Objective speech measures capture depressive symptoms and associated cognitive difficulties.

Area of Science:

• Psychiatry
• Computational Linguistics
• Neuroscience

Background:

• Current depression assessment relies on subjective scales like the Hamilton Depression Rating Scale (HAMD-17).
• Objective biomarkers are needed for accurate diagnosis and monitoring of major depressive disorder (MDD).
• Speech patterns may reflect underlying emotional and cognitive states in psychiatric conditions.

Purpose of the Study:

• To investigate speech features as objective biomarkers for depressive symptoms.
• To explore the association between speech characteristics and cognitive difficulties in major depressive disorder (MDD).
• To determine if speech analysis can differentiate individuals with MDD from controls.

Main Methods:

• Sixty-six individuals with MDD and 54 controls provided speech samples in response to a standardized prompt.
• Linguistic (valence, emotional intensity, agency) and acoustic (pitch, speech rate, pausing) features were extracted using natural language processing.
• Analysis included ANCOVA for group comparisons and linear regression for associations with depression severity and executive function.

Main Results:

• Individuals with MDD exhibited more negative valence, higher emotional intensity, and less agency in their speech compared to controls.
• Depressed participants also showed lower pitch, slower speech rate, and increased pausing.
• Several speech features were significantly associated with executive function performance, though not with depression severity (HAMD-17).

Conclusions:

• Speech features, including linguistic and acoustic characteristics, can objectively differentiate individuals with major depressive disorder (MDD).
• These speech markers show potential for detecting depressive symptoms and associated executive function deficits.
• Speech analysis presents a scalable and objective approach for psychiatric assessment.