Unraveling the Associations Between Voice Pitch and Major Depressive Disorder: A Multisite Genetic Study

Area of Science:

• Neuroscience
• Psychiatry
• Computational Linguistics

Background:

• Major depressive disorder (MDD) lacks definitive diagnostic biomarkers.
• Objective voice analysis may offer potential for identifying MDD.
• Distinguishing MDD predisposition from current symptoms requires specific markers.

Purpose of the Study:

• To identify voice biomarkers for major depressive disorder (MDD).
• To differentiate between biomarkers for MDD predisposition and current depressive symptoms.
• To assess the diagnostic utility of identified voice features in MDD classification.

Main Methods:

• A two-stage meta-analytic design was employed to identify vocal pitch features associated with MDD.
• A multisite case-control cohort study of Chinese women with recurrent depression was conducted.
• Genotype data was utilized to investigate the genetic correlation with MDD.

Main Results:

• Sixteen vocal features were robustly associated with MDD, with moderate to large effect sizes.
• Eleven voice features significantly correlated with current depressive symptoms.
• Voice features reflecting slower pitch change and lower pitch achieved an AUC-ROC of 0.90 for MDD classification.

Conclusions:

• Specific vocal features serve as reliable biomarkers for major depressive disorder (MDD).
• These voice biomarkers can differentiate MDD predisposition from current depressive states.
• Vocal analysis holds significant potential for clinical and research applications in MDD.