Characterizing speech heterogeneity in schizophrenia-spectrum disorders

Area of Science:

• Psychiatry and Neuroscience
• Computational Linguistics
• Clinical Psychology

Background:

• Schizophrenia-spectrum disorders (SSD) exhibit significant heterogeneity in clinical presentation and outcomes.
• Speech anomalies are a recognized symptom of SSD, yet their potential for identifying patient subgroups remains underexplored.
• Understanding speech heterogeneity may offer insights into distinct SSD subtypes.

Purpose of the Study:

• To investigate speech heterogeneity in patients with SSD using acoustic analysis.
• To identify distinct speech clusters within the SSD population.
• To examine the relationship between identified speech clusters and clinical characteristics, including symptom severity and cognitive function.

Main Methods:

• Acoustic speech analysis was performed on samples from 142 patients with SSD and 147 healthy controls.
• Hierarchical clustering was applied to acoustic speech markers to identify patient subgroups.
• Symptom severity was assessed using the Positive and Negative Syndrome Scale, and cognition was evaluated with the Brief Assessment of Cognition for Schizophrenia.

Main Results:

• Three distinct speech clusters were identified in the SSD group: mild disturbances, fragmented speakers, and prolonged pausers.
• Severely impaired speech clusters (fragmented speakers, prolonged pausers) showed greater cognitive dysfunction compared to the mildly impaired group.
• Prolonged pausing was specifically associated with memory deficits, chronic psychosis, and refractory symptoms.

Conclusions:

• Speech clustering identified clinically relevant subgroups within SSD with differing disease trajectories and cognitive profiles.
• Distinct speech patterns in SSD correlate with specific cognitive impairments and symptom characteristics.
• Identifying speech-based subtypes can inform tailored interventions and improve long-term outcomes for individuals with schizophrenia-spectrum disorders.