participating in almost no exercise in daily life, and thus experienced a reduced incidence of fall. In other words, this association could be explained by reverse causation. Supporting this, HADS depression score was higher in patients who had no exercise in daily life than in those who did (P=0.051 by t-test; mean ± SD, 7.6±4.9 vs 5.7±3.1). This data on HADS represents new information showing that HADS score was significantly associated with fall risk for older adults in otolaryngology clinics.

To efficiently predict fall risk in older adults visiting otolaryngology clinics, clinicians need to consider mutual effects between objective data and subjective symptoms of balance. From our data, odds ratio for postural balance (postural sway with eyes closed) for fall risk was markedly higher among older adults with anxiety and depression (odds ratio, 133.1) than in older adults without these psychological conditions (odds ratio, 2.8), even if those odds ratios were not significant. As tentative guidance based on the ROC analysis, we propose that the high fall risk patients with nystagmus and postural sway area with eyes closed >6.1 cm² warrant particular caution and rehabilitation to avoid fall accidents. It has been reported that the mean values of this parameter in healthy subjects (65-69 years old) are 4.9 in females and 5.8 in males ¹⁹.

Most patients in the otolaryngology clinic had cochleovestibular diseases and concurrent sensorineural hearing loss. Hearing loss may be associated with significantly increased odds

ratios of falls in older adults ². In our data, however, mean hearing level of older adults exerted no significant influence on fall risk. A significant finding for mean hearing level was the correlation with OLS; that is, more severe hearing loss was associated with worse results for the OLS. Lee et al. reported that age-related hearing loss correlated significantly with increased risk of sarcopenia (reduced muscle mass in the appendicular skeleton) ²⁰.

A subject with a unilateral hearing loss vs bilateral hearing loss most likely has different etiology and symptoms. For instance, bilateral hearing loss subjects might have better

directional cues than single sided hearing loss subjects ²¹. In our dataset, no difference in the fall risk was observed between the patients with asymmetric hearing loss (hearing loss with

≥15 dB difference between the right and left ears), and those with symmetric hearing loss (P>0.05 by Pearson’s chi-squared test and Fisher’s exact test) (data not tabulated). It needs to be further studied how bilateral hearing loss or unilateral hearing loss affects fall risk in older adults.

The limitation of this study is the relatively small number of subjects included in the study, and the diversity of these subjects recruited in this study. It might be useful to group the subjects into acute and chronic dizziness/vertigo, as these two generally have different diagnosis/prognosis.A larger-scale study is also needed to analyze the data comparing those with multiple falls vs one fall vs no falls. Another limitation was the collection of data on fall incidents by anamnesis, which depends on patient recall. In the interview sheet used at our clinic, it was confirmed from the patients how many times they experienced falls during the last year, and a subject was considered having experienced falls if at least one fall incident was reported by the subject. Ideally, it is more appropriate as a study design to prospectively collect the data of patients who have not experienced any falls yet. To this end, a larger-scale prospective study of fall risk in otolaryngology practices is warranted. Both objective data and subjective symptoms need to be evaluated in such future studies.

In conclusion, poor postural balance is associated with increased fall risk among older

adults at otolaryngology clinics. Conversely, clinicians should keep in mind that the self- perceived dizziness handicap of DHI score is an insufficient tool to evaluate fall risk.

Acknowledgments None.

Declaration of Conflicting Interests

The authors declare that there is no conflict of interest.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

1. Panel on Prevention of Falls in Older Persons AGS, British Geriatrics S. Summary of the Updated American Geriatrics Society/British Geriatrics Society clinical practice guideline for prevention of falls in older persons. J Am Geriatr Soc. Jan 2011;59(1):148-57.

doi:10.1111/j.1532-5415.2010.03234.x

2. Jiam NT, Li C, Agrawal Y. Hearing loss and falls: A systematic review and meta-

analysis. Laryngoscope. Nov 2016;126(11):2587-2596. doi:10.1002/lary.25927 3. Anacker SL, Di Fabio RP. Influence of sensory inputs on standing balance in community-dwelling elders with a recent history of falling. Phys Ther. Aug 1992;72(8):575- 81; discussion 581-4. doi:10.1093/ptj/72.8.575

4. Buatois S, Gueguen R, Gauchard GC, Benetos A, Perrin PP. Posturography and risk of recurrent falls in healthy non-institutionalized persons aged over 65. Gerontology.

2006;52(6):345-52. doi:10.1159/000094983

5. Jacobson GP, Newman CW. The development of the Dizziness Handicap Inventory.

Arch Otolaryngol Head Neck Surg. Apr 1990;116(4):424-7.

doi:10.1001/archotol.1990.01870040046011

6. Mutlu B, Serbetcioglu B. Discussion of the dizziness handicap inventory. J Vestib Res. 2013;23(6):271-7. doi:10.3233/VES-130488

7. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand. Jun 1983;67(6):361-70. doi:10.1111/j.1600-0447.1983.tb09716.x

8. Zigmond AS, Snaith RP. Hospital Anxiety and Depression Scale (Translated into Japanese by Toshinori Kitamura). Seishinkashindangaku (Japanese). 1993;4:371-372.

9. Piker EG, Kaylie DM, Garrison D, Tucci DL. Hospital Anxiety and Depression Scale:

Factor Structure, Internal Consistency and Convergent Validity in Patients with Dizziness.

Audiol Neurootol. 2015;20(6):394-9. doi:10.1159/000438740

10. Williamson V, Stevelink SAM, Greenberg K, Greenberg N. Prevalence of Mental Health Disorders in Elderly U.S. Military Veterans: A Meta-Analysis and Systematic Review.

Am J Geriatr Psychiatry. May 2018;26(5):534-545. doi:10.1016/j.jagp.2017.11.001 11. Menant JC, Meinrath D, Sturnieks DL, et al. Identifying Key Risk Factors for Dizziness Handicap in Middle-Aged and Older People. J Am Med Dir Assoc. Mar 2020;21(3):344-350 e2. doi:10.1016/j.jamda.2019.08.016

12. Yip CW, Strupp M. The Dizziness Handicap Inventory does not correlate with vestibular function tests: a prospective study. J Neurol. May 2018;265(5):1210-1218.

doi:10.1007/s00415-018-8834-7

13. Chiarovano E, Wang W, Reynolds P, MacDougall HG. Imbalance: Objective measures versus subjective self-report in clinical practice. Gait Posture. Jan 2018;59:217- 221. doi:10.1016/j.gaitpost.2017.10.019

14. Rossi-Izquierdo M, Santos-Perez S, Del-Rio-Valeiras M, et al. Is there a relationship between objective and subjective assessment of balance in elderly patients with instability?

Eur Arch Otorhinolaryngol. Sep 2015;272(9):2201-6. doi:10.1007/s00405-014-3122-3 15. Michikawa T, Nishiwaki Y, Takebayashi T, Toyama Y. One-leg standing test for elderly populations. J Orthop Sci. Sep 2009;14(5):675-85. doi:10.1007/s00776-009-1371-6 16. Wiseman T, Betihavas V. The association between unexplained falls and cardiac arrhythmias: A scoping literature review. Aust Crit Care. Sep 2019;32(5):434-441.

doi:10.1016/j.aucc.2018.08.003

17. Davalos-Bichara M, Agrawal Y. Normative results of healthy older adults on standard clinical vestibular tests. Otol Neurotol. Feb 2014;35(2):297-300.

doi:10.1097/MAO.0b013e3182a09ca8

18. Criter RE, Honaker JA. Fall risk screening protocol for older hearing clinic patients.

Int J Audiol. Oct 2017;56(10):767-774. doi:10.1080/14992027.2017.1329555

19. Imaoka K, Murase H, Fukuhara M. Collection of data for healthy subjects in stabilometry. Equilibrium Res Suppl (Japanese). 1997;12:1-84.

20. Lee J, Han K, Song JJ, Im GJ, Chae SW. Sarcopenia and Hearing Loss in Older Koreans: Findings from the Korea National Health and Nutrition Examination Survey (KNHANES) 2010. PLoS One. 2016;11(3):e0150281. doi:10.1371/journal.pone.0150281 21. May BJ, Bowditch S, Liu Y, Eisen M, Niparko JK. Mitigation of informational masking in individuals with single-sided deafness by integrated bone conduction hearing aids. Ear Hear. Jan-Feb 2014;35(1):41-8. doi:10.1097/AUD.0b013e31829d14e8

Legends

Table 1. Clinical characteristics of patients.

Data for the 76 patients (67.1% female) enrolled in this study.

Table 2. Diagnoses of patients.

Diagnoses of the 76 patients enrolled in this study.

Table 3. Correlations between objective data parameters and subjective symptoms associated with dizziness.

Postural balance, represented by postural sway with eyes closed, did not correlate with self- perceived dizziness symptoms (DHI and HADS score). Worse mean hearing level correlated significantly with poor OLS (r=-0.34; P<0.01).

Table 4. Prediction model for fall risk.

Using a logistic regression model adjusted for sex and cardiovascular diseases (chi-square

test for the model: P=0.015), objective data for postural sway area (with eyes closed) significantly predict fall risk. Nystagmus tends to be associated with fall risk with a high odds ratio. In contrast, subjective symptoms represented by total DHI score do not predict fall risk.

Higher HADS depression score was negatively associated with lower fall risk.