Customize your JAMA Network experience by selecting one or more topics from the list below.
Researchers seeking a less invasive, more objective way to classify posttraumatic stress disorder (PTSD) have become interested in speech-based biomarkers, building on previous work associating vocal changes with mood disorders. The current gold standard method for diagnosing PTSD, a structured interview called the Clinician-Administered PTSD Scale (CAPS), is lengthy, challenging for patients, and subject to the clinician’s interpretation.
“Because measuring voice qualities is noninvasive, inexpensive, and might be done over the phone, many labs have sought to design speech-based diagnostic tools,” said Charles Marmar, MD, chair of the department of psychiatry at NYU Langone Health. In a recent study Marmar led, an artificial intelligence-trained algorithm differentiated US veterans with and without PTSD with 89.1% accuracy based on 18 voice markers.
The classifier culled these markers from recorded CAPS interviews with 129 male Iraq and Afghanistan war veterans, 52 of whom had clinician-diagnosed PTSD. Slower and more monotonous speech with less change in tonality and excitability emerged as PTSD voice markers in the study, which appeared in the journal Depression and Anxiety. If further validated in a larger independent data set, the panel of voice markers could one day be used to develop a widely accessible tool for assessing PTSD.
Future research will test the classifier’s ability to distinguish PTSD from major depressive disorder, which may have similar voice markers. Research is also needed to determine if people with PTSD demonstrate the same vocal changes outside of a stressful CAPS interview.
Abbasi J. Using Speech Markers to Diagnose PTSD. JAMA. 2019;321(22):2155. doi:10.1001/jama.2019.7550
Create a personal account or sign in to: