Evaluation of the Frequency Following Response as a Predictive Biomarker of Response to Cognitive Training in Schizophrenia

Neurophysiological biomarkers of auditory processing show promise predicting outcomes following auditory-based targeted cognitive training (TCT) interventions in patients with schizophrenia, but the viability of the early frequency following response (FFR) as a biomarker has yet to be examined, despite its ecological and face validity for auditory-based interventions. FFR is an event-related potential (ERP) that reflects the earliest stages of auditory processing. We predicted that schizophrenia patients would show acute- and longer-term FFR malleability in the context of auditory-based TCT. Fifty-two schizophrenia patients were randomized to either TCT (n = 30) or treatment as usual (TAU; n = 22), and electroencephalography was recorded during rapid presentation of an auditory speech stimulus before treatment, after one hour of training, and after 30 hours of training. Patients in the TCT group did not show changes in FFR after training, but FFR measurements in the TAU group diminished over time. FFR was positively associated with performance on a measure of single word-in-noise perception in the TCT group, and with performance on a measure of sentence-in-noise perception in both TCT and TAU groups. Psychometric reliability analyses of FFR scores indicated high internal consistency but low test-rest reliability between sessions one hour and 12 weeks apart. While the present study did not find evidence of FFR malleability or utility in predicting response to auditory-based TCT, significant associations were observed between FFR measures and behavioral measures of speech discriminability. These findings support the dissociation between measures of speech discriminability along the hierarchy of cortical and subcortical early auditory information processing in schizophrenia.