Talking While Walking After Concussion: Acute Effects of Concussion on Speech Pauses and Gait Speed

Abstract

Background

Deficits in dual-tasks (DT) are frequently observed post-concussion (ie, mild Traumatic Brain Injury). However, traditional DT may not be relevant to daily life. Walking while talking elicits DT costs in healthy adults and is part of daily life.

Objective

We investigated the effect of concussion on walking with extemporaneous speech and explored relationships between DT and acute symptoms.

Methods

Participants with recent concussion (<14 days post-injury) and controls completed 3 tasks: single-task gait without speaking (ST_G), single-task speaking without walking (ST_S), and walking while speaking (DT). Silent pauses in speech audio reflected cognitive performance, and gait was quantified using inertial sensors. We used linear mixed models to compare groups and conditions and explored associations with self-reported symptoms.

Results

Both concussion (n = 19) and control (n = 18) groups exhibited longer speech pauses (P < .001), slower walking speeds (P < .001), and slower cadence (P < .001) during the DT compared to ST conditions. There were no group differences or interactions for speech pauses (P > .424). The concussion group slowed down more during DT than the control group (group × task P = .032). Vestibular symptoms strongly associated with ST speech pause duration (ρ = .72), ST gait speed (ρ = −.75), and DT gait speed (ρ = −.78).

Conclusions

Extemporaneous speech is well-practiced but challenging to complete while walking post-concussion. Strong associations between DT outcomes and vestibular-related symptoms suggest DT deficits vary with post-concussion symptomology. DT deficits may be deleterious to daily tasks post-concussion.

Keywords

mild traumatic brain injury mobility ecological dual-tasks speech production locomotion

Introduction

Cognitive and motor problems, such as balance and gait deficits, are common following concussions.^1
-3 Performing cognitive-motor dual-tasks (DT)—completing simultaneous cognitive and motor tasks at the same time—elicits slower gait speeds, shorter strides, and longer stride times in people after a concussion.^2,4
-7 DT deficits arise from an inability to handle competing cognitive and motor demands and are thought to result from deficits in attentional control and a loss of automaticity—defined as an increased reliance on cognitive resources for relatively automatic tasks (eg, walking).^8,9 After a concussion, DT deficits can remain for months and persist long after the resolution of other signs and symptoms.^5,10,11 Further, DT deficits are reported in people with chronic, persisting post-concussion symptoms; people with persisting concussion-related symptoms prioritize the cognitive task, exhibiting higher costs to gait performance than cognitive task performance while carrying out both simultaneously, referred to as a “posture second” strategy.^12,13 The “posture second” is a compensation that reflects an inability to appropriately allocate resources to motor performance when faced with a cognitive challenge.^14,15 Yet, the consequences of DT deficits during daily life remain unclear because prior work has typically studied cognitive tasks that are irrelevant to daily life.¹⁶

Ideally, studies of cognitive-motor DT effects should quantify both the cognitive and the motor task with equivalent resolution to examinate the potential for bidirectional interference.¹⁷ Unfortunately, prior DT paradigms have used contrived arithmetic, alphabet recitation, memorization, visuospatial, and auditory/visual Stroop tasks when testing concussion populations that are not measured continuously.¹⁶ Although these tasks are easy to administer and score, they may not represent the cognitive demands of everyday life, like walking and talking. Further, laboratory-based cognitive tests may be affected by individual differences in education/socio-economic background, mathematics-related anxiety or skill-level, comfort with public speaking, and task-engagement.¹⁸ Typical contrived tasks are vulnerable to compensatory strategies that may not reflect the function of targeted cognitive behaviors,¹⁹ and boast learning and practice effects that are difficult to control for in an experimental setting.²⁰

In contrast to traditional arithmetic or recitation tasks, extemporaneous speech is a cognitive task that is highly practiced, relevant to daily living, commonly performed during other motor tasks such as walking, and elicits DT costs (DTC) in healthy young adults.^18,21
-23 Extemporaneous speech involves complex cognitive processes (eg, language formulation, executive function, and processing speed) and demands ongoing cognitive processing and continuous language retrieval. Similar to gait, the cognitive load of extemporaneous speech can be obtained from measures of speed and fluency; speech rate per syllable decreases under high cognitive-linguistic demands, with longer and more prevalent silent pauses.²⁴ A natural pause during extemporaneous speech is around 150 to 250 ms,²⁵ and longer silent speech pauses serve as an indicator of language fluency.^26,27 Silent pauses are associated with cognitive planning and memory retrieval and reflect syntactic complexity in language formulation, and more frequent pauses at prosodic boundaries (ie, clauses and sentence breaks) are associated with reduced syntactic complexity.²⁷ Therefore, the frequency and duration of silent pauses during extemporaneous speech reflect cognitive-linguistic demands.^27,28

Speech deficits such as slower articulation rate have been documented in moderate-to-severe traumatic brain injuries (TBIs)²⁹ and speech pauses may be an indicator of mild TBI (ie, concussion). One study observed individuals exhibited more pauses and filler words during a picture description task acutely after concussion compared to their pre-concussion performance.³⁰ Separately, a machine learning-based analysis of speech patterns revealed high diagnostic accuracy in a small preliminary sample of athletes after concussion.³¹ Further, clinical care patterns highlight the role of speech and language deficits post-concussion; a chart review indicated 43% of pediatric patients were referred from a specialty concussion clinic to speech language pathologists to treat issues with communication, attention, and memory.³² Yet, it remains unclear how these concussion-related deficits in speech fluency, defined here as the frequency and duration of silent pauses in extemporaneous speech, interact with competing motor tasks, such as walking.

The purpose of our study was to investigate the effect of concussion on extemporaneous speech production during single-task (ST) and DT walking paradigms. Our primary hypotheses were that people recovering from concussion would exhibit impaired speech production, defined by more frequent and longer silent pauses during speech; that both groups would exhibit a decline in speech production when walking compared to sitting; and that those with a concussion would exhibit greater DTC to gait and speech compared to the control group, defined as slower DT gait speed or longer and more frequent DT speech pauses. As a secondary aim, we explored the association between self-reported post-concussion symptoms and the DT effects of walking and talking.

Materials and Methods

Participants

As part of a larger study, 22 participants with recent concussions (<14 days post-injury) and 19 healthy control participants were recruited and provided informed written consent for the study. All protocols were approved by the Institutional Review Board, and participants provided informed written consent in advance. Concussion participants were identified and recruited using electronic medical records indicating a recent concussion-related injury. Control participants were recruited from the local community using flyers, online postings, and public dissemination, and were age- and gender-matched to the concussion participants. Candidates were included if they had no history of neurological illness (eg, stroke), history of a major neurological condition (eg, epilepsy), major psychiatric disorders that required in-patient hospitalization, orthopedic conditions that would explain balance or gait issues, history of vestibular or orthostatic blood pressure problems, and no more than 3 concussions in their lifetime. Additionally, healthy control participants had no history of concussion or concussion symptoms in the 5 years prior to participation. Due to other components of the larger study, participants also had no contraindications for an magnetic resonance imaging.

Procedures

Participants completed a series of cognitive, balance, mobility, autonomic, and neuroimaging assessments as part of the larger study; however, the current analysis is focused on a subset of balance and cognitive tasks. All participants completed the Neurobehavioral Symptom Inventory (NSI) to identify self-reported symptoms.³³ Participants also completed the National Institutes of Health (NIH)-Toolbox Cognition battery as a computerized assessment of several aspects of cognition³⁴ including: Picture Vocabulary, Flanker Inhibitory Control and Attention, List Sorting, Dimensional Change Card Sort, Pattern Comparison, Picture Sequence Memory, Oral Reading Recognition. All NIH Toolbox tests were administered via an iPad with the test administrator in an isolated room. The Total Cognition Composite summary score was extracted as a global measure of cognition.

The cognitive-motor assessment included 3 tasks: single-task speaking while seated (ST_S), single-task gait without speaking (ST_G), and walking while speaking (DT). By design, continuous measures of both cognitive and motor tasks with equivalent resolution were selected to gain insight into dual-task effects on both cognitive and motor function. Participants were given a list of pre-designed topics and asked to select 5 topics from a list of 20 that they felt most comfortable discussing (eg, favorite childhood memory). Participants were instructed that they should select topics that they could talk about for 1 minute continuously and that it did not matter what they said, just that they kept talking the entire time. Next, participants were seated and given 1 prompt from the 5 selected topics (eg, “Tell me about your favorite childhood memory”). A timer was used to ensure all participants spoke for 1 minute. Following the ST_S, participants were instructed to walk back and forth between 2 lines spaced 20 m apart at their comfortable pace for 1 minute (ST_G). Following both ST conditions, participants completed a DT condition where they were assigned a new topic and instructed to repeat the walking task while talking about their given topic. Participants were not provided explicit instructions to prioritize either task.

Audio from each task was recorded using a lapel microphone and wireless transmitter (WMX-1, Movo Photo, Los Angeles CA, USA) connected to an iPad (eighth generation, Apple Inc.) such that audio and video were recorded simultaneously. Standard spatiotemporal measures of gait speed and cadence were obtained for each walking trial using inertial sensors (APDM Inc., Portland OR, USA) placed bilaterally on the feet, lumbar spine, forehead, and sternum. Inertial sensors recorded tri-axial acceleration and angular velocity, which was processed using validated, automated algorithms (Mobility Lab v2, APDM Inc. Portland OR, USA) to obtain gait speed and cadence.³⁵ To account for differences in stature, gait speed was normalized by dividing the speed by participant height.

Data Analysis

All audio recordings were trimmed to exactly 1 minute long, removing all noise before and after the cues to start and stop talking. Audio files were then imported into MATLAB and processed using a custom script to identify speech and measure silent pause frequency and duration for each participant. The audio signal for each recording was filtered using a fourth order bandpass filter with a passband of 100 to 5000 Hz to isolate frequencies and their harmonics associated with speech. A moving variance window of 50 ms was then applied to the signal, and a “speech threshold” equal to the 40th percentile of the windowed variance was used to identify the presence or absence of speech, based on agreement with manual identification of silent pauses. Durations longer than 250 ms without speech (windowed variance >40th percentile) were designated as silent speech pauses. The 250 ms threshold was selected based on prior work that indicated an increase in speech pauses were related to increased cognitive demand during walking.^29,30,36
-38 The total duration of silent pauses was selected as the primary outcome. Secondary outcomes included the total number of pauses.

Statistical Analysis

To investigate the effect of concussion on extemporaneous speech production during ST and DT walking, we implemented linear mixed effect regression models for each speech and gait outcome. Models included fixed effects for group, task (single vs dual), and the group × task interaction. Models were adjusted for covariates of age, sex, and global cognition using the NIH Toolbox Total Cognition Composite score. Random intercepts by subject were included to account for within-subject correlations across tasks. Between-group effect sizes were calculated for both single- and DT conditions using Hedges’s g.³⁹ A significance level of .05 was used for each analysis.

To explore the associations between self-reported symptoms and the DT effects of walking and talking, Pearson correlation coefficients were computed between the primary outcomes (total pause duration and gait speed) and the symptom categories (affective, cognitive, somatosensory, vestibular, and total) of the NSI within the concussion group. In addition to primary outcomes during ST and DT conditions, DTC were calculated for both Dual-task Costs on (DTC_S) and Dual-task Costs on gait (DTC_G) outcomes using $DTC = \frac{DT - ST}{ST} 100 %$ , where ST and DT represent the ST and DT outcome. To ensure all outcomes remained in the same direction (ie, more negative value indicates worse performance), all speech outcomes were multiplied by a negative sign for DTC_S calculations. Correlations were computed between self-reported symptoms and all ST, DT, and DTC outcomes. As this was an exploratory analysis, correlation coefficients were interpreted using guidelines proposed by Rowntree (Rowntree, 1981): values less than .2 indicate a very weak relationship, .20 to .39 indicate a weak relationship, .40 to .69 indicates a moderate relationship, .70 to .89 indicates a strong relationship, and .90 or greater indicates a very strong relationship. All analyses were conducted in MATLAB using the Statistics and Machine Learning Toolbox (r2020a, The MathWorks, Inc.).

Results

Speech recordings from 4 participants were excluded due to poor audio quality that prohibited analysis, leaving a total of 19 adults with concussion and 18 healthy controls included. Full demographic information is provided in Table 1.

Table 1.

Demographic Information About Subjects in Each Group.

Demographic variable	Concussion	Control
N (M/F)^a	19 (15/4)	18 (9/9)
Age (yrs)	31.4 (9.3)	27.7 (7.1)
Height (cm)	178.1 (8.7)	174.1 (12.8)
Mass (kg)	80.5 (19.7)	70.4 (12.4)
Race^a
American Indian or Alaska native	0	0
Asian	1	3
Black or African American	0	0
Native Hawaiian or other pacific Islander	1	0
White	17	14
More than 1 race	0	1
Time since concussion (days)	8.6 (3.1)	N/A
NSI total symptom score	31.6 (21.6)	5.8 (6.0)
NSI-affective	9.2 (7.8)	3.2 (3.9)
NSI-vestibular	4.6 (3.3)	0.1 (0.3)
NSI-cognitive	7.4 (4.4)	0.8 (1.2)
NSI-somatosensory	9.2 (6.3)	1.4 (1.3)
NIH toolbox total cognition composite (fully-adjusted t-score)	48.39 (8.51)	57.53 (9.86)

Abbreviations: NSI, Neurobehavioral Symptom Inventory; NIH, National Institutes of Health.

All quantities are reported as mean (standard deviation) unless otherwise noted.

Reported as count.

Both groups exhibited bidirectional interference based on the DTC_G and DTC_S, though there was more variability in the DTC_S outcomes across participants (Figure 1). Both concussion and control groups exhibited longer total pause durations (P < .001), slower speeds (P < .001), and slower cadence (P < .001) during the DT compared to ST_G and ST_S conditions (Tables 2 and 3). There was no significant difference in total speech pause duration between groups (P = .408) or significant group × task interaction (P = .983). However, there were notable between-group effect sizes for both ST_S and DT conditions (g = 0.42 and 0.57, respectively; Table 2). There was also a significant group × task interaction for gait speed (P = .032) and cadence (P = .047), indicating that participants with concussion exhibited greater reductions in speed and cadence from ST_G to DT compared to control subjects. While there were moderate to large between group effect sizes for gait speed and cadence (see Table 2), there were no main effects of group on gait speed or cadence after adjusting for global cognition, which was significantly associated with faster gait speed (β = 0.6 cm/s/m, SE = 0.2, P = .002, Table 3) and cadence (β = .33 steps/min SE = 0.15, P = .032, Table 3). No group, task, or group × task effects were detected for the number of pauses per minute, but notable between-group effect sizes were observed for the number of pauses in the DT condition (g = 0.57, Table 2).

Figure 1.

Scatter histograms depicting the dual-task (DT) effect for speech (y-axis) and gait (x-axis) as a percentage of single-task (ST) performance for participants with concussion (orange) and healthy controls (blue). Dual-task effects were calculated as $D T C = \frac{(D T - S T)}{S T} 100 % .$ Kernal density plots are presented adjacent to the scatter plot indicating the distributions of each group.

Table 2.

Univariate Descriptive Statistics of Speech and Gait Outcomes by Group and Task (Single-task [ST] and Dual-task [DT]).

Variable	Concussion		Control		Between-group effect sizes Hedges’ g
	ST	DT	ST	DT	ST	DT
Speech fluency
Total pause duration (s)	14.2 (2.6)	16.1 (2.7)	13.0 (3.1)	14.5 (3.1)	0.42	0.57
Pauses/min	24.2 (2.9)	26.1 (5.1)	23.9 (4.5)	23.6 (3.1)	0.06	0.57
Gait
Normalized gait speed (cm/s/m)	65.2 (8.4)	57.7 (8.9)	73.1 (12.2)	68.7 (12.3)	0.74	1.00
Cadence (steps/min)	106.5 (8.3)	102.4 (8.5)	111.7 (8.8)	108.7 (8.4)	0.60	0.73

Gait speed was normalized to the participant height.

Table 3.

Linear Mixed Model Results for Each Outcome, Including Beta Coefficients, 95% Confidence Intervals (CIs), and P-values.

Independent variables within each model	Beta	SE	95% CIs	P value
Total pause duration (s)
Group (concussion/control)	−.41	0.49	[−0.58, 1.40]	.408
Condition (ST/DT)	0.91	0.15	[0.61, 1.20]	<.001
Group × condition	−.00	0.15	[−0.29, 0.29]	.983
Age	.11	0.05	[0.00, 0.21]	.045
Sex	1.18	0.42	[0.33, 2.03]	.007
Total cognition	.00	0.15	[−0.28, 0.29]	.983
Pauses/min
Group (concussion/control)	−.17	0.53	[−0.88, 1.23]	.742
Condition (ST/DT)	.32	0.41	[−0.51, 1.15]	.440
Group × condition	.35	0.41	[−1.18, 0.48]	.400
Age	.05	0.06	[−0.07, 0.16]	.416
Sex	1.88	0.45	[0.97, 2.79]	<.001
Total cognition	−.08	0.05	[−0.19, 0.02]	.116
Normalized gait speed (cm/s/m)
Group (concussion/control)	−1.1	1.9	[−2.7, 4.8]	.570
Condition (ST/DT)	−2.9	0.3	[−3.6, −2.2]	<.001
Group × condition	−.8	0.3	[0.1, 1.5]	.032
Age	−.1	0.2	[−0.5, 0.3]	.744
Sex	−5.4	1.6	[−8.6, −2.1]	.002
Total cognition	.6	0.2	[0.2, 0.9]	.003
Cadence (steps/min)
Group (concussion/control)	.09	1.51	[−3.12, 2.93]	.951
Condition (ST/DT)	−1.68	0.29	[−2.26, −1.10]	<.001
Group × condition	−.59	0.29	[0.01, 1.17]	.047
Age	−.11	0.16	[−0.44, 0.21]	.478
Sex	−5.13	1.30	[−7.74, −2.52]	<.001
Total cognition	.33	0.15	[0.03, 0.63]	.032

Models were contrast-coded effects and adjusted for sex, age, and global cognition from the National Institutes of Health Toolbox Total Cognition Composite Score. Gait speed was normalized to the participant height. Associations between each outcome and individual National Institutes of Health Toolbox Cognitive Subscores are presented in the Supplemental Material.

Within the concussion group, longer total pause durations in the ST_S condition were strongly associated with greater vestibular-related symptom scores (ρ = .72) and moderately associated with all other symptom subscores (ρ = .55-.66, Table 4). The strength of the associations between symptoms and total pause duration decreased during walking; weak-to-moderate associations were observed for DT total pause duration and DTC_S (Table 4) for all symptom scores. Conversely, ST_G and DT gait speed was strongly associated with vestibular symptoms (ρ = −.75 and −.78 respectively, see Table 4 and Figure 2), and moderately associated with all other symptom subscores (ρ = −.50 and −.63).

Table 4.

Correlation Coefficients Between Self-reported Symptoms on the Neurobehavioral Symptom Inventory (NSI) and Primary Speech (Total Pause Duration) and Gait (Gait Speed) Outcomes Within the Concussion Group for Single-task (ST), Dual-task (DT), and Dual-task Costs on Speech (DTC_S) and Gait (DTC_G).

Variable	NSI-Vestibular	NSI-Cognitive	NSI-Somatosensory	NSI-Affective	NSI-Total
Total pause duration (s)
ST	0.72	0.55	0.66	0.60	0.66
DT	0.37	0.26	0.41	0.30	0.37
DTC_S	−0.48	−0.40	−0.36	−0.41	−0.40
Normalized gait speed (m/s/m)
ST	−0.75	−0.50	−0.58	−0.57	−0.63
DT	−0.78	−0.53	−0.55	−0.57	−0.62
DTC_G	−0.45	−0.34	−0.21	−0.29	−0.30

Bold coefficients indicate strong associations (0.7-0.89).

Figure 2.

Scatter plots depicting the association between the vestibular subscore on the Neurobehavioral Symptom Inventory with total pause duration and normalized gait speed during single- and dual-task conditions in participants with a concussion.

Discussion

Our study investigated the effects of concussion on extemporaneous speech production during ST (seated) and DT (walking) paradigms. In agreement with prior work, we observed extemporaneous speech elicited DTC on walking, indicated by slower gait speeds and slower cadence. Our results did not fully support our hypothesis about group differences; individuals with a concussion did not uniformly exhibit longer pauses during extemporaneous speech and did not demonstrate significantly larger DTC in speech compared to controls, although moderate effect sizes warrant further study. However, our results indicated that those with a concussion had larger DTC in gait speed than controls. Eliciting greater DTC to gait using an ecologically relevant cognitive DT complements well-established deleterious effects of concussion on DT gait using standardized cognitive tasks.^{2,4
-7,9,10,13,16} Further, variable association of DTC with differing symptom domains suggests that DT effects are worst in symptomatic individuals, and more importantly that DT deficits are heterogeneous across people with concussion and may be dependent on symptomology.

While we did not observe any significant group differences in the total duration of pauses or the number of pauses during speech, the effect sizes observed complement prior work suggesting concussion-related symptoms may increase pauses during speech. Prior work in acutely concussed athletes (0-6 days post-concussion) observed increased pause and time fill (ie, filler word) errors after concussion compared to pre-injury baseline performance.³⁰ The difference between these 2 studies may be attributed to different task demands, symptomology, and time since injury. While prior work used a visual scanning task (describing a picture), the talking task here did not require visual processing. It is possible that speech deficits when describing a picture are compounded by ocular-motor deficits that affect visual scanning. Additionally, participants ranged from 8 to 14 days post-concussion with varying symptom burdens in the present study. A relatively longer time since injury and the heterogeneous symptom burden at the time of testing revealed a strong relationship between self-reported symptoms, particularly vestibular-related symptoms, and the total duration of pauses during extemporaneous speech. While the mechanisms underlying the association between vestibular symptoms and ST speech, ST gait, and DT gait remain unclear, we speculate that such results originate from attention / rumination on symptoms. Since the strongest associations were with vestibular symptom scores, rather than cognitive scores, we posit that individuals may have been allocating attention to minimize head motion / rotation to avoid exacerbating vestibular symptoms.⁴⁰ Dedicating attention to limiting aversive vestibular stimulation would effectively add a third motor task and another implicit goal to the walking and talking tasks. Further, while we failed to detect statistically significant differences for total speech pause duration, we observed notable between-group effect sizes, especially in DT walking conditions, supporting differences in attention allocation after concussion that warrant further study. Given the associations between vestibular symptoms and our outcomes, future studies should work to understand how concussion affects talking while standing as an intermediary postural challenge to measure cognitive-motor status as a function of vestibular demands (eg, seated and speaking, standing and speaking, and walking and speaking).

While we did not observe a significant group difference in DTC in speech pauses (DTC_S), we did detect a difference in DTC to walking (DTC_G), indicated by the group × task interaction, for gait speed. Increased DTC_G have been well-documented after concussion, especially in symptomatic individuals within 14 days post-injury.⁵ Previous work used standardized cognitive tasks such as reciting months of the year in reverse, spelling a 5-letter word backwards, or serial subtraction, but such tasks lack ecological relevance to daily life. Both cognitive and motor task complexity affect the ability of dual-tasking to differentiate those with concussion from healthy controls,^41,42 with generally more difficult tasks eliciting larger concussion-related dual-task deficits. Our results extend prior studies and demonstrate that even during ecologically relevant cognitive tasks such as extemporaneous speech, individuals with concussion demonstrate greater motor deficits during a DT compared to healthy controls. While large between-group effects for gait speed and cadence were explained by differences in global cognition, the dual-task effects on gait speed and cadence remained when adjusting for global cognition. However, it remains unclear whether dual-task performance would degrade if extemporaneous speech were paired with more complex motor tasks (eg, obstacle crossing and turning). Motor complexity drives task prioritization and the combination of cognitive demand plus added motor challenge could elicit different results.^15,18 Plummer D’Amato et al⁴³ used an obstacle crossing task and revealed that a clock-monitoring task elicited less dual-task costs than spontaneous speech in impaired older adults. Therefore, pairing extemporaneous speech with ecologically relevant gait tasks of increasing motor complexity might further elucidate the impact of concussion-related dual-task deficits on daily life.

Despite the lack of group differences in speech-related outcomes, our results indicate that an extemporaneous speech task elicits a strong dual-task effect overall, in agreement with prior work ^18,23,43. Both concussion and control groups exhibited greater speech pause durations, slower gait speeds, and slower cadences during the DT compared to the ST condition. Thus, the walking-while-talking task successfully elicited mutual interference on both extemporaneous speech and gait in the majority of participants, regardless of group (see Figure 1). The consistent effect of walking while talking further supports its use as an ecologically relevant cognitive task to assess dual-task gait in young adults.

Limitations

The primary limitations of this work include the relatively modest sample size and the reliance on silent speech pauses to infer cognitive task performance. Despite the sample size, the consistent chronicity (8-14 days post-injury) and symptomatic nature of the participants is a strength of the study. The reliance on silent speech pauses was grounded in prior work using walking and talking tasks. However, it is possible that examining the syntactic or lexical complexity of the speech may better elucidate differences between groups or tasks compared to only using pauses. Finally, participants were only tested cross-sectionally at a single point in time. Without a baseline from the concussion participants, we cannot determine if their current speech performance was directly caused by the concussion.

Conclusions

Extemporaneous speech offers an ecologically relevant cognitive DT that elicited larger gait DTC in people after a concussion. While we did not observe significant group differences in speech performance, effect sizes, and strong associations between vestibular-related symptoms, total pause duration, and gait speed suggest DTC are variable based on a person’s post-concussion symptomology. Further studies should probe compensatory recruitment of different cognitive processes during DT walking after a concussion, and how such effects may differ based on symptomology. Overall, these results suggest that DT deficits after concussion may have deleterious effects on tasks that are essential to daily life.

Supplemental Material

sj-docx-1-nnr-10.1177_15459683251317184 – Supplemental material for Talking While Walking After Concussion: Acute Effects of Concussion on Speech Pauses and Gait Speed

Supplemental material, sj-docx-1-nnr-10.1177_15459683251317184 for Talking While Walking After Concussion: Acute Effects of Concussion on Speech Pauses and Gait Speed by Shu Yang, Paula K. Johnson, Colby R. Hansen, Elisabeth A. Wilde, Melissa M. Cortez, Leland E. Dibble, Peter C. Fino and Tiphanie E. Raffegeau in Neurorehabilitation and Neural Repair

Footnotes

Acknowledgements

The authors would like to especially thank Cecilia Martindale, Sarah Hill, Dr. Ryan Pelo, Gabrielle Gaudette, Emma Nilsson Read, and Elizabeth Hovenden for their assistance with participant recruitment and data collection on this project.

Author Contributions

Shu Yang: Conceptualization; Data curation; Investigation; Software; Visualization; Writing—original draft; and Writing—review & editing. Paula K. Johnson: Data curation; Investigation; Methodology; Supervision; and Writing—review & editing. Colby R. Hansen: Funding acquisition; Methodology; Resources; and Writing—review & editing. Elisabeth A. Wilde: Funding acquisition; Methodology; Project administration; and Writing—review & editing. Melissa M. Cortez: Funding acquisition; Methodology; Project administration; and Writing—review & editing. Leland E. Dibble: Funding acquisition; Methodology; Project administration; Resources; Supervision; and Writing—review & editing. Peter C. Fino: Conceptualization; Formal analysis; Funding acquisition; Methodology; Project administration; Resources; Software; Supervision; Validation; Visualization; Writing—original draft; and Writing—review & editing. Tiphanie E. Raffegeau: Conceptualization; Formal analysis; Investigation; Methodology; Validation; Writing—original draft; and Writing—review & editing.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Research reported in this publication was supported by the Eunice Kennedy Shriver National Institute of Child Health & Human Development of the National Institutes of Health under Award Number R21HD100897 (PI: Fino) and the National Center for Advancing Translational Sciences under Award Number UL1TR002538. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

ORCID iDs

Paula K. Johnson

Colby R. Hansen

Melissa M. Cortez

Leland E. Dibble

Peter C. Fino

Tiphanie E. Raffegeau

Supplementary material for this article is available on the Neurorehabilitation & Neural Repair website along with the online version of this article.

References

Manaseer

Gross

Dennett

Schneider

Whittaker

JL.

Gait deviations associated with concussion: a systematic review. Clin J Sport Med. 2020;30 Suppl 1:S11-S28. doi:10.1097/JSM.0000000000000537

Parker

Osternig

VAN Donkelaar

Chou

LS.

Gait stability following concussion. Med Sci Sports Exerc. 2006;38(6):1032-1040. doi:10.1249/01.mss.0000222828.56982.a4

Wood

Hsieh

Ballard

Sosnoff

JJ.

Balance and gait alterations observed more than 2 weeks after concussion: a systematic review and meta-analysis. Am J Phys Med Rehabil. 2019;98(7):566-576. doi:10.1097/PHM.0000000000001152

Catena

van Donkelaar

Chou

LS.

Altered balance control following concussion is better detected with an attention test during gait. Gait Posture. 2007;25(3):406-411. doi:10.1016/j.gaitpost.2006.05.006

Fino

Parrington

Pitt

, et al. Detecting gait abnormalities after concussion or mild traumatic brain injury: a systematic review of single-task, dual-task, and complex gait. Gait Posture. 2018;62:157-166. doi:10.1016/j.gaitpost.2018.03.021

Martini

Sabin

DePesa

, et al. The chronic effects of concussion on gait. Arch Phys Med Rehabil. 2011;92(4):585-589. doi:10.1016/j.apmr.2010.11.029

Parker

Osternig

Lee

van Donkelaar

Chou

LS.

The effect of divided attention on gait stability following concussion. Clin Biomech. 2005;20(4):389-395. doi:10.1016/j.clinbiomech.2004.12.004

Jain

Graci

Beam

, et al. Neurophysiological and gait outcomes during a dual-task gait assessment in concussed adolescents. Clin Biomech Bristol Avon. 2023;109:106090. doi:10.1016/j.clinbiomech.2023.106090

Martini

Mancini

Antonellis

, et al. Prefrontal cortex activity during gait in people with persistent symptoms after concussion. Neurorehabil Neural Repair. 2024;38(5):364-372. doi:10.1177/15459683241240423

10.

Büttner

Howell

Ardern

, et al. Concussed athletes walk slower than non-concussed athletes during cognitive-motor dual-task assessments but not during single-task assessments 2 months after sports concussion: a systematic review and meta-analysis using individual participant data. Br J Sports Med. 2020;54(2):94-101. doi:10.1136/bjsports-2018-100164

11.

Howell

Kirkwood

Provance

Iverson

Meehan

WP.

Using concurrent gait and cognitive assessments to identify impairments after concussion: a narrative review. Concussion Lond Engl. 2018;3(1):CNC54. doi:10.2217/cnc-2017-0014

12.

Bryk

Passalugo

Chou

, et al. Dual task cost in adults with persistent concussion symptoms. Gait Posture. 2023;101:120-123. doi:10.1016/j.gaitpost.2023.02.008

13.

Martini

Parrington

Stuart

Fino

King

LA.

Gait performance in people with symptomatic, chronic mild traumatic brain injury. J Neurotrauma. 2021;38(2):218-224. doi:10.1089/neu.2020.6986

14.

Yogev-Seligmann

Hausdorff

Giladi

The role of executive function and attention in gait. Mov Disord. 2008;23(3):329-342; quiz 472. doi:10.1002/mds.21720

15.

Yogev-Seligmann

Hausdorff

Giladi

Do we always prioritize balance when walking? Towards an integrated model of task prioritization. Mov Disord. 2012;27(6):765-770. doi:10.1002/mds.24963

16.

Kleiner

Wong

Dubé

Wnuk

Hunter

Graham

LJ.

Dual-task assessment protocols in concussion assessment: a systematic literature review. J Orthop Sports Phys Ther. 2018;48(2):87-103. doi:10.2519/jospt.2018.7432

17.

Dromey

Jarvis

Sondrup

Nissen

Foreman

Dibble

LE.

Bidirectional interference between speech and postural stability in individuals with Parkinson’s disease. Int J Speech Lang Pathol. 2010;12(5):446-454. doi:10.3109/17549507.2010.485649

18.

Raffegeau

Haddad

Huber

Rietdyk

Walking while talking: young adults flexibly allocate resources between speech and gait. Gait Posture. 2018;64:59-62. doi:10.1016/j.gaitpost.2018.05.029

19.

Miyake

Friedman

NP.

The nature and organization of individual differences in executive functions: four general conclusions. Curr Dir Psychol Sci. 2012;21(1):8-14. doi:10.1177/0963721411429458

20.

Lovett

MC.

A strategy-based interpretation of stroop. Cogn Sci. 2005;29(3):493-524. doi:10.1207/s15516709cog0000_24

21.

Kubose

Bock

Dell

Garnsey

Kramer

Mayhugh

The effects of speech production and speech comprehension on simulated driving performance. Appl Cogn Psychol. 2006;20(1):43-63. doi:10.1002/acp.1164

22.

Lee

AMC

Cerisano

Humphreys

Watter

. Talking is harder than listening: the time course of dual-task costs during naturalistic conversation. Can J Exp Psychol Rev Can Psychol Exp. 2017;71(2):111-119. doi:10.1037/cep0000114

23.

Raffegeau

Brinkerhoff

Clark

, et al. Walking (and talking) the plank: dual-task performance costs in a virtual balance-threatening environment. Exp Brain Res. 2024;242(5):1237-1250. doi:10.1007/s00221-024-06807-w

24.

Mitchell

Hoit

Watson

PJ.

Cognitive-linguistic demands and speech breathing. J Speech Hear Res. 1996;39(1):93-104. doi:10.1044/jshr.3901.93

25.

Love

Starbuck

Christensen

JM.

Pause frequency in fluent and nonfluent speech. J Acoust Soc Am. 1972;51(1A_Supplement):122. doi:10.1121/1.1981351

26.

Krivokapić

Styler

Parrell

Pause postures: the relationship between articulation and cognitive processes during pauses. J Phon. 2020;79:100953. doi:10.1016/j.wocn.2019.100953

27.

Lee

Huber

Jenkins

Fredrick

Language planning and pauses in story retell: evidence from aging and Parkinson’s disease. J Commun Disord. 2019;79:1-10. doi:10.1016/j.jcomdis.2019.02.004

28.

Huber

Darling

Effect of Parkinson’s disease on the production of structured and unstructured speaking tasks: respiratory physiologic and linguistic considerations. J Speech Lang Hear Res. 2011;54(1):33-46. doi:10.1044/1092-4388(2010/09-0184)

29.

Wang

Kent

Duffy

Thomas

JE.

Dysarthria associated with traumatic brain injury: speaking rate and emphatic stress. J Commun Disord. 2005;38(3):231-260. doi:10.1016/j.jcomdis.2004.12.001

30.

Patel

Grabowski

Dayalu

Testa

AJ.

Speech error rates after a sports-related concussion. Front Psychol. 2023;14:1135441. doi:10.3389/fpsyg.2023.1135441

31.

Wall

Powell

Young

, et al. A deep learning-based approach to diagnose mild traumatic brain injury using audio classification. PLoS ONE. 2022;17(9):e0274395. doi:10.1371/journal.pone.0274395

32.

Oldham

Lent

Peretiatko

Dec

Predictors of speech language pathology referral after pediatric concussion using a speech language checklist. Am J Phys Med Rehabil. 2023;102(10):919-922. doi:10.1097/PHM.0000000000002293

33.

Cicerone

Kalmar

Persistent postconcussion syndrome: the structure of subjective complaints after mild traumatic brain injury. J Head Trauma Rehabil. 1995;10(3):1.

34.

Hodes

Insel

Landis

SC.

The NIH toolbox. Neurology. 2013;80(11 Suppl 3):S1. doi:10.1212/WNL.0b013e3182872e90

35.

Morris

Stuart

McBarron

Fino

Mancini

Curtze

Validity of MobilityLab (version 2) for gait assessment in young adults, older adults and Parkinson’s disease. Physiol Meas. 2019;40(9):095003. doi:10.1088/1361-6579/ab4023

36.

Bosker

Pinget

Quené

Sanders

de Jong

NH.

What makes speech sound fluent? The contributions of pauses, speed and repairs. Lang Test. 2013;30(2):159-175. doi:10.1177/0265532212455394

37.

Khawaja

Ruiz

Chen

Think before you talk: an empirical study of relationship between speech pauses and cognitive load. Paper presented at: Proceedings of the 20th Australasian Conference on Computer-Human Interaction: Designing for Habitus and Habitat. OZCHI ’08; December 8-12, 2008; Cairns Australia. Association for Computing Machinery; 2008:335-338. doi:10.1145/1517744.1517814

38.

Noufi

Lammert

Mehta

, et al. Vocal biomarker assessment following pediatric traumatic brain injury: a retrospective cohort study. Interspeech. 2019;2019:3895-3899. doi:10.21437/Interspeech.2019-1200

39.

Hedges

Olkin

Statistical Methods for Meta-Analysis. Academic Press; 1985.

40.

Loyd

Dibble

Weightman

, et al. Volitional head movement deficits and alterations in gait speed following mild traumatic brain injury. J Head Trauma Rehabil. 2023;38(3):E223-E232. doi:10.1097/HTR.0000000000000831

41.

Howell

Osternig

Koester

Chou

LS.

The effect of cognitive task complexity on gait stability in adolescents following concussion. Exp Brain Res. 2014;232(6):1773-1782. doi:10.1007/s00221-014-3869-1

42.

Howell

Myer

Grooms

Diekfuss

Yuan

Meehan

WP.

Examining motor tasks of differing complexity after concussion in adolescents. Arch Phys Med Rehabil. 2019;100(4):613-619. doi:10.1016/j.apmr.2018.07.441

43.

Plummer-D’Amato

Altmann

LJP

Reilly

Dual-task effects of spontaneous speech and executive function on gait in aging: exaggerated effects in slow walkers. Gait Posture. 2011;33(2):233-237. doi:10.1016/j.gaitpost.2010.11.011

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB