A vocally healthy adult scores an average of just 2.83 out of 40 on the Voice Handicap Index-10, while patients walking into a multidisciplinary voice clinic average 19.5, close to seven times higher (19.5 divided by 2.83 is about 6.9) (Arffa et al., Journal of Voice 2012; Misono et al., Journal of Voice 2014). That gap is what makes the VHI-10 such a common patient-reported outcome in voice care: in the same clinic cohort, 80% of patients scored above the abnormal threshold of 11. Dysphonia affects nearly one-third of the population at some point in life (AAO-HNS, Clinical Practice Guideline: Hoarseness 2018), and the minimal change that patients actually feel ranges from 4 to 6 points depending on the population (Young et al., Laryngoscope 2018; Misono et al., JAMA Otolaryngology 2017). We aggregated data from the original VHI and VHI-10 development papers, normative studies in the Journal of Voice, the University of Sydney Voice Lab, validation studies in seven languages and occupational surveys of teachers and singers. Where our acoustic voice analysis roundup covers what instruments measure, this one covers what patients report.
TL;DR
- Mean VHI-10 in adults without voice complaints: 2.83 (SD 3.93) out of 40 (Arffa et al., Journal of Voice 2012).
- A VHI-10 score above 11 is the most widely used abnormal threshold (University of Sydney Voice Lab, PROMs Normative Data Reference Sheet 2025).
- Mean VHI-10 at an academic voice clinic: 19.5, with 80% above 11 and 24.3 for vocal fold motion abnormality vs 16.0 for benign lesions (Misono et al., Journal of Voice 2014).
- Mixed voice disorder patients in Montenegro averaged 21.1 versus 2.3 for matched healthy controls (Zvrko and Kadic, Arh Hig Rada Toksikol 2022).
- The VHI-10 correlated with the 30-item VHI at greater than .90 in 819 patients (Rosen et al., Laryngoscope 2004).
- Minimal clinically important difference: a 4-point decrease in vocal fold paralysis (Young et al., Laryngoscope 2018) and 6 points in general voice clinic patients (Misono et al., JAMA Otolaryngology 2017).
- Singers need a bigger shift: the SVHI-10 MCID is 9.5 points, rounded to 10 or more (Perrin et al., Laryngoscope 2025).
- Healthy singers average 9.13 on the SVHI-10, with 20 or above considered abnormal (Gochman et al., Journal of Voice 2025).
- Language-specific VHI-10 cut-offs range from 4.5 in Persian to 12.5 in Algerian Arabic (Nazari et al. and Kaddour et al., Journal of Voice 2026).
- 41.0% of 400 university staff had an abnormal VHI-10 (Younis et al., BMC Public Health 2026).
- Healthy children average 0.26 on the Children VHI-10, with a cut-off of 2.5 (De Maio et al., Journal of Voice 2024).
1. From 85 Items to 10: How the Voice Handicap Index Was Built
The Voice Handicap Index started as a deliberately oversized questionnaire. An 85-item draft was given to 65 consecutive patients at the Henry Ford Hospital Voice Clinic and cut to 30 items through internal-consistency analysis (Jacobson et al., American Journal of Speech-Language Pathology 1997). The authors built it to fill a gap: no instrument then measured the psychosocial consequences of a voice disorder, as opposed to the acoustic signal itself.
Seven years later, a second team asked whether a third of the items could carry the same information. Item analysis on 100 voice patients and 159 controls produced the 10-item version, and in an 819-patient comparison the short form tracked the long form almost perfectly (Rosen et al., Laryngoscope 2004). The practical consequence: a 10-question form now does the screening work that once took 30 questions.
| Metric | Value | Source |
|---|---|---|
| Items in the original VHI draft | 85 items, tested on 65 patients | Jacobson et al., AJSLP 1997 |
| Items in the final VHI | 30 | Jacobson et al., AJSLP 1997 |
| Patients in the VHI test-retest study | 63 | Jacobson et al., AJSLP 1997 |
| VHI-10 item-analysis sample | 100 patients with voice problems and 159 controls | Rosen et al., Laryngoscope 2004 |
| VHI-10 validation comparison sample | 819 patients | Rosen et al., Laryngoscope 2004 |
| Correlation between VHI and VHI-10 | greater than .90 (P = .01) | Rosen et al., Laryngoscope 2004 |
| VHI-10 score range | 0 to 40 | Misono et al., JAMA Otolaryngology 2017 |
Context note: Rosen and colleagues also found that VHI-10 scores ran higher than the expected 33% of the full VHI score, so a VHI-10 total is not simply one-third of the matching VHI-30 total. Most recent available data for instrument development: Jacobson et al., 1997, and Rosen et al., 2004.
2. What a Normal VHI-10 Score Looks Like
The normative benchmark is strikingly low. In 158 analyzed VHI-10 questionnaires from adults without voice complaints, the mean was 2.83 (SD 3.93), so a typical healthy speaker uses about 7% of the 40-point scale (2.83 divided by 40) (Arffa et al., Journal of Voice 2012). The same study set the abnormal threshold at a score above 11, after defining outliers as more than 3 SD above the mean.
That 2012 threshold has aged well. The University of Sydney Voice Lab still lists a VHI-10 above 11 as abnormal in its April 2025 reference sheet (University of Sydney Voice Lab, PROMs Normative Data Reference Sheet 2025), and independent control groups in other countries land in the same low range.
| Metric | Value | Source |
|---|---|---|
| Mean VHI-10 in healthy adults | 2.83 (SEM 0.31, SD 3.93), n = 158 of 190 enrolled | Arffa et al., Journal of Voice 2012 |
| Mean full VHI-30 in healthy adults | 6.86 (SEM 0.79, SD 9.88), n = 156 | Arffa et al., Journal of Voice 2012 |
| Abnormal VHI-10 threshold | score above 11 | Arffa et al., Journal of Voice 2012 |
| Abnormal VHI-10 threshold in current clinical reference | score above 11 | University of Sydney Voice Lab 2025 |
| Mean VHI-10 in Montenegrin vocally healthy controls | 2.3 ± 2.5 (n = 50) | Zvrko and Kadic, Arh Hig Rada Toksikol 2022 |
| Median VHI-10 in Algerian vocally healthy controls | 4.0 (n = 90) | Kaddour et al., Journal of Voice 2026 |
Outlier note: the full-VHI standard deviation (9.88) is larger than its mean (6.86), a sign that a handful of healthy respondents report real handicap. Normal is a distribution, not a zero. Most recent available data for US norms: Arffa et al., 2012.
3. VHI-10 Scores by Dysphonia Diagnosis
Diagnosis shapes the patient’s experience more than the presence of a lesion does. Patients with vocal fold motion abnormality averaged 24.3 on the VHI-10, compared with 16.0 for those with likely benign lesions, and the difference against muscle tension dysphonia was significant in both directions (Misono et al., Journal of Voice 2014). A visible benign lesion can come with less self-reported handicap than a voice problem with no lesion at all, which is where muscle tension dysphonia lands: in the middle of the range, above the lesion group.
Women scored higher than men in every diagnostic group of that cohort, and the same pattern shows up in Parkinson’s disease, where women averaged 27.5 against 17.5 for men before deep brain stimulation (Tuysuz et al., European Archives of Oto-Rhino-Laryngology 2026). The patients behind these scores also carry a psychological load: 32% of the 2014 clinic cohort met strict case criteria for depression, anxiety and/or somatic concerns.
| Metric | Value | Source |
|---|---|---|
| Mean VHI-10, all voice clinic patients (share above 11) | 19.5 (SD 9.4), n = 191 (80% above 11) | Misono et al., Journal of Voice 2014 |
| Muscle tension dysphonia | 18.0 (SD 9.6), n = 87 | Misono et al., Journal of Voice 2014 |
| Vocal fold motion abnormality | 24.3 (SD 8.4), n = 38 | Misono et al., Journal of Voice 2014 |
| Vocal fold lesion, likely benign | 16.0 (SD 7.8), n = 42 | Misono et al., Journal of Voice 2014 |
| Mixed voice disorder patients (Montenegro) | 21.1 ± 7.6, n = 50 | Zvrko and Kadic, Arh Hig Rada Toksikol 2022 |
| Parkinson’s disease before subthalamic DBS | 22.7 ± 7.92, n = 105 | Tuysuz et al., Eur Arch Otorhinolaryngol 2026 |
| Ulcerative laryngitis at presentation | average 25 (range 6-38), n = 23 | Young et al., Journal of Voice 2020 |
Context note: the 2014 diagnostic groups overlap because a patient could carry more than one diagnosis. Most recent available data for multi-diagnosis comparison: Misono et al., 2014.
4. Minimal Clinically Important Difference: How Much Change Counts
A statistically significant drop is not the same as a drop a patient notices. In 281 patients with unilateral vocal fold paralysis, a VHI-10 decrease of 4 points best separated improvement from no improvement (sensitivity 0.62, specificity 0.88, area under the ROC curve 0.80) (Young et al., Laryngoscope 2018). That is one point below the decrease of 5 or more that clinicians had used by consensus, and it matters for anyone reading vocal fold paralysis treatment outcomes.
In a broader voice clinic population the bar was higher. Across 273 patients retested after 2 weeks, a difference of 6 points was proposed as the minimal important difference, but the correlation between global change and VHI-10 change was only 0.32 (95% CI, 0.12-0.49), and the authors called for more studies (Misono et al., JAMA Otolaryngology-Head and Neck Surgery 2017). Measured against these thresholds, some recent trial results look modest.
| Metric | Value | Source |
|---|---|---|
| Significant change on the original 30-item VHI | 18 points between two administrations | Jacobson et al., AJSLP 1997 |
| VHI-10 MCID in unilateral vocal fold paralysis | decrease of 4 | Young et al., Laryngoscope 2018 |
| Prior clinical consensus threshold | decrease of 5 or more | Young et al., Laryngoscope 2018 |
| VHI-10 MID in general voice clinic patients | 6 points | Misono et al., JAMA Otolaryngology 2017 |
| Mean VHI-10 change when patients rated a large change | 9 (SD 13) | Misono et al., JAMA Otolaryngology 2017 |
| VHI-10 change in an online MTD intervention, perceived-control arm | -2.44 (95% CI, -4.25 to -0.63) | Nguyen-Feng et al., JAMA Otolaryngology 2025 |
| VHI-10 change in the education arm of the same trial | -2.94 (95% CI, -4.12 to -1.76) | Nguyen-Feng et al., JAMA Otolaryngology 2025 |
| Average VHI-10 decrease in ulcerative laryngitis | 15 points (95% of patients improved) | Young et al., Journal of Voice 2020 |
Outlier note: in the Parkinson’s cohort from section 3, the 0.9-point drop after deep brain stimulation (22.7 to 21.8) was statistically significant (p = 0.031) yet well below every MCID listed here. Even after an average 15-point drop in ulcerative laryngitis, only 50% of patients finished within normal limits (Young et al., Journal of Voice 2020).
5. Reliability and Validity: How Stable Is the Score?
For a 10-question self-report, the VHI-10 family is unusually consistent. Test-retest intraclass correlation reached 0.98 in Montenegrin voice patients, with a Cronbach’s alpha of 0.94 (Zvrko and Kadic, Arhiv za higijenu rada i toksikologiju 2022), and alpha values above 0.90 repeat across the Lithuanian, Kannada and Persian versions (see the table below and section 6). Acoustic measures tell clinicians how the voice sounds; reliability figures like these explain why patient-reported scores sit beside them in the acoustic and perceptual workup rather than being replaced by them.
The singing version shows the same pattern. The SVHI-10 reached an alpha of .94 and a test-retest Spearman correlation of 0.86 at launch (Cohen et al., Laryngoscope 2009), and a 2025 prospective retest after 30 days still found r = 0.879.
| Metric | Value | Source |
|---|---|---|
| Cronbach’s alpha and ICC, Montenegrin VHI-10 (patients) | alpha 0.94, ICC 0.98 | Zvrko and Kadic, Arh Hig Rada Toksikol 2022 |
| Cronbach’s alpha and ICC, Lithuanian VHI-10 | alpha 0.906, ICC 0.87 (2-3 week retest) | Polianskis et al., Journal of Voice 2026 |
| Cronbach’s alpha and ICC, Kannada VHI-10 | alpha 0.93, ICC 0.933 | Benoy et al., Journal of Voice 2026 |
| ICC, Algerian Arabic VHI-10 (combined sample) | 0.95 | Kaddour et al., Journal of Voice 2026 |
| SVHI-10 internal consistency and test-retest | alpha .94, Spearman 0.86 | Cohen et al., Laryngoscope 2009 |
| SVHI-10 test-retest after 30 days | r = 0.879, p < 0.001 | Perrin et al., Laryngoscope 2025 |
Context note: the review of translated 30-item VHI versions found high test-retest reliability and internal consistency overall, but only moderate construct validity across its three domains (Seifpanahi et al., Iranian Journal of Public Health 2015).
6. Cross-Language Validations and Local Cut-Offs
Translation moves the threshold. Validated VHI-10 cut-offs now span from 4.5 in Persian to 12.5 in Algerian Arabic, which means a score of 8 is abnormal on the Brazilian Portuguese, European Portuguese and Lithuanian versions but normal on the Algerian Arabic one (Kaddour et al., Journal of Voice 2026). A US threshold applied to a translated questionnaire can misclassify patients in either direction.
Discrimination is excellent almost everywhere. The largest single dataset, 975 Brazilian speakers (486 with dysphonia, 489 vocally healthy), returned a VHI-10 efficiency of 0.991 at a cut-off of 7.5 (Behlau et al., Journal of Voice 2016), and the European Portuguese version reached an area under the curve of 0.998 (Guimaraes et al., Journal of Voice 2026).
| Metric | Value | Source |
|---|---|---|
| Brazilian Portuguese VHI-10 cut-off (VHI-30 cut-off) | 7.5 (19); sensitivity 0.981, specificity 1 | Behlau et al., Journal of Voice 2016 |
| European Portuguese VHI-10 cut-off (VHI-30 cut-off) | 5.5 (13.5); sensitivity 0.977, specificity 0.955 | Guimaraes et al., Journal of Voice 2026 |
| European Portuguese VHI-10 area under the curve | 0.998 (95% CI: 0.995-0.999), n = 350 | Guimaraes et al., Journal of Voice 2026 |
| Lithuanian VHI-10 cut-off | 7 or above; AUC 0.987, n = 205 | Polianskis et al., Journal of Voice 2026 |
| Algerian Arabic VHI-10 cut-off | 12.5; sensitivity 88.3%, specificity 91.1%, AUC 0.94 | Kaddour et al., Journal of Voice 2026 |
| Kannada VHI-10 cut-off | 6.5 or above; AUC 0.926, n = 273 | Benoy et al., Journal of Voice 2026 |
| Persian VHI-10 cut-off | 4.5; AUC 0.989, accuracy 97.33%, alpha 0.961 | Nazari et al., Journal of Voice 2026 |
| Translated VHI-30 versions in the systematic review | 12 articles (1997-2014) | Seifpanahi et al., Iranian Journal of Public Health 2015 |
Context note: the Persian VHI-10 abstract assigns its group means in an order that looks reversed (higher mean for healthy speakers), so we report only its cut-off and accuracy figures (Nazari et al., Journal of Voice 2026). Most recent available data for a cross-language systematic review: Seifpanahi et al., 2015.
7. Singers, Teachers and Children: Population-Specific Benchmarks
The general-population norm does not transfer to professional voice users. Vocally healthy singers average 9.13 out of 40 on the singing-specific SVHI-10, and the abnormal line sits at 20 or above (Gochman et al., Journal of Voice 2025). A singer’s baseline is not zero, and the change needed to matter is larger too (9.5 points, see section 4 for the general-population figures) (Perrin et al., Laryngoscope 2025).
Teachers carry a visibly heavier voice burden than other workers. A UK survey covering 210 primary and 244 secondary school teachers and 304 non-teachers found voice problems reported by 30.0% of teachers versus 9.0% of non-teachers (Gadepalli et al., Clinical Otolaryngology 2019). For the structural side of the same occupational risk, see our roundup on vocal fold nodules in teachers and singers. At the other end of the age range, children score close to zero (De Maio et al., Journal of Voice 2024).
| Metric | Value | Source |
|---|---|---|
| Mean SVHI-10 in healthy singers | 9.13 (SD 5.15), n = 149 | Gochman et al., Journal of Voice 2025 |
| SVHI-10 abnormal threshold (mean + 2 SD = 19.43) | 20 or above | Gochman et al., Journal of Voice 2025 |
| SVHI-10 abnormal threshold in clinical reference sheet | 14 or above | University of Sydney Voice Lab 2025 |
| Persian SVHI-10 cut-off | 14.5; sensitivity 84.85%, specificity 94.59% | Nazari et al., Journal of Voice 2026 |
| Teachers vs non-teachers reporting voice problems | 30.0% vs 9.0% | Gadepalli et al., Clinical Otolaryngology 2019 |
| Teachers vs non-teachers missing work for voice | 12.8% vs 2.0% | Gadepalli et al., Clinical Otolaryngology 2019 |
| University staff with an abnormal VHI-10 | 41.0% (164/400) | Younis et al., BMC Public Health 2026 |
| Mean Children VHI-10 in healthy children (cut-off) | 0.26, SD 0.74 (cut-off 2.5), n = 201 | De Maio et al., Journal of Voice 2024 |
Context note: in the university staff study, drinking fewer than three cups of water per day carried an adjusted odds ratio of 5.09 (2.39-10.87) for voice disorder, and teaching-staff job rank an aOR of 2.67 (1.19-5.96) (Younis et al., BMC Public Health 2026). The parent-proxy Children VHI-10 averaged 0.15 (SD 0.49) with a cut-off of 1.5.
Summary: Voice Handicap Index by the Numbers
| Metric | Value | Source |
|---|---|---|
| Mean VHI-10, adults without voice complaints | 2.83 (SD 3.93) | Arffa et al., Journal of Voice 2012 |
| Abnormal VHI-10 threshold (US norm) | above 11 | Arffa et al., Journal of Voice 2012 |
| Mean VHI-10, academic voice clinic (share above 11) | 19.5 (80%) | Misono et al., Journal of Voice 2014 |
| Mean VHI-10, Parkinson’s disease before DBS | 22.7 | Tuysuz et al., Eur Arch Otorhinolaryngol 2026 |
| Lifetime share of population with dysphonia | nearly one-third | AAO-HNS, Hoarseness Guideline 2018 |
| Items: original VHI vs VHI-10 | 30 vs 10 | Jacobson et al., AJSLP 1997; Rosen et al., Laryngoscope 2004 |
| VHI vs VHI-10 correlation | greater than .90 | Rosen et al., Laryngoscope 2004 |
| Significant change on the 30-item VHI | 18 points | Jacobson et al., AJSLP 1997 |
| VHI-10 MCID, vocal fold paralysis | decrease of 4 | Young et al., Laryngoscope 2018 |
| VHI-10 MID, general voice clinic | 6 points | Misono et al., JAMA Otolaryngology 2017 |
| SVHI-10 MCID | 9.5 (10 or more meaningful) | Perrin et al., Laryngoscope 2025 |
| Mean SVHI-10, healthy singers | 9.13 | Gochman et al., Journal of Voice 2025 |
| SVHI-10 internal consistency | alpha .94 | Cohen et al., Laryngoscope 2009 |
| Montenegrin VHI-10 test-retest ICC | 0.98 | Zvrko and Kadic, Arh Hig Rada Toksikol 2022 |
| Brazilian Portuguese VHI-10 cut-off | 7.5 (975 subjects) | Behlau et al., Journal of Voice 2016 |
| European Portuguese VHI-10 cut-off | 5.5 | Guimaraes et al., Journal of Voice 2026 |
| Algerian Arabic VHI-10 cut-off | 12.5 | Kaddour et al., Journal of Voice 2026 |
| University staff with abnormal VHI-10 | 41.0% | Younis et al., BMC Public Health 2026 |
| Healthy children, Children VHI-10 mean | 0.26 | De Maio et al., Journal of Voice 2024 |
Methodology and Sources
All figures come from peer-reviewed abstracts, open-access full texts or institutional reference documents read in October 2026. Medical content is reported as published and is not clinical advice.
- Jacobson et al., The Voice Handicap Index (VHI): Development and Validation, American Journal of Speech-Language Pathology 1997
- Rosen et al., Development and Validation of the Voice Handicap Index-10, Laryngoscope 2004
- Arffa et al., Normative Values for the Voice Handicap Index-10, Journal of Voice 2012
- University of Sydney Voice Lab, Patient Reported Outcome Measures Normative Data Reference Sheet v3.0, April 2025
- Misono et al., Psychosocial Distress in Patients Presenting With Voice Concerns, Journal of Voice 2014
- Misono et al., Minimal Important Difference in Voice Handicap Index-10, JAMA Otolaryngology-Head and Neck Surgery 2017
- Young et al., Minimal Clinically Important Difference of VHI-10 in Vocal Fold Paralysis, Laryngoscope 2018
- Young et al., Characteristics and Voice Outcomes of Ulcerative Laryngitis, Journal of Voice 2020
- Nguyen-Feng et al., Online Intervention for Muscle Tension Dysphonia: A Pilot Randomized Clinical Trial, JAMA Otolaryngology 2025
- Tuysuz et al., Impact of Subthalamic Nucleus Deep Brain Stimulation on Voice and Speech in Parkinson’s Disease, European Archives of Oto-Rhino-Laryngology 2026
- Cohen et al., Development and Validation of the Singing Voice Handicap-10, Laryngoscope 2009
- Gochman et al., Prospective Determination of Normative Value of the SVHI-10, Journal of Voice 2025
- Perrin et al., Singing Voice Handicap Index-10 Minimal Clinically Important Difference, Laryngoscope 2025
- Zvrko and Kadic, Validity, Internal Consistency, and Test-Retest Reliability of the Montenegrin VHI-10, Arhiv za higijenu rada i toksikologiju 2022
- Behlau et al., Efficiency and Cutoff Values of Self-Assessment Instruments on the Impact of a Voice Problem, Journal of Voice 2016
- Guimaraes et al., Voice Handicap Index: Cut-Off Points for Screening European Portuguese With Voice Disorders, Journal of Voice 2026
- Polianskis et al., Validation of the Lithuanian Version of the VHI-10, Journal of Voice 2026
- Kaddour et al., Validation of the Algerian Arabic VHI-10, Journal of Voice 2026
- Benoy et al., Psychometric Properties of the VHI-10 in Kannada, Journal of Voice 2026
- Nazari et al., Adaptation and Validation of the Persian VHI-10, Journal of Voice 2026
- Nazari et al., Adaptation and Validation of the Persian SVHI-10, Journal of Voice 2026
- Seifpanahi et al., Translated Versions of Voice Handicap Index (VHI)-30 across Languages: A Systematic Review, Iranian Journal of Public Health 2015
- Gadepalli et al., Voice Burden in Teachers and Non-Teachers in a UK Population, Clinical Otolaryngology 2019
- Younis et al., Understanding Voice Disorders Among University Staff: Prevalence and Associated Factors, BMC Public Health 2026
- De Maio et al., Normative Values for the Children Voice Handicap Index-10 (CVHI-10 and CVHI-10-P), Journal of Voice 2024
- Stachler et al., AAO-HNS Clinical Practice Guideline: Hoarseness (Dysphonia) (Update), Otolaryngology-Head and Neck Surgery 2018
- Data watch: the two VHI-10 MCID estimates (4 points in vocal fold paralysis, 6 points in general clinic patients) come from different populations and should not be averaged; the general-population MID authors themselves called for more studies. The SVHI-10 abnormal threshold conflicts between sources: 20 or above in the 2025 prospective normative study versus 14 or above in the University of Sydney Voice Lab sheet, which cites the 2009 development paper. The Sydney sheet dates the Rosen VHI-10 paper to 2009, while the journal record is 2004. Several 2026 language validations (Lithuanian, Algerian Arabic) were published ahead of print and may change slightly in final form. The US normative values (2012) and the original VHI-10 validation (2004) have not been re-derived in a new US sample, and the only cross-language systematic review covers studies up to 2014.
Last updated: October 18, 2026. We update this roundup quarterly, and the next refresh is expected when the Lithuanian and Algerian Arabic VHI-10 validations appear in a final Journal of Voice issue.