Development of rhythmic accent training for speech fluency in Mandarin-speaking individuals with post-stroke non-fluent aphasia: a pilot study

Background Rhythm-based approaches may facilitate speech production in non-fluent aphasia, but intervention protocols adapted to the linguistic characteristics of Mandarin remain underdeveloped.Aims To develop a rhythmic accent training (RAT) protocol for Mandarin speakers with post-stroke non-fluent aphasia and to preliminarily evaluate its immediate effects on fluency-related performance.Methods Twenty-four Mandarin-speaking participants with post-stroke non-fluent aphasia completed three sequential sessions within a repeated-measures protocol-development study. Session 1 compared vowel-accented and syllable-accented training. Session 2 examined whether rhythm-synchronized left-hand tapping enhanced the effects of accent training. Based on these findings, a RAT protocol was developed. Session 3 examined immediate changes in speech rate, articulation rate, pause measures, and articulation accuracy associated with RAT.Results Syllable-accented training yielded larger improvements than vowel-accented training in fluency-related measures and articulation accuracy. Adding rhythm-synchronized left-hand tapping further enhanced fluency outcomes. In the final session, participants showed significant increases in speech rate, articulation rate, and articulation accuracy, together with significant reductions in total pauses, unnatural pauses, mean pause duration, and mean unnatural pause duration after RAT.Conclusions These findings support the stepwise development of RAT and suggest that rhythm-based training incorporating syllable-level linguistic units and left-hand tapping may improve immediate fluency-related performance in structured repetition tasks in Mandarin speakers with post-stroke non-fluent aphasia. Because RAT combines meaningful language practice with rhythmic speech-motor cueing, the relative contributions of these two components cannot be separated by the present outcome measures. The preliminary results are limited to the immediate effects on trained sentences and should not be generalized to spontaneous speech or functional communication without further validation.