Music-based interventions for aphasia could act through a motor-speech mechanism: a systematic review and case–control analysis of published individual participant data

Background: Melodic Intonation Therapy, a music-based intervention for the recovery of oral language production in aphasia, has been shown to be particularly effective in patients with Broca's aphasia compared to other aphasia subtypes. It has been suggested that this therapy might improve language output by acting on motor-speech deficits often associated with Broca's aphasia. In this article, we examine the relevance of a motor-speech mechanism for music-based interventions designed to improve verbal expression in patients with any type of aphasia. Aim: To test the association between the presence of motor-speech disorders (MSDs) and improvement with music-based protocols targeting verbal expression in participants with aphasia. Methods and procedures: We conducted a systematic review of publications reporting language production outcomes following a music-based intervention in participants with aphasia and performed a case-control analysis on extracted individual participant data (IPD). The databases PubMed, MEDLINE (1800 to 9 March 2018), and PsycINFO (1806 to March 2018) were screened, followed with cross-referencing. We recorded data at the level of study and, when possible, at the IPD level. When not explicitly reported, we applied a series of heuristics to infer the presence/absence of an MSD in participants. Binomial logistic regressions were performed to ascertain the effects of the presence of an MSD, aphasia severity, treatment duration (in weeks), and treatment intensity (hours/week) on the likelihood that participants would show a speech or a language improvement following intervention. Outcomes & Results: Forty original articles were included in this review. Twenty-two reported sufficient details to be included in our IPD analysis, for a total sample of 105 participants. Most interventions included some sort of singing as their primary music-based facilitation technique for language production. For speech improvement, statistically significant predictor variables were the presence of an MSD and treatment intensity. For language improvement, statistically significant predictor variables were the presence of an MSD, treatment intensity, and duration. Severity of aphasia was not associated with the likelihood of speech or language improvement. Conclusion: Music-based interventions for language production in aphasia may act via a motor-speech mechanism. We suggest that music and singing-based therapies might be further investigated as treatment options for patients with MSDs, whether associated with aphasia or not.