“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Note. This article will be published in a forthcoming issue of the

Motor Control. The article appears here in its accepted, peer-reviewed

form, as it was provided by the submitting author. It has not been

copyedited, proofread, or formatted by the publisher.

Article Title: Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke

Survivors

Authors: Cliff S. Klein1,2

, Geoffrey A. Power3, Dina Brooks

1,2 and Charles L. Rice

3,4

Affiliations: 1Department of Physical Therapy, University of Toronto, Toronto, ON, Canada;

2Toronto Rehabilitation Institute, Toronto, ON, Canada;

3Canadian Centre for Activity and

Aging, School of Kinesiology, Faculty of Health Sciences, The University of Western Ontario,

London, ON, Canada; 4

Department of Anatomy and Cell Biology, Schulich School of Medicine

and Dentistry, The University of Western Ontario, London, ON, Canada.

Journal: Motor Control

Acceptance Date: March 15, 2013

©2013 Human Kinetics, Inc.

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Cliff S. Klein1,2

, Geoffrey A. Power3, Dina Brooks

1,2 and

Charles L. Rice

3,4

1Department of Physical Therapy, University of Toronto, Toronto, ON, Canada;

2Toronto

Rehabilitation Institute, Toronto, ON, Canada; 3 Canadian Centre for Activity and Aging, School

of Kinesiology, Faculty of Health Sciences, The University of Western Ontario, London, ON,

Canada; 4

Department of Anatomy and Cell Biology, Schulich School of Medicine and Dentistry,

The University of Western Ontario, London, ON, Canada

Abbreviated title: Post-stroke dorsiflexor weakness

Date of submission: March 8, 2013

Address correspondence and submit galley proofs to:

Cliff S. Klein PhD

Rehabilitation Institute of Chicago, 345 E. Superior St., Chicago, IL, USA, 60611

Phone:312-238-2695, cklein@ric.org

Geoffrey A. Power, School of Kinesiology, University of Western Ontario,

411D Arthur & Sonia Labatt Health Sciences Bldg, London ON, N6A 5B9

Phone: 519-661-2111 ext. 81628, geoffpowernl@gmail.com

Dina Brooks PhD, Dept of Physical Therapy, University of Toronto, 160-500 University Ave.,

Toronto, Ontario, M5G 1V7, Phone: 416-978-1739, dina.brooks@utoronto.ca

Charles L. Rice PhD, School of Kinesiology, University of Western Ontario,

411D Arthur & Sonia Labatt Health Sciences Bldg, London ON, N6A 5B9

Phone: 519-661-2111 ext. 81628, crice@uwo.ca

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Neural and muscular determinants of dorsiflexor weakness in chronic stroke survivors

Cliff S. Klein1,2

, Geoffrey A. Power3, Dina Brooks

1,2 and Charles L. Rice

3,4

1Department of Physical Therapy, University of Toronto, Toronto, ON, Canada;

2Toronto

Rehabilitation Institute, Toronto, ON, Canada; 3Canadian Centre for Activity and Aging, School

of Kinesiology, Faculty of Health Sciences, The University of Western Ontario, London, ON,

Canada; 4

Department of Anatomy and Cell Biology, Schulich School of Medicine and Dentistry,

The University of Western Ontario, London, ON, Canada

Abbreviated title: Post-stroke dorsiflexor weakness

Address correspondence to:

Cliff S. Klein PhD

Rehabilitation Institute of Chicago

345 E. Superior St.,

Chicago, IL, USA, 60611

312-238-2695

cklein@ric.org

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Abstract:

Few examined the contribution of neural and muscular deficits to weakness in the same stroke

subject. We determined maximal voluntary contraction (MVC) and 50 Hz torques, activation

(twitch interpolation), electromyographic (EMG) amplitude and antagonist coactivation, and

muscle volume using magnetic resonance imaging (MRI) of the dorsiflexors bilaterally in 7

chronic stroke subjects (40-67 y). Recordings of MVC and 50 Hz torque were also done in 7

control subjects (24-69 y) without stroke. The MVC torque was smaller in the contralesional

than ipsilesilesional limb (29.8±21.3 Nm vs. 42.5±12.0 Nm, P =0.04), and was associated with

deficits in activation (r2

=0.77) and EMG amplitude (r2

=0.71). Antagonist coactivation

percentage was not significantly different between limbs. Muscle volume, 50 Hz torque, and

specific torque (50Hz torque/muscle volume) were also not different between sides. The concept

that atrophy is commonplace after stroke is not supported by the results. Our findings indicate

that dorsiflexor weakness in mobile stroke survivors is not explained by atrophy or reduced

torque generating capacity suggesting an important role for central factors.

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Chronic stroke survivors (>1 y post-stroke) have varying degrees of weakness and

impaired motor control that compromises daily function (Arene & Hidler, 2009). During walking

the dorsiflexors help to stabilize the ankle during stance and optimize foot-ground clearance

during swing (Dubo et al., 1976). These dorsiflexion functions may be impaired by weakness

and thereby contribute to the slow gait and falls following stroke (Knutsson & Richards, 1979).

Weakness, defined as the inability to generate normal levels of torque during a maximal

voluntary contraction (MVC), predominates in the contralesional (paretic) limb after hemiparetic

stroke (Adams, Gandevia, & Skuse, 1990; Frontera, Grimby, & Larsson, 1997; Horstman et al.,

2008; Klein, Brooks, Richardson, McIlroy, & Bayley, 2010; Knorr, Ivanova, Doherty, Campbell,

& Garland, 2011; Landau & Sahrmann, 2002; Levin & Hui-Chan, 1994). Whether the loss of

strength arises solely from neural impairments or also reflects altered intrinsic muscle properties

is uncertain (Frontera, et al., 1997; Jakobsson, Edstrom, Grimby, & Thornell, 1991; Landau &

Sahrmann, 2002; Odajima, Ishiai, Okiyama, Furukawa, & Tsukagoshi, 1987; Ramsay, Barrance,

Buchanan, & Higginson, 2011; Scelsi, Lotta, Lommi, Poggi, & Marchetti, 1984; Slager, Hsu, &

Jordan, 1985).

Impaired muscle activation likely contributes to post-stroke weakness (Frontera, et al.,

1997; Horstman, et al., 2008; Klein, et al., 2010; Knorr, et al., 2011; Newham & Hsiao, 2001).

For instance, during a dorsiflexor MVC, the surface electromyographic (EMG) amplitude or

motor unit discharge rates of the tibialis anterior (TA) are often reduced in the contralesional

limb (Frontera, et al., 1997; Knorr, et al., 2011). In addition, activity of the antagonist muscles

during a MVC may reduce MVC torque. Antagonist coactivation in stroke survivors has been

shown to be excessive during an MVC in some cases (Klein, et al., 2010; Levin & Hui-Chan,

1994; Yusevich, 1968), although other reports show that coactivation is similar between the

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

contralesional and ipsilesional limb (Clark, Condliffe, & Patten, 2006; Lum, Patten, Kothari, &

Yap, 2004; Newham & Hsiao, 2001; Wagner, Dromerick, Sahrmann, & Lang, 2007).

The contribution of post-stroke changes in muscle properties to dorsiflexor weakness is

uncertain (Frontera, et al., 1997; Jakobsson, et al., 1991; Landau & Sahrmann, 2002; Odajima, et

al., 1987; Slager, et al., 1985). The study by Landau and Sahrmann (2002) compared the side-to-

side differences in dorsiflexor MVC force and maximal force generating capacity evoked by 50

Hz stimulation over the TA. In the 14 subjects with chronic stroke, the mean MVC force was

significantly less in the contralesional than ipsilesional limb, but mean 50 Hz forces were not

different. However, contralesional 50 Hz force was 20-70% smaller than the ipsilesional value in

four persons, and was unexpectedly larger by 30-150% in five others, but the cause of these

differences was not evaluated (Landau & Sahrmann, 2002). A reduction in 50 Hz force may

result from muscle atrophy or reduced specific tension (ie., torque per unit area or volume), or

both, whereas an increase could indicate opposite responses. Post-stroke dorisflexor (Odajima, et

al., 1987; Ramsay, et al., 2011) or TA muscle fiber (Frontera, et al., 1997; Jakobsson, et al., 1991;

Scelsi, et al., 1984; Slager, et al., 1985) morphology has been determined, but the findings are

not always in agreement. One study reported smaller dorsiflexor cross-sectional area (CSA) in

the contralesional than ipsilesional limb based on computer tomography images (Odajima, et al.,

1987), but another found no differences in TA volume determined by MRI (Ramsay, et al.,

2011). At the fiber level, Scelsi and coworkers (1984) reported progressively smaller type II

fiber diameter in the contralesional TA with increased time (1-17 months) since stroke, but

Jakobsson et al., (1991) found that fiber CSA was not different between chronic stroke subjects

and healthy controls.

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Although central neural activation, muscle size, and maximal muscle force generating

capacity are primary determinants of voluntary strength, all three determinants have not been

assessed concurrently in the same stroke subjects; and thus the purpose of this study. More

specifically, we addressed whether weakness arises solely from neural impairments or also

results from atrophy and reduced maximal force generating capacity. In addition to this

comprehensive assessment of the determinants of weakness, the findings should provide some

insight into the conflicting results noted above pertaining to dorsiflexor (TA) size and force

generating capacity. We hypothesized that most of the contralesional dorsiflexor weakness

would reflect reduced activation, but that intrinsic changes in muscle properties also would

contribute in some cases.

METHODS

Participants

Five men and two women (56.9 ± 9.0 years, mean±SD) who had a stroke at least 2 years

before the study were recruited (Table 1). These individuals participated in a previous study of

plantar flexion weakness (Klein, et al., 2010), and showed significant limb differences in

strength, activation, and muscle volume. Therefore the same subjects were tested in the present

study providing a sufficient sample size for these comparisons. Motor impairment of the leg

and foot (Gowland et al., 1993) ranged from mild to moderate (Table 1). Three subjects (no. 4-6)

walked with a more pronounced reduction in knee flexion than the others. Two subjects (no. 4

and 7) used a single point cane when walking, and two others (no. 3 and 6) wore an ankle

orthotic. All participants could walk without assistance, but their gait speed varied threefold, and

the average speed (0.83±0.33 m/s) was about 30% slower than healthy adults (Klein, et al., 2010).

Four subjects (subjects 1, 3, 5, and 6) engaged in 20-30 min of aerobic exercise about two times

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

per week at a fitness club over the previous year. We also determined bilateral MVC and evoked

torque in 7 men with no history of neurological impairment (38±16 years, range, 24-69;

175.2±14.7 cm; 82±13 kg). Informed written and verbal consent were obtained in all subjects

and the study was approved by the local research ethics committees of our institution.

Experimental set-up

Only one leg was assessed during each of the two visits to the laboratory in order to

minimize general fatigue of the participants. The testing order of the contralesional and

ipsilesional limb, and right-left side of the controls, was randomized. Subjects were seated with

their leg positioned in a custom torque measuring device (Klein, et al., 2010). The knee and hip

joints were flexed to 90°, and a clamp stabilized the knee from above. The foot rested on a plate

and was secured with straps across the instep and proximal to the great toe. The ankle was

plantar flexed 30° from the neutral (0°) position. At this angle, the peronei are shortened and are

therefore less effective in counteracting the evoked torque of the dorsiflexors. Also, 30° is near

optimal for MVC torque (Marsh, Sale, McComas, & Quinlan, 1981) and optimal for 50 Hz

torque (van Schaik, Hicks, & McCartney, 1994).

The compound muscle action potential (M-wave) and MVC EMG were recorded with

disposable silver-silver chloride disc electrodes positioned in a bipolar configuration (Bortec

Biomedical Ltd., Calgary, Alberta). The recording electrodes were placed over the distal third of

lateral gastrocnemius (LG) medial gastrocnemius (MG), and midline of the soleus (2 cm below

the distal end of the MG) using a 2 cm center-to-center spacing. For the TA, the distal electrode

was placed at the border of the muscle belly and tendon, and the other electrode was placed 4-5

cm proximally.

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

The EMG was amplified (x 500) and filtered between 30 Hz and 1 kHz (Astro-Medical,

Model P511, West Warwick, RI, USA). Torque was sensed by a transducer beneath the ankle

joint (LCDA-500, Omega Engineering Inc, Stamford, CT, USA), and amplified (x 400) and

filtered (DC-100 Hz). The EMG and torque signals were digitally converted by a 12-bit

converter (1401 Plus, Cambridge Electronic Design, Cambridge, UK) at sampling rates of 2000

Hz and 500 Hz, respectively, and analyzed with Spike 2 software (Spike 2, version 7.03,

Cambridge Electronic Design).

In order to find the optimal site for stimulation of the common fibular nerve, a bar

electrode (1 cm diameter discs, 3 cm spacing; Medtronic, Skovlunde, Denmark) was placed

immediately inferior to the fibular head and manually adjusted until M-waves were evoked

during submaximal stimuli. For all subsequent recordings, 3 cm diameter round carbon rubber

electrodes (Empi, St. Paul, Minnesota, USA), coated in conductive gel, were positioned at the

optimal site and secured with tape. These electrodes minimized subject discomfort during 50 Hz

stimulation.

Protocol. Single pulses (50 µs) of progressively greater current intensity (5 mA

increments) were applied by a constant current stimulator ( Model DS7AH, Digitimer) every 5 s

until the TA M-wave peak-to-peak amplitude did not increase further (ie., Mmax). Subjects

generated a few practice voluntary contractions of the dorsiflexors at ~ 50% effort. Three to four

MVCs, each lasting about 4-5 s followed by a two min rest, were recorded. Strong

encouragement was provided during the MVC and a display of the on-going torque trace was

provided. Before and after each MVC, single supramaximal (120 % of Mmax current) pulses

were applied so that percentage activation could be determined (Belanger & McComas, 1981).

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

The stimuli were triggered manually by the investigator about 2-4 s after the beginning of the

MVC, when there was a plateau in the torque trace, and also about 1 s after the MVC finished.

Three minutes after recording the MVCs, two practice 50 Hz contractions (each 0.5 s)

were evoked at progressively greater submaximal current intensities. A 50 Hz train (0.5 s) was

then evoked using the supramaximal current. Supramaximal stimulation of the dorsiflexors at 50

Hz is well tolerated and the evoked response reflects the maximal torque that can be generated

by the dorsiflexor muscles (independent of subject volition) (Marsh, et al., 1981). Inclusion of

this measure together with the assessment of muscle volume allowed us to determine whether

changes in muscle properties contribute to the loss of MVC strength. Specifically, smaller

contralesional than ipsilesional 50 Hz torque would indicate a loss in torque generating capacity,

either because of atrophy or reduced specific tension (force per unit area) of the muscle fibers, or

both. Specific torque (50 Hz torque/muscle volume) was calculated in the present study as an

indirect measure of specific tension. After recording the 50 Hz torque, subjects completed three

to four plantar flexor MVCs, each lasting 4-5 s.

MRI. The participants were placed in a supine position in a 3T whole body scanner with

the shank in a quad head coil (Signa Excite HDx, General Electric Company). The knee was

extended, the leg supported with padding, and the ankle was stabilized in ~20 ° of plantar flexion.

Images were acquired from each leg separately using serial T1-weighted spin-echo sequences.

Initially, scout images of the leg were acquired in the coronal and sagittal planes. After the scout

images, about 40-50 axial slices was aligned perpendicular to the tibial shaft in the coronal and

sagittal planes between the knee and ankle joints. The axial images were acquired in an

interleaved manner using the following parameters; echo time 18 ms, repetition time 850 ms,

matrix 256 x 256, field of view 200 mm, slice thickness 10 mm, interslice gap 0 mm, number of

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

acquisitions 2. Our results in the stroke subjects were compared with the ample age-matched

control data reported for this muscle group from our lab (McNeil, Vandervoort, & Rice, 2007)

and others (Kent-Braun & Ng, 1999).

Data analysis

MVC. The peak MVC torque was determined for each trial. The TA root-mean-square

EMG over a 500 ms period was determined just prior to the interpolated twitch, and was

normalized to the Mmax (ie. RMS EMG/peak-to-peak amplitude of the M-wave). The

interpolated torque was equal to the torque at the onset of the stimulus minus the evoked peak

torque. Activation percentage was equal to the following: [1- (Ts/Tr)] x 100 %, where Ts was the

interpolated torque evoked by the stimulus during the MVC, and Tr was the torque of twitch after

the MVC.

Antagonist coactivation percentage. The root-mean-square EMG of the soleus, LG, and

MG over 500 ms of the dorsiflexor MVC was expressed as a percentage of the muscle’s

maximum EMG recorded during the plantar flexor MVC.

Dorsiflexor volume. The MR images were converted to TIFF files from their original

DICOM format. All images were analyzed by one investigator using available software (ImageJ,

ver 1.38, National Institutes of Health, Bethesda, MD). The border of the dorsiflexor muscle

group that included the tibialis anterior, extensor hallucis longus, extensor digitorum longus, and

peroneus tertius, was manually traced for all axial slices from muscle origin to insertion and the

corresponding CSAs determined. Intramuscular fat in each image was subtracted from the

muscle CSA with a user-defined threshold. Dorsiflexor volume was calculated as the sum of fat-

free CSA area times slice thickness for all slices of the muscle group. Specific torque was

calculated as maximal 50 Hz torque divided by dorsiflexor volume.

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Statistics. Paired and unpaired t-tests were used to compare side-to-side differences

within a group and between groups, respectively, for all measures except voluntary activation

percentage, for which the Mann-Whitney U-test was used. The relationship between measures

was determined with Pearson’s product-moment correlation coefficients. Data are presented as

means±SD, and differences were considered significant when P<0.05.

RESULTS

MVC torque. Representative MVC recordings in two stroke subjects are shown in Fig. 1.

In one, peak MVC torque was similar in both legs, although contralesional torque declined

slightly toward the end of the contraction (Fig. 1A, uppermost two traces). Also, there was no

obvious interpolated twitch in either leg. In the other subject, peak contralesional MVC torque

was 30% of the ipsilesional side (Fig.1B). Reduced contralesional activation was apparent,

shown by the interpolated twitch and sparse TA EMG. In six of the seven subjects, the MVC

torque was less in the contralesional than ipsilesional limb, and the mean values were different

(P =0.04, Fig. 2, Table 2).

MVC EMG amplitude and activation percentage. The TA Mmax was not significantly

different between the contralesional and ipsilesional limb (P =0.70, Table 2). The MVC root-

mean-square EMG, normalized to the Mmax, was less in the contralesional than ipsilesional limb

in all subjects, and the mean values were different (P =0.01, Fig. 2, Table 2). Activation

percentage via twitch interpolation was less in in the contralesional than ipsilesional limb, but the

mean values were not different (P =0.2, Table 2). Those with greater weakness tended to have

larger activation deficits. Hence, relative contralesional MVC (% ipsilesional side) was

associated with relative EMG (r2=0.71, P=0.01) and activation percentage (r

2=0.77, P=0.01).

Because subject 7 could not generate any voluntary dorsiflexor torque, a value of zero was

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

assigned for each of his contralesional MVC torque, EMG amplitude, and activation scores and

these zero values were included in the means presented in table 2. The reported statistically

significant differences between the contralesional and ipsilesional means for MVC torque, EMG

amplitude, and percent activation were unchanged when data were reanalyzed with subject 7

excluded (N=6); 34.8±18.3 vs. 43.3±13 vs. Nm (P <0.05), 0.050±0.022 vs. 0.120±0.046 (P

<0.05), and 91.1±17.2 vs. 98.9±1.0 (P= 0.30), respectively.

Antagonist coactivation percentage. The mean antagonist EMG amplitude (mV) was

less in the contralesional than the ipsilesional leg in the soleus (0.015±0.005 vs. 0.27±0.004,

P=0.03) and gastrocnemii (0.013±0.012 vs. 0.023±0.016, P=0.005). The soleus coactivation

percentages ranged from 10.6%-110.7% and 11.4%-21.7% in the contralesional and ipsilesional

limb, respectively. The corresponding values for the gastrocnemii were 6.8%-70.3% and 4.3%-

26.4%. The mean soleus and gastrocnemii coactivation percentages were larger in the

contralesional than ipsilesional limb, but the differences were not significant (P= 0.2, Table 2).

The larger contralesional means were mainly due to the values recorded in subject 6 (110% and

70% in the soleus and gastrocnemii, respectively). The augmented coactivation percentage in this

subject reflected a larger absolute EMG coactivation (mV) in the gastrocnemii, but not the soleus,

combined with a dramatically smaller absolute MVC EMG of both muscles; 12% and 5% of the

ipsilesional side, respectively. Similar to the findings of others (Clark, et al., 2006; Lum, et al.,

2004; Newham & Hsiao, 2001; Wagner, et al., 2007), our results reveal little evidence of

excessive coactivation in the contralesional limb.

50 Hz torque, muscle volume, and specific torque. The 50 Hz torque recordings of two

subjects are shown in Fig 3 and demonstrate that the evoked torques were similar between sides.

In all subjects except one (no. 4), contralesional 50Hz torque was equal to or larger than the

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

ipsilesional value, and the means were not different (P =0.3, Table 2). The lack of a difference in

mean 50 Hz torque generating capacity between limbs supports the findings of a previous study

(Landau & Sahrmann, 2002), but, unlike their large side-to-side 50 Hz differences in individual

subjects (up to 150%), differences in the present subjects were ≤ 25% (see discussion). The total

dorsiflexor volume (cm3), fat volume (cm

3), and percentage of total volume that was fat, was not

significantly different between the contralesional and ipsilesional limbs; 264.8±69.0, 4.6±2.3,

1.9±1.1% vs 255.5±81.2, 5.6±2.5, 2.5±1.7% (P=0.4, P=0.2, P=0.1, respectively). Contralesional

fat-free volume (and CSA) was similar to or larger than ipsilesional volume across subjects (Fig.

2), and the means were not different (P =0.4, Table 2). The differences in the 50 Hz

torque/volume ratio between limbs were not different (P =0.9, Table 2).

Comparisons to control subjects. The MVC torque (P=0.04) and MVC EMG (P=0.003)

were significantly less in the contralesional limb compared to the right limb of the 7 controls, but

TA Mmax, activation, coactivation, and 50 Hz torque did not differ significantly (Table 2). There

were no significant differences between ipsilesional and control means for any parameter (Table

2). Right-left differences in mean TA Mmax, control MVC torque, MVC EMG, and 50 Hz

torque were modest (<16%) and not significant (P >0.05).

DISCUSSION

In this study, we assessed both central and peripheral determinants of weakness in

chronic stroke subjects. Our findings indicate that most of the weakness reflects a reduction in

central neural drive to the motoneuron pool, with little contribution from antagonist muscle

activity. The idea among some clinicians that atrophy is commonplace after stroke is not

supported by the present results.

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Because muscles receive some of their innervation from the ipsilateral motor cortex, the

ipsilesional side may show impairments in activation and weakness following stroke (Adams, et

al., 1990; Newham & Hsiao, 2001). However, in the present study, MVC and 50 Hz torques did

not differ between the ipsilesional and control limbs (Table 2). Although the controls were

younger on average than the stroke subjects, previous work showed little age-related differences

in dorsiflexor properties between 30 and 70 y of age. Hence, mean ipsilesional MVC, 50 Hz

torque, and dorsiflexor CSA of the five men with stroke (47 Nm, 34 Nm, and 14.2 cm2) are

similar to the means reported for older (65 y) healthy men (44 Nm, 30 Nm and 14.3 cm2, 65y)

(Kent-Braun & Ng, 1999; McNeil, et al., 2007; van Schaik, et al., 1994). The preserved

ipsilesional dorsiflexor motor function may reflect overuse of this limb to compensate for

contralesional impairments (Olney, Griffin, Monga, & McBride, 1991).

Contralesional weakness was evident in six of the seven subjects and was associated with

deficits in EMG amplitude and activation percentage. The recorded weakness, particularly

evident in subjects 4, 6, and 7, may be partially attributed to a diminished capacity to recruit and

fully activate the dorsiflexor motoneuron pool. Indeed, one study of stroke subjects found that

the motor unit firing rates of the contralesional TA did not rise above 15 Hz during an MVC

(Frontera, et al., 1997). This level of motor unit firing is less than half of the normal rate and

would be expected to generate smaller peak forces. Others have reported contralesional

dorsiflexor weakness (Adams, et al., 1990; Eng, Lomaglio, & Macintyre, 2009; Landau &

Sahrmann, 2002) and reduced EMG amplitude (Frontera, et al., 1997; Jakobsson, et al., 1991;

Knorr, et al., 2011), but the contribution of changes in whole muscle morphology was not

evaluated.

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Our results indicate that reduced central neural drive is the primary determinant of

contralesional dorsiflexor weakness in these subjects because of a lack of atrophy or any

contractile impairment combined with their deficits in activation and EMG amplitude. However,

the percent contribution of reduced neural drive to weakness may be underestimated when based

on twitch interpolation data and overestimated when based on the EMG amplitude results (Fig.

2) (Klein, et al., 2010). Although twitch interpolation is considered to be one of the better

methods to assess voluntary activation, it is not without limitations. The sensitivity of the

method, or the ability to detect small deficits in dorsiflexor activation, is reduced at higher (> 80

% MVC) torques, possibly because the fibular (peroneal) muscles can act as antagonists when

activated during electrical stimulation (Belanger & McComas, 1981). In healthy adults, the

relationship between relative (%MVC) EMG amplitude and dorsiflexor torque is curvilinear

(Garland, Garner, & McComas, 1988). Hence, at a contraction strength equal to 50% MVC

torque, the corresponding EMG is 30% of the MVC EMG. This difference between relative

EMG and relative torque may widen after stroke because of a reduction in maximal motor unit

firing rate in the contralesional limb (Frontera, et al., 1997), and could explain the relatively

larger deficits in contralesional EMG than MVC torque (Fig. 2).

Dorsiflexor atrophy was not evident in the present subjects and supports the concept of

little to no loss in dorsiflexor mass after stroke. Our results confirm previous studies of chronic

stroke subjects that found no evidence of TA atrophy based on the determination of muscle

volume derived from magnetic resonance images (Ramsay, et al., 2011) and fiber area of

biopsies (Jakobsson, et al., 1991). In one study, dorsiflexor CSA, derived from computer

tomography images, was about 12% smaller in the contralesional than the ipsilesional limb

(Odajima, et al., 1987). Another reported evidence of contralesional TA fiber atrophy in some

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

stroke subjects, but they may have been less mobile than the present stroke subjects since they

were to undergo corrective foot surgery (split anterior tibial tendon transfer) to improve

ambulation (Slager, et al., 1985).

The lack of atrophy and contractile impairment in the present subjects is evident despite

deficits in EMG and activation (ie. no. 4 and 7). Subject 7 was unable to generate any EMG or

torque during the MVC, and his TA activity during gait was also found to be negligible.

Immediately after the torque recordings of the contralesional and ipsilesional limbs, we had

subject 7 walk back and forth a distance of 5 m in the laboratory at his preferred speed while

recording the associated EMG activity using the same electrodes (data not shown). During this 5

m walk, his contralesional mean TA EMG amplitude at heel strike was found to be about 5% of

the ipsilesional TA MVC EMG amplitude; much less than typically recorded in healthy controls

(Jakobsson, et al., 1991). Our whole muscle results support and extend the findings of a previous

study in which there was no significant contralesional TA fiber atrophy, even though the TA

EMG amplitude during walking was less than 10% of adult controls (Jakobsson, et al., 1991).

Together, these observations suggest that recruitment of a large portion of the dorsiflexors is not

a necessary precondition for maintenance of muscle mass and force generating capacity

following stroke. Rather, factors independent of muscle activation, including passive mechanical

tension or neurotrophic support may play important roles (McComas, Sica, Upton, & Aguilera,

1973; Yucesoy & Huijing, 2007).

In subject 6, hypertrophy was apparent in the contralesional dorsiflexors (+28%) and the

peronei (+42%) but with atrophy of the gastrocnemii (-50%) (Klein, et al., 2010), indicating that

responses may be muscle-dependent in some cases (Odajima, et al., 1987; Ploutz-Snyder, Clark,

Logan &Turk 2006; Prado-Medeiros et al., 2012). Scattered fiber hypertrophy has been noted in

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

the TA and other muscles after stroke (Jakobsson, et al., 1991; Slager, et al., 1985). Interesting,

in some individuals with anterior cruciate ligament injury, TA volume was also unexpectedly

larger in the injured limb compared to the non-injured limb (Binder-Macleod & Buchanan, 2006).

Like these types of injuries, peculiar changes in gait mechanics following stroke (Perry, Waters,

& Perrin, 1978) may lead to stretch-induced hypertrophy of some muscles and disuse atrophy of

others. In subject 6, the equinovarus foot position during gait may have resulted in stretch-

induced hypertrophy of the TA and peronei.

We explored intrinsic contractile function by comparing side-to-side 50 Hz torque

differences using maximal stimulation of the common fibular nerve and which could be related

to muscle quantity. In one previous study of chronic stroke, contralesional 50Hz torque, tested

by using pad stimulation over only the TA, was found to differ widely (~20% to 150% of the

ipsilesional value in one-half of the subjects), which was not easy to explain (Landau &

Sahrmann, 2002). Compared to their data (Landau & Sahrmann, 2002) we found that the side-to-

side 50 Hz torque differences were more modest in the stroke subjects (≤ 25%) and controls (≤

9%). The design to isolate the current field to the TA in the prior study (Landau & Sahrmann,

2002) may not have accounted for variable activation spread to other local dorsiflexors and

plantar flexors (peronei). Supramaxmial nerve stimulation, as applied in the present study, is

likely a more precise method for assessing limb differences in torque generating capacity

because all dorsiflexor and peronei muscles are maximally activated.

Study limitations. With a limited sample size and inclusion of only persons with chronic

neurological impairments who are mobile and physically active, these findings are not

necessarily representative of the wider stroke population. Dorsiflexor atrophy and weakness

would be expected to be greater in those who are less mobile and inactive (Eng, et al., 2009;

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Odajima, et al., 1987). In addition, weakness and antagonist coactivation may be more

pronounced when measured during dynamic contractions as opposed to isometric conditions

employed here (Lum, et al., 2004).

Conclusions. Despite weakness and reduced EMG amplitude, muscle volume, 50 Hz

torque, and specific torque were well preserved in the stroke subjects. Our results suggest that

much of the weakness can be attributed to a loss of central neural drive to the agonist

motoneuron pool. The preservation of contralesional dorsiflexor muscle properties may be

explained in part by their relatively preserved mobility and regular physical activity.

Acknowledgements

The authors thank the subjects who volunteered for this study and the Medical Imaging staff at

Toronto Western Hospital for their technical support. This research was supported by the Dean’s

Fund, Faculty of Medicine, University of Toronto.

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

References

Adams, R. W., Gandevia, S. C., & Skuse, N. F. (1990). The distribution of muscle weakness in

upper motoneuron lesions affecting the lower limb. Brain, 113 ( Pt 5), 1459-1476.

http://www.ncbi.nlm.nih.gov/pubmed/2245306

Arene, N., & Hidler, J. (2009). Understanding motor impairment in the paretic lower limb after a

stroke: a review of the literature. Top Stroke Rehabil, 16(5), 346-356.

http://www.ncbi.nlm.nih.gov/pubmed/19903653

Belanger, A. Y., & McComas, A. J. (1981). Extent of motor unit activation during effort. J Appl

Physiol, 51(5), 1131-1135. http://www.ncbi.nlm.nih.gov/pubmed/7298453

Binder-Macleod, B. I., & Buchanan, T. S. (2006). Tibialis anterior volumes and areas in ACL-

injured limbs compared with unimpaired. Med Sci Sports Exerc, 38(9), 1553-1557.

http://www.ncbi.nlm.nih.gov/pubmed/16960514

Clark, D. J., Condliffe, E. G., & Patten, C. (2006). Activation impairment alters muscle torque-

velocity in the knee extensors of persons with post-stroke hemiparesis. Clin Neurophysiol,

117(10), 2328-2337. http://www.ncbi.nlm.nih.gov/pubmed/16926111

Dubo, H. I., Peat, M., Winter, D. A., Quanbury, A. O., Hobson, D. A., Steinke, T., & Reimer, G.

(1976). Electromyographic temporal analysis of gait: normal human locomotion. Arch

Phys Med Rehabil, 57(9), 415-420. http://www.ncbi.nlm.nih.gov/pubmed/962568

Eng, J. J., Lomaglio, M. J., & Macintyre, D. L. (2009). Muscle torque preservation and physical

activity in individuals with stroke. Med Sci Sports Exerc, 41(7), 1353-1360.

http://www.ncbi.nlm.nih.gov/pubmed/19516167

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Frontera, W. R., Grimby, L., & Larsson, L. (1997). Firing rate of the lower motoneuron and

contractile properties of its muscle fibers after upper motoneuron lesion in man. Muscle

Nerve, 20(8), 938-947. http://www.ncbi.nlm.nih.gov/pubmed/9236783

Garland, S. J., Garner, S. H., & McComas, A. J. (1988). Reduced voluntary electromyographic

activity after fatiguing stimulation of human muscle. J Physiol, 401, 547-556.

http://www.ncbi.nlm.nih.gov/pubmed/3171996

Gowland, C., Stratford, P., Ward, M., Moreland, J., Torresin, W., Van Hullenaar, S., . . . Plews,

N. (1993). Measuring physical impairment and disability with the Chedoke-McMaster

Stroke Assessment. Stroke, 24(1), 58-63. http://www.ncbi.nlm.nih.gov/pubmed/8418551

Horstman, A. M., Beltman, M. J., Gerrits, K. H., Koppe, P., Janssen, T. W., Elich, P., & de Haan,

A. (2008). Intrinsic muscle strength and voluntary activation of both lower limbs and

functional performance after stroke. Clin Physiol Funct Imaging, 28(4), 251-261.

http://www.ncbi.nlm.nih.gov/pubmed/18355344

Jakobsson, F., Edstrom, L., Grimby, L., & Thornell, L. E. (1991). Disuse of anterior tibial

muscle during locomotion and increased proportion of type II fibres in hemiplegia. J

Neurol Sci, 105(1), 49-56. http://www.ncbi.nlm.nih.gov/pubmed/1795169

Kent-Braun, J. A., & Ng, A. V. (1999). Specific strength and voluntary muscle activation in

young and elderly women and men. J Appl Physiol, 87(1), 22-29.

http://www.ncbi.nlm.nih.gov/pubmed/10409554

Klein, C. S., Brooks, D., Richardson, D., McIlroy, W. E., & Bayley, M. T. (2010). Voluntary

activation failure contributes more to plantar flexor weakness than antagonist

coactivation and muscle atrophy in chronic stroke survivors. J Appl Physiol, 109(5),

1337-1346. http://www.ncbi.nlm.nih.gov/pubmed/20724561

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Knorr, S., Ivanova, T. D., Doherty, T. J., Campbell, J. A., & Garland, S. J. (2011). The origins of

neuromuscular fatigue post-stroke. Exp Brain Res, 214(2), 303-315.

http://www.ncbi.nlm.nih.gov/pubmed/21847644

Knutsson, E., & Richards, C. (1979). Different types of disturbed motor control in gait of

hemiparetic patients. Brain, 102(2), 405-430.

http://www.ncbi.nlm.nih.gov/pubmed/455047

Landau, W. M., & Sahrmann, S. A. (2002). Preservation of directly stimulated muscle strength in

hemiplegia due to stroke. Arch Neurol, 59(9), 1453-1457.

http://www.ncbi.nlm.nih.gov/pubmed/12223033

Levin, M. F., & Hui-Chan, C. (1994). Ankle spasticity is inversely correlated with antagonist

voluntary contraction in hemiparetic subjects. Electromyogr Clin Neurophysiol, 34(7),

415-425. http://www.ncbi.nlm.nih.gov/pubmed/7859670

Lum, P. S., Patten, C., Kothari, D., & Yap, R. (2004). Effects of velocity on maximal torque

production in poststroke hemiparesis. Muscle Nerve, 30(6), 732-742.

http://www.ncbi.nlm.nih.gov/pubmed/15468340

Marsh, E., Sale, D., McComas, A. J., & Quinlan, J. (1981). Influence of joint position on ankle

dorsiflexion in humans. J Appl Physiol, 51(1), 160-167.

http://www.ncbi.nlm.nih.gov/pubmed/7263411

McComas, A. J., Sica, R. E., Upton, A. R., & Aguilera, N. (1973). Functional changes in

motoneurones of hemiparetic patients. J Neurol Neurosurg Psychiatry, 36(2), 183-193.

http://www.ncbi.nlm.nih.gov/pubmed/4350702

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

McNeil, C. J., Vandervoort, A. A., & Rice, C. L. (2007). Peripheral impairments cause a

progressive age-related loss of strength and velocity-dependent power in the dorsiflexors.

J Appl Physiol, 102(5), 1962-1968. http://www.ncbi.nlm.nih.gov/pubmed/17303705

Newham, D. J., & Hsiao, S. F. (2001). Knee muscle isometric strength, voluntary activation and

antagonist co-contraction in the first six months after stroke. Disabil Rehabil, 23(9), 379-

386. http://www.ncbi.nlm.nih.gov/pubmed/11394588

Odajima, N., Ishiai, S., Okiyama, R., Furukawa, T., & Tsukagoshi, H. (1987). [CT findings of

leg muscles in the hemiplegics due to cerebrovascular accidents--correlation to disuse

atrophy]. Rinsho Shinkeigaku, 27(9), 1154-1162.

http://www.ncbi.nlm.nih.gov/pubmed/3440361

Olney, S. J., Griffin, M. P., Monga, T. N., & McBride, I. D. (1991). Work and power in gait of

stroke patients. Arch Phys Med Rehabil, 72(5), 309-314.

http://www.ncbi.nlm.nih.gov/pubmed/2009047

Perry, J., Waters, R. L., & Perrin, T. (1978). Electromyographic analysis of equinovarus

following stroke. Clin Orthop Relat Res(131), 47-53.

http://www.ncbi.nlm.nih.gov/pubmed/657644

Ploutz-Snyder, L. L., Clark, B. C., Logan, L., & Turk, M. (2006). Evaluation of spastic muscle in

stroke survivors using magnetic resonance imaging and resistance to passive motion.

Arch Phys Med Rehabil, 87(12), 1636-1642.

http://www.ncbi.nlm.nih.gov/pubmed/17141645

Prado-Medeiros, C. L., Silva, M. P., Lessi, G. C., Alves, M. Z., Tannus, A., Lindquist, A. R., &

Salvini, T. F. (2012). Muscle atrophy and functional deficits of knee extensors and

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

flexors in people with chronic stroke. Phys Ther, 92(3), 429-439.

http://www.ncbi.nlm.nih.gov/pubmed/22135704

Ramsay, J. W., Barrance, P. J., Buchanan, T. S., & Higginson, J. S. (2011). Paretic muscle

atrophy and non-contractile tissue content in individual muscles of the post-stroke lower

extremity. J Biomech, 44(16), 2741-2746.

http://www.ncbi.nlm.nih.gov/pubmed/21945568

Scelsi, R., Lotta, S., Lommi, G., Poggi, P., & Marchetti, C. (1984). Hemiplegic atrophy.

Morphological findings in the anterior tibial muscle of patients with cerebral vascular

accidents. Acta Neuropathol, 62(4), 324-331.

http://www.ncbi.nlm.nih.gov/pubmed/6730908

Slager, U. T., Hsu, J. D., & Jordan, C. (1985). Histochemical and morphometric changes in

muscles of stroke patients. Clin Orthop Relat Res(199), 159-168.

http://www.ncbi.nlm.nih.gov/pubmed/4042473

van Schaik, C. S., Hicks, A. L., & McCartney, N. (1994). An evaluation of the length-tension

relationship in elderly human ankle dorsiflexors. J Gerontol, 49(3), B121-127.

http://www.ncbi.nlm.nih.gov/pubmed/8169329

Wagner, J. M., Dromerick, A. W., Sahrmann, S. A., & Lang, C. E. (2007). Upper extremity

muscle activation during recovery of reaching in subjects with post-stroke hemiparesis.

Clin Neurophysiol, 118(1), 164-176. http://www.ncbi.nlm.nih.gov/pubmed/17097340

Yucesoy, C. A., & Huijing, P. A. (2007). Substantial effects of epimuscular myofascial force

transmission on muscular mechanics have major implications on spastic muscle and

remedial surgery. J Electromyogr Kinesiol, 17(6), 664-679.

http://www.ncbi.nlm.nih.gov/pubmed/17395489

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Yusevich, Y. S. (1968). The significance of "global" electromyography for analysing the

pathological mechanisms of spastic paralysis. Electromyography, 8(2), 135-157.

http://www.ncbi.nlm.nih.gov/pubmed/5705819

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Figure 1. Ipsilesional and contralesional dorisflexor MVC torque (uppermost two traces) and

raw EMG of the TA, soleus, LG and MG in two stroke subjects (contralesional EMG is the

lower trace of each pair). (A) In this subject (no. 2), peak MVC torque is similar between sides

(B) In this subject (no. 4), MVC torque was much less in the contralesional than the ipsilesional

limb. Note the interpolated twitch and sparse TA EMG in the contralesional limb.

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Figure 2. Individual contralesional dorsiflexor voluntary activation (▼), and relative

contralesional values (% ipsilesional side) for MVC torque (●), MVC EMG (▲), 50 Hz torque

(♦), and fat-free muscle volume (■) for subjects 1-7.

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Figure 3. Evoked dorsiflexor torque and EMG during supramaximal 50 Hz stimulation in

subject 2 (A) and subject 7 (B). For each subject, the torque and EMG are shown for the

contralesional limb (dotted line and lower EMG trace) and ipsilesional limb (continuous line and

upper EMG trace).

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Table 1. Characteristics of the stroke subjects

Subject

no.

Sex Age

(yr)

Ht

(cm)

Wt

(kg)

CMSA

Leg/foot

Speed

(m/s)

TSO

(months)

1

M 67 185.4 104.5 5/5 1.35 36

2 M 63 180.3 81.8 4/4 1.0 43

3 F 61 162.6 61.4 4/3 0.63 37

4 M 56 177.8 68.2 4/3 0.49 27

5 M 51 185.3 95.4 3/3 0.58 24

6 F 40 167.6 68.2 3/3 0.82 51

7 M 60 178.0 61.4 4/2 0.91 48

Abbreviations: M, male; F, female; Ht, height, Wt, weight; CMSA, Chedoke-McMaster Stroke

Assessment Score; Speed, walking speed; TSO, time since onset of stroke

“Neural and Muscular Determinants of Dorsiflexor Weakness in Chronic Stroke Survivors”

by Klein CS, Power GA, Brooks D, Rice CL

Motor Control

© 2013 Human Kinetics, Inc.

Table 2. Dorsiflexor properties in stroke and control subjects

Measure

Contralesional Ipsilesional Control

TA Mmax (mV)

7.8 ± 1.2 8.0 ± 2.8 7.4 ± 2.6

MVC (Nm)

29.8 ± 21.3* 42.5 ± 12.0 49.7 ± 7.1

MVC EMG

0.043 ± 0.028* 0.118 ± 0.045 0.09 ± 0.01

MVC activation (%)

78.1 ± 37.8 98.7 ± 1.0 98.2 ± 1.5

Soleus coactivation (%)

35.4 ± 37.6 16.3 ± 3.9 19.1 ± 13.2

Gastrocnemii coactivation (%) 25.8 ± 26.6 10.1 ± 8.1 18.0 ± 6.1

50 Hz torque (Nm)

32.3 ± 12.6 30.4 ± 10.5 35.7 ± 6.6

Dorsiflexor volume (cm3)

260 ± 69 250 ± 82 _

Dorsiflexor CSA (cm2)

13.2±3.0 12.9 ±4.1 _

50Hz/volume (Nm/cm3)

0.12 ± 0.03 0.12 ± 0.02 _

Note the MVC EMG values were normalized to the TA Mmax.

* Significantly different compared to the ipsilesional limb and control limb.