Development and Validation of a Brief Fall Questionnaire (FALL‐HD) for Patients with Huntington's Disease

ABSTRACT

Background

Huntington's disease (HD) is an autosomal dominant neurodegenerative disorder caused by a mutation in the huntingtin gene on chromosome 4, leading to progressive cognitive decline, motor impairment, and functional disability. Falls represent the leading cause of nursing home placement in HD. Despite the significant clinical implications, validated instruments for fall assessment, including fall frequency, are lacking in HD.

Objectives

To construct and evaluate the psychometric properties of a questionnaire for assessing fall status, including fall frequency, in ambulatory patients with HD.

Methods

We developed a concise 3‐item questionnaire based on a comprehensive literature review and initial pilot study. We evaluated the responses of 107 HD gene‐positive participants in the main study. Reliability (test–retest, internal consistency) and validity (criterion, convergent, divergent) were assessed.

Results

The FALL‐HD Questionnaire demonstrated reliability with a test–retest correlation coefficient of r = 0.85 and an internal consistency (Cronbach's alpha) of 0.71. Convergent validity was established using Spearman correlations with the UHDRS Total Motor Score (r = 0.75, p < 0.001), Total Maximal Chorea Score (r = 0.65, p < 0.001), and Total Functional Capacity (r = 0.53, p < 0.001), in addition to Total Body Sway (TBS, r = 0.62, p < 0.001), a measure of postural stability from the BTrackS™ Balance System. Using ROC curves, the FALL‐HD outperformed (AUC = 0.90) existing measures of balance such as the Total Body Sway (AUC = 0.85), Chair Sit‐to‐Stand Test (AUC = 0.65) and Timed Up and Go (AUC = 0.58).

Conclusions

The FALL‐HD is a brief, valid and reliable, screening instrument of fall status, including fall frequency, in ambulatory patients with HD.

Huntington's disease (HD) is an autosomal dominant neurodegenerative disorder caused by a mutation in the huntingtin gene on chromosome 4, leading to progressive cognitive decline, motor impairment, and functional disability. ¹ Balance impairment begins early in HD and progresses as the disease worsens. ² Contributing factors include bradykinesia, ataxia, chorea, as well as cognitive and behavioral disturbances. This leads to falls in these patients, which substantially diminishes quality of life and represents the leading cause of nursing home placement. Despite the significant clinical implications, validated instruments for fall assessment, including fall frequency, are lacking in HD. ³

In clinical research on neurodegenerative diseases, falls have been captured using several methods including fall diaries ⁴ , ⁵ questionnaires such as the Falls Efficacy Scale (FES) ⁶ recall surveys, ⁷ wearable and sensor‐based technologies, ⁸ documentation in medical records, and even video and environmental monitoring. ⁹ Fall diaries or wearable and sensor‐based technologies require substantial commitment from both patients and caregivers and can only be used prospectively ⁷ , ¹⁰ It has been shown that more progressed neurodegenerative patients, who are more likely to fall, have a lower rate of compliance with maintaining a fall diary. ² , ¹⁰

In the absence of a validated tool for tracking falls in HD, clinical researchers have focused on clinical and observational assessments of mobility, gait and balance to evaluate fall risk. ¹¹ A standardized assessment such as the Unified Huntington's Disease Rating Scale (UHDRS) Total motor score (TMS) provides only an estimate of a patient's balance issues. These scales do not record actual falls or their outcomes and require considerable expertise to administer. For example, a cutoff score of ≥14 s on the Timed Up and Go (TUG) and ≤40 s on the Berg Balance Scale (BBS) is associated with increased fall risk in HD. ³ On the Tinetti Mobility Test (TMT), ¹² , ¹³ , ¹⁴ a score of <21 identified fallers with 74% sensitivity and 60% specificity. ¹² Although physical balance measures might be useful in predicting fall risk, performing these assessments can be difficult in participants needing ambulatory support; additionally, their low sensitivity and specificity make them less advantageous.

Thus, the purpose of this study was to develop and examine the psychometric properties of a new measure of fall status, including fall frequency, in HD. Our priority was a brief, yet reliable and validated instrument that was easy to administer.

METHODS

Participants

Participants were recruited from the Huntington's Disease Clinical Research Center (HDCRC) at the University of California, San Diego (UCSD). All participants were provided written and informed consent in accordance with the Institutional Review Board (IRB) and all procedures were approved by the University of California San Diego.

The FALL‐HD and UHDRS [including TMS, Total Maximal Chorea (TMC) Score, and Total Functional Capacity (TFC)] were administered to all participants, in addition to the Mini‐Mental State Examination (MMSE), Symbol Digit Modality Test (SDMT), Stroop Word Reading (SWR), TUG, Total Body Sway (TBS) from the BTrackS™ Balance Plate, and the CST. Based on the UHDRS TMS, all HD participants received a diagnostic confidence level (DCL) of 4, meaning that they exhibited indisputable symptoms of HD, while all pre‐manifest HD (PM) participants received a DCL of <4.

Statistical Analyses

Analyses were conducted with SPSS Statistics for Macintosh. Since data were not normally distributed, Mann–Whitney U test was used to calculate differences between the HD and PM groups. Spearman's rank order correlation (r) was utilized for all comparisons. Cronbach's alpha was employed to examine internal consistency. Study power was calculated using a Wilcoxon Mann–Whitney test. Receiver Operating Characteristic (ROC) analysis was used to classify PM and manifest HD participants by the measures of balance (FALL‐HD, TBS, TUG, CST).

Construct Development

A pool of questions were generated based on a comprehensive literature review of known risk factors for falls. Falls were defined as events that unexpectedly led to the participant unintentionally coming to the ground, floor or other lower level. ⁷ , ¹⁵ A recall period of “the past year” was chosen to maximize accurate information collection and reduce bias. Responses to the questions were sought from the participant in corroboration with the caregiver or other collateral source.

Piloting the Measure

The initial item pool of six questions was administered to 30 gene positive participants as part of a pilot assessment. These individuals were chosen with a convenience sampling design for their availability to participate in additional assessments. They were specifically asked to identify areas that were not well covered by the current items and their thoughts on the questionnaire's readability, format, and applicability. Participant feedback from the pilot study led to the removal of one question regarding mental status.

To provide data for psychometric evaluation, the item pool of five questions was then administered to the 107 gene‐positive participants. Table 1 provides additional details on the demographics and clinical characteristics of the test sample. Internal consistency of the items was evaluated using Cronbach's alpha. Item‐total statistics were examined to identify and remove questions that detracted from the scale's internal consistency. To optimize internal consistency, items were removed if they demonstrated a low corrected item‐total correlation, indicating weak association with the overall scale, and if their deletion resulted in a higher Cronbach's alpha.

TABLE 1.

Demographics and Clinical Characteristics for Pre‐Manifest (PM) and Manifest HD (HD) participants, Mean (Standard Deviation)

	PM	HD	p‐value
N	54	53	‐
Age (years)	40.04 (11.09)	51.40 (8.85)	<0.001
Gender (M/F)	28/26	23/30	0.56
Education (years)	15.47 (2.87)	15.34 (2.61)	0.93
MMSE	28.05 (2.34)	26.25 (3.30)	<0.001
SDMT	50.76 (11.28)	33.43 (12.52)	<0.001
SWR	92.90 (18.99)	71.91 (21.92)	<0.001
TMS	3.13 (5.15)	22.60 (10.34)	<0.001
TMC	0.53 (1.07)	4.71 (2.78)	<0.001
TUG	9.55 (1.74)	10.30 (2.53)	0.14
TBS	9.25 (2.75)	20.41 (18.25)	<0.001
CST	11.55 (3.83)	13.25 (4.13)	0.02
TFC	12.80 (0.66)	10.70 (1.89)	<0.001
Total FALL‐HD	0.42 (0.71)	2.21 (1.20)	<0.001

Abbreviations: PM = pre‐manifest; HD = manifest Huntington disease; yrs = years; M = male; F = female; MMSE = Mini‐Mental State Examination; SDMT = Symbol Digit Modality Test; SWR = Stroop Word Reading; TMS = Total Motor Score; TMC = Total Maximal Chorea; TUG = Timed Up and Go; TBS = Total Body Sway; CST = Sit to Stand; TFC = Total Functional Capacity. Mann–Whitney U test was used to calculate differences between the two groups.

Validation Process

The FALL‐HD underwent psychometric analyses to examine the reliability and validity of the measure. We established the reliability of the FALL‐HD by first examining the test–retest reliability of the scores. This test–retest analysis was performed on a sample of 15 gene positive participants with the second administration performed after 2–3 months had lapsed. Internal consistency, the stability across responses to all items in a measure, was then analyzed to examine the reliability of the FALL‐HD using Cronbach's alpha.

To establish the validity, we examined face validity, criterion validity, and convergent/divergent validity. Face validity was established through feedback from fellow colleagues who were not involved in the development process. Criterion validity is usually established using sensitivity and specificity analysis to determine how well a measure performs against a gold standard assessment which, in our case, would be the extent to which the FALL‐HD relates to known balance measures. Thus, a ROC curve was used to compare the sensitivity and specificity of FALL‐HD, with known measures of balance impairment, specifically the TUG, TBS, and CST. Convergent validity was evaluated via examination of the relationship of the FALL‐HD with the TBS, while divergent validity was assessed by the relationship of the FALL‐HD with a cognitive measure, the MMSE. Finally, a post‐hoc examination of study power was conducted using a Wilcoxon Mann–Whitney test.

RESULTS

Item Reduction and Internal Consistency

The initial 5‐item questionnaire demonstrated Cronbach's alpha of 0.66. Removal of the question regarding difficulty getting out of a chair, which had a corrected item‐total correlation of 0.07, increased Cronbach's alpha to 0.69. Subsequent removal of the question regarding the use of mobility aids, which had a corrected item‐total correlation of 0.35, further increased Cronbach's alpha to 0.71. The item reduction process resulted in the retention of three questions for the final questionnaire (FALL‐HD). The final 3‐item FALL‐HD questionnaire demonstrated Cronbach's alpha of 0.71, indicating acceptable internal consistency. Item‐total correlation analysis confirmed that each item in the final FALL‐HD contributed meaningfully to the scale's performance. The remaining three questions assessed: Stability when walking (“How stable do you feel when walking without any support?”); Fear of falling (“Do you feel fearful about falling?”); Fall history (“Within the past year, have you fallen?”). Corrected item‐total correlations were: stability while walking (r = 0.557), fear of falling (r = 0.626), and fall history (r = 0.467). Item responses were coded and summed to create a total score ranging from 0 to 5. Figure 1 shows the final version of FALL‐HD.

Figure 1.

FALL‐HD measure.

Reliability

Reliability was assessed by measuring the test–retest reliability. The test–retest reliability was strong, with a correlation coefficient (r) of 0.85.

Validity

Criterion Validity

ROC curve analysis assessed the 3‐item FALL‐HD's ability to evaluate balance impairment compared with other widely used measures, including the TUG, TBS, and CST. The Fall‐HD outperformed all three comparators (Fig. 2). The highest Area Under the Curve (AUC) was for the FALL‐HD (AUC = 0.907) compared to the TBS (AUC = 0.85), the CST (AUC = 0.65), and the TUG (AUC = 0.58).

Figure 2.

Receiver Operating Characteristic (ROC) curve classifying pre‐manifest and manifest HD participants by measures of balance. FALL‐HD AUC = 0.91; Total Body Sway (TBS) AUC = 0.86; Chair Sit‐to‐Stand (CST) AUC = 0.65; Timed Up and Go (TUG) test AUC = 0.59.

Convergent and Divergent Validity

The 3‐item FALL‐HD was significantly positively correlated with the TBS (r = 0.62, P < 0.001), the UHDRS TMS (r = 0.754, P < 0.001), and the UHDRS TMC Score (r = 0.653, P < 0.001), demonstrating convergent validity. Concerning divergent validity, the 3‐item FALL‐HD was only weakly associated with the MMSE (r = −0.234, P = 0.028). The weak association with a cognitive measure supports the discriminant (divergent) validity of the FALL‐HD, given that its construct is balance/motoric function.

Study Power

Post‐hoc analysis revealed that the study power to detect true statistical effects was 1.0, indicating excellent power.

DISCUSSION

FALL‐HD is a novel measure designed to screen for falls in individuals with HD. This questionnaire's brevity and ease of scoring address the need for efficient fall screening, including frequency of falls, within this population. The final version of the FALL‐HD comprises three items, assessing stability while walking, fear of falling, and fall history, with total scores ranging from 0 to 5, where higher scores indicate greater risk of falling.

Until now, there has been no validated fall status questionnaire for use specifically in HD populations. Instead, clinicians have collected fall history either by patient interview or with the use of fall diaries and/or questionnaires used in parkinsonian disorders, ataxia, or neuromuscular conditions. ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ Fall diaries require substantial commitment from both patients and caregivers. They can be complex to maintain, require consistent detailed recording, and often lead to participant fatigue during long studies. As a result, this can lead to data gaps. Studies indicate that a higher fall risk is associated with lower odds of returning completed diaries, indicating that those at greater risk may struggle more with diary maintenance. ²⁰ , ²¹ Not surprisingly, in the absence of a validated fall instrument created for HD, prior research has focused on mobility and balance assessment tools such as the TMT, BBS, TUG, CST, and the motor section of the UHDRS to infer fall risk in subjects with HD.

Conversely, several fall risk assessment questionnaires have been developed and validated specifically for Parkinson's disease, including the Dresden Fall Questionnaire (DREFAQ), Parkinson's Disease Fall Questionnaire (PDF‐Q), and Fall Risk Perception Scale for Parkinson's Disease (FRPS‐PD), The DREFAQ, ²² which demonstrates strong reliability and validity in evaluating falls is a 5‐item questionnaire designed to assess falls, fear of falling, fall‐related injuries, and causes of falls in PD. The PDF‐Q ²³ captures falls and their consequences. The FRPS ²⁴ assesses individuals’ perception of their fall risk. Even tools like the FES, ⁶ which have been used for fall assessment in HD, lack HD‐specific validation. FALL‐HD addresses this gap as an HD‐specific fall risk screening instrument.

The development of the FALL‐HD involved a systematic process of item generation and reduction. An initial pool of five questions was created based on a comprehensive literature review, drawing upon recommendations from the World Guidelines for Falls Prevention and Management for Older Adults: A Global Initiative ¹⁵ and key elements from a highly sensitive fall risk tool, the “three key questions” (3KQ). ¹⁵ , ²⁵ The 3KQ encompasses the domains of fall history, unsteadiness while standing or walking, and fear of falling. To optimize the FALL‐HD's psychometric properties, item reduction was performed using internal consistency analysis, resulting in the removal of two questions due to their low corrected item‐total correlations and negative impact on Cronbach's alpha.

The removal of the ambulatory assistance question was supported by both statistical and theoretical considerations. Statistically, the inclusion of this question reduced the scale's internal consistency. Theoretically, it was inconsistent with the FALL‐HD's goal of screening for falls across the spectrum of the disease. Because balance problems are an early manifestation of HD, ²⁶ the FALL‐HD is intended to identify individuals at risk of falls throughout the disease course, not just those already dependent on assistive devices. In other patient populations, it has been found that there is a lower incidence of falls and injuries in individuals using wheelchairs as compared to ambulatory controls. ¹⁷ , ¹⁸

Notably, the three questions retained in the final FALL‐HD are remarkably similar to those included in the 3KQ, suggesting that the FALL‐HD effectively captures key domains relevant to falls. Given the FALL‐HD's brevity, this similarity is a notable strength, potentially allowing for a quick yet informative assessment of fall history and frequency. It is important to emphasize that the FALL‐HD, in its current form, is designed for ambulatory patients with the HD gene, aiding the clinician in making a quick judgment about fall history and frequency. Inquiring about steadiness while walking and fear of falling provides valuable insight into the patient's stability status. This information is particularly important in ambulatory HD patients, as early intervention based on fall screening can potentially prevent fall‐related complications in this population. ²⁷ , ²⁸ Beyond informal clinical questioning, FALL‐HD provides standardized scoring that reduces assessment variability between providers, enables objective tracking over time, and ensures systematic evaluation of three validated fall domains that may be inconsistently addressed in routine care.

The 3‐item FALL‐HD demonstrated evidence of both reliability and validity in this study. Test–retest reliability was strong (r = 0.85), indicating good stability of scores over a 2–3 month period. The questionnaire also showed good criterion validity, evidenced by its superior performance in ROC curve analysis compared to established balance measures (TUG, TBS, CST). Furthermore, the FALL‐HD demonstrated convergent validity through significant positive correlations with the TBS, UHDRS‐TMS, and UHDRS‐TMC, suggesting that it measures constructs related to balance and motor function. Given this correlation with known balance measures, the FALL‐HD may hold utility in future research for the prediction of falls in this population. Divergent validity was supported by the weak association with the MMSE, indicating that the FALL‐HD is relatively independent of cognitive function.

This study is not without limitations. Results obviously relied on a single HD research center. The measure was developed on a group of English‐speaking individuals with an English‐speaking caregiver; thus, the FALL‐HD is an English‐language measure with no translations available. We are unable to assess the reliability and validity among distinct cultural groups in our sample. Further examination of the cross‐cultural reliability and validity of the FALL‐HD is needed. Another possible limitation is how anosognosia, a lack of awareness of deficits, may have impacted the self‐report data. Anosognosia can have motor, cognitive, and psychiatric implication in HD. ²⁹ , ³⁰ , ³¹ , ³² We attempted to limit this by asking that questionnaires be filled out by patient and caregivers together. Since there is anosognosia even in the early stages of HD, information provided by the participant was corroborated by the caregiver. However, certain questions like those on stability and fear of falls are questions that the participant might be able to answer better than the caregiver, so one needs to use both the caregiver and participant's judgment in answering the questionnaire. Finally, while the current study provides evidence for the reliability and validity of the FALL‐HD, further investigation into its predictive validity (ie, its ability to predict future falls) would be valuable.

In conclusion, the FALL‐HD is a brief, validated measure for evaluating fall status, including fall frequency, in ambulatory patients with HD. It is a short three‐item questionnaire with a total maximal worst score of 5 points that performs better than other physical measures of balance in assessing falls in HD. The FALL‐HD demonstrated good test–retest reliability, internal consistency, and validity in a well‐characterized HD population; thus, should be a useful tool in both clinical and research HD settings.

Author Roles

(1) Research Project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript Preparation: A. Writing of the First Draft, B. Review and Critique.

J.K.: 1A, 1B, 1C, 2A, 2B, 2C, 3A.

D‐Y.L.: 1B, 1C, 2C, 3A, 3B.

K.B.: 1B, 1C, 3B.

N.Y.: 1B, 1C, 3B.

A.H.: 1B, 1C, 3B.

P.G.: 1A, 2C, 3B.

D.G.: 1A, 2C, 3B.

J.C.B.: 1A, 1B; 2B, 2C, 3A, 3B.

Disclosures

Ethical Compliance Statement: Institutional review board approval was obtained from the Institutional review board of the University of California, San Diego (Protocol no. #170038). Written informed consent was obtained from all participants before participation. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.

Funding Sources and Conflicts of Interest: No specific funding was received for this work. The authors declare that there are no conflicts of interest relevant to this work.

Financial Disclosures for the Previous 12 Months: DJG is eligible for royalties from a patent (US Patent 10,660,558, 2020) related to the technology used in this study. In addition, he has an equity stake (stock options) in Balance Tracking Systems, Inc. This financial conflict of interest is mitigated by a management plan put in place by his academic institution to ensure the integrity of his research.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.