Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2012 Oct 1.
Published in final edited form as: Muscle Nerve. 2011 Jul 13;44(4):509–517. doi: 10.1002/mus.22098

Tibialis anterior architecture, strength and gait in individuals with cerebral palsy

Daniel C Bland 1, Laura A Prosser 2, Lindsey A Bellini 2, Katharine E Alter 2, Diane L Damiano 2
PMCID: PMC3175274  NIHMSID: NIHMS279299  PMID: 21755515

Abstract

Introduction

The relationship of tibialis anterior (TA) muscle architecture including muscle thickness (MT), cross-sectional area (CSA), pennation angle (PA) and fascicle length (FL) to strength and ankle function was examined in ambulatory individuals with CP and unilateral foot drop.

Methods

Twenty individuals with CP participated in muscle ultrasound imaging, unilateral strength testing, and 3D gait analysis.

Results

Muscle size (MT and CSA) were positively related to strength, fast gait velocity, and ankle kinematics during walking. Higher PA was related to a more dorsiflexed ankle position at initial contact and inversely with fast gait velocity. FL was related to strength, fast velocity, and step length at self-selected speed.

Discussion

Muscle architecture partially explains the degree of impairment in strength and ankle function in CP. Treatments to increase TA size and strength may produce some gait improvement, but other factors that may contribute to ankle performance deficits must be considered.

Keywords: cerebral palsy, tibialis anterior, gait, muscle architecture, strength

Introduction

Cerebral palsy (CP) refers to a group of permanent motor disorders resulting from a non-progressive injury to the brain during intrauterine or early post-natal development. Activity limitations result from impairments in the development of movement and posture, which often include spasticity, dystonia, muscle weakness, a lack of motor control, poor coordination, and decreased range of motion.1 CP is the most common neuromuscular disorder among children2 and the most common cause of physical disability in childhood. Its prevalence has been increasing in past decades.3 While CP has diverse manifestations that affect each individual differently, by definition all affected individuals experience some type of motor dysfunction. There is typically more distal than proximal extremity motor impairment,4 and the ankle is the most commonly affected joint in individuals who are ambulatory. While CP is caused by a non-progressive injury, the nature of the musculoskeletal impairments may worsen with time and growth due to the reinforcement of altered movement patterns and reduced physical activity.5

Advances in muscle imaging have led to an increasing number of studies relating non-invasive measurements of muscle architecture to muscle performance during functional activities in trained and untrained adults and more recently in neurological and other impaired populations. Physiological Cross Sectional Area (PCSA) has a well-established relationship with the maximal force production capability of a muscle, but it requires measurement of the entire muscle so that the total volume can be divided by fiber length6 and is most accurately and easily captured with magnetic resonance imaging (MRI). Muscle volume alone, which is one component of the PCSA, has also been shown to have a strong linear relationship with force. It may even be a more accurate predictor variable than PCSA in certain sub-populations such as the elderly. The anatomical, or largest, cross-sectional area of a muscle from a single static image can be captured with any muscle ultrasound device (assuming the muscle is not too large to fit within the field of view), but the relationship of this parameter with maximal force is not as strong as with the PCSA, which is a mean rather than maximal value.6

Muscle thickness (MT) is the most easily and readily obtained ultrasound measure of muscle size and has also shown moderately strong relationships to maximal torque. In contrast, the length of the muscle fibers, or fascicle length (FL) as measured on ultrasound, typically has its strongest and most direct relationship to muscle contraction velocity.7

Interpretation of pennation angle (PA) differences across groups or in response to interventions is not as straightforward as the other parameters. A higher PA per unit volume typically relates to greater torque production, and lower angles are associated with greater contraction velocity.8 In fact it has been shown that resistance versus velocity training regimens produce opposite effects on pennation angle.9

Muscle size and length measurements can be more easily understood at the level of the contractile unit of muscle, the sarcomere. Basically, the number of sarcomeres in parallel relates physiologically to strength and anatomically to muscle area, and the number in series relates to velocity and muscle length.10

Considerable individual variation exists in the composition of muscle tissues and the neural control of muscle activation in healthy populations, which has a marked effect on the strength of the relationships between muscle architecture and muscle performance, as shown through inconsistent results from the analyses of different subgroups.11 In cerebral palsy, alterations have been found at all levels of the neuromotor system that introduce even greater variability and distort these relationships even more. These include changes in neural activation patterns and efficiency, decreased fiber diameter, selective atrophy and alterations in fiber types, and collagen infiltration in muscle tissue, among others.12

This study aims to inform the relationship between muscle structure and function in CP by examining a single muscle in this population that has a very specific effect on gait function. Common ankle impairments in CP are insufficient dorsiflexion during the swing phase and excessive plantarflexion during stance (refer to Figure 1). Typical sagittal plane ankle motion during walking involves swing phase dorsiflexion to clear the toes from the ground as the limb advances. Dorsiflexion also occurs during the stance phase of gait as the body moves forward over the standing limb while the opposite limb advances. Plantarflexion typically occurs at the end of the stance phase when the limb is pushing off to propel the body forward and upon lowering of the foot after initial contact. Excessive plantarflexion in stance is believed to be caused or exacerbated by excessive or inappropriate muscle activity (spasticity, dystonia or muscle imbalance) of the gastrocnemius and soleus muscles. Foot drop is a swing phase problem which may be due to either persistent equinus or decreased muscle activation and/or weakness in the dorsiflexors. A paradigm shift has occurred over time in the treatment of the motor impairments in CP and now posits that weakness has a far greater effect on functional disability than spasticity which was long thought to be the primary culprit.13 Therefore, the focus here is on the role of dorsiflexor (TA) muscle size and structure through the presumed association of muscle size with maximal force and how this relates to ankle motion in swing phase and overall gait function as measured by self-selected and fast velocity.

Figure 1.

Figure 1

Open in a new tab

Representative mean ankle angle data during gait from one participant who was GMFCS Level II (black line). The gray band represents one standard deviation around the mean for typical gait. Stance phase is the period when the limb is in contact with the ground; swing phase is the period when the limb is not in contact with the ground and is advancing forward. Gait variables of interest are noted: ICDf, dorsiflexion at initial contact; DfMxSt, maximum dorsiflexion in stance; DfTO, dorsiflexion at toe-off; DfMxSw, maximum dorsiflexion in swing.

The ultrasound measurements in this study were only performed on the major dorsiflexor of the ankle, the tibialis anterior (TA) muscle, which lifts and inverts the foot. Three other muscles also contribute to ankle dorsiflexion: the extensor digitorum longus, extensor hallicis longus and peroneus tertius. The TA is a bipennate muscle, meaning that muscle fibers branch out at a specific angle, termed pennation angle (PA), from a central intramuscular septum to the epimysium. A recent study reported that TA muscle size was significantly smaller in the more affected extremity of children with asymmetric CP.14 However, the relationships among TA muscle architecture, strength, and walking function have not yet been investigated.

Specifically, the goal of this study was to examine the relationship of TA muscle thickness (MT), pennation angle (PA), fascicle length (FL) and cross-sectional area (CSA) to strength and ankle function during walking in individuals with asymmetric CP. We hypothesized that TA muscle architecture would be related to strength on the more affected side and that both would be related to ipsilateral ankle function and temporal spatial gait parameters with smaller size, pennation angles, and fascicle lengths related to poorer functioning. To further show the relationship between tibialis anterior structure and function, we hypothesized that all muscle architectural measures would have lower values on the more affected side. While comparisons are not made to a healthy population here, comparing parameters across sides of those with asymmetrical involvement provides one context for helping to understand some of these relationships.

Materials and Methods

Participants

Twenty individuals (age range 8.3–19.9 years; mean 13.1±SD 3.5, 10 male, 10 female) with asymmetric CP and a diagnosis of unilateral foot drop participated in this prospective, observational cohort study. All were part of a larger, ongoing intervention trial of a functional electrical stimulation orthotic device to augment ankle dorsiflexion, and the data for this study were obtained during their baseline visits. The majority (17 participants) had hemiplegia, but three had diplegic CP with one side more affected than the other. All participants demonstrated reduced ankle dorsiflexion at initial foot contact and during the swing phase of the gait cycle on clinical observation of barefoot walking. Passive range of motion data were available from the Biodex testing set-up procedures for 19 of the 20 participants. Mean passive dorsiflexion with the knee flexed to 60° was 12.4° (SD = 9.1) with a range of 1–31°. All participants functioned at Gross Motor Function Classification System (GMFCS) levels I (n=14) and II (n=6),15 indicating that they could walk independently without the use of assistive devices, but with varying levels of difficulty performing more advanced motor skills or navigating challenging terrain. Electromyographic (EMG) evidence of stretch responses (spasticity) in the medial gastrocnemius in response to repeated passive dorsiflexion at a constant velocity of 90°/second was seen in 14 of the 20 participants, nine of whom were GMFCS I and five of whom were GMFCS II.

Exclusion criteria were botulinum toxin injection to the gastrocnemius or soleus muscles within the four months prior to testing or orthopedic surgery to the legs in the previous year, since these interventions might result in changes to muscle architecture or gait that had not stabilized. For example, changes in muscle volume and length have been observed after gastrocnemius recession.16 The study was approved by the institutional review board (IRB) at the National Institutes of Health in Bethesda, MD, USA. Written informed consent was obtained from participants over 18 years of age and parents of minors. Written assent was obtained from each child under 18 years of age.

Data Procurement

Real-time musculoskeletal ultrasound imaging and muscle strength testing were performed to examine muscle architecture and strength of the TA. Both lower extremities were scanned with the ultrasound, but in the interest of time only the more affected lower extremity was tested for strength. Participants were seated with the ankle joint aligned with a multi-mode dynamometer (Biodex System 3, Biodex Medical Systems, Inc., Shirley, NY, USA) in 80° hip flexion, 60° knee flexion, and 0° (neutral) ankle position measured with a goniometer. Range of motion (ROM) for isokinetic and EMG spasticity assessment was determined individually by moving the device manually in each direction until initial resistance was felt, then more slowly to an endpoint that was close to the maximum range but still tolerable to the participant.

For ultrasound imaging (SonixTOUCH, Ultrasonix Medical Corporation, Richmond, BC, Canada), participants were instructed to relax fully during the scans. Muscle contractions were easily detected, and a resting muscle state was verified visually in real-time. Ultrasound gel (Aquasonic CLEAR, Parker Laboratories, Fairfield, NJ, USA) was applied to the skin to provide acoustic coupling and to reduce compression or deformation of the muscle. At each visit, one of two ultrasound operators took TA images at the thickest region of the muscle belly as determined by visual inspection of real-time ultrasound, and the locations were marked and recorded as a percentage of the distance from the fibular head to the medial malleolus. A two-dimensional B-mode ultrasound device with a 5–14 MHz linear array transducer was used to record images from which measurements of TA MT, PA, FL and CSA could be obtained. Bilateral longitudinal images were taken with the probe oriented in the sagittal plane perpendicular to the skin for viewing thickness and pennation angle. Bilateral cross-sectional images were taken with the probe oriented in the transverse plane perpendicular to the skin. Two images were obtained in each view.

Intrasession reliability of ultrasound measures generated from both investigators who collected ultrasound data was assessed with a healthy volunteer under a different IRB protocol. Intraclass correlation coefficients (ICC) were used to evaluate intrarater reliability for each operator as well as interrater reliability between the two operators and between each operator and an expert ultrasonographer.

For force generation (strength) testing, participants performed three maximum voluntary isometric contractions (MVIC) of ankle dorsiflexion on the more affected side at 0° ankle position while attempting to hold each contraction approximately five seconds before relaxing. Only the more affected side was tested, because the protocol for the larger study focused on the side receiving the intervention. Ultrasound of the less affected side was additionally collected in order to provide information for the symmetry analysis. Strength data were collected with a custom-written program (LabVIEW 8, National Instruments, Austin, TX, USA) at 1000 Hz. Ultrasound video in B-mode was recorded concurrently during the second and third MVIC trials at 24 frames per second.

Three-dimensional lower extremity kinematic data were collected with a ten-camera motion capture system (Vicon MX, Vicon, Lake Forest, CA, USA). This system tracked the movement of 34 reflective markers placed on pelvic and lower extremity landmarks at a sampling frequency of 120 Hz. Markers were strategically placed to create a six degrees-of-freedom seven segment (pelvis, upper legs, lower legs, and feet) model. Specifically, the pelvis was defined using the anatomical locations of the left and right anterior superior iliac spine (ASIS) and the posterior superior iliac spine (PSIS). Participants walked barefoot over-ground down the middle of the gait laboratory, approximately eight meters in length, at self-selected (SSW) and fast (FAST) speeds. They were instructed to look straight ahead and walk at a “normal, comfortable pace” and “as fast as possible,” respectively.

Data Analysis

Analysis of the ultrasound images was performed with the Medical Image Processing, Analysis and Visualization (MIPAV) software (National Institutes of Health, Bethesda, MD, USA) by one investigator (DB). The image from each pair that showed the clearest picture for analysis of all measures was chosen. Thickness was calculated at the midpoint of the ultrasound image as the distance between the superficial and deep aponeuroses at a 90° angle from the deep aponeurosis. Pennation angle was measured as the positive angle between the central intramuscular septum and the line of the clearest fascicle (Fig. 2). The perimeter of the TA was manually traced in the software for each cross-sectional image, from which the CSA was calculated (Fig. 2). Fascicle Length was estimated by dividing the muscle thickness value by the sine of the pennation angle.17

Figure 2.

Figure 2

Open in a new tab

Longitudinal (top) and transverse views (bottom) of the tibialis anterior muscle on the more (left) and less (right) affected side for one participant. Thickness and pennation angle were measured in the longitudinal views, and cross-sectional area in the transverse views. The top of each image is the skin and fascial layer. The muscle directly below the TA in the longitudinal image is the tibialis posterior.

Strength data were processed in MATLAB (MathWorks, Natick, MA, USA), and the peak force of each of three trials was recorded. The maximum of the three trials was used for analysis to represent MVIC. The ultrasound videos recorded during the strength testing trials were synchronized to the MVIC force time series by first calculating the duration of time between the initiation of force generation and the peak force from the time series. The ultrasound frame at which the peak force occurred was identified. This frame was analyzed with MIPAV in the fashion described above to measure and analyze CSA during MVIC. MVIC data from one patient could not be obtained, because she exhibited dystonic posturing of her more affected leg during exertion for MVIC and had to be removed from the device, but this did not significantly affect performance on any other measures.

Marker data and anthropometric measurements were used to create a subject-specific anatomical model in Visual 3D (C-Motion, Inc., Gaithersburg, MD) from which gait variables were computed (Fig. 1). Variables chosen to represent general and ankle-specific function in this study included temporal-spatial parameters such as walking speed and step length; kinematic variables such as maximum dorsiflexion (angle) in stance, maximum and mean dorsiflexion in swing, dorsiflexion at toe-off, ankle joint angular velocity in the direction of dorsiflexion, and dorsiflexion at initial foot contact (ICDf). The average values of five gait cycles were used for our analysis. The minimum number of trials recommended in the literature for children with CP who are GMFCS Level I is four, and for those in Levels II–III, it is six. Given the mild involvement of our subjects with most in GMFCS I and none in Level III, we determined that five trials were sufficient here.18

Pearson correlation coefficients were calculated to examine relationships between muscle architecture, muscle strength, and walking function (α-level = 0.05). Strength of the correlations was determined from the guidelines presented by Portney and Watkins,19 which state that values from 0.25–0.50 suggest a fair relationship, 0.50–0.75 a moderate to good relationship and 0.75 and above a good to excellent relationship. MT, CSA, and MVIC20,21 were normalized by body weight for correlation tests to ankle kinematic variables to control for variations in the sizes of the children (ages 8–19 years). For the same reason, temporal-spatial data were normalized for the statistical analyses to a dimensionless value according to Hof.21 Paired t-tests (two-tailed; α-level = 0.05) were used to analyze differences in ultrasound and gait measures between more and less affected legs.

Results

Assessment of intrasession reliability and validity of ultrasound measures for both investigators who obtained ultrasound data yielded ICC values of ≥0.99 and ≥0.98 for the primary and secondary operator, respectively. Interrater reliability between the two scanning investigators was 0.98. Comparison of each of these investigators with the ultrasonography expert resulted in ICC values of ≥0.99 for the primary operator and ≥0.96 for the secondary.

On the more affected side, the thickest region of the TA muscle belly was on average 31.4% (4.1%) of the distance from the fibular head to the medial malleolus, while the corresponding area of maximum thickness on the less affected side was 34.8% (4.7%) of the distance. The more affected TA was also shorter than the less affected TA (p<0.001); the average length was 31.8 cm (3.1 cm) and 33.0 cm (2.6 cm) on the more and less affected sides, respectively. After controlling for TA length, there was no longer a statistical difference in the location of maximum MT between sides (p=0.07).

Within the muscle architectural measures, CSA and MT were significantly related to each other (r=0.89, p<0.001), and CSA at rest was highly correlated with CSA during MVIC (r=0.97, p<0.001). Because TA CSA was not significantly different at rest and during MVIC (p=0.61), only relationships between CSA at rest and other measures are reported here. FL was inversely related to pennation angle (−0.81, p<0.01).

Table 1 shows the correlation coefficients and p-values for the relationships between strength and muscle architecture and muscle architecture and function. Table 2 shows the correlation coefficients and p-values for the relationships between muscle architecture and ankle joint kinematics. Only those correlations and p values not shown in the table will be included in the text. The first hypothesis was supported as shown by moderate correlations between muscle size and strength on the more affected side. MT, CSA and FL were significantly related to muscle strength, while PA was not. Support for the relationship of muscle structure to gait function was also found. During self-selected walking, larger MT was related to greater maximum dorsiflexion in swing; larger PA was related to greater dorsiflexion at initial foot contact; and larger CSA was related to greater maximum and mean dorsiflexion in swing, and greater dorsiflexion at toe-off. During FAST walking, larger MT was related to a higher velocity and greater ankle dorsiflexion velocity during swing; larger CSA was related to a higher velocity and greater maximum and mean dorsiflexion in swing. PA was inversely related to fast walking speed; however, it was also inversely related to fascicle length, which was directly related to fast gait speed. MVIC had similar relationships with kinematic parameters as those with muscle size. In the SSW condition, a stronger TA muscle was related to greater maximum (r=0.56, p=0.02) and mean dorsiflexion (r=0.57, p=0.01) during swing, and greater dorsiflexion at toe off (r=0.48, p=0.05). During FAST walking, greater muscle strength was related to greater maximum (r=0.61, p=0.01) and mean (r=0.61, p=0.01) dorsiflexion in swing, and fast velocity (r=0.53, p=0.02) and step length (r=0.51, p=0.03).

Table 1.

Relationships between muscle architecture measures and maximal muscle strength and walking function.

MVIC Walking
speed		 Step
length
SSW MT r=0.57* r=0.13 r=0.39
p=0.01 p=0.58 p=0.09
PA r=−0.38 r=−0.16 r=−−0.36
p=0.12 p=0.50 p=0.12
FL r=0.51* r=0.32 r=0.46*
p=0.03 p=0.17 p=0.04
CSA r=0.56* r=0.05 r=0.20
p=0.02 p=0.84 p=0.39
FAST MT r=0.67* r=0.38
p=0.01 p=0.10
PA r=−0.47* r=−0.05
p=0.04 p=0.82
FL r=0.58* r=0.23
p=0.01 p=0.34
CSA r=−0.51* r=0.25
p=0.02 p=0.28
Open in a new tab
*

denotes significance at α=0.05. SSW, self-selected walking; FAST, fast walking; MT, muscle thickness; PA, pennation angle; FL, fascicle length; CSA, cross-sectional area; MVIC, maximum voluntary isometric contraction.

Table 2.

Relationships between muscle architecture measures and ankle kinematics.

ICDf DfMxSt DfTO DfMxSw DfMn velDf
SSW MT r=−0.01 r=−0.02 r=0.29 r=0.47* R=0.36 r=0.38
p=0.95 p=0.93 p=0.22 p=0.04 p=0.11 p=0.10
PA r=0.45* r=0.13 r=−0.07 r=0.01 R=0.11 r=−0.27
p=0.04 p=0.58 p=0.79 p=0.98 p=0.64 p=0.25
FL r=−0.35 r=0.08 r=0.15 r=0.23 R=0.17 r=0.35
p=0.13 p=0.73 p=0.54 p=0.34 p=0.48 p=0.13
CSA r=0.13 r=0.21 r=0.52* r=0.64* R=0.60* r=0.24
p=0.59 p=0.38 p=0.02 p=0.002 p=0.01 p=0.30
FAST MT r=0.10 r=−0.30 r=−0.04 r=0.40 R=0.25 r=0.50*
p=0.67 p=0.20 p=0.88 p=0.08 p=0.28 p=0.02
PA r=0.14 r=0.21 r=−0.27 r=−0.26 R=−0.23 r=−0.13
p=0.54 p=0.38 p=0.25 p=0.26 p=0.33 p=0.57
FL r=0.05 r=−0.04 r=0.29 r=0.40 R=0.37 r=0.28
p=0.84 p=0.86 p=0.22 p=0.08 p=0.11 p=0.23
CSA r=0.30 r=0.11 r=0.27 r=0.61* R=0.54* r=0.41
p=0.19 p=0.65 p=0.25 p=0.004 p=0.01 p=0.07
Open in a new tab
*

denotes significance at α=0.05. SSW, self-selected walking; FAST, fast walking; MT, muscle thickness; PA, pennation angle; FL, fascicle length; CSA, cross-sectional area; DfMxSt, max dorsiflexion in stance; DfMxSw, max dorsiflexion in swing; DfMn, mean dorsiflexion in swing; DfTO, dorsiflexion at toe-off; velDf, dorsiflexion angular velocity; ICDf, dorsiflexion at initial contact

The TA muscles on the more affected sides had lower MT values than those on the less affected sides by 25.6%. Similarly, the CSA of the more affected TA was 36.2% smaller than the less affected TA. No difference was seen in PA, but normalized FL was 28% smaller on the more affected side. Group means and standard deviations for all muscle architecture and self selected walking gait parameters with p values for the side-to-side comparison are listed in Table 3. Group means and standard deviations for the fast walking gait parameters with p values for the side-to-side comparison are listed in Table 4. The visual gait observation to determine which side was ‘worse’ was corroborated by the gait analysis in that many of the ankle kinematic measures of foot drop during walking were significantly more impaired on the more affected side. During self-selected pace walking, the more affected side demonstrated significantly less dorsiflexion through the majority of the gait cycle (maximum in stance and swing, mean in swing, and at initial foot contact), and a 53.6% slower ankle dorsiflexion velocity during swing. Step length and dorsiflexion at toe-off, on the other hand, were not significantly different between more and less affected sides. In the FAST walking condition, similar differences were observed. The more affected side demonstrated less dorsiflexion at the same gait cycle events (maximum in stance and swing, mean in swing, and at initial foot contact) and a 46.2% slower ankle dorsiflexion velocity during swing. Again, no significant differences were found in step length or dorsiflexion at toe-off.

Table 3.

Side to side comparisons of muscle architecture and self-selected gait parameters.

Muscle architecture LA MA p
TA muscle thickness (cm) 2.2 (0.3) 1.6 (0.3) <0.001*
TA pennation angle (°) 9.7 (2.1) 10.6 (2.7) 0.131
TA cross-sectional area (cm2) 4.6 (1.1) 3.0 (0.7) <0.001*
FL (cm) 13.4 (2.4) 9.6 (3.7) <0.001*
SW
Gait LA MA p
Step length (m) 0.6 (0.1) 0.5 (0.1) 0.57
Max dorsiflexion in stance (°) 15.1 (3.5) 11.3 (4.7) 0.02*
Max dorsiflexion in swing (°) 7.9 (3.1) −0.1 (4.7) <0.001*
Mean dorsiflexion in swing (°) 2.0 (2.9) −4.7 (4.4) <0.001*
Dorsiflexion at toe-off (°) −6.5 (4.8) −5.6 (5.4) 0.52
Dorsiflexion velocity (° /s) 0.9 (0.2) 0.4 (0.2) <0.001*
Dorsiflexion at initial contact (deg) 0.1 (3.1) −6.8 (3.8) <0.001*
Open in a new tab

Values are means (SD).

*

denotes significant difference between sides (α=0.05). TA, tibialis anterior; SSW, self-selected pace walking; FAST, fast walking; LA, less affected; MA, more affected; deg, degrees

Table 4.

Side to side comparisons of fast gait parameters.

FAST
Gait LA MA p
Step length (m) 0.6 (0.1) 0.6 (0.1) 0.29
Max dorsiflexion in stance (°) 13.9 (3.8) 7.7 (5.1) <0.001*
Max dorsiflexion in swing (°) 8.9 (3.7) −0.3 (5.8) <0.001*
Mean dorsiflexion in swing (°) 1.7 (3.9) −5.7 (5.3) <0.001*
Dorsiflexion at toe-off (°) −9.3 (6.3) −8.9 (6.6) 0.83
Dorsiflexion velocity (° /s) 1.2 (0.3) 0.6 (0.4) <0.001*
Dorsiflexion at initial contact (deg) 0.6 (3.2) −6.1 (3.6) <0.001*
Open in a new tab

Values are means (SD).

*

denotes significant difference between sides (α=0.05). TA, tibialis anterior; SSW, self-selected pace walking; FAST, fast walking; LA, less affected; MA, more affected; deg, degrees

Discussion

All TA muscle measurements were obtained from a single site to represent the whole muscle, but it must be acknowledged that different muscle shapes might exist (i.e., may be more or less tapered across individuals). The CSA measurement here was an anatomical one rather than a physiological one, which requires measurement of muscle volume. A potential source of variability that we attempted to minimize was imaging the TA at the thickest part of the muscle belly. Obtaining ultrasound images at a standardized location across all patients and limbs would likely yield even more varied results, especially if that standardized point were considerably different than the region of maximum MT. It has been common practice to measure the TA at a standardized location (e.g., 20–25% of the distance from the fibular head to the medial malleolus), which is more proximal than the average region of maximum MT in our sample (31.4% and 34.8% of TA length on the more and less affected sides, respectively). In cases where the largest cross-sectional area location is visible, we would recommend that size measurements be taken at that location. It is also known that variations in other architectural measures can occur throughout the muscle length.22 The point at which the measurements were taken must therefore be noted and considered when comparing architecture and size across studies.

The strong relationship between TA MT and resting CSA in our sample of children with CP was expected and supports previous work.14,23 An inverse relationship was found between PA and FL. There was also a strong relationship of MT to CSA during maximal contraction, which is logical given that CSA at rest and during MVIC was statistically equivalent and muscle volume remains constant. The work of Narici10 partially explains these findings; when contracted, the fibers of a pennate muscle pivot about their origin, shortening in length and increasing their PA, while maintaining a constant distance between the aponeuroses.

MT, CSA and fascicle length were moderately correlated with TA MVIC. These relationships suggest that muscle structure influences function, or vice versa, a concept well-recognized in normal skeletal muscles.24 However, there are inconsistencies across studies, and a recent review suggests that this relationship is far more complex than has been previously acknowledged, even in healthy populations.11 Major factors identified in the literature that may modify this relationship include age and gender as well as individual differences in genetic make up, neural recruitment and muscle tissue composition.11 Each of these could act alone or in combination to influence the outcomes of training programs or therapeutic interventions aimed at improving force output in healthy and patient populations.

The link between muscle size and strength is even less well documented or understood in disorders such as CP. Given the brain and muscle pathology that has been identified in CP, there is reason to suspect that it may differ from normal as was seen in the study by Moreau and colleagues.25 In both the CP and control groups, the vastus lateralis MT was the best predictor of knee extension MVIC. That plus age accounted for 91% of the variance in strength in the control group but only 63% of the variance in the group with CP. The addition of GMFCS Level in the group with CP increased this to 82%.

In this study, only 25–35% of the variance in MVIC was explained by either TA MT or CSA. This is considerably lower than previously reported values in young healthy adults, which range from 44% to more than 75% of the variance.25 However, in the absence of an age-related control group, the effect of age versus pathology cannot be determined.

Finally, the unique contribution of our study was the finding of significant correlations of weak to moderate strength between muscle architecture and ankle function during walking. In general, there are significant positive relationships of TA muscle size to gait velocity and the amount of ankle dorsiflexion during walking, while PA has fewer relationships with ankle function and appears to be inversely related to gait velocity. It is also inversely related to FL, which shows a positive relationship with fast velocity. All of the above findings further suggest that a muscle’s structure impacts its performance during functional movement in CP. Ohata and colleagues20 previously reported a relationship between muscle size and function in the quadriceps and longissimus muscles in this population, showing that MT differed significantly according to functional level (GMFCS) in patients with CP.

The strengths of the correlations between size and function and the lack of relationships with some gait variables also likely indicates that additional factors other than muscle architecture may contribute to the ability of the ankle to function during walking. For example, studies have quantified a disruption in voluntary neuromuscular activation in children with CP.26 Children with CP also have a lack of selective voluntary motor control or an inability to perform isolated joint movements by activating only specific patterns of muscle groups upon request, and this impairment is more severe in distal joints like the ankle.4 Further, the composition of a muscle affects its function. Collagen accumulation has been shown to occur in the muscles of children with CP. It replaces the force-producing fibers necessary for movement of the ankle,27 and variations in fiber size and abnormal distributions of Type I and Type II fibers have been seen.28 A combination of all of these factors may determine the degree of limitation in ankle function observed in CP as well as the potential response to strength training.

In addition to the correlations, differences in muscle size, strength and gait parameters across sides further support, albeit less directly, the proposed link between muscle structure and function. It is generally assumed that muscles of children with CP are relatively smaller than those of their peers, but few studies have examined this hypothesis.29 In hemiplegic CP, the more affected leg has been found to have smaller MT and CSA of the ankle plantarflexors30,31 and smaller MT of the rectus femoris/vastus intermedius group.20 Likewise, Bandholm et al.14 reported that the dorsiflexors of the more affected leg in children with hemiplegic CP are significantly smaller than on the less affected leg. Thus, our results are consistent with the existing evidence of smaller muscle size in more impaired limbs, while showing no significant differences in PA. The side-to-side comparisons here of FL and PA are similar to the results of a study comparing two groups of sprinters. In that study, the faster of the two groups had consistently lower but non-significant differences in pennation angles with significantly greater fascicle lengths.32

In individuals with hemiplegic CP, abnormalities in sagittal plane ankle kinematics have been observed in the less affected limb as well as the more affected limb, as these individuals attempt to achieve a more symmetric gait pattern.33 However, Cimolin and authors34 reported that the anomalous compensation patterns on the less affected side generally affect the more proximal joints, leaving the ankle kinematics close to the normative range and hence different from the more affected side, as was seen in our results. The addition of a control group would likely have made the structure to functional relationships even more apparent.

Limitations

There are several potential limitations to this study. An apparent one was that we only measured muscle size in one of four muscles that dorsiflex the ankle. A similar decision has been made in other studies, and we compared this to measures of strength and performance of the entire muscle group. In addition, all measures were performed with participants in a single, neutral (0°) ankle position despite the fact that the children had differences in plantarflexion range of motion. Although the TA was assumed to be in a slackened position for all children, differences in passive muscle tension may alter muscle shape and sarcomere length, which could affect the length-tension properties and force production.35 This is especially important to consider in pennate muscles, which have shorter fiber lengths compared to parallel-fiber muscles. For the same absolute increase in fiber length under tension, greater relative stretch per sarcomere will result.10 Furthermore, a larger sample size may have yielded more power to detect statistically significant relationships between the measures investigated.

Another limitation is the primary focus on sagittal plane ankle motion without consideration of motion in other planes, in particular inversion and eversion. Although we collected three-dimensional movement of the ankle joint during walking, we did not use a detailed foot model here which is necessary for accurately quantifying motion of the other joints within the foot such as the subtalar joint. In the model used here, the foot is considered a rigid segment and any frontal plane motion occurring at the subtalar joint or elsewhere in the foot is recorded instead as ankle rotation.

Another possible limitation was the fact that we only related barefoot gait data to muscle size even though limb size asymmetry was demonstrated. We chose to study barefoot walking over other walking conditions, because fewer variables are introduced (e.g. different shoes or orthotics) between subjects or sides. However, in children with asymmetric injuries, leg length discrepancies in barefoot may significantly alter gait parameters. Tibial length as a proxy for leg length was compared between the more and less affected sides and was found to be statistically different (p<.05) with the more affected side being on average 7.7 mm shorter. A literature review conducted by Knutson found that, in general, a leg length inequality of less than 20 mm is unlikely to require treatment.36 Unfortunately, femur length was not measured here, but given the small difference found in tibial length, it is unlikely that more than a few of our subjects, if any, exceeded this clinical threshold. Thus, barefoot assessments are acceptable here.

Sample heterogeneity inherent to the CP population may also have affected the power of the study. The participant who demonstrated dystonic movement during strength testing, although an extreme case, illustrates the point that individuals with CP may exhibit considerable variability in the type as well as extent of motor impairments that may have differential effects on function.

Clinical Significance

Ultrasound imaging is well-established as a less invasive, more affordable alternative to techniques like magnetic resonance imaging and biopsies, which is readily adaptable for clinical or laboratory use. Its reliability has been demonstrated in studies, including ours, and it has been recommended by experts for analyzing muscle architecture in children with CP in clinical settings.37,29 However, the most relevant muscle architecture measure has not been identified. Ohata et al20 suggest the use of MT to determine function, but CSA has traditionally been used due to its directly proportional relationship to the number of sarcomeres in parallel and hence to force production in normal muscles.10 However, CSA for larger muscles requires a large field of view that is not possible on all US machines. The study by Moreau, et.al.25 also recommended the use of vastus lateralis MT as a surrogate measure of voluntary strength in children with and without CP. In our work, due to its somewhat stronger correlation than CSA to MVIC and its similarity to CSA in relationship to ankle function during walking, MT may be a more clinically useful measure than CSA, particularly given its greater availability and ease of image collection and analysis relative to CSA. Future research should focus on deciphering the contributions of different factors (i.e., muscle architecture, muscle composition, neuromuscular activation, and selective control) to decreased force production and impaired ankle function. Our primary goal is to determine whether physical training programs that may target one or more of these factors (i.e., resistance training, functional electrical stimulation (FES), or intense motor practice) can improve or restore lost ankle function during walking in children with cerebral palsy

Acknowledgments

The authors thank Chris Stanley, MS and Cristiane Zampieri-Gallagher, PT, PhD for their assistance with data collection. This research was supported by the Clinical Research Training Program, a public-private partnership between the NIH and Pfizer Inc (via a grant to the NIH Foundation from Pfizer Inc) and was also supported in part by the Intramural Research Program at the NIH Clinical Center.

Abbreviations

CP

cerebral palsy

CSA

cross-sectional area

FAST

“as fast as possible” walking

FL

Fascicle length

GMFCS

Gross Motor Function Classification System

MT

muscle thickness

MVIC

maximum voluntary isometric contraction(s)

PA

pennation angle

SSW

self-selected walking

TA

tibialis anterior

REFERENCES

  • 1.Rosenbaum P, Paneth N, Leviton A, Goldstein M, Bax M. A report: the definition and classification of cerebral palsy. 2006:8–14. [PubMed] [Google Scholar]
  • 2.Pakula AT, Van Naarden Braun K, Yeargin-Allsopp M. Cerebral palsy: classification and epidemiology. Phys Med Rehabil Clin N Am. 2009;20:425–452. doi: 10.1016/j.pmr.2009.06.001. [DOI] [PubMed] [Google Scholar]
  • 3.Koman LA, Smith BP, Shilt JS. Cerebral palsy. Lancet. 2004;363:1619–1631. doi: 10.1016/S0140-6736(04)16207-7. [DOI] [PubMed] [Google Scholar]
  • 4.Fowler EG, Staudt LA, Greenberg MB. Lower-extremity selective voluntary motor control in patients with spastic cerebral palsy: increased distal motor impairment. Dev Med Child Neurol. 2010;52:264–269. doi: 10.1111/j.1469-8749.2009.03586.x. [DOI] [PubMed] [Google Scholar]
  • 5.Damiano DL. Activity, activity, activity: rethinking our physical therapy approach to cerebral palsy. Phys Ther. 2006;86:1534–1540. doi: 10.2522/ptj.20050397. [DOI] [PubMed] [Google Scholar]
  • 6.Fukunaga T, Miyatani M, Tachi M, Kouzaki M, Kawakami Y, Kanehisa H. Muscle volume is a major determinant of joint torque in humans. Acta Physiol Scand. 2001;172:249–255. doi: 10.1046/j.1365-201x.2001.00867.x. [DOI] [PubMed] [Google Scholar]
  • 7.Zajac FE. How musculotendon architecture and joint geometry affect the capacity of muscles to move and exert force on objects: a review with application to arm and forearm tendon transfer design. J Hand Surg Am. 1992;17:799–804. doi: 10.1016/0363-5023(92)90445-u. [DOI] [PubMed] [Google Scholar]
  • 8.Manal K, Roberts DP, Buchanan TS. Can pennation angles be predicted from EMGs for the primary ankle plantar and dorsiflexors during isometric contractions? J Biomech. 2008;41:2492–2497. doi: 10.1016/j.jbiomech.2008.05.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Abe T, Kumagai K, Brechue WF. Fascicle length of leg muscles is greater in sprinters than distance runners. Med Sci Sports Exerc. 2000;32:1125–1129. doi: 10.1097/00005768-200006000-00014. [DOI] [PubMed] [Google Scholar]
  • 10.Narici M. Human skeletal muscle architecture studied in vivo by non-invasive imaging techniques: functional significance and applications. J Electromyogr Kinesiol. 1999;9:97–103. doi: 10.1016/s1050-6411(98)00041-8. [DOI] [PubMed] [Google Scholar]
  • 11.Jones EJ, Bishop PA, Woods AK, Green JM. Cross-sectional area and muscular strength: a brief review. Sports Med. 2008;38:987–994. doi: 10.2165/00007256-200838120-00003. [DOI] [PubMed] [Google Scholar]
  • 12.Mockford M, Caulton JM. The pathophysiological basis of weakness in children with cerebral palsy. Pediatr Phys Ther. 2010;22:222–233. doi: 10.1097/PEP.0b013e3181dbaf96. [DOI] [PubMed] [Google Scholar]
  • 13.Levitt S. Treatment of cerebral palsy and motor delay. 5th Edition. Oxford; Malden, Mass: Wiley-Blackwell; 2010. p. 356. [Google Scholar]
  • 14.Bandholm T, Magnusson P, Jensen BR, Sonne-Holm S. Dorsiflexor muscle-group thickness in children with cerebral palsy: relation to cross-sectional area. NeuroRehabilitation. 2009;24:299–306. doi: 10.3233/NRE-2009-0482. [DOI] [PubMed] [Google Scholar]
  • 15.Palisano R, Rosenbaum P, Walter S, Russell D, Wood E, Galuppi B. Development and reliability of a system to classify gross motor function in children with cerebral palsy. Dev Med Child Neurol. 1997;39:214–223. doi: 10.1111/j.1469-8749.1997.tb07414.x. [DOI] [PubMed] [Google Scholar]
  • 16.Fry N, Gough M, McNee A, Shortland A. Changes in the volume and length of the medial gastrocnemius after surgical recession in children with spastic diplegic cerebral palsy. Journal of Pediatric Orthopedics. 2007;27:769–774. doi: 10.1097/BPO.0b013e3181558943. [DOI] [PubMed] [Google Scholar]
  • 17.Shortland AP, Harris CA, Gough M, Robinson RO. Architecture of the medial gastrocnemius in children with spastic diplegia. Dev Med Child Neurol. 2002;44:158–163. doi: 10.1017/s0012162201001864. [DOI] [PubMed] [Google Scholar]
  • 18.Klejman S, Andrysek J, Dupuis A, Wright V. Test-retest reliability of discrete gait parameters in children with cerebral palsy. Arch Phys Med Rehabil. 2010;91:781–787. doi: 10.1016/j.apmr.2010.01.016. [DOI] [PubMed] [Google Scholar]
  • 19.Portney L, Watkins M, editors. Foundations of Clinical Research, Application to Practice. Stamford: Appleton & Lange; 1993. [Google Scholar]
  • 20.Ohata K, Tsuboyama T, Ichihashi N, Minami S. Measurement of muscle thickness as quantitative muscle evaluation for adults with severe cerebral palsy. Phys Ther. 2006;86:1231–1239. doi: 10.2522/ptj.20050189. [DOI] [PubMed] [Google Scholar]
  • 21.Hof A. Scaling gait data to body size. Gait & Posture. 1996;4:222–223. [Google Scholar]
  • 22.Blazevich A, Gill N, Zhou S. Intra-and intermuscular variation in human quadriceps femoris architecture assessed in vivo. J Anat. 2006;209:289–310. doi: 10.1111/j.1469-7580.2006.00619.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Akagi R, Kanehisa H, Kawakami Y, Fukunaga T. Establishing a new index of muscle cross-sectional area and its relationship with isometric muscle strength. J Strength Cond Res. 2008;22:82–87. doi: 10.1519/JSC.0b013e31815ef675. [DOI] [PubMed] [Google Scholar]
  • 24.Lieber R, Friden J. Functional and clinical significance of skeletal muscle architecture. Muscle Nerve. 2000;23:1647–1666. doi: 10.1002/1097-4598(200011)23:11<1647::aid-mus1>3.0.co;2-m. [DOI] [PubMed] [Google Scholar]
  • 25.Moreau NG, Simpson KN, Teefey SA, Damiano DL. Muscle Architecture Predicts Maximum Strength and Is Related to Activity Levels in Cerebral Palsy. Phys Ther. 2010 doi: 10.2522/ptj.20090377. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Stackhouse SK, Binder-Macleod SA, Lee SC. Voluntary muscle activation, contractile properties, and fatigability in children with and without cerebral palsy. Muscle Nerve. 2005;31:594–601. doi: 10.1002/mus.20302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Booth CM, Cortina-Borja MJ, Theologis TN. Collagen accumulation in muscles of children with cerebral palsy and correlation with severity of spasticity. Dev Med Child Neurol. 2001;43:314–320. doi: 10.1017/s0012162201000597. [DOI] [PubMed] [Google Scholar]
  • 28.Rose J, Haskell WL, Gamble JG, Hamilton RL, Brown DA, Rinsky L. Muscle pathology and clinical measures of disability in children with cerebral palsy. J Orthop Res. 1994;12:758–768. doi: 10.1002/jor.1100120603. [DOI] [PubMed] [Google Scholar]
  • 29.Moreau NG, Teefey SA, Damiano DL. In vivo muscle architecture and size of the rectus femoris and vastus lateralis in children and adolescents with cerebral palsy. Dev Med Child Neurol. 2009;51:800–806. doi: 10.1111/j.1469-8749.2009.03307.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Elder G, Kirk J, Stewart G, et al. Contributing factors to muscle weakness in children with cerebral palsy. Dev Med Child Neurol. 2003;45:542–550. doi: 10.1017/s0012162203000999. [DOI] [PubMed] [Google Scholar]
  • 31.Mohagheghi A, Khan T, Meadows T, Giannikas K, Baltzopoulos V, Maganaris C. Differences in gastrocnemius muscle architecture between the paretic and non-paretic legs in children with hemiplegic cerebral palsy. Clin Biomech (Bristol, Avon) 2007;22:718–724. doi: 10.1016/j.clinbiomech.2007.03.004. [DOI] [PubMed] [Google Scholar]
  • 32.Stafilidis S, Arampatzis A. Muscle - tendon unit mechanical and morphological properties and sprint performance. J Sports Sci. 2007;25:1035–1046. doi: 10.1080/02640410600951589. [DOI] [PubMed] [Google Scholar]
  • 33.Allen P, Jenkinson A, Stephens M, O'Brien T. Abnormalities in the uninvolved lower limb in children with spastic hemiplegia: the effect of actual and functional leg-length discrepancy. J Pediatr Orthop. 2000;20:88–92. [PubMed] [Google Scholar]
  • 34.Cimolin V, Galli M, Tenore N, Albertini G, Crivellini M. Gait strategy of uninvolved limb in children with spastic hemiplegia. Eura Medicophys. 2007;43:303–310. [PubMed] [Google Scholar]
  • 35.Lieber R, Murray W, Hentz V, Friden J. Biomechanical properties of the brachioradialis muscle: implications for surgical tendon transfer. Journal of Hand Surgery. 2005;30A:273–282. doi: 10.1016/j.jhsa.2004.10.003. [DOI] [PubMed] [Google Scholar]
  • 36.Knutson GA. Anatomic and functional leg-length inequality: a review and recommendation for clinical decision-making. Part I, anatomic leg-length inequality: prevalence, magnitude, effects and clinical significance. Chiropr Osteopat. 2005;13:11. doi: 10.1186/1746-1340-13-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Damiano D. Muscle size matters. Dev Med Child Neurol. 2009;51:415–419. doi: 10.1111/j.1469-8749.2009.03287.x. [DOI] [PubMed] [Google Scholar]

RESOURCES