Upper Limb Movement Performance in Individuals Sustaining Mild Traumatic Brain Injuries

Abstract

Background

Individuals sustaining a mild traumatic brain injury (TBI) continue to have suboptimal upper limb (UL) involvement in daily life. Intensity of task practice is one key factor to promote better motor improvement. Task-practice intensity metrics include number of repetitions/sessions, and this value is currently unknown in people with mild TBI. Kinematic analysis can help estimate the number of repetitions/sessions.

Purpose

We estimated the minimal number of repetitions for a plateau in performance in an UL pointing task in 10 individuals who had sustained a mild TBI and seven age-matched controls.

Methods

All participants performed 45 repetitions and pointed to a central target at arm’s length. The TBI group underwent assessments of UL motor impairment, spasticity and activity limitations. The primary outcome was the number of trials to achieve an asymptote in endpoint error. Secondary outcomes included movement speed, straightness, trunk and UL joint ranges of motion.

Results

Clinical assessments revealed absence of motor impairment or activity limitations. However, individuals with mild TBI required more trials (28.5) to reach an asymptote in the pointing movement performance compared to controls (18; p = .005, effect size [ES] = −0.66). They also had more curved movements (1.11 ± 0.06 vs 1.06 ± 0.01; p = .036, ES = 0.64), used more trunk displacement (13.1 ± 3 vs 10.2 ± 2.1 mm; p = .044, ES = 1.09) and had lower ranges of motion in wrist extension (24.8 ± 3.8 vs 17.3 ± 3.4; p = .006, ES = 1.60), elbow extension (144.7 ± 6.8 vs 152.3 ± 6.8°, p = .025, ES = 1.22), shoulder flexion (60.5 ± 5.2 vs 66.6 ± 6.4, p = .046, ES = 1.07) and shoulder horizontal adduction (77.7 ± 5.0 vs 87.4 ± 9.6, p = .014, ES = 1.35).

Conclusion

After sustaining a mild TBI, individuals have deficient UL movement performance. Use of kinematic analyses can help uncover latent deficits in those with perfect scores on clinical assessments.

Keywords

Arm concussion rehabilitation kinematics outcomes intensity

Introduction

Traumatic brain injury (TBI) is a leading cause of worldwide adult morbidity and mortality, with the majority of injuries occurring in young individuals.¹ Individuals with TBIs have motor impairments in the lower and upper limbs (ULs), which affect performance of functional activities of daily living (ADLs).² These impairments tend to persist beyond the completion of rehabilitation. The majority of motor rehabilitation intervention studies in individuals post-TBI focus on remediation of balance and mobility.³ Only a few studies have addressed UL issues in individuals post-TBI.⁴

Issues with UL functioning tend to be more diffuse and persist for extended periods of time in individuals post-TBI.⁵ Deficits in UL functioning identified in this population include less precise, slower and more curved movements.⁶ However, most of the commonly available clinical assessments focus primarily on the ability to perform movements. Very few measures focus on the quality of movement performance, or how the movement is performed,⁷ thus limiting the amount of information available to a clinician. This is an important aspect, as many individuals sustaining TBIs often report issues with the finer aspects of controlling movements. This lack of control affects performance of daily life activities including driving, dressing and grooming to get ready for work.^{8, 9} Improved UL motor functioning is beneficial for individuals post-TBI to perform ADLs such as eating, dressing, self-care and community reintegration post-TBI.¹⁰ Improvements in motor functioning after a TBI are attributable to adaptive neuroplasticity and motor learning.¹¹ Factors influencing adaptive neuroplasticity and motor learning include provision of intense, variable, salient and task-specific practice.¹²

One factor that has received considerable attention is the intensity of task practice. Many different metrics can quantify intensity of task practice including time and/or effort spent per rehabilitation session,¹³ and the number of repetitions performed each session.¹⁴ Recent work has focussed upon the number of repetitions, as one of the factors to help reduce the observed variability in studies involving rehabilitation interventions. Knowledge about the number is essential as exercises, prescribed by rehabilitation personnel, are repeated many times in each session. Only one previous study has reported that individuals with TBI used a mean of ≈26 active repetitions of UL movements/sessions.¹⁵

In individuals with stroke, kinematic analyses have helped estimate the minimal number of repetitions necessary to achieve a performance asymptote in an UL pointing task.¹⁶ Kinematic analyses provide detailed descriptions of the task at the motor performance and movement pattern levels.¹⁷ Motor performance can provide information in terms of accuracy, speed, movement straightness and smoothness. Movement patterns provide information on the individual joint movements as well as interjoint and intersegmental coordination.¹⁸ Kinematic analyses are additionally more sensitive in the detection of deficits in movement performance compared to routinely used clinical assessments.¹⁹

The minimal number of active repetitions/sessions to achieve performance asymptote or motor adaptation in individuals with TBI in general, and mild TBI specifically, is currently unknown. This information is essential to better understand whether, and to what extent, motor task performance and motor learning are impaired in this population. Given the previous use of kinematic analyses for this purpose, we used this approach in individuals with mild TBI. We estimated the minimal number of repetitions necessary for an asymptote in accuracy (no further change in error) in a pointing task in individuals with mild TBI compared to controls. We hypothesised that individuals sustaining a mild TBI would need a greater number of repetitions to achieve an asymptote in accuracy of a pointing task compared to controls. As a secondary aim, we also compared other motor performance and movement pattern variables between groups. Preliminary results have previously been published as an abstract.²⁰

Methods

Participants

Ten individuals with mild TBI and seven healthy controls participated in this pilot cross-sectional study. We used a convenience sample. Inclusion criteria for individuals with TBI were: (a) one or more TBIs ≥6 months and ≤15 years prior to study enrolment; (b) aged 18–50; (c) able to understand and follow two-step verbal commands in English; and (d) able to sit for at least 45–60 min. Participants were excluded if they had (a) significant health problems (e.g., respiratory disease, neurological, cardiovascular, musculoskeletal, etc.) that impacted study participation; (b) a history of significant psychotic disorder, alcoholism or substance abuse; and (c) visual impairment (e.g., ‘legally blind’—eyesight uncorrectable with spectacles; <30° off centre line peripheral vision). Age-equivalent healthy controls were recruited if they had no orthopaedic or neurological impairments interfering with task performance. Both groups of participants signed consent forms approved by the Institutional Review Board of the University of Texas Health Science Center at San Antonio. The mild TBI group underwent a clinical assessment prior to the kinematic assessment. Both groups participated in one session where we assessed pointing movement performance using kinematic analysis.

Clinical Assessment

In individuals with mild TBI, information was collected on initial hospitalisation status (yes/no), any loss of consciousness and/or duration of post-traumatic amnesia. We also assessed levels of UL motor impairment, spasticity and limitations in ADL performance using psychometrically sound outcomes. UL motor impairment was quantified using the Chedoke–McMaster Assessment arm sub-scale.²¹ The Chedoke is based on Brunnstrom’s seven stages and is scored out of seven points. A score of 1 indicates flaccid paralysis and seven perfect movements. The presence of spasticity was assessed using the Composite Spasticity Index (CSI). The CSI evaluates both tonic and phasic components of spasticity. The excitability of the biceps brachii tendon reflex, wrist clonus and resistance to moderate velocity (~100°/s) passive elbow flexion are scored. With a total of 16 points, scores of 4, 5–9, 10–12 and 13–16 indicate normal tone, mild, moderate and severe spasticity respectively.²² ADL performance limitations were quantified using the Chedoke Arm and Hand Activity Inventory, CAHAI-9.²³ The CAHAI-9 involves a series of unimanual and bimanual UL movements. It is scored on a seven-point scale (1–7), for a total of 63 points. An experienced clinician conducted all assessments.

Kinematic Assessment

Kinematic assessment involved a pointing task. Participants pointed to a target placed in front of them in the central UL workspace at arm’s length (Figure 1). The target was placed at a height of 90° (or maximal possible) shoulder flexion and aligned to the sternum. The target required a combination of shoulder flexion, shoulder horizontal adduction and elbow extension movements for successful pointing. With eyes closed, we asked participants to point ‘as fast and as accurately as possible’ to the target centres. Both groups of individuals pointed to the target 45 times, in three sets of 15 repetitions. These numbers were based on our pilot testing results, which indicated that 45 trials were a useful number to help ascertain the point of asymptote.

Figure 1.

Illustration of the Body Segment Positions at the Starting Position (Black) and End (Grey Segments) of the Pointing Motion to the Target Placed at Arm’s Length from the Right Side. The Circles Represent Markers on the Ipsilateral Shoulder, Elbow, Wrist, Endpoint and Target.

To avoid fatigue, we provided a short break (1–2 min) at the end of five trials with a longer break (5–10 min) in between sets. All participants received summary knowledge of results (KR) feedback after five trials. We provided only KR feedback, based on previous results involving individuals sustaining a TBI.⁴ Six Osprey digital real-time cameras (Motion Analysis Corp.) were used to record the movements. Recording frequency was 100 Hz and trial duration was 3–4 s. We placed eight reflective markers on the index fingertip (endpoint), wrist styloid, lateral epicondyle (elbow), ipsilateral and contralateral acromion processes, midsternum and femur (greater trochanter) and on the target.

Data Analysis

Custom programmes written in MATLAB (MathWorks, Natick, MA, USA) were used to analyse kinematic data. Data were low-pass filtered using a 10-Hz Butterworth filter. Endpoint tangential velocities were computed from x, y and z displacements of the marker on the fingertip. Times when the velocity exceeded and continued to increase, or decreased and remained below 10% of peak velocity, denoted movement onset and offset. Endpoint error was quantified as the root mean-squared error between the endpoint position and the target centre at movement offset. Movement straightness was measured using the index of curvature (IC). The IC is the ratio of the length of the actual path travelled by the endpoint to the length of a straight line joining the initial and final endpoint positions. A straight line has an IC of 1, while a semicircle has an IC of 1.57.¹⁶

Movement pattern outcomes were calculated as the values at movement offset. Trunk displacement was denoted by the movement of the midsternal marker. Wrist movement was calculated using vectors between the endpoint to wrist styloid and wrist styloid to lateral epicondyle. Elbow extension range was computed using the vectors between the wrist styloid and elbow and the elbow and ipsilateral acromion (full elbow extension: 180°). Vectors between the elbow and ipsilateral acromion, and between the acromion and greater trochanter, were used to quantify shoulder flexion range (arm alongside body: 0°). Shoulder horizontal adduction range was defined using the vectors formed by the ipsilateral acromion and elbow markers and a vertical line projected horizontally between the two acromions (full abduction: 0°). Estimation of the various kinematic parameters is similar to what has previously been reported in the literature.

Statistical Analysis

Descriptive statistics highlighted the main demographic characteristics. For the primary outcome, we grouped the trials into triads (each triad consisted of three trials)¹⁶ and calculated the mean and 95% CI for each triad. We considered the triad in which the upper end of the 95% CI was smaller than the lower end of the first triad as the point of asymptote. Given the ordinal nature of these data, we used non-parametric statistics for this outcome. Shapiro–Wilk tests were used to verify the assumptions of normality for the data distributions of the remaining kinematic variables. Data with non-normal and normal distributions were analyzed using Mann–Whitney U-tests and independent t-tests, respectively.

For data analysed with non-parametric statistics, effect sizes (ESs) were estimated by dividing the Z value by the square root of the sample size. For all parametric comparisons, Hedges’ G²⁴ was used to estimate ESs, given the different sample sizes in both groups. ESs were categorised as small (0.08–0.18), medium (0.19–0.40) and large (≥0.41) based on recent recommendations.²⁵ Given earlier suggestions of error influencing learning,²⁶ we analysed whether the error in the first triad differed between groups. In addition, we estimated the strength of the association between error in the first triad and the number of trials taken to achieve an asymptote in learning. Since both values could vary, we used an orthogonal regression approach.²⁷ Significance was set at p < .05. We applied Bonferroni–Holm corrections²⁸ to account for multiple comparisons. Analyses were conducted using SPSS v25 (IBM Corp., Armonk, NY, USA) and Minitab v17 (Minitab LLC, State College, PA, USA).

Results

The control group (n = 7) included four females, and their age (mean ± SD) was 26 ± 2.1 years. The mild TBI group (n = 10) included five females and their age was 28.5 ± 5.5 years (Table 1).

Table 1.

Demographic Characteristics of Participants Sustaining a TBI.

Subject	Age/Sex	Duration of PTA (min)	Mechanism of Injury	Chedoke– McMaster Arm  Subscale Score (/7)	Composite Spasticity Index Score (/16)	Chedoke Arm and Hand Activity  Inventory Score (/63)
S1	25/F	<30	Sports-related injury/while playing football	7	4	63
S2	30/F	<30	Sports-related injury/while playing football	7	4	63
S3	35/M	<30	Fall from a ladder	7	4	63
S4	24/F	<30	Sports-related injury/while playing football	7	4	63
S5	26/F	<30	Sports related injury/while playing basketball	7	4	63
S6	27/F	<30	Sports-related injury/while playing football	7	4	63
S7	25/M	<30	Car crash	7	4	63
S8	28/M	<30	Sports-related injury/while playing football	7	4	63
S9	41/M	<30	Sports-related injury/while kickboxing	7	4	63
S10	24/M	<30	Sports-related injury/while playing basketball	7	4	63

Note: F: Female, M: Male, PTA: Post-traumatic amnesia, TBI: Traumatic brain injury.

All 17 participants completed all the clinical and kinematic assessments. The duration of post-traumatic amnesia (a measure of TBI severity) was less than 30 min for all participants. Clinical assessment scores revealed the absence of motor impairment or limitations in performing ADLs (perfect scores on all assessments indicated the absence of clinically assessed motor impairment, spasticity or limitations in ADL performance).

Examples of movement asymptote estimation for one control and one participant with mild TBI are shown in Figures 2A and 2B, respectively. Individuals with mild TBI required more trials [median: 28.5 (IQR: 25.5; 30)] to reach an asymptote compared with controls [median: 18 (IQR: 18; 21)], and this difference was statistically significant (Mann–Whitney U = 7.5, Z = −2.733, p = .005, ES = −0.66; Figure 3A). Individuals with mild TBI also had significantly greater error [81.8 mm (SD: 29.7)] in the first triad compared to controls [53.8 mm (SD: 14.1); t₁₆ = −2.303, p = .036, ES = 1.13; Figure 3B]. However, we found no association between the number of trials needed to reach an asymptote and error in the first triad in both groups (Figure 4). Additionally, endpoint speeds did not differ significantly between groups [mild TBI—median: 2,012.5 mm/s (IQR: 1,666.4; 3,033.1); control—median: 2,288.9 mm/s (IQR: 2,021; 2,889.9)].

Figure 2.

Note: CI: Confidence interval, TBI: Traumatic brain injury.

Figure 3.

Note: *p < .05, **p < .01.

Figure 4.

Note: mTBI: Mild traumatic brain injury.

Individuals with mild TBI had significantly more curved movements [1.11 (SD: 0.06)] compared to controls [1.07 (SD: 0.01); t₁₆ = −2.303, p = .036, ES = 0.64; Figure 5A]. The mild TBI group also used more trunk displacement [13.1 mm (SD: 3.0)] compared to controls [10.2 mm (SD: 2.1); t₁₆ = −2.195, p = .044, ES = 1.09; Figure 5B]. Similar results were found for other movement pattern outcomes. The mild TBI group used less wrist extension [19.2° (SD: 3.5)] compared to controls [26.2° (SD: 5.4); t₁₆ = 3.239, p = .006, ES = 1.60; Figure 5C]. In comparison with controls, they also used less elbow extension [144.7° (SD: 5.8), controls 152.3°(SD: 6.8); t₁₆ = 2.492, p = .025, ES = 1.22; Figure 5D], less shoulder flexion [60.5° (SD: 5.2), controls 66.6° (SD: 6.4); t₁₆ = 2.172, p =.046, ES = 1.07; Figure 5E] and less shoulder horizontal adduction [77.7° (SD: 5.0), controls 87.4° (SD: 9.6); t₁₆ = 2.768, p = .014, ES = 1.35; Figure 5F].

Figure 5.

Notes: *p < .05, **p < .01; mTBI: Mild traumatic brain injury.

Discussion

We sought to estimate the number of trials required to reach an asymptote in motor performance in individuals with mild TBI. Our results revealed that individuals with mild TBI need about 29 trials to reach an asymptote compared to controls who reached an asymptote in 18 trials. This finding supports our main hypothesis. To our knowledge, this is the first study to estimate the number of trials necessary to achieve an asymptote in motor performance (i.e., motor adaptation) in a sample of adults with mild TBI. Our results suggest that the number of active repetitions used in every session needs to be higher than numbers reported previously (≈26) in the literature. In addition, our results in the control group are similar to those found previously¹⁵ and provide evidence of reproducibility of the approach used. Whether the use of a greater number of repetitions, based on motor learning principles, results in better levels of UL functioning, similar to previous work in stroke,²⁹ needs to be investigated.

As stated above, the number of trials needed to reach a performance asymptote was lower in controls. As can be seen in Figure 2A, endpoint error decreased in subsequent triads and stabilised. In contrast to this behaviour, individuals with mild TBI managed to decrease error for a few triads before the error increased again or no change was noted (Figure 2B). It took a larger number of trials to achieve an asymptote. Our results agree with previous findings that the capacity to learn and improve performance is influenced by injury to the brain.³⁰ The variability in error seen suggests that individuals with mild TBI seem to have difficulty in reproduction of the motor command (forward model³¹ or referent body configuration³²) necessary to achieve the necessary error correction.

Our results on error in the first triad agree with other findings of greater initial error in individuals sustaining a mild TBI.¹⁹ However, no relationship was found between the error in the first triad and the number of trials needed to correct error in both controls and individuals with mild TBI. This finding could be attributed to the fact that we did not explicitly instruct on error reduction, but rather focussed on performance asymptote. In addition, we did not have information from imaging on cerebellar injury in any of the participants in the mild TBI group. It remains to be determined whether individuals sustaining mild TBIs with cerebellar involvement have deficits in motor learning tasks involving motor adaptation, similar to previous results seen in stroke.³³ Absence of significant differences in endpoint speed between groups, despite the noted variability, also indicates that all participants most probably followed instructions provided ‘to point as fast as possible’. Our results, however, do not help assess whether participants with mild TBI sacrificed speed for greater accuracy.

In terms of other kinematic measures, individuals with mild TBI had more curved movements and used more trunk displacement. They also used less wrist and elbow extension, as well as less shoulder flexion and horizontal adduction. Our results on IC values in the mild TBI group are similar to those found previously,³⁴ and agree with previous work suggesting that individuals with mild TBI have more curved movements.⁶ The mild TBI group used ≈30% more trunk displacement compared to the healthy group. This corresponds to previous findings that individuals with mild levels of UL hemiparesis used 33% more trunk displacement compared to controls.¹⁸ Values for shoulder flexion³⁵ and shoulder horizontal adduction³⁴ for the mild TBI group are similar to previous results. Our findings regarding the range of shoulder horizontal adduction in controls are in agreement with those obtained for a similar pointing task.¹⁷

Despite having perfect scores on clinical assessments, individuals with mild TBI still had deficient performance on pointing tasks. Our results agree with previous suggestions that the use of kinematic assessments can better quantify UL deficits in this population.¹⁹ Similarities in range of motion to previous studies^{17, 34, 35} suggest that injury to the brain, whether acquired (stroke) or traumatic, interferes with successful movement production, especially in the chronic stages. Whether the outcomes differ based on injury severity (i.e., mild, moderate and severe) and time since injury (i.e., acute, sub-acute and chronic stages) remains to be determined.

Our results also highlight the need to use more focussed outcomes in individuals sustaining a mild TBI. This is particularly important for assessing UL issues. As mentioned previously, UL issues affect the performance of daily life activities including driving, dressing and grooming to get ready for work. There are very few approved measures for assessment of UL issues in the TBI population,³⁶ and they primarily cater to those with moderate-to-severe injuries. In addition, measures such as the Action Research Arm Test (a recommended measure in TBI) can have a ceiling effect,³⁷ especially in those with mild injuries. Our results suggest that it is important to assess UL functioning in those with mild TBI. Such information can help guide the decision-making process on clinical management as well as allocation of resources. Although kinematic outcomes are best suited for this purpose, they require specialised equipment and resources that are unavailable in regular clinical practice. Outcomes based on observational kinematics, including the Reaching Performance Scale in Stroke^{38, 39} and Comprehensive Coordination Scale⁴⁰ are suggested as alternatives.

The applicability of our results is limited by the small sample size. Our participants were in the chronic stage of mild TBI. Whether the number of trials necessary to reach an asymptote in performance varies depending upon time since injury and injury severity is currently unknown. We had a younger group of individuals who had sustained a mild TBI. We included them as the majority of injuries are sustained in this age group. However, whether and to what extent the kinematic values differ in older individuals sustaining a mild TBI from healthy age and sex-matched controls needs to be investigated.

Conclusion

Individuals with chronic mild TBIs required more trials to reach an asymptote in learning and had deficient performance on a pointing task compared with controls. Results suggest that the use of kinematic analyses can provide more information about motor deficits that can be identified even in individuals with no clinically identified deficits. These deficits can then be targeted in therapy to improve UL function and enhance motor improvement in this population.

Abbreviations

ADL: Activities of daily living, CAHAI: Chedoke Arm and Hand Activity Inventory, CI: Confidence interval, CSI: Composite Spasticity Index, ES: Effect size, IC: Index of curvature, IQR: Interquartile range, KR: Knowledge of results, SD: Standard deviation, TBI: Traumatic brain injury, UL: Upper limb

Footnotes

Acknowledgements

We would like to acknowledge the participants and Dr Jericho Barela, PT, DPT, Dr Laura Ensminger, PT, DPT and Dr. Bryan Sheffield, PT, DPT for their help in the initial phase of the data collection, and Parisa Hemmat and Venkata Pullabhotla for their help with kinematic data analyses.

Authors’ Contribution

SKS: Conceptualisation, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, supervision, writing—review and editing.

EAG, LV, MDC and DSM: Data curation, conceptualisation, investigation, methodology, writing—original draft.

MVG: Conceptualisation, methodology, writing—review and editing.

Consent to Participate

Written consent was obtained from all participants. All participants signed informed consent forms in accordance with the Declaration of Helsinki.

Data Availability

De-identified data can be obtained upon reasonable request from the corresponding author.

Statement of Ethics

Ethics approval was sought from the Institutional Review Board of the University of Texas Health Science Center at San Antonio (No. 17-508H).

Declaration of Conflicting Interests

The authors declare no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This project was supported by a seed grant from the School of Health Professions, University of Texas Health Science Center at San Antonio.

ICMJE Statement

The authors of this manuscript adhere to the authorship criteria established by the International Committee of Medical Journal Editors (ICMJE). The corresponding author takes public responsibility for the content.

ORCID iD

Sandeep K. Subramanian

References

Maas

, Menon

, Manley

, . Traumatic brain injury: Progress and challenges in prevention, clinical care, and research. Lancet Neurol 2022; 21: 1004–1060.

Capizzi

, Woo

and Verduzco-Gutierrez

Traumatic brain injury: An overview of epidemiology, pathophysiology, and medical management. Med Clin North Am 2020; 104: 213–238.

Alashram

, Annino

, Raju

, . Effects of physical therapy interventions on balance ability in people with traumatic brain injury: A systematic review. NeuroRehabilitation 2020; 46: 455–466. DOI: 10.3233/NRE-203047.

Subramanian

, Fountain

, Hood

, . Upper limb motor improvement after traumatic brain injury: Systematic review of interventions. Neurorehabil Neural Repair 2022; 36: 17–37.

Jang

, Kim

, Son

, . Clinical application of diffusion tensor tractography for elucidation of the causes of motor weakness in patients with traumatic brain injury. NeuroRehabilitation 2009; 24: 273–278.

Schafer

, and Ustinova

KI.

Does use of a virtual environment change reaching while standing in patients with traumatic brain injury?

J Neuroeng Rehabil 2013; 10: 76. DOI: 10.1186/ 1743-0003-10-76.

Levin

, Kleim

, and Wolf

SL.

What do motor “recovery” and “compensation” mean in patients following stroke?

Neurorehabil Neural Repair 2009; 23: 313–319.

van Velzen

, van Bennekom

, van Dormolen

, . Factors influencing return to work experienced by people with acquired brain injury: A qualitative research study. Disabil Rehabil 2011; 33: 2237–2246.

Lefkovits

, Hicks

, Downing

, . Surviving the “silent epidemic”: A qualitative exploration of the long-term journey after traumatic brain injury. Neuropsychol Rehabil 2021; 31: 1582–1606.

10.

Soeker

, and Pape

The use of the model of occupational self-efficacy for work retraining: A multiple case study. Occup Ther Int 2019; 2019: 3867816. DOI: 10.1155/2019/3867816.

11.

Breceda

, and Dromerick

AW.

Motor rehabilitation in stroke and traumatic brain injury: Stimulating and intense. Curr Opin Neurol 2013; 26: 595–601.

12.

Kleim

, and Jones

TA.

Principles of experience-dependent neural plasticity: Implications for rehabilitation after brain damage. J Speech Lang Hear Res 2008; 51: S225–S239.

13.

Banina

, Molad

, Solomon

, . Exercise intensity of the upper limb can be enhanced using a virtual rehabilitation system. Disabil Rehabil Assist Technol 2022; 17: 100–106.

14.

Lang

, Lohse

, and Birkenmeier

RL.

Dose and timing in neurorehabilitation: Prescribing motor therapy after stroke. Curr Opin Neurol 2015; 28: 549–555.

15.

Kimberley

, Samargia

, Moore

, . Comparison of amounts and types of practice during rehabilitation for traumatic brain injury and stroke. J Rehabil Res Dev 2010; 47: 851–862.

16.

Cirstea

, Ptito

and Levin

MF.

Arm reaching improvements with short-term practice depend on the severity of the motor deficit in stroke. Exp Brain Res 2003; 152: 476–488.

17.

Subramanian

, and Levin

MF.

Viewing medium affects arm motor performance in 3D virtual environments. J Neuroeng Rehabil 2011; 8: 36. DOI: 10.1186/1743-0003-8-36.

18.

Subramanian

, Yamanaka

, Chilingaryan

, . Validity of movement pattern kinematics as measures of arm motor impairment poststroke. Stroke 2010; 41: 2303–2308.

19.

Logan

, Semrau

, Debert

, . Using robotics to ruantify impairments in sensorimotor ability, visuospatial attention, working memory, and executive function after traumatic brain injury. J Head Trauma Rehabil 2018; 33: E61–E73.

20.

Subramanian

, Chavez

, Gonzalez

, . Estimation of task practice intensity in individuals with mild traumatic brain injury. Arch Phys Med Rehabil 2021; 102:e52–e53.

21.

Gowland

, Stratford

, Ward

, . Measuring physical impairment and disability with the Chedoke–McMaster Stroke Assessment. Stroke 1993; 24: 58–63.

22.

Levin

, and Hui-Chan

Are H and stretch reflexes in hemiparesis reproducible and correlated with spasticity?

J Neurol 1993; 240: 63–71.

23.

Johnson

, Harris

, Stratford

, . Inter-rater reliability of the Chedoke Arm and Hand Activity Inventory. NeuroRehabilitation 2017; 40: 201–209.

24.

Farmus

, Beribisky

, Martinez Gutierrez

, . Effect size reporting and interpretation in social personality research. Curr Psychol 2023; 42: 15752–15762.

25.

Kinney

, Eakman

, and Graham

JE.

Novel effect size interpretation guidelines and an evaluation of statistical power in rehabilitation research. Arch Phys Med Rehabil 2020; 101: 2219–2226.

26.

Skidmore

ER.

Training to optimize learning after traumatic brain injury. Curr Phys Med Rehabil Rep 2015; 3: 99–105.

27.

Carr JR. Orthogonal regression: A teaching perspective. Int J Math Educ Sci Technol 2012; 43: 134–143.

28.

McLaughlin

and Sainani

KL.

Bonferroni, Holm, and Hochberg corrections: Fun names, serious changes to p values. PM R 2014; 6: 544–546.

29.

Subramanian

, Chilingaryan

, Sveistrup

, . Depressive symptoms influence use of feedback for motor learning and recovery in chronic stroke. Restor Neurol Neurosci 2015; 33: 727–740.

30.

Bourassa

M-E

, Dumel

, Charlebois-Plante

, . Persistent implicit motor learning alterations following a mild traumatic brain injury sustained during late adulthood. J Clin Exp Neuropsychol 2021; 43: 105–115.

31.

Kitago

and Krakauer

JW.

Motor learning principles for neurorehabilitation. In: Barnes MP and Good DC (eds) Handbook of clinical neurology . Amsterdam: Elsevier, 2013, pp. 93–103.

32.

Subramanian

, Feldman

, and Levin

MF.

Spasticity may obscure motor learning ability after stroke. J Neurophysiol 2018; 119: 5–20.

33.

Boyd

, and Winstein

CJ.

Cerebellar stroke impairs temporal but not spatial accuracy during implicit motor learning. Neurorehabil Neural Repair 2004; 18: 134–143.

34.

Knaut

, Subramanian

, McFadyen

, . Kinematics of pointing movements made in a virtual versus a physical 3-dimensional environment in healthy and stroke subjects. Arch Phys Med Rehabil 2009; 90: 793–802.

35.

Cirstea

, and Levin

MF.

Compensatory strategies for reaching in stroke. Brain 2000; 123(Pt 5): 940–953.

36.

McCulloch

, De Joya

, Hays

, . Outcome measures for persons with moderate to severe traumatic brain injury: Recommendations from the American Physical Therapy Association Academy of Neurologic Physical Therapy TBI EDGE Task Force. J Neurol Phys Ther 2016; 40: 269–280.

37.

Lin

, Hsu

, Sheu

, . Psychometric comparisons of 4 measures for assessing upper-extremity function in people with stroke. Phys Ther 2009; 89: 840–850.

38.

Subramanian

, Banina

, Turolla

, . Reaching performance scale for stroke—test–retest reliability, measurement error, concurrent and discriminant validity. PM R 2022; 14: 337–347.

39.

Subramanian

, Margolese

, Turolla

, . Responsiveness of the Reaching Performance Scale for stroke. Arch Phys Med Rehabil 2023; 104: 1588–1595.

40.

Molad

, Alouche

, Demers

, . Development of a comprehensive outcome measure for motor coordination, step 2: Reliability and construct validity in chronic stroke patients. Neurorehabil Neural Repair 2021; 35: 194–203.