Effects of quadriceps femoris muscle strengthening on crouch gait in children with spastic diplegia.

Background and Purpose. Despite evidence of weakness in children with cerebral palsy, the use of strength training in this population remains controversial. Subjects. Fourteen children with spastic diplegia, ranging in age from 6 to 14 years (X = 9.1, SD = 2.5), participated in a bilateral quadriceps femoris muscle strengthening program. Methods. The children exercised three times per week for 6 weeks using ankle weights at loads of approximately 65% of each child's maximum isotonic force production. The maximal voluntary contraction of the quadriceps femoris muscles was measured before, at the mid-point of, and immediately following the exercise program at 30, 60, and 90 degrees of knee flexion. Gait analyses were performed before and after the strengthening program to determine whether quadriceps femoris muscle strengthening influenced gait. Results. Children with spastic diplegia can increase quadriceps femoris muscle strength through heavy resistance exercise. Repeated-measures analysis of variance and multivariate analysis of variance procedures were used to assess changes in force and in gait variables. Improvement in the degree of crouch at initial floor contact at the freely selected speed and an increase in stride length at free and fast speeds were found. Conclusion and Discussion. These findings suggest that resistance exercise is an effective treatment strategy and as such should be considered as one component in the habilitation of children with cerebral palsy.

The impairment of voluntary motor control is the hallmark of cerebral palsy (CP). Cerebral palsy is defined as a nonprogressive insult to a developing or immature central nervous system (CNS), particularly to those areas that affect motor function.[1] Spastic diplegia, the most prevalent form of CP,[2] is characterized by motor incoordination, primarily in the lower extremities, that impairs many functional abilities, most notably ambulation.[3] The ability to walk is a major concern of the parents of children with CP, and improving or maintaining this ability is often considered to be the primary focus of most therapeutic interventions addressing the motor problems seen in children with spastic diplegia.[4] Although the majority of children with spastic diplegia are eventually able to ambulate,[2] the acquisition of this skill is delayed and differs qualitatively from normal pediatric gait.[5]

Abnormal gait patterns in spastic diplegia are common. The lesions associated with spastic diplegia may produce any or all of the following symptoms: (1) increased muscle tone, or a velocity-dependent resistance to passive muscle stretch, in synergistic muscle groups; (2) a loss of selective muscle control; (3) deficient equilibrium reactions; and (4) relative imbalance of muscle forces across the joints in the lower extremities.[1] Our study focuses on the symptom of muscle imbalance, which implies that the normal physiological relationship of muscle groups acting across a joint is altered, thereby affecting normal function at that joint.

Absolute muscle strength of antagonistic muscle groups is rarely equal because opposing muscles have different task requirements and are usually anatomically quite distinct.[6] Relative strength of antagonists acting across a given joint and total passive muscle excursion, however, are fairly consistent in normal joints,[6] and weakness, decreased excursion, or inappropriate muscle activity can distort this relationship. Evidence of imbalanced muscle function in spastic diplegic gait includes exaggerated hip flexion, adduction, and medial (internal) rotation in stance; excessive knee flexion or "crouch" during stance; and increased ankle equinus.[1] Overactivity of the hamstring muscles is a basic finding in the electromyographic (EMG) patterns during gait in children with spastic diplegia,[7] and these muscles' antagonistic action to the hip flexors and the knee extensors results in reduced sagittal-plane motion at the knee and hip.[1] Without intervention, and often despite intervention, muscle imbalance in the lower extremities is thought to become more pronounced over time. This is hypothesized to cause farther weakness, muscle atrophy, joint contractures, and eventual joint deformity.

A treatment commonly used to minimize the functional effects of knee joint contractures is a surgical lengthening of the distal hamstring tendon that, in effect, shifts the range of joint excursion to a more extended position at the knee.[8] Lengthening a musculotendinous unit, however, is thought to weaken the more dominant muscle,[9] which could prove problematic in children who am already weak.[10] We examined the effects of an additive approach to altering the balance of muscle forces at the knee by strengthening the quadriceps femoris muscles through resistance exercise. This is in contrast to a subtractive approach where there is weakening of the hamstring muscles through surgery.

Weakness and muscle imbalance have been identified in children with CP,[11-13] but these findings have not been quantified precisely and the etiologies of weakness and imbalance are still poorly understood. Berger and colleagues[14] presented descriptive EMG evidence of reduced force output in lower-extremity muscles during gait in children with CP that could not be accounted for merely by the children's slower speed of walking. These results should be interpreted cautiously, however, because differences in the magnitude of EMG signals across muscles within subjects or in the same muscle across subjects are not indicative of differences in force output.[15]

Inadequate power production in the affected lower extremity of children with hemiplegia, particularly in the ankle plantar-flexor and knee extensor muscles, has been reported by Olney and colleagues.[10] Perhaps the most compelling clinical evidence of weakness in spastic CP is the short-term postoperative outcome of children who underwent selective dorsal rhizotomies.[12,16] Reduction of "spasticity" did not restore normal motor control, but instead uncovered profound weakness.[12] Physiological evidence of weakness in CP also exists in that selective muscle fiber atrophy has been reported in children with spastic CP.[17] Bobath[18] proposed that hyperactivity of antagonistic muscles produces prolonged inhibition of the agonists during active movements, eventually producing weakness in those muscles. Although restraint by overactive antagonists can be present, diminished force output can still occur in its absence,[19] and support exists for a primary rather than a secondary dysfunction in the agonist.[20] The CNS lesion is static in CP, but clinical manifestations are not, and soft tissue contractures typically worsen over time. Ambulation becomes increasingly difficult, as indicated by increased energy costs, in children with primary muscle weakness as their growth outpaces their muscle strength.[11] A similar deterioration in gait with increasing age in persons with CP21 suggests that they are similarly weak.[22]

Strengthening exercises have been routinely used in persons with orthopedic problems and athletes to increase force production or minimize muscle imbalance; however, such exercises have not been advocated for persons with CP despite identification of muscle weakness and restricted force production in this population.[12,23-25] Contemporary motor control research seeks to examine movement on several different levels with the assumption that many different subsystems, of which neurophysiology is only one aspect, contribute to the development of motor control.[23] Gordon[26] suggests that physical therapists have concentrated too long on the reduction of positive symptoms in CNS disorders such as spasticity, while virtually ignoring the negative symptoms of weakness and loss of function.

Increasing dissatisfaction with current therapeutic approaches, both clinically and theoretically, has stimulated reevaluation of the premise that strengthening of muscles shown to be weak is not appropriate in adults and children with CNS pathology.[23,27,28] After reviewing the scientific literature, Guiliani[12] uncovered no evidence that strengthening exercises are detrimental in this population, and the literature indicated that these exercises may be beneficial.[13,29]

Although the number of studies is very limited, strength has been shown to increase in CP through resistance exercise programs.[24,30,31] Healy[30] compared isometric and isotonic exercise regimens in five boys with spastic diplegia, with each child performing a different type of resistive exercise on each leg. Comparable improvements were noted in strength and range of motion for both types of exercise. McCubbin and Shasby[31] examined the rate of torque development and the speed of movement of elbow extension in adolescents with CP who had participated in a 6-week program of isokinetic training versus a program of repetitive exercise in which subjects performed the same type and amount of motion but without resistance. The authors found improvements in torque production and speed only in the isokinetic training group. They further noted that the effects of isokinetac training on neuromuscular performance of children with CP were similar to those found in children without CP. Horvat[24] compared free weights and isokinetic training of equivalent intensity in an individual with CP and found that each method produced similar torque gains that were greater than those found in persons without CP and similar to those seen in deconditioned persons or those with primary muscle weakness.

In summary, muscle weakness and imbalance are frequent clinical findings in children with spastic diplegia. Not only have children with CP been found to be weak, but reduced strength of the knee extensors has been shown to be related to diminished functional capacity in adolescents with CP, as evidenced by lower scores on the Gross Motor Function Measure and increased energy expenditure during gait in the weaker children.[25] Furthermore, it is possible to increase muscle strength in these children through traditional strength training programs. A causal link between resistance training and improved gross motor function, however, has not previously been established and was addressed by our study. More specifically, the contribution of quadriceps femoris muscle weakness to the degree of crouch during gait has not been well documented and remains unresolved.[32] Sutherland and Cooper[33] propose that the two principles of treatment to overcome crouch gait are to reduce the knee flexion contracture and to restore quadriceps femoris muscle strength. The central hypothesis in our study was that a quadriceps femoris muscle strength training program in the absence of knee flexion contractures would produce improvements in gait performance by reducing the degree of knee crouch during stance.

Method

Subjects

The participants in this study were 14 children with spastic diplegia, 10 boys and 4 girls, ranging in age from 6 to 14 years (X = 9.1, SD = 2.5). Two of the children required a posterior walker to ambulate, and 1 child used bilateral quad canes. The rest of the children were independent ambulators. Letters describing the study were mailed to the families of all children followed at the Kluge Children's Rehabilitation Center (Charlottesville, Va) at the time of the study, and interested families then contacted the Motion Analysis Laboratory to receive further information or volunteer for participation. Volunteers were then screened for the existence of crouch gait as verified on the pretraining gait analysis, and any child who demonstrated less than 10 degrees of knee flexion at initial floor contact was excluded from the study. Any child without full passive knee extension in a supine position was also excluded from the study. Previous lower-extremity surgery, such as prior hamstring muscle lengthening, did not preclude participation, nor did the existence of static hip or ankle deformities. In Table 1, each participant is listed and described by age, gender, need for assistive devices, and prior surgical history.

[TABULAR DATA OMITTED]

Procedure

The initial assessment consisted of isometric strength testing and gait analysis. Strength is classically defined as the ability to produce isometric force,[34] and a hand-held dynamometer(*) was used to quantify force in the quadriceps femoris and hamstring muscle groups bilaterally. Subjects were seated in a straight-backed chair with armrests, which they were encouraged to use to maximize stability. For the isometric strength tests of the quadriceps femoris muscles, the knee was positioned at 90, 60, and 30 degrees of knee flexion. These positions were chosen so that force could be assessed at different muscle lengths. The 90-degree position was needed for comparison with similar studies that used this as the test position for quadriceps femoris muscle force measurements. The 60-degree position was chosen because it is thought to the position in which there is an optimal length-tension relationship for generating force in the quadriceps femoris muscles.[35] We used the 30-degree position to yield a force value nearer to full knee extension, which could hypothetically affect or be affected by persistent knee crouch. Full extension force was not tested because of the possible confounding factor of mechanical locking of the knee, which could distort the contribution of muscle strength at that position.

An electrogoniometer was taped to the lateral aspect of the leg prior to these measurements to verify and maintain the angle position during the force testing. Hand-held dynamometers, when used correctly, have documented reliability in children[36] and in persons with neuromotor deficits.[37] Stabilization of the child, the supporting surface, and the examiner is necessary to ensure that no motion of the lower limb occurs during the testing so that force values can be accurate and consistent. A single physical therapist familiar with the device took all force measurements. The maximal voluntary contraction (MVC) was measured at each position, with two trials administered and the larger of the two measurements recorded. Instructions regarding the testing procedure and strong verbal encouragement during the trials were used to produce maximal effort, and in the rare instances When children did not appear to comprehend or comply as evidenced by lack of increased physical effort in the trunk or upper extremities during the maximal effort, these instructions were repeated and the trials were readministered. During the 6-week training program, quadriceps femoris muscle force was assessed weekly at 90 degrees only, except for the midpoint of the training when the quadriceps femoris muscle force was also measured at 30 and 60 degrees of flexion. A potential adverse effect of weight training in persons with spasticity is the strengthening of antagonistic muscle through co-contraction; therefore, hamstring muscle force was measured before and after the training program in the supported sitting position as described for testing with the knee flexed to 90 degrees.

Kinematic or joint-angle displacement data were captured at 60 Hz using a three-dimensional ExpertVision[TM] system provided by the Motion Analysis Corporation (Santa Rosa, Calif).(dagger) A set of 15 retroreflective body surface markers were taped over the sacrum and symmetrically on both lower extremities, including the anterior superior iliac spines, knee joint centers, lateral malleoli, between the second and third metatarsal heads, and the heels. Nonlinear "outrigger" markers were attached to the lateral aspect of the thighs and calves with Velcro[R](double dagger] bands. The system consisted of four NEC CCD cameras(sections) located in the four comers of the Motion Analysis Laboratory in order to capture the trajectories of the body surface markers in the sagittal, transverse, and coronal planes. Once markers were placed, each child walked barefoot along a carpet strip in the center of the room. Data were collected for 5 seconds per trial at 60 samples per second. Each child completed three trials at a freely selected speed and three trials at a fast speed (the child was instructed to "walk as fast as possible without running"). The trajectories of the hip, knee, and ankle were digitized with a video image processor (VP320),(dagger) and a Sun Microsystems 3/110 Workstation(parallel) was utilized as the central controller for data procurement and processing. For each 5-second trial collected, a single gait cycle consisting two consecutive foot strikes for each leg was extracted for analysis.

The kinematic or joint-angle displacement data were then transferred to three-dimensional coordinates using the ExpertVision[TM] software. After the data were in the proper format, Orthotrak software[dagger] was used to calculate joint angles and produce printouts of joint motion at the pelvis and the right and left hip, knee, and ankle across a representative gait cycle. The pattern of motion of a specific joint at each point in the gait cycle as well as the total excursion of a joint during an entire cycle can be determined from these printouts. Temporal and spatial information was also calculated for a specified gait cycle, including gait speed, stride length, and cadence.

Strength Training Protocol

Each subject participated in a 6-week, bilateral, progressive resistance quadriceps femoris muscle strengthening program at a frequency of three times per week. For the training, each subject was seated in a straight-backed chair with the hip in at least 90 degrees of flexion, the knee joint just anterior to the front edge of the chair with a small cloth roll under the knee to position it in 90 degrees of flexion, and the foot hanging free above the supporting surface. Velcro[R]-attached ankle weights were used in the program. To determine the maximum weight that a child could lift through his or her full passive range of motion without fatiguing, an approximation was first obtained by measuring the MVC of the quadriceps femoris muscle at 90 degrees of knee flexion with a hand-held dynamometer and then doubling this value. This measure was based on the finding that isometric force typically underestimates maximum isotonic strength by half, as determined by one-repetition maximum effort.[38] The child then attempted to lift a weight of 65% of this load, and if the child could successfully lift this amount of weight throughout the full active range of motion, that became the training weight for the week. The MVC was reassessed weekly for each subject, and the training weight was adjusted proportionally with gains in force production.

Each child performed quadriceps femoris muscle strengthening exercises 3 days per week with at least 1 day of rest between sessions. Prior to beginning a weight training session, each child first performed a 5-minute warm-up consisting of lower-extremity stretching exercises followed by a brief walk around the room if that child had been sitting for more than 15 minutes immediately prior to the exercise session. Once properly positioned, the weights were attached, and the child fully extended one knee at a time while lifting 65% of the one-repetition maximum as determined for that leg. The child was instructed to slowly lower the leg back to the starting position so that the quadriceps femoris muscles performed concentric and eccentric muscle activity. The child performed four sets of five repetitions per set, resting a minute between sets, with the weighted leg fully supported during rest periods. The child then repeated the same procedure for the opposite leg. The training continued for 6 consecutive weeks. The strength training protocol utilized in this project was a progressive resistance program in which the percentage of maximum was fixed throughout the study, but the actual weight lifted increased proportionately with gains in isometric strength. The protocol was established based on the strength training literature.[34,39,40] The load intensity of 65% has been cited as the minimum value needed to produce strength force gains,[39] repetitions should be kept within single figures to minimize fatigue,[40] and a frequency of three to five exercise sessions per week for 4 weeks has been shown to be effective in producing strength gains.[34]

Each subject was evaluated weekly by the physical therapist, at the Motion Analysis Laboratory, at the child's home, or at school as decided by the parents to measure the MVC of the quadriceps femoris muscle group using the hand-held dynamometer, to readjust the one-repetition maximum, and to monitor the training program. All subjects, assisted by their families, kept a log of each of the exercise sessions, including the amount of weight lifted per leg, the number of sets and repetitions completed, and any additional subjective comments. The children were instructed to alternate the starting leg each session.

All 14 children who agreed to participate in the program completed the entire 6-week program. Twelve of the subjects completed all 18 exercise sessions, and 2 children missed 1 session. Six of the 18 exercise sessions for each child were conducted by the same physical therapist, and the rest were supervised by family members or school therapists. As instructed, none of the children altered their normal physical activities during participation in the program. Seven of the children were enrolled in weekly or bimonthly physical therapy before and during the program, and their strength results did not differ significantly from those of the children who were not currently receiving physical therapy services. None of the children who were receiving physical therapy performed any type of resistance exercise in their regular therapy sessions.

Data Analysis

A factorial repeated-measures analysis of variance (ANOVA) procedure was used to determine the effect of a strengthening program on increasing quadriceps femoris muscle force, as measured at three different angles of knee flexion at three assessment times. The statistical package used for all analyses was the SPSS/PC+(#) for IBM-compatible computers. This software package uses a multivariate analysis of variance (MANOVA) program for this type of analysis, as shown in Table 2. Multiple repeated-measures ANOVA and MANOVA procedures were also used to compare statistically the mean changes in hamstring muscle force; time and distance variables such as speed, stride length, and cadence; and kinematic data, including knee angle at floor contact and mid-stance and total hip and knee excursion as a result of the weight training program (P<.05). Follow-up Scheffe F tests were performed when indicated to identify which variables showed significant differences.

Table 2. Multivariate F Tests for
Effect of Measurement Times and Angle
of Knee Flexion on Quadriceps Femoris
Muscle Force

Effect            df           F

Right leg
  Angle           2,12   80.44(a,b)
  Times           2,12   93.93(a,b)
  Angle X times   4,10    2.04(c)
Left leg
  Angle           2,12   26.76(a,b)
  Times           2,12   23.40(a,b)
  Angle X times   4,10    3.43(c)

(a) P<.001.
(b) Critical F = 3.89.
(c) Critical F = 3.48.

Results

Means and standard deviations for quadriceps femoris and hamstring muscle force for each participant are presented in Table 3. Statistical analyses revealed force values differed among angles and that force increased across at least one of the assessment intervals, as shown in Table 2. The interaction effect of angles X times was not significant for either leg, indicating that the pattern of change in force was similar across the different angles of knee flexion that were tested. No changes were found m the hamstring muscle force measurements over the intervention period for the right and left legs. Follow-up univariate F tests were performed to determine more specifically where force changes occurred (Tab. 4).

Table 4. Univariate Post Hoc Scheffe
F Tests for Determining Significance of
Changes in Quadriceps Femoris Muscle
Force Measurements Before, at the
Midpoint of, and After the Strength
Training Program

                              F

Comparison                    Right       Left

90[degrees]
  Pretraining-midtraining     2.72(a)     1.50
  Midtraining-posttraining    1.04        0.75
  Pretraining-posttraining    7.14(a,b)   3.76

60[degrees]
  Pretraining-midtraining     7.04(a,b)   2.94(a)
  Midtraining-posttraining    0.61        0.74
  Pretraining-posttraining   11.80(a,b)   6.62(a,b)

30[degrees]
  Pretraining-midtraining     7.94(b)     4.45
  Midtraining-posttraining    0.53        0.20
  Pretraining-posttraining   12.58(a,b)   6.50(a,b)

(a) P<.05.
(b) Critical F=4.67.

In summary, force gains in the right and left quadriceps femoris muscles were achieved as a result of the 6-week strength training program at all three angles of knee flexion. Except for the pretraining to posttraining comparison of the 90-degree measurements on the left, an increase in quadriceps femoris muscle force was found at the completion of the 6-week program. At all three angles, gains were more pronounced in the first 3 weeks than in the last 3 weeks of the program. Neither the right nor left hamstring muscles increased in force production as a result of the quadriceps femoris muscle strengthening program.

Crouch gait is manifested by increased knee flexion at initial floor contact, with maintenance of a flexed knee posture throughout the stance phase of gait. To determine whether the degree of crouch improved to a more extended position after the training program, the amount of flexion at floor contact and the maximum value of knee extension attained during stance were obtained for each subject across the two walking speeds at the preassessment and postassessment periods (Tab. 5).

[TABULAR DATA OMITTED]

The mean knee flexion at floor contact was 32 degrees at the free speed and 34 degrees at the fast speed prior to the strength training. After the strength training program, 10 of the 14 children had less knee flexion at floor contact at the free speed, with a mean decrease of more than 5 degrees (Tab. 6).

Table 6. Repeated-Measures Analysis of Variance Summary for Effect
of Training Program on Knee Flexion at Floor Contact for
Free Speed

Source             df   SS      MS   F       P

Between subjects   13   3636   280   5.63   .0014
Within subjects    14    696    50
  Training          1    297   207   5.0    .0355(a)
  Residual         13    489    38
Total              27   4332

(a) P<.05.

The mean maximum amount of knee extension achieved during mid-stance prior to strengthening was 13 degrees of flexion at the free speed and 14 degrees of flexion at the fast speed and decreased to 11 and 12 degrees, respectively, after the training program, neither of which was statistically significant. Although no mean change was found, the individual responses to this variable produced some interesting results. The 4 children who demonstrated no reduction in the degree of crouch at initial floor contact at the free speed also showed no improvement in maximum extension in mid-stance. Of the remaining 10 subjects with improved crouch at initial contact, 7 showed improved knee extension in stance with a mean reduction of 9 degrees of knee flexion, and 2 showed essentially no change. one undesir-able result for 5 of the 10 children was the development of mild knee hyperextension or the exaggeration of an already hyperextended knee position during the mid-stance. Three of the 5 children developed hyperextension of less than 3 degrees. Only 1 of the 5 children had undergone a prior hamstring muscle lengthening, and her degree of recurvatum after the training was the most severe. Four of the 5 children who subsequently developed recurvatum lacked 10 degrees or less of knee extension at mid-stance prior to the strengthening program, although the initial degree of crouch at floor contact was greater than 10 degrees in those 5 subjects. Considering the 2 children who had the greatest improvements in crouch at initial contact, 1 child had less than 5 degrees of knee flexion at mid-stance initially and remained in the normal range after the program. The other child shifted the entire knee sagittal-plane curve toward greater extension, with an equivalent reduction in flexion at both floor contact and mid-stance.

Total hip and knee excursion in the sagittal plane throughout the gait cycle was also measured before and after the program, and the means and standard deviations are listed in Table 5. Even though slight increases in the amount of excursion were found across speeds and assessment times, these changes were not statistically significant. Speed, stride length, and cadence measurements obtained in each trial were averaged to produce a single value per subject for each variable at free and fast speeds, respectively (Tab. 5). Multiple repeated-measures ANOVA procedures performed on the time and distance variables across speeds indicate that the changes in speed and cadence within each of the speeds were not significant. Mean stride length increased by 8 cm in the free-speed condition and by 14 cm in the fast-speed condition (Tab. 7).

[TABULAR DATA OMITTED]

Discussion

The purpose of this study was to explore the use of muscle strengthening as a potential therapeutic option for children with spastic CP. The results of this study verify that children with CP were able to increase isolated muscle strength through a resistance training program. All of the children who participated in the study had consistent and proportionately large increases in quadriceps femoris muscle force production without a concurrent increase in hamstring muscle force production. As mentioned earlier, one of the supposed contraindications for the use of strength training in children with CNS involvement is that it increases the recruitment of other muscles that are already overactive and could inadvertently strengthen those muscles at the same time. The lack of an increase in hamstring muscle force in this group of children with spastic diplegia after heavy resistance training provides evidence negating this assumption.

This research reconfirms the finding that children with CP have the capacity to increase force production of their quadriceps femoris muscles with an exercise program of relatively short duration. The magnitude of the force increases is greater than that normally expected for children without CP, and these force increases are consistent with those described for children with primary muscle weakness.[36] There was improved knee extension at initial floor contact when walking at a self-selected free speed for 10 of the 14 children studied. Maximum extension during stance did not change, although the finding that nearly all of the children with decreased crouch at floor contact also had greater extension in mid-stance suggests a possible effect. Two of the children in this study, subjects 2 and 5, shifted from a forefoot strike to a true heel-strike after the strengthening program, and their improvement in knee extension at floor contact was approximately 15 degrees for each child. A possible explanation is that greater quadriceps femoris muscle force enabled this muscle group to counteract the hamstring muscles as they acted to decelerate the thigh in later swing. The resulting increase in knee extension elongated the stride, decreased the flexion angle in terminal swing, and unproved positioning of the foot prior to floor contact. Kinematic results for fast walking showed no change in crouch at either point in the gait cycle. The most likely explanation for the different results at free and fast speeds can be deemed by looking at the individual responses to fast walking trials where positive and negative changes effectively cancel each other. For some children, the gait pattern deteriorates or moves further from normal kinematic patterns when the children are challenged by being asked to walk faster, whereas other children more closely approximate normal when they walk faster.

Although mean changes in maximum extension were not found, the fact that several children developed knee hyperextension in mid-stance may be of clinical importance. Knee hyperextension is also a potential complication of surgical hamstring muscle lengthenings, and because this procedure shifts the knee motion curve to a more extended position throughout the cycle, perhaps quadriceps femoris muscle strengthening alone may be contraindicated in those children without limitations in knee extension during mid-stance. Recent evidence of proportionately greater weakness in the hamstring muscles than in the quadriceps femoris muscles in children with CP[41] suggests that perhaps these muscles should be strengthened at the same time.

The changing interrelationship between stride length and cadence after the training program was particularly interesting. Because both stride length and cadence are linearly related to speed in children without CP, both variables would be expected to increase concurrently. After the strengthening program, the increase in speed from the pretraining fast-speed condition to the posttraining fast-speed condition was accomplished by an increase in stride length, with a relative decrease in cadence, even though the speed at the posttraining time was slightly higher. Typically, children with spastic diplegia have restricted stride lengths as well as diminished ability to alter their stride length to maintain or increase their speed,[42] so they overcompensate by increasing their cadence. Increased force enables them to take longer strides, so they can begin to relax their exaggerated cadence.

Although mean force gains of more than 50% over initial values were produced, a similarly large improvement in gait was not observed in this study for all participants. For example, the mean decrease in crouch at the free speed was only 5 degrees. Possible explanations include the fact that the duration of the program was relatively short, targeted only one muscle group, and on completion all children did not attain normal force values.[41] Another potential explanation is that although quadriceps femoris muscle weakness is identified in these children, and may more severely limit some of their other motor capabilities such as rising from a chair or stair climbing, less force may be required to ambulate on level surfaces. Yet another possibility is that more stringent patient selection criteria might have produced more definitive results by identifying only those children who demonstrated crouch as a result of hamstring muscle overactivity and limited flexibility in the absence of a fixed contracture. Sussman[32] defined crouch gait as the lack of knee extension during stance regardless of foot position, but hip flexion contractures and excessive ankle dorsiflexion can exacerbate or initiate knee crouch. Therefore, eliminating those children with other factors that may cause crouch secondarily would have potentially strengthened our study.

Weakness, like spasticity, is a concurrent symptom of a CNS lesion and is not a primary cause of the motor dysfunction seen in these children. Therefore, alleviating only one of the symptoms may have a minimal effect on motor function. In our study, children with the least motor involvement initially appeared to show the greatest degree of improvement in their gait patterns, whereas those with more obvious gait limitations generally improved very little in comparison. This result seemed counterintuitive at first, but a logical explanation may exist. We believe there are three basic prerequisites for normal movement: range of motion within normal limits, adequate force production, and normal motor control. If any one of these is deficient, the quality of the motor function is compromised. Perhaps the major problem in the children with minimal motor involvement was centrally mediated muscle imbalance that produced weakness secondarily, although they demonstrated fairly good selective motor control. Increasing force production of the quadriceps femoris muscles in these children appeared to produce spontaneous improvements in their gait. In the children with greater gait defects, although weakness was identified and could be improved, their lack of selective motor control may have interfered with their ability to utilize these muscles for functional activities. More effective remediation of the motor deficits seen in CP must address motor control problems as well as weakness and static contractures.

We measured multiple variables, and the possibility exists that some of the findings, particularly in the gait variables, might have occurred by chance. A MANOVA incorporating all of the gait variables was performed, and the variables that emerged as significant in the univariate ANOVAs and the MANOVAs on subgroups of variables also demonstrated very low probability values on these analyses, barely above the arbitrary cutoff level of P<.05. Given that the statistical analyses were performed on a small, variable sample, effects needed to be quite large to reach significance on multivariate analyses. Therefore, the chances of finding no difference when in reality a difference exists appeared more likely than uncovering nonexistent differences, which provided the rationale for the types of analyses reported here.

Conclusion

A complex multifaceted disorder such as CP most likely requires a multifaceted approach to rehabilitation. Factors that can be remedied need to be identified, and more effective strategies for producing positive functional outcomes must be developed. Our study focused on quadriceps femoris muscle weakness as one component of the motor dysfunction seen in children with CP, and utilizing a traditional orthopedic approach rarely recommended with this population demonstrated clinical improvement in strength and ambulatory ability. Although strengthening is an effective treatment option in CP, more evidence of functional improvements due to strengthening that have a significant impact on the quality of life of these children needs to be produced through the use of motion analysis or other objective outcome measures prior to widespread advocation of this approach in the clinic.

(*) Chatillon, 7609 Business Park Dr, Greensboro, NC 27409. (dagger) Motion Analysis Corp, 93 Stony Cir, Santa Rosa, CA 95401. (double dagger) Velcro USA Inc, 406 Brown Ave, Manchester, NH 03108. (sections) NEC Corp, NEC Bldg, 33-1, Shiba 5-chrome, Minato-ku, Tokyo 108, Japan (distributed by NEC Corp, BRoadcas Equipment Dept, 1555 w Walnut Hill Ln, Irving, TX 75038). (parallel) Sun Microsystems Inc, 2550 Garcia Ave, Mountain View, CA 94043. (#) SPSS, 444 N Michigan Ave, Chicago, IL 60611.

References

[1] Gage JR. Gait Analysis in Cerebral Palsy. Oxford, England: Mac Keith Press; 1991. [2] Watt J, Sims D, Harckham F, et al. A prospective study of inhibitive casting as an adjunct to physiotherapy for cerebral-palsied children. Dev Med Child Neurol. 1986;28:480-488. [3] Binder H, Eng GD. Rehabilitation management of children with spastic diplegic cerebral palsy. Arch Phys Med Rehabil. 1989;70:482-489. [4] Bleck EE. Management of the lower extremities in children who have cerebral palsy. j Bone Joint Surg [Am]. 1990;72:140-144. [5] Rosenbaum PL, Russell DJ, Cadman DT, et al. Issues in measuring change in motor function in children with cerebral palsy: a special communication. Phys Ther. 1990;70:125-131. [6] Silver RL, de la Garza J, Rang M. The myth of muscle balance. J Bone Joint Surg [Br]. 1985;67:432-437. [7] Perry J, Newsam C. Function of the hamstrings in cerebral palsy. In: Sussman MD, ed. Diplegic Child.. Evaluation and Management. Rosemont, Ill: American Academy of Orthopaedic Surgeons; 1992:299-308. [8] Thometz J, Simon S, Rosenthal R. The effect on gait of lengthening the medial hamstrings in cerebral palsy. J Bone Joint Surg [Am]. 1989;71:345-353. [9] Moseley CF. Physiologic effects of soft-tissue surgery. In: Sussman MD, ed. Diplegic Child: Evaluation and Management. Rosemont, Ill: American Academy of Orthopaedic Surgeons; 1992:259-270. [10] Olney SJ, MacPhail HEA, Hedden DM, Boyce WF. Work and power in hemiplegic cerebral palsy gait. Phys Ther. 1990;70:431-438. [11] Bar-Or O. Pediatric Sports Medicine for the Practitioner. New York, NY: Springer-Verlag New York Inc; 1983. [12] Guiliani CA. Dorsal rhizotomy for children with cerebral palsy: support for concepts of motor control. Phys Ther, 1991;71:248-259. [13] Leonard EL. Early motor development and control: foundations for independent walking. In: Smidt GL, ed. Gait in Rehabilitation. New York, NY: Churchill Livingstone Inc; 1990: 121-140. [14] Berger W, Quintern J, Dietz V. Pathophysiology of gait in children with cerebral palsy. Electroencephalogr Clin Neurophysiol. 1982;53: 538-548. [15] Perry J. Gait Analysis: Normal and Pathological Function. New York, NY: Slack Inc; 1992. [16] Oppenheim ML, Staudt LA, Peacock WJ. The rationale for rhizotomy. In: Sussman MD, ed. Diplegic Child: Evaluation and Management. Rosemont, Ill: American Academy of Orthopaedic Surgeons; 1992:271-286. [17] Castle ME, Reyman TA, Schneider M. Spasticity of spastic muscle in cerebral palsy. Clin Orthop. 1979;142:223-233. [18] Bobath B. Adult Hemiplegia: Evaluation and Treatment. London, England: William Heinemann Medical Books Ltd; 1978. [19] Knutsson E, Martensson A. Dynamic motor capacity in spastic paresis and its relation to prime mover dysfunction, spastic reflexes and antagonist co-activation. Scand J Rehabil Med. 1980;12:93-106. [20] Sahrmann SA, Norton BJ. The relationship of voluntary movements to spasticity in the upper motor neuron syndrome. Ann Neurol. 1977;2:460-465. [21] Campbell J, Ball J. Energetics of walking in cerebral palsy. Orthop Clin North Am. 1978;9: 374-377. [22] Rose J, Gamble JG. Human Walking. 2nd ed. Baltimore, Md: Williams & Wilkins; 1994. [23] Corcos DM. Strategies underlying the control of disordered movement. Phys Ther. 1991; 71:25-38. [24] Horvat M. Effects of a progressive resistance training program on an individual with spastic cerebral palsy. American Corrective Therapy Journal. 1987;41:7-11. [25] Kramer JF, MacPhail HEA. Relationships among measures of walking efficiency, gross motor ability, and isokinetic strength in adolescents with cerebral palsy. Pediatric Physical Therapy. 1994;6(1):3-8. [26] Gordon J. Assumptions underlying physical therapy intervention: theoretical and historical perspectives. In: Carr JH, Shepherd RB, eds. Movement Science: Foundations for Physical Therapy in Rehabilitation. London, England: Heinemann Physiotherapy; 1987: 1-30. [27] Bohannon RW. Is the measurement of muscle strength appropriate in patients with brain lesions? a special communication. Phys Ther. 1989;69:225-236. [28] Bourbonnais D, Vanden Noven S. Weakness in patients with hemiparesis. Am J Occup Ther. 1989;43:313-319. [29] Kolobe THA. The use of upper extremity proprioceptive neuromuscular facilitation techniques with children with spastic diplegia. Pediatric Physical Therapy. 1990;1:186. [30] Healy A. Two methods of weight training for children with spastic type of cerebral palsy. Res Q. 1957;29:389-395. [31] McCubbin JA, Shasby GB. Efects of isokinetic exercise on adolescents with cerebral palsy. Adapted Physical Activity Quarterly. 1985;2:56-64. [32] Sussman MD, ed. Diplegic Child: Evaluation and Management. Rosemont, Ill: American Academy of Orthopaedic Surgeons; 1992. [33] Sutherland DH, Cooper L. The pathomechanics of progressive crouch gait in spastic diplegia. Orthop Clin North Am. 1978;9:143-154. [34] Atha J. Strengthening muscle. Exerc Sport Sci Rev. 1981;9:1-74. [35] Kues J, Rothstein JM, Lamb RL. The relationships among knee extensor torques produced during maximum voluntary contractions under various test conditions. Phys Ther. 1994;74:674-683. [36] Hoskings GP, Bhat US, Dubowitz V, Edwards RHT. Measurements of muscle strength and performance in children with normal and diseased muscle. Arch Dis Child. 1976;51:957-963. [37] Riddle DL, Finucaine SD, Rothstein JM, Walker ML. Intrasession and intersession reliability of hand-held dynamometer measurements taken on brain-damaged patients. Phys Ther. 1989;69:182-194. [38] Nunn KD, Mayhew JL. Comparison of three methods of measuring strength imbalances at the knee. J Orthop Sports Phys Ther. 1988;10:134-137. [39] Berger RA. Comparison of the effects of various weight training loads on strength. Res Q. 1965;36:141-146. [40] Berger RA. Optimum repetitions for strength development. Res Q. 1961;33:334-338. [41] Abel MF, Damiano DL. Strategies for increasing walking speed in cerebral palsy. J Pediatr Orthop. In press. [42] Damiano DL, Vaughan CL, Abel MF. Muscle response to heavy resistance exercise in children with spastic diplegia. Dev Med Child Neurol. In press.

DL Damiano, PhD, PT, is Assistant Professor of Orthopaedics, University of Virginia, Charlottesville, VA 22903. Address all correspondence to Dr Damiano at Motion Analysis Laboratory, Kluge Children's Rehabilitation Center, University of Virginia, 2270 Ivy Rd, Charlottesville, VA 22903 (USA) (DLD2S @ Virginia.edu).

LE Kelly, PhD, is Associate Professor of Education, University of Virginia.

CL Vaughan, PhD, is Professor of Orthopaedics and Biomedical Engineering, University of Virginia.

This study was approved by the Human Investigation Committee, University of Virginia.