Objective: To perform a meta-analysis evaluating acoustic speech features as potential biomarkers of Huntington’s disease by comparing individuals with manifest HD and premanifest HD to healthy control participants.

Background: Speech abnormalities are increasingly recognized as potential biomarkers of neurodegeneration in Huntington’s disease (HD). Acoustic measures such as harmonics-to-noise ratio (HNR), shimmer, and speech rate may capture subtle motor and phonatory deficits associated with striatal and cortical dysfunction. However, findings across studies remain heterogeneous and have not been quantitatively synthesized.

Method: A meta-analysis was conducted examining three commonly reported acoustic speech measures: harmonics-to-noise ratio (HNR), shimmer, and speech rate. Effect sizes were calculated comparing (1) manifest HD vs healthy controls and (2) premanifest HD vs healthy controls. Random-effects models were used, and heterogeneity was assessed using the I² statistic.

Results: For manifest HD vs controls, all three acoustic measures demonstrated significant differences. HNR showed a moderate pooled effect size (g = 0.57, 95% CI 0.55–0.60, p = 0.002), indicating reduced voice quality in HD compared with controls. Shimmer demonstrated a large negative effect size (g = −1.15, 95% CI −1.82 to −0.49, p = 0.029), reflecting increased amplitude instability in HD. Speech rate showed a large effect size (g = 1.45, 95% CI 1.165–1.744, p = 0.002), with controls consistently exhibiting faster speech than individuals with HD. Heterogeneity was negligible across analyses (I² = 0%).

For premanifest HD vs controls, none of the acoustic measures showed statistically significant differences. HNR demonstrated a small-to-moderate pooled effect (g = 0.469, 95% CI −3.754 to 4.692, p = 0.392), shimmer showed a small negative effect (g = −0.30, 95% CI −1.26 to 0.65, p = 0.156), and speech rate showed a very small effect (g = 0.11, 95% CI −0.58 to 0.80, p = 0.289).

Conclusion: Acoustic speech measures, particularly shimmer and speech rate, demonstrate robust differences between individuals with manifest HD and healthy controls and may serve as promising digital biomarkers of disease expression. However, current evidence does not support reliable detection of premanifest HD using these measures alone. Larger longitudinal studies are needed to determine whether speech biomarkers can detect early prodromal changes and track disease progression.

« Back to 2026 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/speech-changes-as-biomarkers-in-huntingtons-disease-a-meta-analysis/