Secondary Logo

Journal Logo

Academy News & Notes

Academy of Neurologic Physical Therapy 2018 Combined Sections Meeting Platform Presentations: Scheduling & Abstracts

(Searchable Platform Abstracts are published on-line in PDF format at

Journal of Neurologic Physical Therapy: January 2018 - Volume 42 - Issue 1 - p 33-43
doi: 10.1097/NPT.0000000000000205
  • Free

THURSDAY, FEBRUARY 22, 2018, 11 am–1 pm
Parkinson Disease and Balance Disorders

Rate of Progression in Activity and Participation Outcomes in Exercisers With Parkinson Disease

S. Combs-Miller, M. Mayol, Elizabeth S. Moore, V. Nicholos, B. Ragano, A. Heron, M. Mayol, E. Moore.

Purpose/Hypothesis: The purpose this study was to describe the annual rate of progression in activity and participation domains over 5 years in community-based exercisers with Parkinson disease (PD). We hypothesized that the rate of progression in this group of exercisers would be less than reported rates for general samples of people with PD.

Number of Subjects: Seventy-three self-reported exercisers with idiopathic PD were enrolled in this prospective, longitudinal cohort study. Participants who completed the 5-year testing session were included in the analysis (n = 36; mean age 64.7 [7.0] years; male 69%; mean months postdiagnosis 67.7 [46.1], median Hoehn & Yahr 1.5 [1.0], median minutes of exercise/week 225.0 [120.0]). Demographic and PD characteristics were not significantly different between analyzed and excluded participants at baseline.

Materials/Methods: Participants tested at baseline, 1, and 5 years. Exercisers were defined as having scores of 4 to 5 on the Stages for Readiness to Exercise Scale and reported an average of 60+ minutes of exercise/week within 6 months of each testing session. Participant demographics and PD characteristics were collected at baseline.

Hoehn & Yahr stage, exercise behavior and outcomes within International Classification of Functioning, Disability and Health (ICF) domains of activity (comfortable 10-m walk test [CWT], 6-minute walk test [6MWT], Activities-specific Balance Confidence Scale [ABC], MiniBEST, and Unified Parkinson Disease Rating Scale—activities of daily living subscale), and participation (Parkinson Disease Questionnaire-39 summary index) were assessed at each testing session during the ON cycle of anti-Parkinson medications.

Progression rates for each outcome were based on maximum or baseline scores and calculated as the percentage difference per year between follow-up and baseline values.

Results: Median rate of progression at 1 year from baseline ranged from 0% to 7.5%, with improvement in all outcomes except the ABC (−0.6% decline). Median annual rates of progression over 5 years showed deteriorations of less than 1% per year in all outcomes. Hoehn & Yahr stages remained stable over 1 and 5 years from baseline (0% change). The CWT and 6MWT showed significantly different progression rates per year measured over 1 year compared with annually over 5 years (P < 0.02).

Conclusions: While most outcomes showed improvements 1 year from baseline, all scores deteriorated by less than 1% per year when measured over 5 years. Rate of progression over a longer period (5 vs 1 years) more accurately depicted change in gait-related activities. Yearly rates of progression by our sample of exercisers with PD were less than previous reports from general samples of people with PD that range from 2.4% to 7.4% annual declines in motor and disability measures. Regular exercise may be a critical factor in reducing progression rates in our sample.

Clinical Relevance: Clinicians should encourage their patients with PD to participate in regular exercise as a potential means of slowing disease progression and improving quality of life.

Association of APOE and GBA Variants With Postural Instability/Gait Disturbance in People With Parkinson Disease

V. Kelly, C. Johnson, J. Leverenz, S. Factor, L. Rosenthal, T. Dawson, M. Albert, Z. Wszolek, O. Ross, D. Dickson, J. Quinn, K. Chung, A. Peterson-Hiller, J. Goldman, G. Stebbins, B. Bernard, B. Cholerton, A. Espay, F. Revilla, J. Devoto, I. Mata, K. Edwards, T. Montine, F. Horak, C. Zabetian.

Purpose/Hypothesis: Postural instability and gait disturbances (PIGDs) are common and consequential aspects of Parkinson disease (PD), but PIGD severity and progression are highly variable. Genetic variants are associated with the clinical presentation of PD and may specifically influence PIGD. The apolipoprotein E (APOE) ɛ4 allele is associated with poorer cognitive function in people with PD and with poorer mobility in older adults without PD.

Glucocerebrosidase (GBA) variants are associated with earlier age of onset, more severe cognitive impairment, and more rapid cognitive and motor decline in PD. This study examined the association of PIGD with APOE ɛ4 and GBA variants in a large PD cohort. We hypothesized that those carrying APOE ɛ4 or GBA variants would have greater PIGD severity than noncarriers.

Number of Subjects: 1112.

Materials/Methods: A secondary analysis was performed on cross-sectional data collected at Cleveland Clinic, Emory, Rush, University of Cincinnati, Johns Hopkins Udall Center, Mayo Clinic Udall Center, and the Pacific Udall Center (Portland [Oregon Health & Science University/Portland VA] and Seattle [University of Washington/VA Puget Sound] sites). Demographic characteristics included age, sex, disease duration, and medications. Genomic DNA was extracted from blood or saliva for genotyping. The Movement Disorder Society-Unified Parkinson Disease Rating Scale (MDS-UPDRS) Part III, Motor Subscale (in the “on” state, if taking meds) assessed motor severity, with total PIGD scores calculated as the sum of gait (3.10), freezing of gait (3.11), and postural stability (3.12) items used in determining PIGD phenotype. Associations between genetic variants (APOE ɛ4 carrier vs noncarrier; GBA pathogenic mutation or E326K carrier vs noncarrier of both) and PIGD were assessed using linear regression, adjusted for age, sex, site, disease duration, and levodopa equivalent dose.

Results: Participants had an average (standard deviation) age of 67.9 (9.7) years, disease duration of 8.8 (6.3) years, mild-moderate disease severity (median Hoehn & Yahr stage of 2), and 70% were male. Of those successfully genotyped for APOE, 271 carried 1 ɛ4 allele and 14 carried 2 ɛ4 alleles. For GBA, 107 were pathogenic mutation and/or E326K carriers. In adjusted models, there was no association of PIGD scores with the APOE ɛ4 allele (β = 0.12; 95% confidence interval [CI], −0.13 to 0.38; P = 0.35). Higher PIGD scores were associated with GBA mutation/E326K carrier status (β = 0.41; 95% CI, 0.03-0.79; P = 0.03).

Conclusions: Variants of GBA, but not APOE, were associated with more severe PIGD symptoms as measured by the MDS-UPDRS. This suggests that people with GBA variants, in addition to earlier-onset and more rapid progression, have a clinical profile characterized by poorer balance and gait.

Clinical Relevance: An improved understanding of the genetic factors associated with the severity and progression of PIGD can inform the timing, type, and intensity of physical therapy. Instrumented measures of balance and gait may reveal the influence of genetic risk factors on specific domains to better guide intervention.

Low Back Pain in Parkinson Disease: Impact on Functional Activities, Motor Severity, Physical Activity, and Quality of Life

R. Duncan, L. Van Dillen, J. Garbutt, G. Earhart, J. Perlmutter.

Purpose/Hypothesis: We aimed to (1) describe functional activities limited by low back pain (LBP) in people with Parkinson disease (PD) and (2) determine the relationships between functional limitations resulting from LBP and PD motor severity, physical activity level, and quality of life (QOL) in people with PD. We hypothesized that functional limitations associated with LBP would be frequently reported during standing and walking and would be related to physical activity level and QOL in people with PD.

Number of Subjects: Forty-four subjects with idiopathic PD (mean age = 65.4 ± 9.3 years; 43% female; mean Movement Disorder Society-Unified Parkinson Disease Rating Scale III [MDS-UPDRS III] score = 28.1 ± 10.8).

Materials/Methods: A convenience sample of subjects with PD completed the Revised Oswestry Disability Questionnaire (RODQ), a disease-specific, self-report measure of limitations in function due to LBP. Items in the RODQ assess LBP intensity and its impact on functional activities (eg, standing and walking). We measured motor severity using the MDS-UPDRS III, physical activity using the Physical Activity Scale for the Elderly (PASE) and Life Space Questionnaire (LSQ), and QOL using the Parkinson Disease Questionnaire-39 (PDQ-39). Descriptive statistics characterized LBP intensity and the impact of LBP on functional activities. Spearman correlations were calculated to determine the relationships between the RODQ and (1) MDS-UPDRS III, (2) PASE, LSQ, and (3) PDQ-39.

Results: Of the 44 subjects, 32 (72%) indicated they currently had functional limitations attributed to LBP. LBP was rated to be of at least moderate intensity in 20 (62.5%) participants. The functional activities most frequently reported to be limited due to LBP were standing (75.2%), sleeping (56.3%), walking (53.1%), and lifting (53.1%). RODQ scores were significantly related (P < 0.05) to scores on the MDS-UPDRS III (r = 0.38), PASE (r = −0.40), LSQ (r = −0.47), PDQ-39 Summary Index (r = 0.56), PDQ-39 Mobility subscale (r = 0.56), and PDQ-39 Bodily Pain subscale (r = 0.45).

Conclusions: People with PD frequently experience functional limitations attributed to LBP. A greater degree of functional limitations due to LBP was associated with greater motor sign severity, lower physical activity level, and lower QOL in PD. Future work should substantiate these findings in a larger sample of people with PD with a wider range of disease severity. Further, investigators should explore biomechanical characteristics of functional activity performance to determine their contribution to LBP in people with PD.

Clinical Relevance: Functional limitations associated with LBP are prevalent in people with PD and may negatively impact physical activity level and QOL in this population. Physical therapists should be prepared to screen for and treat LBP and associated functional limitations in people with PD.

Does the Sway Balance Mobile Application Predict Future Falls in People With Parkinson Disease Beyond Other Accepted Clinical Measures or Fall History?

C. Fiems, N. Buchanan, B. Larson, E. Knowles, R. Snow, S. Combs-Miller.

Purpose/Hypothesis: The Sway Balance Application (SWAY) uses smartphone technology equipped with triaxial accelerometers to examine postural stability through the direct measurement of postural sway in an objective and portable fashion. The purpose of this study was to determine whether the SWAY predicts falls and to evaluate its discriminatory sensitivity and specificity relative to the Mini-BESTest, Movement Disorder Society-Unified Parkinson Disease Rating Scale (MDS-UPDRS) motor section, and Activities-specific Balance Confidence (ABC) Scale in identifying fallers versus nonfallers in individuals with Parkinson disease (PD).

Number of Subjects: Fifty-nine subjects with idiopathic PD who completed baseline balance testing were included in the analysis (mean age 68.2 ± 7.8 years; Hoehn & Yahr level 1 = 20%, level 2 = 53%, level 3 = 15%, level 4 = 12%; mean months since diagnosis 83.6 ± 67.3; female 51%).

Materials/Methods: Participants completed a Modified Clinical Test of Sensory Integration and Balance protocol by the SWAY, consisting of holding each of 4 balance conditions for 30 seconds. A score of 0 to 100 (100 indicates the least postural sway) is calculated by the SWAY for each condition and then averaged for an overall stability score. Participants also completed testing for the MDS-UPDRS, Mini-BESTest, ABC and reported 6-month fall history. Participants were then asked to complete a fall diary for each of the following 6 months. Participants were classified as fallers if they reported more than one fall in the 6 months prior to or subsequent to baseline. Binary logistic regression was used to identify significant predictors of future fall status. Cutoff points, sensitivity, and specificity were based on receiver operating characteristic plots.

Results: Of the 5 predictor variables, only past falls and ABC were statistically significant (P < 0.02). The most predictive logistic regression model included falls, the ABC, and the SWAY (c2(3) = 33.623, P < 0.001). This model explained 61% (Nagelkerke R2) of the variance in fall prediction and correctly classified 85% of fallers. Using this model, participants were 32 times more likely to fall in the future if they fell in the past. The ABC and MiniBESTest demonstrated greater accuracy than the SWAY (area under the curve 0.76, 0.72, and 0.65, respectively). Cutoff scores to identify fallers were 85% for the ABC and 21/29 for the MiniBEST.

Conclusions: Our results suggest that the SWAY did not improve the accuracy of predicting future fallers beyond common clinical measures and fall history. The combination of knowledge of previous falls and the ABC was highly predictive of identifying future fallers with PD.

Clinical Relevance: Physical therapists can use the combination of past falls and the ABC to predict future falls in their patients with PD. Mobile applications like the SWAY will require further modification in order to accurately predict future falls.

Perceived Stability Across Conditions of the Modified Clinical Test of Sensory Integration for Balance Differs Based on History of Falling

E. Anson, P. Sparto, Y. Agrawal.

Purpose/Hypothesis: Individuals with balance problems often report feeling unsteady, but some individuals appear to have normal sway during balance tests despite feeling unsteady (eg, persistent postural perceptual dizziness, mal de debarquement syndrome). Perception of postural stability (ie, “How stable did you feel?”) has been related to sway area during standing, such that as sway area increases perceived stability decreases. Fingertip touch contact, stance width, and availability of vision have all been shown to modulate perception of postural stability and sway area. It is not clear how perception of postural stability is related to sway area when sensory information is systematically manipulated, or whether perceived stability differs based on a history of prior falls. We hypothesized that the decline in perceived stability across balance conditions with progressive sensory challenges would be negatively impacted by prior falls.

Number of Subjects: 442 community-dwelling healthy adults (mean age 72.5 [12.4] years, range 24-93 years) participating in the Baltimore Longitudinal Study of Aging were tested.

Materials/Methods: Center of mass sway area was measured using BalanSens (BioSensics, LLC) during a balance test consisting of eyes open and eyes closed conditions on the floor and on a foam cushion. Each individual reported whether they had fallen in the past year. Participants were asked to rate how stable (0 [completely unstable] to 10 [completely stable]) they felt after each balance test condition. Mixed model analyses were used to determine the relationship between perceived stability and sway area while controlling for age, gender, and accounting for repeated balance tests (up to 4 conditions) across individuals. Logistic regression was used to determine whether the odds of having fallen in the past year depend on how perceived stability changed across conditions.

Results: Perceived stability was negatively correlated with sway area such that individuals who swayed more felt less stable (β = −0.084, P < 0.001). Perceived stability declined significantly across each of the 4 balance tests (Ps < 0.005). Independent of sway area, perceived stability decreased with increasing age (β = −0.011, P < 0.001). Individuals with an average reduction in perceived stability of greater than −0.6 across the balance conditions were 3.4 times more likely to have reported at least 1 fall in the past year (P < 0.001).

Conclusions: A history of falling negatively impacts how stable an individual feels during standing balance testing with progressively higher challenging conditions. Subjective perception of postural stability during balance testing may provide additional information useful for estimating fall risk.

Clinical Relevance: Clinicians should consider the impact of perceived stability when interpreting changes in sway area across progressively challenging balance tests.

The Effect of Aging on Brain Activity During Di fferent Visual Sensory Integration Balance Conditions—A Functional Near-infrared Spectrometry Study

C. Blanton, C. Lin, R. Hudson, T. Huppert, P. Sparto.

Purpose/Hypothesis: Functional near-infrared spectrometry (fNIRS) has been utilized to investigate brain hemodynamic changes during dynamic postural control tasks. It is not clear how a change in visual reference affects the activation in prefrontal cortex and vestibular cortex during the sensory integration process. The purpose of this study is to examine the effect of aging on the hemodynamic changes by using different visual inputs (fixed, sway-referenced, and optokinetic stimulation).

Number of Subjects: 17 young adults aged 18 to 35 years and 12 older adults aged 65 to 85 years were recruited in this study. All subjects were right-handed.

Materials/Methods: Screening examinations were performed to confirm the participants' health status. Participants were excluded from this analysis if they scored less than 19 on the Dynamic Gait Index, less than 21 on the Functional Gait Index, less than 67% on the Activities-specific Balance Confidence Scale, or more than 20 on the dizziness handicap inventory. An fNIRS device (NIRSport, NIRx, Germany) was used to detect hemodynamic changes on the left hemisphere over the dorsolateral prefrontal cortex (DLPFC) and temporoparietal (vestibular) cortex (VESTC). An A1-B-A2-B-A2-B-A2 block design was used to elicit changes in vision inputs and somatosensory inputs during 8 pairs of sensory integration balance conditions: (1) sway-referenced visual surround (SV)-fixed visual surrounding (FV)-SV with fixed platform (FP); (2) FV-SV-FV with sway-referenced platform (SP); (3) optokinetic stimulation (OS)-no-optokinetic stimulation (NOS)-OS with FP; and (4) NOS-OS-NOS with SP. Each block was 20 seconds. The 8 test conditions were randomly presented. fNIRS data were analyzed based on a spatial-temporal version of a general linear model. Group-level analysis across the subjects was performed using a random-effects model of brain activity. The level of significance is set at P < 0.05.

Results: Older adults significantly increased activation in the DLPFC compared with younger adults in Condition 2 when changing from FV to SV. Older adults also showed greater activity in the DLPFC during Condition 1 when comparing FV after SV (A2) to baseline FV (A1) on a sway-referenced platform (P < 0.05). Older adults had greater activity in the VESTC during Condition 3 when comparing OS after NOS (A2) to baseline OS (A1) on a fixed platform (P < 0.05).

Conclusions: Our data suggested young and older adults process visual information differently. DLPFC activities increased in older adults when challenged with changing visual input. Older adults were not able to return baseline brain activities after changing or acquired visual information as quickly as younger adults.

Clinical Relevance: Balance training using changing visual inputs is commonly used during vestibular rehabilitation. By understanding the underlying neural mechanisms, current balance interventions can be assessed for value and refined and new interventions may be created.

Use of Near-infrared Spectroscopy to Examine Cerebral Activation During Optic Flow in Individuals With and Without Complaints of Visual Vertigo

C. Hoppes, P. Sparto, S. Whitney, J. Furman, T. Huppert.

Purpose/Hypothesis: Individuals with visual vertigo describe symptoms of dizziness, disorientation, and/or impaired balance in environments with conflicting visual and vestibular information or complex visual stimuli. Physical therapists often prescribe habituation exercises using optic flow as part of a rehabilitation regimen to treat these symptoms, but there are no evidence-based guidelines for delivering optic flow. While beneficial and often prescribed, it is unclear how the brain processes the visual stimuli. The purposes of this study were to use functional near-infrared spectroscopy (fNIRS) to explore cerebral activation during optic flow, and determine whether the support surface had a modulating effect on brain activity.

Number of Subjects: Thirty participants (15 patients with visual vertigo and 15 age- and gender-matched healthy controls).

Materials/Methods: Participants stood on a force plate in a virtual reality environment and viewed two types of anterior-posterior optic flow (single sine and sum of sines) while standing on a fixed or sway-referenced surface. Changes in cerebral activation were recorded from the bilateral fronto-temporo-parietal and occipital lobes using fNIRS.

Results: Cerebral activation, indicated by change in oxyhemoglobin concentration, was reduced in patients with visual vertigo compared with healthy controls in the bilateral anterior frontotemporal regions during optic flow when standing on a fixed floor (single sine optic flow: T = −3.95 left anterior, T = −4.59 right anterior; sum of sines optic flow: T = −3.71 left anterior, T = −3.16 right anterior; all P < 0.02, false-discovery rate corrected). Cerebral activation was also reduced in patients with visual vertigo compared with healthy controls in the right anterior frontotemporal region during optic flow when standing on a sway-referenced floor (sum of sines optic flow: T = −2.02, P = 0.05, false-discovery rate corrected).

Conclusions: Individuals with visual vertigo show less cerebral activation in regions associated with multisensory integration in comparison to healthy controls. This decreased activation may represent an altered ability to perform sensory reweighting of visual and vestibular information, leading to symptoms of dizziness and imbalance. The decreased activation of the temporal regions in our patients with visual vertigo is consistent with the decreased activation of the right superior temporal region observed in patients with unilateral vestibular neurectomy and chronic subjective dizziness. Decreased functional connectivity with the right superior temporal region has also been observed in patients with visually induced dizziness during resting-state fMRI.

Clinical Relevance: Our findings suggest that people with visual vertigo integrate optic flow information differently from healthy subjects. Future work should explore whether this cortical activation pattern can be modified with physical therapy intervention, such as habituation exercises using optic flow.

Gaze Stability in Healthy Elite Level Athletes

S. Massingale, A. Alexander, E. McQueary, H. Kisana, C. Leddon, J. Pardini.

Purpose/Hypothesis: Vestibular symptoms including dizziness and visual motion sensitivity following mild traumatic brain injury (mTBI) are commonly reported. Research suggests that functional impairments with the vestibular ocular reflex (VOR) could be one cause of these symptoms. The VOR is used to stabilize the visual field as the head is in motion, both actively and passively, by moving the eyes equally and opposite of the head, to stabilize the image on the retina. One way to assess the function of the VOR is with computerized gaze stabilization testing (GST). Current research suggests that normal GST for a general adult sample ranges from 90° to 150°/s in the yaw plane; however, there is limited literature on GST performance of elite athletes. Using typical normative velocities may be inaccurate for elite athletes when deciding whether gaze stability has recovered after injury. The purpose of this study is to establish gaze stability ranges of head velocity for elite athletes.

Number of Subjects: 100 male professional baseball players (age range: 17-31 years; mean = 23.08 years).

Materials/Methods: Subjects were tested using the Bertec gaze stability testing protocol, including static visual acuity (SVA), visual processing speed, and gaze stability tests in the high performance range. Each subject sat upright, 5 ft from a monitor, with his head moved passively in the yaw plane by a trained therapist. Velocity of head movement was determined using an accelerometer on the forehead. Left (L) and right (R) head velocity was recorded while testing for correct acuity. The optotype used for the visual target was a capital letter E, sized at 2.0 logMAR above SVA, which flashed on the computer screen when the target head velocity was achieved for 3 head turns within a prescribed range. The athlete's answers for the orientation of the letter were recorded by the therapist. Computer algorithms were used to determine the highest velocity achieved while still correctly identifying the orientation of the letter.

Results: Average head velocity for this group of athletes at a normal state was 208.15°/s (L) and 222.79°/s (R). GST ranged from 130 to 335 (L) and 130 to 365 (R). Trials for L and R GST resulted in high consistency markers for all trials.

Conclusions: Performance of elite athletes on GST measures was observed to be much higher than that reported in nonathlete samples. Thus, it is important to establish separate norms for specialized populations to ensure appropriate clinical management.

Clinical Relevance: Understanding normal performance in high-level athletes via specialized norms allows the provider to better determine recovery of vestibular function after injury and contribute meaningful information to be used as part of a safe return to play process after mTBI.

FRIDAY, FEBRUARY 23, 2018, 8 am–10 pm
Spinal Cord Injury and Measurement

Effects of Locomotor Training Intensity on Walking Performance in Individuals With Chronic Spinal Cord Injury

G. Brazg, M. Fahey, C. Holleran, M. Connolly, J. Woodward, P. Hennessy, B. Schmit, T. Hornby.

Purpose/Hypothesis: Many physical interventions can improve locomotor function in individuals with motor incomplete spinal cord injury (iSCI), although the training parameters that maximize recovery are not clear. Previous studies in individuals with other neurologic injuries suggest the intensity of training may positively influence walking outcomes. However, the effects of intensity during locomotor training (LT) provided to individuals with iSCI have not been tested. The purpose of this pilot, blinded-assessor randomized crossover trial was to evaluate the effects of LT intensity on walking outcomes in individuals with iSCI.

Number of Subjects: Seventeen subjects with iSCI at the neurological level of T10 or higher and could walk at least 10 m at less than 1.0 m/s without physical assistance completed the study.

Materials/Methods: Participants with chronic (>1 year) iSCI performed either high- or low-intensity LT for up to 20 sessions over 4 to 6 weeks. Four weeks following completion, the training interventions were alternated. Targeted intensities focused on achieving specific heart rate (HR) ranges or ratings of perceived exertions (RPE). The high-intensity LT paradigm targeted 70% to 85% of predicted maximum HR (HRmax) and an RPE of 15 to 17 (“hard” to “very hard”). The low-intensity LT paradigm targeted 50% to 65% HRmax and an RPE of 11 to 13 (below “somewhat hard”).

Intensity was manipulated by increasing speeds or applying loads to the trunk or legs, and LT included speed- and skill-dependent treadmill and overground training, and stair training. Participants were assessed prior to and following each LT paradigm. Primary measures included the 6-minute walk test and peak treadmill (TM) speed, with simultaneous collection of cardiorespiratory function (K4b2, CosMed, Chicago, IL).

Results: Mixed model analyses of variance using both time and intensity as repeated main factors and order as an independent factor revealed significantly greater increases in changes in peak TM speeds and secondary measures of overground speed following high- versus low-intensity LT, with no interactions with order of training. Greater improvements in combined peak aerobic capacity and efficiency were also observed. To minimize carryover effects, analyses of the initial training conditions revealed similar significant differences. Moderate to high correlations were observed between selected differences in locomotor improvements and differences in HR or RPE during high- versus low-intensity training.

Conclusions: This pilot randomized trial provides the first evidence that the intensity of stepping practice may be an important determinant of locomotor outcomes in individuals with iSCI. Whether such training is feasible in larger patient populations and contributes to improved locomotor outcomes deserves further consideration.

Clinical Relevance: Intensity of stepping practice is an important determinant to consider when providing task- specific LT with patients with iSCI in order to optimize functional outcomes.

Myelin Increases Accompany Eccentric Motor Learning in Chronic SCI

T. Faw, B. Lakhani, L. Worthen-Chaudhari, T. Thaxton, R. Deibert, L. Fisher, M. Bjelac, H. Nguyen, P. Schmalbrock, J. Schmiedeler, D. McTigue, L. Boyd, D. Basso.

Purpose/Hypothesis: Activity-dependent myelin plasticity is a novel mechanism of neuroplasticity with the potential to support recovery after central nervous system (CNS) damage. In fact, animal studies indicate that new myelin formation is required for motor skill learning in the intact CNS. While traditional activity-based rehabilitation produces modest gains after spinal cord injury (SCI), eccentric deficits persist, making independent ambulation impossible for most individuals. Downhill (DH) training is a novel, challenging intervention that promotes greater skill learning of locomotion in rodent SCI. Therefore, the purpose of these studies was to employ a translational paradigm to determine the extent and mechanism of DH training-induced myelin plasticity in humans and rodents with SCI.

Number of Subjects: Human participants (n = 5)

–SCI DH (n = 4)

–Uninjured control (n = 1)

mT/mG mice (n = 9)

–Naive (n = 6)

–SCI DH (n = 3)

–SCI unexercised (n = 0)

Myrf(+/−,−/−) mice (n = 6)

–Myrf(+/−) DH (n = 1)

–Myrf(−/−) DH (n = 5)

Materials/Methods: In chronic, incomplete human SCI, multicomponent T2 relaxation imaging (MCRI) of the brain and spinal cord occurred before and after a 12-week DH training intervention (10% decline; 20 minutes/day; 5-minure bouts separated by 5-minute seated rest; 3 days/week). To determine the mechanisms underlying myelin plasticity, we performed experiments in 2 transgenic mouse lines with and without DH training after chronic, severe T9 contusion SCI. mT/mG mice identify new myelin produced in untrained or DH-trained animals histologically. Brain and spinal cord MCRI pre-/post-training and immunohistochemistry was used to identify training-induced new myelin in mice. Myrf(−/−) mice prevent new myelin formation through genetic deletion of myelin regulatory factor, thus were used to determine whether new myelin is required for eccentric motor learning after SCI. DH training in mice was performed daily (10% decline; 20 min/day; 10-minure bouts separated by 10-minure rest) from 5 to 7 weeks postinjury.

Results: Human MCRI data show training-induced increases in myelin water fraction (MWF), a histopathologically validated measure of myelin. One participant, who responded well, progressed from a rolling walker to cane for ambulation and had greater MWF in brain regions associated with motor learning (primary motor and sensory, motor planning, and visuomotor association). A second participant had right unilateral motor control gains after DH training. In this participant, MWF increases occurred in the contralateral left hemisphere only, suggesting that MCRI is sensitive and specific to motor relearning after SCI. Translational mouse experiments are currently underway.

Conclusions: These preliminary data suggest that eccentric-focused DH training targeting persistent locomotor deficits promotes myelin plasticity in humans with SCI. Ongoing translational experiments in mice will provide valuable insight into the mechanisms underlying this response.

Clinical Relevance: Understanding and exploiting the neurobiology underlying recovery from SCI are critical. Myelin plasticity is one such mechanism accessed via DH training.

Constraints on Ground Reaction Force Production During Overground Walking in Persons With Chronic Incomplete Spinal Cord Injury

D. Peters, Y. Thibaudier, J. Deffeyes, G. Baer, H. Hayes, R. Trumbower.

Purpose/Hypothesis: Persons with incomplete spinal cord injury (iSCI) often face chronic mobility impairments such as slow walking speed, increased reliance on assistive devices (ADs), and difficulty adapting gait to environmental demands. Forces associated with control of gait and body weight support, as measured by ground reaction forces (GRFs), are likely altered following iSCI due to lower extremity/trunk weakness and AD use, but remain largely unexplored. The aim of this study was to examine GRF production during overground walking after iSCI, as greater insight into GRF constraints is important for refining mobility interventions. We hypothesized that persons with iSCI would exert smaller GRFs and impaired GRF modifications to increased cadence compared with able-bodied (AB) persons, especially when using an AD.

Number of Subjects: Fifteen persons with chronic iSCI (age: 47.6 ± 15.5 years; time post-iSCI: 14.0 ± 12.4 years) stratified into no-AD (n = 7) and AD groups (n=8), and 15 age-matched AB controls participated in this study.

Materials/Methods: All participants walked across a walkway with 2 embedded force plates at self-selected and fast (115% self-selected) cadences. To control for cadence and AD on GRF production, AB controls performed additional trials at the self-selected cadence of their iSCI match, and the self-selected/fast cadences of their iSCI match using the matching AD. Independent sample t tests, or Mann-Whitney U tests for nonnormal distributions, were used to compare fore-aft (transition point, braking impulse, and propulsion impulse) and vertical (maximum vertical loading during weight acceptance, push-off, and vertical modulation) GRFs between persons with iSCI and AB controls; paired sample t tests, or Wilcoxon signed rank tests for nonnormal distributions, were used to compare GRFs at different cadences.

Results: Fore-aft and vertical GRFs are reduced in persons with iSCI as compared with AB controls, with greatest deficits in those dependent on an AD. When controlling for cadence and AD, propulsive impulse was still lower in persons with iSCI (no-AD group: 257.3% ± 97.3%; AD group: 176.2% ± 67.5%) as compared with AB controls (no-AD group: 371.5% ± 77.6%; AD group: 279.8% ± 79.9%, P ≤ 0.014). In addition, our results suggest that AD use constrains GRF production, even in AB controls, with braking impulse particularly reduced. In contrast to AB controls, persons with iSCI demonstrated impaired GRF modifications to an increase in cadence.

Conclusions: Reduced GRF generation and altered GRF modification occur after iSCI. While compensatory strategies involving slower cadence and use of an AD can impact GRF production, propulsive forces are still reduced after controlling for these variables.

Clinical Relevance: These results highlight the importance of therapeutic strategies that target GRF improvements. Incorporating dynamic neuromuscular control strategies, decreased reliance on ADs, and/or biofeedback for force production during walking may enhance GRF production following iSCI.

Effect of Acute Intermittent Hypoxia on Upper Limb Function in Individuals With Spinal Cord Injury

S. Anastasopoulos, W. Rymer, M. Sandhu.

Purpose/Hypothesis: Spinal cord injuries (SCIs) disrupt the pathways between brain and spinal cord, resulting in impairment of motor control and loss of independent mobility. Spontaneous plasticity, which occurs over time in the spared spinal pathways, is often insufficient to restore normal function. One unique approach to augment plasticity in spinal networks is via exposure to brief and mild bouts of hypoxia (ie, low oxygen). This approach is known as acute intermittent hypoxia (or AIH), and has been shown to induce rapid plasticity and enhance volitional somatic motor output in the lower extremity in persons with incomplete SCI. Whether AIH induced neuroplasticity is equally prevalent in spinal motor pathways, regulating upper limb muscular function, is not known. Accordingly, in ongoing studies our aim is to test the hypothesis that AIH will augment hand strength and function in persons with quadriplegia.

Number of Subjects: 10.

Materials/Methods: We used a randomized, double-blinded, placebo-controlled, and crossover study design. Ten adults with chronic motor-incomplete SCI between C3 and T1 received a single session of AIH or sham AIH, via a washout period of at least 1 week. During AIH, subjects received 15, 90-second episodes of 10% O2, interspersed with 21% O2 for 90 seconds. During sham AIH, subjects received episodes of normoxia (normal room air) while wearing the breathing masks. Maximal grip strength, pinch strength, 9-hole peg test, and box and block test were assessed prior to administration of AIH or sham AIH, every half hour for up to 5 hours post-AIH.

Results: Grip strength increased by 26% ± 4% from baseline to 60 minutes after AIH treatment (21 ± 5 lb at baseline; 26 ± 5 lb at 60 minutes post-AIH). Sham AIH treatment resulted in a 5% ± 3% increase in strength (21 ± 4 lb at baseline; 22 ± 4 lb 60 minutes after sham treatment). The changes in pinch strength were more variable and did not result in significant differences between the 2 groups. Similarly, we did not see significant differences in 9-hole peg test and box and block test between the 2 groups.

Conclusions: These preliminary observations demonstrate the potential of AIH to rapidly increase grip strength in persons with incomplete SCI. This modality could eventually be developed to induce spinal plasticity, either as an alternative to prevailing therapies, and as an adjunct to bolster the effectiveness of superimposed rehabilitative training in individuals with chronic incomplete SCI.

Clinical Relevance: While repetitive AIH may serve as a stand-alone therapy, we expect it will be most effective when paired with rehabilitation interventions, as facilitating spinal neural networks with high-intensity task-specific training will have a synergistic effect on upper limb function.

Impairments in Dynamic Stability of Gait are Associated With the Deficits in Somatosensory Conduction and Cerebellar Involvement in People With Multiple Sclerosis

G. Gera, P. Fino, B. Fling, F. Horak.

Purpose/Hypothesis: Dynamic stability of gait reflects ability of the system to flexibly adapt to perturbations during walking. Previous work from our laboratory has shown that people with multiple sclerosis (PwMS) have impaired local dynamic stability as compared with control individuals for a steady-state walking. However, we do not know how deficits in somatosensory conduction affect the local dynamic stability. We also do not understand the relationship of local dynamic stability and cerebellar white matter tract integrity (especially middle and inferior cerebellar peduncles, which receive afferent inputs to the cerebellum and contribute to gait and balance function). We hypothesized that deficits in the local dynamic stability will be related to (1) the postural response latency and (2) reduced white matter tract integrity of the middle and inferior cerebellar peduncles.

We also performed secondary analyses where we investigated the relationship of local dynamic stability with balance (MiniBESTest) and disease severity (Expanded Disability Severity Scale; EDSS).

Number of Subjects: Sixteen PwMS; EDSS (2-4).

Materials/Methods: Subjects underwent 20 discrete, backward translations of a support surface. Onset of postural response latency of agonist (medial-gastrocnemius) muscle was assessed. Local dynamic stability was estimated using maximum finite time Lyapunov exponents (LyEs). Triaxial accelerations of the trunk were obtained for 2 minutes of self-paced steady-state gait. The 2 minutes of gait was segmented into multiple bouts by removing turns. Five consecutive strides were extracted from each bout and time normalized. Maximum LyEs were calculated for each bout of 5 strides using a 9D state space composed of the 3D acceleration signals and their twice time-lagged copies. Diffusion-weighted images of brain were also acquired. Radial diffusivity (RD), an indirect neural marker of myelination, of cerebellar peduncles was calculated for each participant. Lower RD is interpreted as being indicative of better white matter tract microstructure.

Results: We found that deficits in local dynamic stability were associated with increased postural response latency (r = 0.6, P < 0.05) and reduced white matter tract integrity of both middle (r = 0.5, P < 0.05) and inferior (r = 0.52, P < 0.05) cerebellar peduncles. Our secondary analysis also revealed that impairments in the local dynamic stability were associated with poor balance (MiniBEST: r = −0.69, P < 0.01) and disease severity (EDSS: r = 0.58, P < 0.05).

Conclusions: Findings of this study suggest that the dynamic stability is impaired in PwMS and is related to the impaired somatosensory conduction, reduced integrity of the cerebellar white matter tracts, poor balance, and disease severity.

Clinical Relevance: Reduced dynamic stability of gait can potentially go unnoticed in mild individuals with MS and can make them susceptible to falls. Thus, improvements in dynamic gait stability should be considered and/or targeted during gait rehabilitation.

Revisiting Perry's Gait Speed Categories

G. Fulk, Y. He, P. Boyne, K. Dunning.

Purpose/Hypothesis: In 1995, Perry and colleagues established widely used gait speed (GS) values to categorize walking ability in people poststroke: household walker (GS < 0.40 m/s), limited community walker (GS: 0.40-0.80 m/s), and community walker (GS > 0.80 m/s). Assignment of participants to these categories was based on clinician expert opinion of the participant's capability. The purpose of this study was to reexamine these GS values using walking ability categories based on real-world home and community walking activity.

Number of Subjects: Data from 441 participants 1-year poststroke were analyzed.

Materials/Methods: Cross-sectional, secondary data analyses of home and community walking activity from 2 stroke rehabilitation trials. Household (100-2499 steps/day), limited community (2500-7499 steps/day), and full community (≥7500 steps/day) walking categories were developed based on population-based studies. Participants' walking activity was measured using an activity monitor, the StepWatch Activity Monitor (SAM). GS was measured over a 10-m walk. Data were analyzed using receiver operator characteristic curve to identify GS cutoff values.

Results: GS less than 0.49 m/s distinguished between household walkers and limited community/community walkers (area under the curve [AUC] = 0.799, sensitivity = 0.87, specificity = 0.61). GS more than 0.92 m/s distinguished between limited community walkers and community walkers (AUC = 0.742, sensitivity = 0.60, specificity = 0.80).

Conclusions: GS values developed by Perry and colleagues may overestimate walking activity. Limitations of this study are that the SAM data do not provide information on where the participants were walking or the goals for their walking activity in the community.

Clinical Relevance: The GS values identified in this study (GS < 0.49 m/s = household walker, GS between 0.49 and 0.92 = limited community walker, and GS > 0.92 m/s = community walker) may be useful for goal setting and monitoring change in walking activity in people with stroke.

Accuracy of Three Commercially Available Activity Monitors to Measure Stepping Activity During Physical Therapy in Individuals Post-stroke

C. Henderson, T. Hornby.

Purpose/Hypothesis: Interventions aimed at addressing locomotor deficits in individuals poststroke are frequent components of physical therapy (PT) sessions. Recent evidence suggests the amount of walking practice may be an important modifiable variable that can contribute to improved walking performance, making accurate measurement during clinical interventions a priority. Many available activity monitors are designed to measure stepping activity in healthy adults, and recent evidence suggests it may be possible to adapt their use for individuals poststroke. As these studies only evaluated short bouts of forward walking, their accuracy during PT sessions, during which a variety of walking and nonwalking tasks are performed, is uncertain. The purpose of this study was to evaluate the accuracy of these devices during actual and simulated PT sessions in individuals poststroke.

Number of Subjects: We recruited 30 individuals with a history of stroke participating in inpatient (n = 23) or outpatient (n = 7) PT, as well as 12 healthy adults.

Materials/Methods: Over 100 PT sessions, the Fitbit One, Stepwatch 3, and Actigraph wGT3X-BT (with and without low frequency extension [LFE]) monitors were worn on both the paretic and nonparetic ankles while stepping was counted manually. In addition, 12 healthy adults completed 33 walking and 29 nonwalking tasks commonly performed during PT sessions. Device and manual step counts were compared via intraclass correlation coefficients (ICCs) and Bland-Altman analyses. Absolute and percent error differences between devices were evaluated via Friedman tests with post hoc Wilcoxon signed rank tests.

Results: A median of 743 steps was taken during observed PT sessions, with ICCs varying from 0.78 to 0.98, median absolute errors from 57 to 352 steps, and median percent errors from 8.7% to 61%. Bland-Altman analyses identified consistent undercounting of the Fitbit and Actigraph in sessions with low step counts. Post hoc analyses suggest superior accuracy of the Stepwatch and Actigraph LFE and low accuracy of the Fitbit regardless of limb placement. Analysis of data from healthy adults indicates excellent median agreement during walking tasks (99.7%-101.2%) with the exception of Actigraph (54%) while the Actigraph LFE was more likely than all other devices to report false stepping during nonstepping activities.

Conclusions: The Stepwatch and Actigraph LFE are similarly accurate in measuring stepping activity during PT sessions regardless of which limb the monitor is placed. The LFE decreases the threshold acceleration required for stepping detection by the Actigraph, making it more susceptible to excessive step counts in sessions with higher portions of nonstepping activities.

Clinical Relevance: Accurately counting stepping activity during PT sessions is possible with selected activity monitors. These data may be of assistance for both researchers and clinicians to identify the amount of walking practice provided in an intervention.

The Use of a Commercial Physical Activity (PA) Monitor to Quantify Physical Activity Levels in Individuals With Parkinson Disease (PD)

S. Pradhan, V. Kelly.

Purpose/Hypothesis: Individuals with Parkinson disease (PD) are less active than healthy older adults (HOAs), with decline in moderate-intensity activity being particularly important. With the surge in the use of commercial physical activity (PA) monitors, it is necessary to generate reference data to compare patient-reported PA levels. This study measured the quantity and intensity of PA in individuals with PD compared with HOA using a commercial monitor. Secondarily, we examined factors associated with PA levels in the PD group. We hypothesized that quantity and intensity of PA would be less in individuals with PD than in HOA.

Number of Subjects: 19 individuals with a mild to moderate PD (mean [standard deviation]) Movement Disorder Society-Unified Parkinson Disease Rating Scale (MDS-UPDRS) motor score of 10 (9.1) and 10 HOA participated. All participants were independent community ambulators without the use of an assistive device.

Materials/Methods: An assessment of physical function included the Movement Disorder Society-Unified Parkinson Disease Rating Scale (MDS-UPDRS), gait speed (10-mr walk), endurance (6-minute walk), balance (MiniBEST), functional reach test, Falls Efficacy Scale-International, and Activities-specific Balance Confidence Scale (ABC). Participants were instructed to wear the Fitbit Charge HR continuously day and night for a period of 14 days, except when in water.

Quantity (number of steps) and intensity (sedentary, light, moderate, and vigorous) were extracted for each day and averaged across days. A 1-way analysis of variance examined differences in PA between PD and HOA. Factors associated with PA levels were examined using bivariate correlations.

Results: Compared with HOA, people with PD took fewer steps/day (F = 8.03, P = 0.09), walked a shorter total distance (F = 8.93, P = 0.006), had fewer moderate- or higher-level activity minutes (F = 9.36, P = 0.005), and had more sedentary time (F = 14.4, P = 0.001). Fifty-two percent of the PD group versus 20% of the control group did not meet weekly recommendations for the amount of time spent in moderate- to vigorous-intensity activity. Eighty-nine percent of individuals with PD versus 50% of the HOA did not walk an average of 10 000 steps/day. In the group with PD, more steps/day were associated with faster gait speed (self-selected: r = 0.68, fastest possible: r = 0.60) and better endurance (6-minute walk test: r = 0.49). Time spent in moderate- or vigorous-intensity activity was significantly associated with better balance confidence (ABC: r = 0.49).

Conclusions: Individuals with PD had had lower and less intense levels of PA and spent more time in sedentary activities. A significantly higher proportion of individuals with PD did not meet the recommendations for quantity or intensity of PA compared with HOA.

Clinical Relevance: Decline in PA is seen early in the disease process in individuals with PD, even those with mild motor symptoms who are still active and participating independently in the home and community. The use of commercial PA monitors can be leveraged to provide real-time feedback about current PA levels to people with PD and can enable newly diagnosed PD to better self-monitor and actively manage their symptoms through improved PA.

SATURDAY, FEBRUARY 24, 2018, 11 am– 1 pm
Stroke and Traumatic Brain Injury

Neuroplasticity and Biomechanical Correlates Underlying Post-stroke Gait Rehabilitation

T. Kesar, K. Watson, S. Eicholtz, L. Marquard, L. Snopek, T. Damas, K. Seynders.

Purpose/Hypothesis: Although neuroplasticity is considered a central mechanism underlying rehabilitation, there are very few comprehensive investigations of neuroplasticity processes underlying gait training. This study investigates the neuroplasticity correlates underlying a gait training intervention, combining fast treadmill walking and functional electrical stimulation (FastFES) delivered to dorsiflexor and plantarflexor muscles. We investigated 2 hypotheses: (1) 3 sessions of either Fast or FastFES training are sufficient to induce change in cortical and spinal excitability; (2) 3 sessions of FastFES training produce greater improvements in corticospinal excitability, gait biomechanics, and walking function compared with Fast training.

Number of Subjects: This is an ongoing study planned on 10 stroke participants. To date, data collected on 3 individuals with poststroke hemiparesis (3 females, 57.0 ± 12.2 years; 35.3± 22.0 months poststroke; LE-Fugl-Meyer score 22 ± 4.6; overground gait speed 0.94 ± 0.23 m/s).

Materials/Methods: Participants underwent a repeated-measures AB/BA crossover design with Fast treadmill training and FasftFES treadmill training. The dependent variables were categorized into 3 domains. Paired-pulse transcranial magnetic stimulation was utilized to evaluate intracortical facilitation and short-interval intracortical inhibition (SICI) of the paretic soleus muscle. Soleus spinal excitability was evaluated using Hmax/Mmax. Gait biomechanics variables included peak anterior ground reaction force (AGRF), ankle plantarflexor moment, and peak trailing limb angle for the paretic leg. Clinical function was evaluated using the Timed Up and Go and 10-m walk tests (self-selected and fast speeds). The study is currently ongoing, with descriptive statistics employed for data analysis.

Results: Three sessions of gait training (Fast or FastFES) induced improvements in corticospinal excitability, biomechanics, and walking function. Compared with changes induced by fast training, 3 sessions of FastFES induced greater improvements in the 10-m walk test at fast speed (change score for FastFES = 0.09 ± 0.03 m/s; Fast = 0.05 ± 0.04 m/s), peak AGRF (change score FastFES = 24.99 ± 6.68 N; Fast = 14.95 ± 12.80 N), and %SICI (change score FastFES = 21.00% ± 22.61%; Fast = −16.15% ± 11.91%).

Conclusions: Our ongoing study showcases a unique multimodal dataset spanning 3 domains (corticospinal excitability, gait biomechanics, and clinical outcomes) as well as multiple time points during training. Our preliminary findings elucidate complex interrelationships among changes in these 3 domains that merit more investigation.

Clinical Relevance: The understanding of how Fast and FastFES treadmill training elicits changes in neurophysiologic, biomechanics, and functional outcomes will elucidate mechanisms underlying the intervention's clinical effectiveness. Our study findings can enable development of gait training approaches that maximally harness neuroplasticity and are based on in-depth knowledge of the neural substrates underlying gait rehabilitation.

Contribution of Impairments in Paretic Plantarflexor Force Capacity, Volitional Strength, and Activation to Paretic Propulsion During Hemiparetic Walking

L. Awad, S. Binder-Macleod.

Purpose/Hypothesis: An inability to generate propulsion by the paretic limb is a major contributor to walking-related disability after stroke; however, few interventions have been able to increase propulsion after training. This is, in part, due to limitations in our ability to diagnose the source of propulsion deficits in the heterogeneous poststroke population. Indeed, although the ankle plantarflexor (PF) muscles are the primary generators of propulsive forces during walking, and their function is impaired after stroke, it is unclear whether it is a reduced force-generating capacity of the PFs, reduced volitional access to the available capacity, or a combination of each that contributes to reduced propulsion during walking. The objective of this study was to investigate the relationship among these variables.

Number of Subjects: 36 participants more than 6 months poststroke.

Materials/Methods: PF capacity (ie, maximum force-generating ability [MFGA]), volitional strength (ie, maximum volitional isometric contraction [MVIC]), and activation (ie, MVIC ÷ MFGA) were measured using dynamometry and supramaximal electrical stimulation. Walking was analyzed on an instrumented treadmill with paretic propulsion defined as the peak anterior ground reaction force. Moderated regression was used to quantify the relationships between PF function and paretic propulsion during (i) comfortable speed walking and (ii) speed modulation, while accounting for variation in walking speed across participants.

Results: Deficits in paretic PF activation independently contributed to deficits in paretic propulsion. A walking speed × PF activation interaction was observed, with the final model predicting paretic propulsion during comfortable speed walking, accounting for 81.7% of the variance (R2 = 0.82, F(3,32) = 47.56, P < 0.001). Moreover, PF activation and walking speed were independent predictors of changes in propulsion during speed modulation (R2 = 0.59, F(2,31) = 22.50, P < 0.001).

Conclusions: Impaired activation, not strength, of the paretic PFs underlies paretic propulsion deficits during hemiparetic walking. This finding highlights the need for interventions that target deficits in paretic limb neuromotor control.

Clinical Relevance: Lower extremity-strengthening programs alone may not be sufficient to improve paretic propulsion during walking.

Comprehensive and Targeted Assessment of Walking Adaptability is Warranted for the Stroke Population

C. Balasubramanian, D. Clark, E. Fox.

Purpose/Hypothesis: Walking adaptability (WA), defined as the ability to modify walking to meet task goals and environmental demands, is crucial to safe ambulation in the home and community. Clinicians lack a rigorous assessment to quantify adaptability limitations. Frequently used stroke assessments involve walking at a steady state to predict walking recovery (eg, 10-m walk speed test). These assessments do not account for the full repertoire of walking skills required to reengage in safe and successful ambulation. The purpose of this study was to assess WA in community-dwelling stroke survivors and estimate the need for specific and comprehensive assessment of WA. We hypothesized that WA will be compromised even in those persons who may be classified as community ambulatory.

Number of Subjects: Thirty-two persons (age = 60 ± 13 years; female =11) with stroke (time since stroke = 28 ± 21 months) who walked independently without physical assistance from another person participated.

Materials/Methods: Individuals performed 19 walking tasks and assessment items captured 8/9 domains of WA. A physical therapist graded performance using a 3-point (0-2) ordinal scale. Walking speed was calculated from the 10-m walk test. Ambulation level was classified using the commonly cited Perry's classification (speed < 0.4 m/s = home ambulatory, 0.4-0.8 m/s = limited community ambulatory, and speed > 0.8 m/s = community ambulatory). “Success in walking adaptability (SWA)” for each participant was calculated as the percentage of adaptability tasks that individuals performed successfully (ie, clinical grade = 2). Descriptive statistics analyzed the data.

Results: Sample walking speed was 0.83 ± 0.29m/s (range = 0.14-1.28 m/s). Sample SWA ranged from 0% to 7%. Seventeen participants who were categorized as community ambulatory (1.05 ± 0.10 m/s) had an average SWA of 50% (range = 10%-77%), suggesting that, on an average, these high-functioning persons were successful on only 50% of the assessed WA tasks. Five adaptability tasks (walk to step on a foam, stairs, walk to negotiate 1 obstacle, walk fast and negotiate 2 obstacles, and walk to step up on a small step) seemed to be most difficult based on the high percentage of sample (20%-53%) receiving a clinical grade of “0.”

Conclusions: Even those individuals classified as community ambulatory succeeded in only 50% of the assessed adaptability tasks. Our approach to adaptability assessment was specific since assessment items closely simulated walking tasks that persons might perform in the home and community. Our results also imply that some adaptability tasks may be harder to perform when compared with others. Future work should validate assessment items and investigate their difficulty levels to develop a targeted clinical assessment tool of WA in individuals poststroke.

Clinical Relevance: While steady-state walking speed has been used to classify ambulatory status, it may be unable to reveal specific WA deficits for persons poststroke. A comprehensive and targeted clinical assessment tool of WA is warranted for the stroke population.

Post-stroke Community Ambulators Demonstrate Deficits in Backward Walking

K. Hawkins, A. Vistamehr, C. Balasubramanian, C. Conroy, D. Rose, D. Clark, E. Fox.

Purpose/Hypothesis: The ability to walk or step backward is an important component of safe mobility in the home and community. For instance, backward walking (BW) is needed when negotiating tight spaces or opening a large door. Healthy adults walk backward using step parameters and kinematics that closely resemble time-reversed forward walking (FW). After stroke, gait rehabilitation emphasizes recovery of FW; however, BW is rarely practiced or examined. Accordingly, it is not known how adults poststroke who achieve community mobility perform BW. We hypothesize that, despite achievement of full community ambulator status (based on FW speed > 0.93 m/s), adults poststroke will demonstrate significant impairments in BW as evidenced by altered spatiotemporal and kinematic measures.

Number of Subjects: 15 adults with poststroke hemiparesis (9 left hemiparesis; age: 60 ± 12.9 years, 7 females) and 10 healthy controls (age: 23 ± 2.9 years, 4 females).

Materials/Methods: Participants performed FW and BW at a self-selected pace without orthotic or assistive devices while three-dimensional kinematics were recorded. Mixed-model analyses of variance were used to analyze spatiotemporal gait characteristics and average peak lower extremity sagittal plane joint kinematics. Group × task interaction effects were examined to assess differences between groups in the change from FW to BW.

Results: The changes in spatiotemporal gait characteristics between FW and BW were more pronounced for the stroke group than for the control group. Specifically, the stroke group demonstrated significantly greater reductions in speed (P < 0.001), step length (P < 0.001), and cadence (P = 0.011) during BW compared with FW. The stroke group also exhibited a larger increase in the percentage of the gait cycle spent in double-support phase (P < 0.001). Changes in the average peak kinematic motions between FW and BW were also altered. Relative to healthy controls, the stroke group had significantly greater reductions in paretic and nonparetic hip extension (P = 0.011, P = 0.008) during BW versus FW. The control group demonstrated increased plantarflexion and dorsiflexion during BW, but these increases were attenuated in the stroke group, especially in the paretic leg (P < 0.01).

Conclusions: Compared with healthy controls, individuals with poststroke hemiparesis demonstrate impairments in BW as evidenced by alterations in spatiotemporal gait characteristics and limb kinematics relative to their FW performance.

Clinical Relevance: In poststroke community ambulators, BW is more impaired than FW. Therefore, BW should be directly assessed when determining readiness for community ambulation. Deficits in BW and other walking adaptations may be a contributing factor in the difficulties reported by individuals poststroke when navigating the home and community settings.

The Effects of Ischemic Conditioning on Walking Function in Chronic Stroke Survivors: A Pilot Study

T. Boerger, R. Palarz, J. Nguyen, B. Schmit, M. Durand, A. Hyngstrom.

Purpose/Hypothesis: The purpose of this study was to quantify the effects of ischemic conditioning (IC) on walking speed and paretic leg kinetics in individuals with chronic stroke. Therapeutic adjuncts that enhance paretic muscle activation and force generation in order to improve walking are limited. Ischemic conditioning is a widely studied noninvasive stimulus known to improve muscle performance and vascular function in neurotypical populations; however, the effects of ischemic conditioning to improve motor function in stroke survivors are largely unexplored. Ischemic conditioning occurs when the tissue of interest (eg, the paretic leg) is exposed to repeated, transient (∼5 minutes) bouts of ischemia. Ischemic conditioning may optimize paretic muscle activation and force generations by enhancing the cardiovascular and neural response to exercise. In response to IC, arterial endothelial cells release local vasodilatory factors that improve blood flow to the ischemic muscle. In addition, IC stimulates group III/IV muscle afferents sensitive to ischemia, which engages the autonomic nervous system to release neuromodulators known to excite motoneuron pools. Therefore, we hypothesized that 2 weeks of ischemic conditioning would result in increases in paretic leg force generation during walking and increase walking speed.

Number of Subjects: 5.

Materials/Methods: Five chronic stroke survivors (>6 months poststroke, 58.4 ± 13 years, 2 female) received 7 sessions of ischemic conditioning over a 2-week period. For each ischemic conditioning session, a compression cuff (D. E. Hokanson, Inc, Bellevue, WA) was wrapped around the proximal thigh of the paretic leg and inflated to 225 mm Hg for 5 minutes. The cuff was then deflated for 5 minutes and this sequence was repeated 5 times. Using the Plug-in Gait model in Vicon (Oxford Metrics, Oxford, UK), self-selected walking speed and paretic leg anterior-posterior ground reaction forces at toe-off were measured before and after the 2-week treatment period. Statistical analyses were conducted in SPSS version 24 (SPSS, Inc) using paired t tests with ground reaction force and speed as repeated measures (α = 0.05).

Results: Self-selected walking speed was on average 29.44% greater following the intervention (baseline: 0.76 ± 0.24 m/s; posttest: 0.97 ± 0.30 m/s, P < 0.05). Posttest ground reaction force was on average 22.04% greater than the baseline ground reaction force (baseline: 5.17% ± 3.30% body weight; posttest: 6.21% ± 3.92% body weight, P < 0.05).

Conclusions: Following 2 weeks of ischemic conditioning alone, comfortable walking speed and paretic leg push-off forces increased. This indicates that this treatment may result in the generation of greater propulsive forces to increase walking speed in stroke survivors.

Clinical Relevance: Ischemic conditioning may be a useful adjunct to enhance paretic leg force generation and walking function poststroke.

Therapeutic Volunteering to Improve Participation and Physical Activity for Survivors of Brain Injury: A Community-based Initiative

P. Oza, R. Stark.

Purpose: The purpose is to present the effectiveness of a safe therapeutic volunteering program developed in collaboration with city public library for survivors of brain injury.

Description: IV STEP conference emphasized the roles and responsibilities of physical therapists to improve participation (ie, involvement of people in all areas of life). Fear of falling and social stigma related to challenging behaviors serve as barriers to participation for survivors of acquired brain injuries (ABIs), leading to decreased physical activity in the community. A therapeutic volunteering program was developed to address ABI-specific concerns and promote community integration in the safe, inclusive environment of a public library. Volunteer task list was individualized for each participant based on physical therapy evaluations. Ten participants (9 males and 1 female), from 3 cohorts over 1 year, were evaluated by a physical therapist to assess functional abilities, confidence levels, and fall risks before and after the program, which was administered by the Sacramento Public Library (Sacramento, CA). Physical function was assessed by functional muscle strength, coordination, Berg Balance Scale (BBS), and Timed Up and Go (TUG) test. Cognitive function was assessed by Montreal Cognitive Assessment (MoCA). Participants' confidence levels were assessed by the Activities-specific Balance Confidence Scale and Optimal Instrument. Participants were matched to appropriate volunteer tasks based on their level of function, interest, and self-assessment. Tasks were progressed by the supervising librarian based on prior established criteria. Attendance, satisfaction surveys, and feedback from the library staff were program outcome measures.

Participants perceived the volunteering work as active leisure giving them satisfaction of helping the library staff and patrons. The paired t test of pre- and postprogram scores showed no significant changes in the MoCA scores (1.91% change; P = 0.43), BBS (4.15% change; P = 0.26), and TUG (11.86% change; P = 0.42). Subjectively, participants reported improved memory, balance, and confidence via volunteer survey. Despite initial hesitation of supervising volunteers from special population, staff survey showed that library staff participated enthusiastically and were surprised by reduced workload, as participants completed several library tasks.

Summary of Use: Collaboration with Sacramento public library helped develop a therapeutic volunteering program to meet the specific needs of survivors of brain injury. The program improved functional activity, community participation, and quality of life for the participants.

Importance to Members: Therapeutic volunteering programs are important to community integration after ABI. This project provides an example of collaboration between physical therapists and community members to create therapeutic volunteering opportunities for individuals with ABI. Similar programs can be developed by members in collaboration with area public libraries or similar institutions.

The Use of the Balance Evaluation Systems Test for Individuals With Chronic Traumatic Brain Injury

K. Hays, C. Tefertiller, J. Ketchum, C. Eagye, D. O'Dell, A. Natale, A. Weintraub.

Purpose/Hypothesis: To evaluate the use of the Balance Evaluation Systems Test (BESTest) in a cohort of individuals with chronic traumatic brain injury (TBI), including a comparison of a clinician's judgment of the most impaired balance system in contrast to the lowest subscale score designated by the BESTest and associations between the BESTest, Community Balance and Mobility Scale (CB&M), and the Activities-specific Balance Confidence Scale (ABC).

Number of Subjects: Participants included 59 individuals with a TBI who were taking part in a larger randomized controlled trial to improve balance. All information used in the current analysis was collected at baseline at Craig Hospital.

Materials/Methods: Individuals who were at least 1 year post-TBI, 3 months postcompletion of skilled PT treatment, and ambulated independently within the home (with assistive devices and/or orthotics as needed) were eligible. Trained clinicians administered the BESTest and then provided their clinical judgment of the most impaired subscale based solely on observation during the test. The CB&M and the ABC were also administered.

Results: Thirty-eight males and 21 females with a mean age of 48 years completed the testing. Clinical judgment of the most impaired subscale matched the lowest subscale score on testing 58.1% of the time (κ = 0.45, 95% confidence interval = 0.28-0.63). Those identified as most impaired on the sensory orientation subscale trended toward lower scores on the CB&M, although this was not statistically significant (P = 0.18). Mean scores on the BESTest were significantly associated with stratification of CB&M scores (P < 0.0001), with higher BESTest scores associated with higher category of the CB&M. There was a significant weak to moderate positive correlation between the ABC and BESTest (r = 0.471, P = 0.0002) and between the ABC and CB&M (r = 0.407, P = 0.0015).

Conclusions: This research provides initial data to support the use of the BESTest to measure balance impairment in individuals with chronic TBI. Further investigation is warranted to validate this measure in chronic TBI. Clinical judgment of the most impaired balance subscale on the BESTest matched the actual outcome just over 50% of the time. The ABC demonstrated a significant positive relationship to the CB&M and the BEST. However, given that the correlation was weak to moderate between the self-report measure and the clinical measures, there is likely important information gained from the clinical measures that would be missed by only using a self-report measure.

Clinical Relevance: The BESTest may have relevance in the TBI population. However, clinicians' judgment of area of most severe impairment does not always match the scores of this assessment. There is a correlation between the self-report measure of balance and the clinical assessments, although participants tend to rate themselves higher than the clinical balance scores indicate.

Rehabilitation and Sensory Reweighting in Patients With Chronic MTBI

L. King, L. Parrington, P. Fino, R. Peterka, J. Chesnutt, J. Wilhelm, J. Schlimgen.

Purpose/Hypothesis: Abnormal use of vestibular and visual systems may be linked to postural instability in patients with mild traumatic brain injury (mTBI). The aim of this study is to assess the effects of rehabilitation on sensory contributions to balance in chronic mTBI patients. Here, we present preliminary data from a novel test of central sensorimotor integration (CSMI).

Number of Subjects: 13.

Materials/Methods: Six patients (age = 36 ± 8 years) with chronic mTBI and 7 controls (age = 25 ± 6 years) with no recent history of mTBI were recruited. Patients with mTBI were tested at baseline and then again after a 6-week (2 times/week) balance rehabilitation program. The dizziness handicap inventory (DHI) was used to assess changes in symptoms. Controls were tested 6 weeks apart to assess CSMI test-retest reliability. The CSMI test uses pseudorandom rotations of the surface and/or visual display to elicit anterior-posterior sway. Here, we report on 2 of the relevant visual and vestibular conditions: C1 (2° surface rotations—eyes closed), and C2 (2° surface rotations—eyes open). The CSMI analysis yields measures of sensory weights (W, the relative contribution of vestibular information to balance in C1 and the vestibular + visual contribution to balance in C2), and measures of sensorimotor stiffness (Kp) and damping (Kd) parameters, indicating the corrective ankle torque generated in proportion to the body sway angle and sway angular velocity, respectively. Consistency of the CSMI test was assessed using intraclass correlation coefficients (ICCs). Changes in the mTBI group were assessed using effect sizes. Reported values for W range between 0 and 1, while Kp and Kd were normalized by each person's height and weight.

Results: Good to excellent consistency was found for each measure; C1: W, ICC = 0.92, CI [0.42, 0.99]; Kp, ICC = 0.88, CI [0.12, 0.98]; Kd, ICC = 0.97, CI [0.80, 0.99]; C2: W, ICC = 0.80, CI [0.14, 0.97]; Kp, ICC = 0.93, CI [0.59, 0.99]; Kd, ICC = 0.93, CI [0.60, 0.99]. Following rehabilitation, W increased in both conditions toward control values (C1: pre = 0.41 ± 0.05, post = 0.47 ± 0.08, d = 0.92; C2: pre = 0.53 ± 0.04, post = 0.60 ± 0.05, d = 1.64). Kp increased in C1 (pre = 1.58 ± 0.23, post = 1.66 ± 0.25, d = 0.34) but decreased in C2 (pre = 1.94 ± 0.41, post = 1.77 ± 0.28, d = −0.52), while for Kd, effects indicated an increase in Kd for C1 and C2, respectively (C1: pre = 0.48 ± 0.06, post = 0.55 ± 0.03, d = 1.50; C2: pre = 0.50 ± 0.06, post = 0.52 ± 0.08, d = 0.31). DHI improved in 5/6 patients, with an overall decrease in symptoms (pre = 43 ± 17, post = 36 ± 23, d = −0.36).

Conclusions: CSMI test-retest results were highly consistent, suggesting it may be an effective way to identify specific abnormalities and track changes in balance control. Increased W in the mTBI group toward control values suggests that rehabilitation changed reliance on sensory systems for balance. Differences in Kp and Kd following rehabilitation indicate adaptation in motor response to the stimuli. Decreased DHI suggests improvement in dizziness symptoms after rehabilitation.

Clinical Relevance: Understanding how sensory and sensorimotor contributions to balance change with rehabilitation may help drive evidence-based programs.

© 2018 Academy of Neurologic Physical Therapy, APTA.