Researchers of this study investigated whether quantitative speech characteristics could serve as reliable diagnostic indicators for schizophrenia. Researchers analyzed speech from 43 patients with schizophrenia or schizoaffective disorder and 46 healthy controls, alongside cognitive and symptom measures. Patients with schizophrenia exhibited more aberrant speech patterns across five categories—utterances, single words, speaking rate, conversational turns, and formulation errors—though not pauses. Using two machine learning models, the study identified 21 speech variables that classified schizophrenia with 90% to 100% specificity and 80% to 90% sensitivity. Some speech features were also selectively linked to symptom domains such as positive symptoms, disorganization, excitement, and formal thought disorder.

The findings support the presence of consistent and measurable speech abnormalities in schizophrenia that differ from those in healthy individuals and may aid diagnosis. Quantitative speech assessment offers a promising, objective tool that could complement traditional clinical evaluations and improve diagnostic accuracy. Importantly, the study demonstrated that multiple speech parameters—not just pauses or word use—can distinguish schizophrenia from other conditions. Further research, especially with automated tools and larger samples, is recommended to validate these findings and support their integration into clinical practice.

Reference: Tan EJ, Meyer D, Neill E, Rossell SL. Investigating the diagnostic utility of speech patterns in schizophrenia and their symptom associations. Schizophr Res. 2021;238:91-98. doi: 10.1016/j.schres.2021.10.003.