Investigating the Impact of Voice Impairment on Quality of Life in Stroke Patients: The Voice Handicap Index (VHI) Questionnaire Study

Abstract

The Voice Handicap Index (VHI) is a patient-centered evaluation tool specifically designed for assessing voice-related quality of life. Although the VHI has been extensively used in patients with voice disorders, its applicability in stroke patients has not been fully established. This prospective cross-sectional study aimed to investigate the feasibility of using the VHI questionnaire in identifying stroke patients with voice problems. The study included a cohort of acute to subacute first-ever stroke patients (n = 48), with or without voice problems, as well as other non-stroke patients (n = 31) who agreed to complete the VHI questionnaire. Stroke patients with self-reported voice problems demonstrated significantly higher VHI scores and poorer life quality scores compared to the control groups. These patients also had lower Mini-Mental State Examination (MMSE), Modified Barthel Index (MBI), and Euro-QoL-5D-5L (EQ-5D-5L) scores. Spearman correlation analysis revealed an inverse association between VHI scores and EQ-5D-5L (rho = −0.77, p < 0.001), Korean Mann Assessment of Swallowing Ability (rho = −0.51, p < 0.001), and other functional parameters, including the National Institutes of Health Stroke Scale, MMSE, and MBI scores. Multiple regression analysis indicated that the VHI score was the biggest contributing factor to EQ scores. This is the first study to demonstrate that stroke patients with voice problems may experience reduced quality of life, even after controlling for other confounding factors such as dysphagia or neurological deficits. Future studies are needed whether addressing these issues by implementing the VHI may facilitate the improvement of patients’ quality of life.

Highlights

• • The Voice Handicap Index (VHI) assesses voice-related quality of life.

• • VHI's applicability in stroke patients is not fully established.

• • Stroke patients with voice problems had higher VHI scores and poorer life quality.

• • Stroke patients with voice problems may experience reduced quality of life.

• • Future studies needed to address these issues and improve patients’ quality of life.

INTRODUCTION

A human voice is a complex and intricate mechanism that enables individuals to express a diverse range of emotions, feelings, and verbal communication. In various professions, having a robust and healthy voice is crucial. In addition, voice impairments can significantly impact a patient’s quality of life, compromising their physical, emotional, and functional well-being. Previous studies have shown that voice problems are prevalent among the elderly population, leading to a reduction in their quality of life [1,2,3]. Speech production and phonation is a complex, rapid motor behavior involving precise coordination of numerous laryngeal, orofacial, and respiratory muscles. However, the laryngeal motor cortical control during speech and other voluntary laryngeal behaviors is not fully understood [4].

Stroke patients are known to experience voice problems and swallowing dysfunction frequently due to the involvement of their vocal folds and glottic insufficiency [5]. Despite formal assessments being conducted on stroke patients for disarticulation and language problems [6,7], little attention has been given to whether the neurological involvement of the laryngeal structures may cause voice problems and affect their quality of life during recovery and reintegration into society. Changes in voice quality can have a significant impact on the quality of life of stroke patients, similar to other functional impairments caused by neurological involvement of the oropharyngeal muscles.

The Voice Handicap Index (VHI) is a patient-centered, voice-specific psychometric evaluation tool that was developed by Jacobson in 1997 to assess the impact of voice impairment [8]. It consists of 30 items, distributed equally across 3 domains: functional, physical, and emotional aspects. Despite its widespread use, it has been mainly applied and studied in the area of otolaryngology and voice pathology [9,10,11,12] and there have been no reports on the validity of the VHI for stroke patients, a population in which voice problems are under-researched. Given the importance of voice in communication, implementing a screening tool for voice changes in stroke patients is a novel approach that warrants further investigation.

Therefore, in this questionnaire study, our aim was to investigate whether the application of the VHI can aid in the screening of patients with voice problems, while exploring the potential correlation between the VHI score and other neurological impairments. Additionally, we sought to assess the impact of subjective voice problems on the quality of life of stroke patients and validate the diagnostic properties of the VHI in stroke patients.

MATERIALS AND METHODS

Study design and participants

In this prospective study, we recruited 79 participants (47 males, 32 females) from a university-affiliated hospital between October 2020 and October 2021. The participants included both in and out-patients, with a mean age of 60.9 ± 14.8 years and a range from 19 to 87 years.

Participants were categorized into 3 groups: group A consisted of patients who experienced persistent voice problems after stroke; group B included patients without any new voice problems after stroke; and group C comprised of healthy individuals who visited the department for other musculoskeletal conditions.

In addition to completing the VHI questionnaire, the time of voice use per day prior to the stroke was recorded for all participants. This information was collected to provide a comprehensive understanding of the potential impact of voice use on post-stroke voice problems. Information on the quality of life were assessed through the Eruo-QoL-5D-5L (EQ-5D-5L) [13]. In addition, functional scales related to stroke, such as National Institutes of Health Stroke Scale (NIHSS) [14], Mini-Mental State Examination (MMSE) [15], Berg Balance Scale (BBS) [16], Modified Barthel Index (MBI) [17] were also checked from medical records. In addition, the level of swallowing was recored with the Korean Mann Assessment of Swallowing Ability (K-MASA) [18], Functional Oral Intake Scale (FOIS) [19].

Participants were excluded if they were belonged to any of the following criteria; cognitive deficits who may not understand or complete questionnaire with an MMSE score below 15, moderately severe or severe cognitive impairment, history of any previous otolaryngologist disorder, prior history of tracheostomy or any procedures related to head and neck surgery and pulmonary disease that can cause shortness of breath. This study was ethically approved by the Institutional Review Board of the Catholic University of Korea, Bucheon Saint Mary’s Hospital (HC21QASI0003).

Instruments for data collection

In this study, the questionnaires were administered to patients in a face-to-face setting with the doctor. The patients were asked a series of questions, including their occupation and the amount of time they spent using their voice per day prior to completing the questionnaire.

Iowa Patient's Voice Index (IPVI)

The IPVI, developed by Verdolini and colleagues at the University of Iowa, was used as a tool to document patients’ perceptions of their own voice quality, vocal effort, and the impact of dysphonia on their daily lives. The IPVI comprises 2 questions that are designed to rate voice quality and the impact of voice quality on an equal-appearing interval scale. Patients were asked to rate these questions on a scale of 0 to 6, where 0 indicates normal or no impact and 6 indicates severe or great impact (Supplementary Data 1) [20]. Consistent with previous studies, participants who received a score of 0 to 1 on ‘item 1’ of the assessment were categorized as ‘without voice problem,’ while those who scored 2 points or more were classified into the ‘with voice problem’ group [21].

VHI

The VHI is a validated self-assessment tool that quantifies the functional, physical, and emotional impact of a voice disorder on an individual’s quality of life [8]. The tool consists of 30 questions that are equally distributed across 3 domains: functional, physical, and emotional aspects (Supplementary Data 2).

Euro-QoL-5D-5L (EQ-5D-5L)

The EQ-5D-5L is a general quality of life assessment tool that was developed by the EuroQol Group in 2005 to measure health-related quality of life [13]. This tool has demonstrated excellent psychometric properties across a broad range of populations, conditions, and settings. The EQ-5D-5L consists of 5 dimensions, including exercise ability, self-management, daily activities, pain/discomfort, and anxiety/depression. The advantage of the EQ-5D-5L is that it can be easily used in various clinical situations and provides numerical results to quantify quality of life. Each dimension is scored on a 5-level severity ranking that ranges from “no problems” to “extreme problems.” The EQ-5D-5L index scores range from −0.224 to 1, with 0, 1, and negative values corresponding to death, full health, and health states worse than death, respectively.

Statistical analysis

Statistical analyses were performed using R Statistical Software (version 4.1.2; R Foundation for Statistical Computing, Vienna, Austria). The t-tests and analysis of variance were conducted to compare the age, sex (male %), voice using time per day, and functional scale scores across each group. Post hoc analyses with Bonferroni correction were performed to evaluate significant differences between the groups.

Spearman correlation coefficients were calculated to identify any relationships between the EQ-5D-5L, NIHSS, MMSE, BBS, MBI, K-MASA, FOIS, and VHI scores. Finally, multiple regression analysis was conducted to establish a causal linear relationship between the EQ-5D-5L and other functional outcome scores. Optimal cut-off values were computed from the receiver operating characteristic (ROC) curve for total VHI score. The area under the curve (AUC) values exhibiting values of 0 till 1, which reflect the diagnostic accuracy and predictive ability, were calculated for the parameter. The statistical significance was set at a p value of 0.05.

RESULTS

Forty-eight stroke patients (mean age 65.1 ± 12.4 years old; range 34–87 years old; 31 males; 17 females) and 31 control group individuals (mean age 55 ± 16.2 years old; range 19–84; 16 males; 15 females) who visited our department for other musculoskeletal disorders were included. Table 1 summarizes the mean age, sex distribution, and underlying diseases of the participants. Although the 2 stroke groups had a higher age distribution than the control group, no age difference was observed between those with and without voice problems in the stroke groups.

Table 1. Dermographics, voice using time, EQ-5D-5L and VHI(total) score of 3 groups.

Groups		Stroke with voice problem (n = 30)	Stroke without voice problem (n = 18)	Non-stroke (n = 31)	p value
Age (yr)		64.5 ± 12.3†	66.1 ± 12.9†	55.3 ± 16.2‡	0.013*
Sex (male)		21 (70)	10 (55.6)	16 (51.6)	0.323
Underlying disease
	Hypertension	14 (46.7)	13 (72.2)	8 (25.8)	0.011
	Diabetes mellitus	8 (26.7)	8 (44.4)	6 (19.4)	0.219
	Atrial fibrillation	2 (6.7)	1 (5.6)	0 (0)	0.350
	Psychogenic disease	1 (3.3)	0 (0)	1 (3.2)	0.722
	Rheumatic disease	1 (3.3)	0 (0)	1 (3.2)	0.500
Min voice using time/day (hr)		6.2 ± 3.0	5.1 ± 2.5	4.8 ± 2.6	0.045
Max voice using time/day (hr)		6.2 ± 3.0	5.1 ± 2.5	4.8 ± 2.6	0.045
Average voice using time/day (hr)		5.2 ± 3.0	4.1 ± 2.5	3.8 ± 2.5	0.037
EQ-5D-5L		0.4 ± 0.2	0.8 ± 0.1	0.7 ± 0.2	< 0.001*
VHI(total)		65.1 ± 21.4	7.4 ± 4.8	3.7 ± 4.9	< 0.001*

Values are presented as number (%). Baseline values are presented as mean ± standard error of the mean. Between group 1 and 3: p value = 0.040, between group 2 and 3: p value = 0.035.

EQ-5D-5L, Euro-QoL-5D-5L; VHI(total), Voice Handicap Index (total score).

*The p value is derived from Mann-Whitney U test because of non-parametric distribution. The p value < 0.05 was considered statistically significant.

^†,‡Statistically significant difference is present between a dagger and two barred cross.

Prior to the stroke, the average voice usage time per day showed no significant differences; however, after stroke, the VHI score was higher in those who experienced persistent post-stroke voice problems than in the 2 comparison groups. Moreover, they exhibited lower EQ-5D-5L scores.

Between the 2 stroke groups, no differences were observed in the initial stroke presentation, as reflected by similar levels of stroke severity scores as indicated by NIHSS scores. No differences were observed in the stroke type, stroke lesion or duration (day) from onset. Spearman’s rank correlation analysis was performed to determine if there was a correlation between stroke duration (day) and VHI scores within a stroke with voice problem group, but no correlation was seen with a p value of 0.211. However, those with voice problems showed lower MMSE scores and a greater degree of functional disability, as indicated by lower MBI scores. They also showed higher levels of swallowing impairment, as reflected by their levels of functional oral intake and the degree of swallowing (Table 2).

Table 2. Stroke type, site and functional scale scores of stroke groups.

Groups		Stroke with voice problem (n = 30)	Stroke without voice problem (n = 18)	p value
Stroke type
	Infarction	22 (73.3)	14 (77.8)	1.000
	Hemorrhage	7 (23.3)	2 (11.1)	0.451
	SDH/EDH	1 (3.3)	2 (11.1)	0.547
Lesion site
	Right	18 (60)	10 (55.6)	0.762
	Left	10 (33.3)	8 (44.4)	0.441
	Bilateral	2 (6.7)	0 (0)	0.521
Duration from onset (day)		98.8 ± 153.8	67.5 ± 26.9	0.173
NIHSS		5.8 ± 5.2	3.6 ± 4.2	0.141
MMSE		24.5 ± 2.6	27.2 ± 2.9	0.002*
BBS		26.5 ± 20.9	38.4 ± 19.3	0.055
MBI		49.3 ± 34.5	75 ± 23.9	0.008*
K-MASA		176.4 ± 17.4	193.9 ± 7.1	< 0.001*
FOIS		4.47 ± 2.1	6.4 ± 0.8	< 0.001*

Values are presented as number (%). Baseline values are presented as mean ± standard error of the mean.

SDH, Subdural hemorrhage; EDH, Epidural hemorrhage; NIHSS, National Institutes of Health Stroke Scale; MMSE, Mini-Mental State Examination; BBS, Beg Balance Scale; MBI, Modified Barthel Index; K-MASA, Korean Mann Assessment of Swallowing Ability; FOIS, Functional Oral Intake Scale.

^*The p value < 0.05 was used for statistical significance.

To investigate the relationship between the distress of voice usage and other functional impairments after stroke, a Spearman correlation analysis was performed, which demonstrated that the higher distress of voice usage was significantly and inversely related to poor quality of life with EQ-5D-5L (rho = −0.77, p < 0.001), lower levels of swallowing MASA (rho = −0.51, p < 0.001), and other functional parameters, including the FOIS, MMSE, BBS, and MBI scores (Fig. 1).

Fig. 1. Correlation heatmap showing association factors and VHI scores. The color scale indicates the degree of correlation (blue: strong positive correlation, white: weak correlation, and red: strong negative correlation). The size of the circle indicates strength of relationship. The stronger the correlation, the closer the correlation coefficient comes to ±1 and showing bigger circle size.

NIHSS, National Institutes of Health Stroke Scale; IPVI, Iowa Patient’s Voice Index; VHIT, Voice Handicap Index total score; MMSE, Mini-Mental State Examination; BBS, Beg Balance Scale; MBI, Modified Barthel Index; MASA, Mann Assessment of Swallowing Ability; FOIS, Functional Oral Intake Scale.

In the multiple regression analysis, the predicted value of the dependent variable ‘EQ’ was calculated by the equation: ‘1.44 − 0.01NIHSS − 0.00MASA − 0.01 × VHI(total)’ (Table 3). The adjusted R² of this model was 0.644 (p < 0.001). The predictor variables, ‘NIHSS,’ ‘MASA,’ and ‘VHI(total),’ accounted for 66.76% of the total variation of the dependent variable. Among these 3 variables, the VHI score was the most significant contributor to poor EQ scores (Fig. 2), indicating the crucial role of voice distress in a patient's quality of life.

Table 3. Summary of selected regression model.

	Beta	SE	Stadard β	lwr	upr	SE	t-value	p
(Intercept)	1.44	0.37	0.00	−0.18	0.18	0.09	3.905	< 0.001
NIHSS	−0.01	0.01	−0.22	−0.43	−0.01	0.10	−2.112	0.041
MASA	−0.00	0.00	−0.19	−0.43	0.04	0.12	−1.648	0.107
VHI(total)	−0.01	0.00	−0.83	−1.04	−0.63	0.10	−8.064	< 0.001

R² = 0.6676, _adjR² = 0.6444, F = 28.78, p < 0.001, Akaike’s Information Criterion = −36.43.

NIHSS, National Institutes of Health Stroke Scale; MASA, Mann Assessment of Swallowing Ability; VHI(total), Voice Handicap Index (total score); SE, standard error; lwr, lower limit of the confidence interval; upr, upper limit of the confidence interval.

Fig. 2. Results of multiple regression anlaysis for variables predicting Euro-QoL-5D-5L.

VHI(total), Voice Handicap Index total score; NIHSS, National Institutes of Health Stroke Scale; K-MASA, Korean Mann Assessment of Swallowing Ability.

An ROC analysis showed that a cut-off value of 24 in the VHI showed an AUC of 0.995 (95% confidence interval [CI], 0.982–1) with excellent sensitivity levels of 96.6% (95% CI) (Fig. 3).

Fig. 3. Receiver operating characteristic curve analysis for Voice Handicap Index.

PPV, positive predictive value; NPV, negative predictive value; AUC, area under the curve.

Furthermore, analysis of subcategories within the EQ survey revealed significant differences in all subscores across the 5 domains of the EQ-5D-5L between stroke patients with voice problems compared to the control group (Table 4).

Table 4. Five subgroup analysis of Euro-QoL-5D-5L.

Groups	Stroke with voice problem (n = 30)	Stroke without voice problem (n = 18)	p value
EQ-Exercise ability	3.3 ± 0.9	1.6 ± 0.7	< 0.001*
EQ-Self management	3.4 ± 0.7	1.7 ± 0.8	< 0.001*
EQ-Daily activities	3.4 ± 0.7	2.0 ± 0.8	< 0.001*
EQ-Pain/discomfort	3.7 ± 1.0	2.3 ± 0.9	< 0.001*
EQ-Anxiety/depression	3.9 ± 0.8	1.9 ± 0.9	< 0.001*

Baseline values are presented as mean ± standard error of the mean.

^*The p value < 0.05 was used for statistical significance.

DISCUSSION

The results of this questionnaire-based study using the VHI demonstrate that post-stroke voice changes significantly affect a patient’s quality of life, playing a role comparable to other physical and neurological impairments after stroke. The multiple regression model revealed that the VHI(total) score, in addition to the NIHSS and levels of swallowing, had the most significant negative impact on EQ-5D-5L scores. Subanalysis of 5 domains in EQ-5D-5L suggested that voice problem could affect all aspects in quality of life. Diagnostic validity of the VHI showed excellent diagnostic properties in screening voice problems for post-stroke patients and identifying their impact on quality of life.

Objective methods for assessing voice include laryngoscopy, videostroboscopy, photoglottography, laryngeal electromyography, acoustic analysis, and aerodynamic study, as they provide a measurable criteria for the evaluator [22]. However, the degree of speech problems can vary greatly from the patient’s perspective and is affected by individual circumstances, such as occupation and social activities [23]. Self-measurement of the biopsychosocial impact of voice problems is essential for monitoring therapy effectiveness when assessing and treating patients with voice problems [24]. Stroke patients commonly have difficulty in explain themselves and communicating with other people. The objected methods previously mentioned seem insufficient to address quality of life issues, highlighting the importance of implemeting the VHI in stroke patietns.

Dysphagia, the difficulty or inability to swallow, is a prevalent post-stroke complication, affecting approximately 19%–81% of stroke patients [25]. The anatomical structures responsible for voice production, including those involved in the movement of the vocal cords, overlap significantly with those involved in swallowing. Consequently, it is reasonable to hypothesize that voice issues may co-occur with dysphagia in individuals experiencing swallowing difficulties. For instance, vocal cord palsy is a common complication following stroke, which can lead to communication problems due to dysphonia, and may also result in pulmonary dysfunction and swallowing difficulties [26]. Previous research has also demonstrated the close association between voice and swallowing; for instance, a study found that neuromuscular stimulation administered 5 times a week for 40 sessions as a treatment for dysphagia resulted in an improvement in voice quality, indicated by an increase in voice intensity and a decrease in jitter and shimmer [27]. In accordance, our results showed that those with higher VHI scores to show more severe grade of dysphagia indicated by the lower MASA and FOIS scores.

Though there is limited information on the impact of voice problems in stroke patients, studies in Parkinson’s disease (PD), a neurodegenerative disorder that also leads to swallowing problems, have shown that voice disorders are a common complication, affecting 70%–90% of patients [28]. Abnormalities in acoustic analysis have been detected in early stages of PD, and voice dysfunction has been suggested to be the earliest sign of motor impairment in PD [29]. One study reported that 82% of early PD patients were not satisfied with their speech [30]. Main features of speech problem in PD include monotonous voice, inaccurate articulation and hypophonia [31]. Additionally, studies have shown a positive correlation between the VHI and disease severity in PD [32]. Therefore, it is expected that VHI would also show abnormalities in stroke patients. In fact, changes of acoustic analysis showed same tendencies in PD and stroke. PD patients showed reduced harmonics-to-noise ratio (HNR) due to asymmetric vocal fold position, with increased jitter and shimmer along with increased fundamental frequency with higher pitch [33]. In parallel, a recent study of post stroke voice change using machine learning with mobile devices showed similar tendencies of accoustic analysis results in HNR, jitter, shimmer and fundamental frequency [34].

Previous study showed that voluntary voice production is controlled by the laryngeal motor cortex, and executed through both direct and indirect pathways; limbic vocal contral pathway and laryngeal motor cortical pathway, descending to the brainstem laryngeal motoneurons [35]. Given the complexity and wide distribution of laryngeal motor cortical pathways that connect from the brainstem to cortical and subcortical areas, it is plausible that damage to any part of the brain involving these pathways can cause voice problems after stroke. Additionally, voice problems are part of a broad category of speech changes encompassing dysphonia, dysarthria, and dysarthric-like speech, further contributing to the potential incidence of voice problems following stroke [36]. In consideration of this broad network, both infratentorial and supratentorial stroke groups showed voice changes. When comparing the VHI scores of the 2 groups, there was no significant difference in the VHI(total) scores of supratentorium lesion group (42.2 ± 35.2) and infratentorium lesion group (42.4 ± 28.5).

While the accoustic analysis and laryngoscopy are objective measures, they require equipment or complex signal analysis in order to extract each parameters. By contrast, VHI is an easy, cost-efficient tool for assessing patient’s subjective voice problem with satisfactory reliability and reliability. According to the Agency for Health Care Research and Quality, which surveyed evaluation tools used in speech pathology in 2002, VHI was the only psychometric evaluation tool related to speech disorders that met the reliability and validity criteria [37]. Korean version of VHI was also confirmed the reliability and reliability in 2008 [21].

Our results has demonstrated a close association between the patient’s quality of life and their scores on the VHI, NIHSS, and the MASA, accounting for 66.76% of the total variation of the dependent variable in a multiple regression analysis. Notably, VHI score emerged as the most significant contributing factor to lower quality of life scores, underscoring how post-stroke voice problems can significantly deteriorate patients' overall well-being. It is worth noting that cut-off values of the VHI score in our study were similar to those reported in previous VHI studies [38,39], further supporting the use of the VHI in stroke patients.

In spite of the significant findings in this study, it is important to acknowledge several limitations. First, objective voice evaluation through acoustic analysis, including measurements of frequency, intensity, jitter, and shimmer was not conducted. Additionally, evaluation of voice quality using the Grade, Roughness, Breathiness, Asthenia, Strain (GRBAS) scale was not performed by a speech-language therapist [40]. Disarticulation was not also considered as a contributing factor. Lastly, since this study was a cross sectional study including in and out-patients, stroke duration of the inpatients was uniformly short due to the limited hospitalization period and patients who are admitted to a local medial center after a long time from onset or whose voice problem is insignificant and do not follow up on an outpatient basis are excluded. Therefore it was difficult to find correlation between stroke duration and VHI score.

In future, to invest post-stroke voice problems, it is recommended to incorporate objective voice assessments such as stroboscopy and GRBAS scores in addition to the VHI questionnaire with a larger number of patient with varying stroke duration.

To our knowledge, this is one of the few studies to examine the subjective discomfort of voice in stroke patients. Our results reveal that post-stroke patients with voice issues encounter a considerable decrease in their quality of life, even after controlling for other confounding factors, including neurological deficit and dysphagia. Those with VHI over the cut-off value of 24 may require further detailed assessment and treatment of post-stroke voice problems. Future studies are needed whether addressing these issues by implementing the VHI may facilitate the improvement of patients’ quality of life.

In conclusions, VHI can be used to identify perceptual voice problems in stroke patients with excellent diagnostic properties. Further research is recommended to explore the potential benefits of early identification and treatment of post-stroke voice problems.

SUPPLEMENTARY MATERIALS

Supplementary Data 1

Iowa Patient’s Voice Index

Supplementary Data 2

Voice Handicp Index