Skip to content

Investigating the Effects of Rhythm and Entrainment on Fluency in People With Aphasia

Investigating the Effects of Rhythm and Entrainment on Fluency in People With Aphasia

Status
Completed
Phases
NA
Study type
Interventional
Source
ClinicalTrials.gov
Registry ID
NCT05248295
Enrollment
75
Registered
2022-02-21
Start date
2018-12-04
Completion date
2023-06-28
Last updated
2025-05-18

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Aphasia, Apraxia of Speech

Brief summary

Speaking in unison with another person is included as a part of many treatment approaches for aphasia. It is not well understood why and how this technique works. One goal of this study is to determine who benefits from speaking in unison, and what characteristics of speech are most helpful. Another goal is to investigate a possible mechanism for this benefit: why does speaking in unison help? A possible mechanism for this benefit is examined, by testing whether the degree of alignment of a person's speech with that of another speaker can account for unison benefit.

Detailed description

Aphasia is an acquired language disorder, most commonly due to stroke. It can affect an individual's ability to speak, understand spoken language, read, and write. Many treatments designed to improve spoken language in persons with aphasia (PWAs) use unison speech, having the person with aphasia speak along with the clinician or with a recording. One goal of this study is to determine who benefits from speaking in unison, and what characteristics of speech help them the most. Another goal is to investigate a possible mechanism for this benefit: why does speaking in unison help? This knowledge is important because understanding who benefits from this commonly used and potentially powerful therapy component, under which conditions they benefit, and why they do, is critical for customizing therapy so it can be as effective and efficient as possible. Unison speech is conducted using one of two different timing patterns: (1) a natural conversational pattern, which is used in everyday conversations, or (2) a metrical pattern, which follows a beat-based timing framework, as in songs or some poems. In either case, precisely aligning one's speech with that of another person (i.e., entraining one's speech) requires prediction: each speaker must plan their own speech motor commands before having heard the other speaker say the words they are planning. Natural conversational timing requires the speaker to make use of knowledge about language, particularly grammar, to align with the other person. In contrast, a metrical pattern allows a speaker to predict speech timing without relying heavily on language-based knowledge. Given that many PWAs have impaired grammar, we hypothesize that most PWAs will benefit more from speaking in unison to sentences with metrical vs. conversational timing patterns. However, there is great variation in linguistic, motor speech, and timing skills across PWAs, so metrical and conversational timing patterns are likely to have different degrees of effectiveness for different individuals. Results from this study will demonstrate how individual characteristics and speech timing affect whether or not a person with aphasia benefits from speaking in unison. Results will also indicate whether a speaker's ability to predict speech timing is necessary for a benefit of unison speech. Prediction ability will be measured by how closely the speaker aligns their speech with a spoken model.

Interventions

BEHAVIORALUnison speech (vs. solo)

Participants will repeat sentences in four conditions, in a 2x2 design with the factors unison vs. solo repetition, and metrical vs. conversational speech timing. Measures of speech accuracy and timing will be collected.

BEHAVIORALMetrical timing (vs. conversational)

This is the secondary contrast in the 2x2 design described above.

Sponsors

National Institute on Deafness and Other Communication Disorders (NIDCD)
CollaboratorNIH
MGH Institute of Health Professions
Lead SponsorOTHER

Study design

Allocation
NON_RANDOMIZED
Intervention model
PARALLEL
Primary purpose
BASIC_SCIENCE
Masking
NONE

Eligibility

Sex/Gender
ALL
Age
18 Years to 80 Years
Healthy volunteers
Yes

Inclusion criteria

* Native-speaker fluency in American English (prior to stroke for people with aphasia) * Controls must report no history of speech, language, neurological disorders, or stroke * People with aphasia must be at least 6-months post-stroke, and aphasia must be due to stroke

Exclusion criteria

* Inadequate hearing ability to reliably complete task: fail hearing screen * Inadequate cognitive ability to understand and remember task: fail cognition screening (different measures for controls and people with aphasia) * Inadequate speech repetition ability to complete task, or to be considered a control: fail speech repetition screening (different thresholds for controls and people with aphasia) * Inadequate auditory comprehension ability to understand task: fail auditory comprehension screen (people with aphasia only)

Design outcomes

Primary

MeasureTime frameDescription
Percent Syllables Correct1 day study visitThe percentage of syllables correctly repeated from the target sentences will be computed for each contrast, Unison vs. Solo and Metrical vs. Conversational, capturing all 4 conditions in the 2x2 design (Unison Metrical, Unison Conversational, Solo Metrical, Solo Conversational). A protocol will be used to score syllables for correctness.

Secondary

MeasureTime frameDescription
Mean Syllable Offset From Model1 day study visitOf the syllables attempted by the participant, the mean offset from the corresponding target syllable in the model is computed for each syllable for each participant using the formula (p - d) - s, where p is the timing of voicing onset in the participant response, d is the timing of voicing onset in the model, and s is the line delay in the audio equipment. Means are computed for each participant and each condition. Note this measure can only be computed in the Unison conditions (Unison Metrical and Unison Conversational), since the Solo conditions do not have a simultaneous spoken model to which the participant is trying to align their speech.

Countries

United States

Participant flow

Participants by arm

ArmCount
People With Aphasia
Adults with stroke-based aphasia
23
Controls
Adults without a history of speech, language, or neurological disorder, or stroke
52
Total75

Withdrawals & dropouts

PeriodReasonFG000FG001
Overall StudyData loss: audio recording error11
Overall Studyfound ineligible41
Overall StudyWithdrawal by Subject01

Baseline characteristics

CharacteristicPeople With AphasiaControlsTotal
Age, Continuous59.0 years
STANDARD_DEVIATION 7.6
42.4 years
STANDARD_DEVIATION 18.4
47.5 years
STANDARD_DEVIATION 17.7
Hearing Screen22 Participants52 Participants74 Participants
Race/Ethnicity, Customized
Asian, ethnicity unknown
0 Participants0 Participants0 Participants
Race/Ethnicity, Customized
Asian, Hispanic
0 Participants1 Participants1 Participants
Race/Ethnicity, Customized
Asian, not Hispanic
0 Participants6 Participants6 Participants
Race/Ethnicity, Customized
Black, Hispanic
0 Participants0 Participants0 Participants
Race/Ethnicity, Customized
Black, not Hispanic
3 Participants5 Participants8 Participants
Race/Ethnicity, Customized
Black, unknown ethnicity
0 Participants0 Participants0 Participants
Race/Ethnicity, Customized
race not specified, ethnicity unknown
0 Participants1 Participants1 Participants
Race/Ethnicity, Customized
White, Hispanic
0 Participants1 Participants1 Participants
Race/Ethnicity, Customized
White, not Hispanic
20 Participants36 Participants56 Participants
Race/Ethnicity, Customized
White, unknown ethnicity
0 Participants2 Participants2 Participants
Sex: Female, Male
Female
9 Participants31 Participants40 Participants
Sex: Female, Male
Male
14 Participants21 Participants35 Participants

Adverse events

Event typeEG000
affected / at risk
deaths
Total, all-cause mortality
0 / 75
other
Total, other adverse events
0 / 75
serious
Total, serious adverse events
0 / 75

Outcome results

Primary

Percent Syllables Correct

The percentage of syllables correctly repeated from the target sentences will be computed for each contrast, Unison vs. Solo and Metrical vs. Conversational, capturing all 4 conditions in the 2x2 design (Unison Metrical, Unison Conversational, Solo Metrical, Solo Conversational). A protocol will be used to score syllables for correctness.

Time frame: 1 day study visit

ArmMeasureGroupValue (MEAN)Dispersion
ControlsPercent Syllables CorrectUnison99.22 percentageStandard Deviation 1.08
ControlsPercent Syllables CorrectSolo99.46 percentageStandard Deviation 1.18
ControlsPercent Syllables CorrectMetrical99.65 percentageStandard Deviation 0.67
ControlsPercent Syllables CorrectConversational99.03 percentageStandard Deviation 1.39
People With AphasiaPercent Syllables CorrectConversational56.50 percentageStandard Deviation 28.98
People With AphasiaPercent Syllables CorrectUnison64.26 percentageStandard Deviation 23.13
People With AphasiaPercent Syllables CorrectMetrical60.47 percentageStandard Deviation 29.68
People With AphasiaPercent Syllables CorrectSolo52.71 percentageStandard Deviation 33.57
Comparison: People with aphasia (PWA) were not directly compared to controls since the finding of a lower % syllables correct in any condition in PWA would be trivial. Instead, within-groups analyses were conducted to understand how the experimental variables affected syllable accuracy within each group.p-value: 0.077Regression, Logistic
p-value: >0.1Regression, Logistic
p-value: <0.001Regression, Logistic
Comparison: Separate regression models were used within each group. Here, the results are reported for the Control group.p-value: 0.036Regression, Logistic
p-value: <0.001Regression, Logistic
p-value: >0.1Regression, Logistic
Secondary

Mean Syllable Offset From Model

Of the syllables attempted by the participant, the mean offset from the corresponding target syllable in the model is computed for each syllable for each participant using the formula (p - d) - s, where p is the timing of voicing onset in the participant response, d is the timing of voicing onset in the model, and s is the line delay in the audio equipment. Means are computed for each participant and each condition. Note this measure can only be computed in the Unison conditions (Unison Metrical and Unison Conversational), since the Solo conditions do not have a simultaneous spoken model to which the participant is trying to align their speech.

Time frame: 1 day study visit

Population: Acoustic data were inadequate for 1 participant with aphasia (could not be reliably labeled due to vocal quality, resulting in too few data points for analysis for this participant).

ArmMeasureGroupValue (MEAN)Dispersion
ControlsMean Syllable Offset From ModelMetrical-.016 secondsStandard Deviation 0.028
ControlsMean Syllable Offset From ModelConversational0.006 secondsStandard Deviation 0.04
People With AphasiaMean Syllable Offset From ModelMetrical0.030 secondsStandard Deviation 0.052
People With AphasiaMean Syllable Offset From ModelConversational0.154 secondsStandard Deviation 0.188
Comparison: People with aphasia were not directly compared to Controls since the Control group is a context-providing reference group rather than a true comparator. Instead, within-groups analyses were conducted to understand how the experimental variable affected timing alignment in each group.p-value: <0.001Regression, Linear
Comparison: People with aphasia were not directly compared to Controls since the Control group is a context-providing reference group and not a true comparator. Instead, within-groups analyses were conducted to understand how experimental variables affected timing alignment in each group.p-value: <0.001Regression, Linear

Source: ClinicalTrials.gov · Data processed: Feb 4, 2026