Spastic Paretic Stiff-Legged Gait Joint Kinetics

Authors: D. Casey Kerrigan, MD, MS Mark E. Karvosky, MA Patrick O. Riley, PhD Affiliations: From the Department of Physical Medicine and Rehabilitation (DCK, POR), Harvard Medical School, and Spaulding Rehabilitation Hospital Center for Rehabilitation Science (DCK, MEK, POR), Boston, Massachusetts. Disclosures: Supported, in part, by Public Health Service grant NIH HD01071-05 and the Ellison Foundation. Reprints: All correspondence and requests for reprints should be addressed to D. Casey Kerrigan, MD, MS, Department of Physical Medicine and Rehabilitation, Harvard Medical School, 125 Nashua Street, Boston, MA 02114. 0894-9115/01/8004-0244/0 American Journal of Physical Medicine & Rehabilitation Copyright 2001 by Lippincott Williams & Wilkins Research Article Gait Spastic Paretic Stiff-Legged Gait Joint Kinetics ABSTRACT Kerrigan DC, Karvosky ME, Riley PO: Spastic paretic stiff-legged gait: joint kinetics. Am J Phys Med Rehabil 2001;80:244 249. Objective: The authors previously suggested that spastic paretic stiff-legged gait, defined as reduced knee flexion in swing associated with upper-motor neuron injury, can be attributed to multiple impairments besides spastic quadriceps activity. This study hypothesizes that subjects with spastic paretic stiff-legged gait have altered kinetics not only about the knee but also about the hip and ankle. Design: Joint kinetic data of 20 subjects with spastic paretic stifflegged gait caused by stroke were compared with data obtained from 20 able-bodied subjects. Results: Significant reductions in the subject group were found in both peak knee-joint power absorption (0.42 0.34 vs. 0.99 0.27 W/(kg m m/sec)) and peak ankle-joint power generation (0.74 0.42 vs. 1.51 0.17 W/(kg m m/sec); both P 0.0001). The authors observed increases in peak external-hip flexion torque in stance, hip-power generation in loading response, knee-extension torque in midstance, ankle-dorsiflexion torque, and ankle-power absorption in stance. There was substantial variability in most torque and power values among subjects, which was significantly greater than that observed in the control subjects. Conclusions: These findings, in conjunction with previous studies, support the likelihood of multiple mechanisms for reduced knee flexion in swing. Alternatively, some of the joint kinetic differences could be compensations for or associated with reduced knee flexion in swing. The substantial variability among subjects implies that despite a similar visual appearance of reduced knee flexion among subjects with a spastic paretic stiff-legged gait pattern, each individual has unique mechanisms associated with this observed gait pattern. Key Words: Gait, Biomechanics, Hip Joint, Knee Joint, Ankle Joint, Muscle Spasticity, Paresis 244 Am. J. Phys. Med. Rehabil. Vol. 80, No. 4

Reduced knee flexion during swing is common in patients with spastic paretic gait as a result of upper-motor neuron injury including stroke, cerebral palsy, traumatic brain injury, and spinal cord injury. 1 6 This gait pattern, referred to as stiff-legged gait, historically has been attributed solely to spastic quadriceps activity. 1 3, 6 Treatment of spastic paretic stiff-legged gait has relied on this assumption and has been aimed principally at reducing quadriceps activity through physical modalities, nerve/motor point blocks, 7 or surgical release or transfer of quadriceps muscles. 1, 2, 6, 8 We have entertained other potential causes of reduced knee flexion beyond that of abnormal function of the quadriceps, including impaired dynamic hip flexion 3, 9, 10 and impaired ankle control during gait. 3, 5 Kerrigan et al. 3 demonstrated relationships between reduced knee flexion and inappropriate hamstring activity and between reduced knee flexion and delayed heel off. Kerrigan et al. 9 showed that in some cases, simulating an increase in hip flexion produced in the model substantial improvement in knee flexion. Riley and Kerrigan 10 confirmed the importance of hip flexion in stiff-legged gait using a more sophisticated modeling approach, and also described complex contributions of the rectus femoris and hamstrings to the model in improving knee flexion but also limiting hip flexion. Riley and Kerrigan 13 demonstrated, using a relatively new induced-acceleration analysis technique, 12 that torques about the ankle as well as about the hip and knee affect knee flexion. Moreover, Kerrigan et al. 14 noted a mean 17 reduction in peak knee flexion, when ablebodied subjects were asked to walk on their toes compared with normal heel-toe walking; this observation suggested that foot ankle-function is related in some way to knee flexion in swing. In this study, we sought to examine measured joint kinetics (joint torques and powers) in a group of subjects with spastic paretic stifflegged gait. The study of joint kinetics, which to our knowledge has not been performed in a population of subjects with stiff-legged gait, is gaining broad acceptance as a way to provide greater insight into abnormal gait patterns and is increasingly being used in clinical rehabilitation settings to improve treatment. 15 20 We hypothesized that joint kinetic alterations in spastic paretic stiff-legged gait are confined not only to the knee but throughout the affected lower limb. Furthermore, we hypothesized considerable variability in joint kinetics among patients. Reports have documented the presence of several generalized gait patterns among all patients with hemiplegia, 21, 22 including the presumably homogeneous pattern of reduced knee flexion. 23 However, despite the generalized or homogeneous appearance of reduced knee flexion in subjects with the same diagnosis (stroke), we believed that the joint kinetic values would vary considerably from subject to subject, implying that the mechanisms of this pattern are varied. METHODS The population studied consisted of 20 subjects with stroke and spastic paretic stiff-legged gait and 20 ablebodied adult control subjects. The 20 subjects with spastic paretic stifflegged gait were recruited from the general community via newspaper, physician referral, and advertisements posted in local hospitals. We performed a preliminary screening to ensure that each subject had a unilateral stroke occurring at least 6 mo before the study, was able to walk without an assistive device or brace, and had an observed decrease in knee flexion in the affected limb. The ablebodied control subjects were recruited from the community via posted and newspaper advertisements. The control subjects were all without known musculoskeletal, neurologic, cardiac, or pulmonary pathology. The study was approved by our Institutional Review Board and written informed consent was obtained from each subject and control. Kinematic and kinetic measurements were collected bilaterally over six trials of walking across a 10-meter walkway, three trials for each limb. Measurements of the affected limb in the stroke subjects and the average of the left and right limbs in the control subjects were evaluated. Hip, knee, and ankle joint motion, torque and power in the sagittal planes were analyzed. The procedures used for this collection are based on standard techniques described elsewhere. 24 29 An optoelectronic motion analysis system (Bioengineering Technology Systems, ELITE System, Milan, Italy) was used to measure the three-dimensional position of infrared reflective markers at 100 frames per second, attached to various bony landmarks over the pelvis and lower limbs during walking. Ground reaction forces were measured synchronously with the motion analysis data using two staggered force platforms (Advanced Mechanical Technology, Newton, MA) embedded in the walkway. Joint torques and powers were calculated using a commercial fullinverse dynamic model, SAFLo (Servizio di Analisi della Funzionalita Locomotoria) from Bioengineering Technology Systems. Accordingly, joint torque and power calculations were based on the mass and inertial characteristics of each lower limb segment, the derived linear and angular velocities and accelerations of each lower limb segment, as well as ground reaction force and joint center position estimates. Joint torques were reported as external. Joint torques and powers were normalized for body weight, body height, and gait velocity; they are reported in Newton meters per kilogram meters- April 2001 Spastic Paretic Stiff-Legged Gait 245

TABLE 1 Temporal parameters for the subject and control groups meters per second (N m)/(kg m m/sec); powers are in watts per kilogram meters-meters per second W/(kg m m/ sec), respectively. Gait velocity values were obtained using the force platform and kinematic information to define initial foot contact times and distance parameters. Averaged kinetic data from all subjects were graphed over the walking cycle (0% to 100% at 2% intervals). The terminology of gait cycle events reported by Perry 15 was adopted to define locations in the gait cycle of all peak kinematic and kinetic values. All major peak kinematic values were evaluated and compared among subjects and controls using a Student s unpaired t test and using the Bonferroni adjustment to adjust for multiple 10 comparisons so that statistical significance was defined at P 0.005 (0.05/10). All major peak torque and power values Subject Group during the gait cycle, normalized for velocity, were evaluated statistically using each subject s average peak value from the three trials. A total of 16 peak joint torque and power values were compared. An F test was performed for each joint kinetic parameter to test for statistically significant differences in standard deviations between the two groups. Student s unpaired t tests, assuming unequal variances, among subjects and controls were then performed for each torque and power variable. Statistical significance was determined using the Bonferroni adjustment for multiple analyses at P 0.0031 (0.05/16). RESULTS Control Group Velocity (m/sec) 0.45 0.16 1.29 0.19 Cadence (steps/sec) 119 9 63 14 Stride length (m) 0.68 0.19 1.31 0.16 Step length (m) (affected limb) 0.44 0.14 0.65 0.08 TABLE 2 Mean peak kinematic parameters in degrees The subjects with stiff-legged gait included 14 males and 6 females, and averaged 53.5 12.7 yr in age, Subject Group Affected Limb Control Group P Value Hip flexion 15.9 8.4 25.2 5.4 0.0002 a Hip extension 6.8 5.7 17.5 4.6 0.3740 Knee flexion in loading response 12.5 10.0 18.1 6.7 0.0466 Knee extension midstance 1.7 10.7 1.8 4.1 0.1773 Knee flexion in swing 24.6 10.0 59.9 4.6 0.0001 a Knee extension at terminal swing 5.5 10.6 2.5 4.7 0.2523 Ankle plantar flexion at initial contact 7.0 4.8 8.6 3.6 0.2236 Ankle dorsiflexion in stance 7.1 4.7 7.6 3.4 0.7150 Ankle plantar flexion 5.2 5.9 16.2 5.2 0.0001 a Ankle dorsiflexion in swing 0.4 5.3 0.8 3.8 0.4311 a Significant at P 0.005 (0.05/10). 1.71 0.09 m in height, and 79.4 14.2 kg in weight. The control subjects included 12 males and 8 females, and averaged 52.0 17.6 yr in age, 1.70 0.10 m in height, and 74.0 13.3 kg in weight. There were no statistically significant differences in age, height, or weight between the subject and control groups. The subjects with stiff-legged gait had significantly reduced walking velocity, cadence, stride, and step lengths (each P 0.0001) (Table 1). The averaged values of all kinematic parameters for subjects vs. controls are shown in Table 2. Affected limb peak knee flexion was 24.6 10.0 for the subjects vs. 59.9 4.6 for the controls (P 0.0001). The averaged plots of the joint torque and powers are illustrated in Figure 1. The averaged peak joint torque and power values are shown in Table 3. The mean peak kinetic values in Table 3 are different from the peak values illustrated in Figure 1 for the following reasons: (1) the values in Table 3 are normalized for velocity, and (2) the peaks in the graphs, calculated as an average at each 2% interval in the gait cycle, did not necessarily reflect the individual trials true peaks, which often occurred outside this 2% interval. For all joint kinetic parameters, the standard deviations were greater in the subjects in comparison with those in the control group. The F-test analysis revealed that these differences in standard deviations were statistically significant for all but one parameter (knee-power absorption in preswing). Specifically, the F test P value for knee-power absorption was 0.3234, whereas the F test P value was 0.001 for all other parameters. The t-test analysis revealed statistically significant differences among subjects and controls in 7 of the 16 kinetic parameters (hip-joint torque in stance, hip-power generation in loading response, knee-extension torque in midstance, knee-power absorption, ankle-dorsiflexion torque, 246 Kerrigan et al. Am. J. Phys. Med. Rehabil. Vol. 80, No. 4

Figure 1: Averaged plots over one gait cycle of torques and powers for the hip, knee, and ankle for all subjects with spastic paretic stiff-legged gait. Torques are expressed in (N m)/(kg m) and powers are in W/(kg m) plotted over a normalized 0% to 100% gait cycle. ankle-power absorption in stance, and ankle-joint power generation). These differences are noted in Table 3 with an asterisk. Both peak kneepower absorption and ankle-power generation were decreased significantly in the subjects. Peak external hip-flexion torque in stance, hippower generation in loading response, knee-extension torque in midstance, ankle-dorsiflexion torque, and ankle-power absorption in stance were significantly greater in the subjects. DISCUSSION We observed alterations in peak joint kinetics compared with normal throughout the lower limb in 7 of the 16 parameters tested. We also found increased variability in all but one of the 16 kinetic parameters (peak kneepower absorption). Peak knee-power absorption was reduced significantly in the subjects as compared with that observed in the controls. Normally, this knee-absorption power occurs while the quadriceps absorb energy, providing some control of the rapidly flexing knee joint. 30 Although kneeflexion torque in preswing was less in the subjects compared with that observed in the controls, the difference was not statistically significant when the Bonferroni adjustment was used (P 0.0124). Thus, in some individuals in whom there is both normal knee-flexion torque in preswing and a reduction in knee-joint power absorption, the quadriceps may not be eccentrically or passively stretching as much as normal. In other individuals where knee-flexion torque in preswing is normal, the reduction in knee-joint power may reflect the fact that the knee is not rapidly flexing. We also noted a reduction in peak ankle-power generation. Anklejoint power generation occurs as a result of an increasing external ankle-dorsiflexion torque, along with rapid ankle plantarflexion of the ankle. 30 Although a reduction in external ankle-dorsiflexion torque in the subjects could explain a reduction in ankle-power generation, we observed an average increase in external ankledorsiflexion torque. Our observed decrease in ankle-power generation, along with an increase in external ankle-dorsiflexion torque, implies that the ankle does not rapidly flex at the end of a stance. Peak ankle plantarflexion was on average less in the subject group compared with that in the control group (5.2 5.9 vs. 16.2 5.2, P 0.0001). Another alteration about the ankle was an increase in ankle-power absorption in stance. These alterations about the ankle suggest that perhaps normal foot-ankle mechanics during the stance period allow the knee to flex during the swing period. Previously, we hypoth- April 2001 Spastic Paretic Stiff-Legged Gait 247

TABLE 3 Mean peak sagittal plane joint torques and powers esized that normal foot-ankle mechanics occurring late in stance, during the terminal stance, and preswing phases are necessary to allow knee flexion to begin in preswing. 3 At the hip, we observed increases in peak external hip-flexion torque in stance and in peak hip-power generation in loading response. These findings imply that during loading response, greater than normal hipextensor action is required. 30 We previously suggested that reduced hip-flexor strength results in reduced knee flexion. 3, 5, 9, 10 Here, we found that hip-power generation in preswing, which is associated with the need for hip-flexor activity, 30 is reduced somewhat, but that the difference on average is not statistically significant. Again, in some individuals, hip-power generation in preswing is considerably reduced, whereas in others it may be considerably greater. The increased variability of hip-power generation across subjects is consistent with our previous findings that improving hip flexion via forward dynamic modeling improves knee flexion in some but not in all subjects with Subject Group Affected Limb Control Group stiff-legged gait. 9, 10 The considerable variability in hip-power generation demonstrated here may be clinically significant in that different individuals may respond differently to rehabilitative treatments aimed at the hip flexors with the goal to improve knee flexion. CONCLUSIONS P Value Hip-flexion torque stance 0.85 0.53 0.32 0.07 0.0003 a Hip-extension torque stance 0.47 0.23 0.42 0.09 0.3740 Hip-flexion torque swing 0.13 0.16 0.08 0.03 0.1845 Hip-power generation loading response 0.94 0.57 0.42 0.19 0.0008 a Hip-power absorption 0.40 0.42 0.31 0.14 0.3726 Hip-power generation preswing 0.58 0.36 0.73 0.12 0.0903 Knee-flexion torque loading response 0.22 0.20 0.24 0.09 0.6867 Knee-extension torque midstance 0.61 0.63 0.12 0.08 0.0026 a Knee-flexion torque preswing 0.11 0.14 0.20 0.05 0.0124 Knee-power absorption initial contact 0.46 0.39 0.24 0.13 0.0252 Knee-power generation stance 0.52 0.37 0.32 0.10 0.0293 Knee-power absorption preswing 0.42 0.34 0.99 0.27 0.0001 a Ankle-plantar flexion torque 0.04 0.05 0.05 0.02 0.4142 Ankle-dorsiflexion torque 1.04 0.44 0.62 0.09 0.0004 a Ankle-power absorption stance 0.61 0.24 0.32 0.11 0.0001 a Ankle-power generation stance 0.74 0.42 1.51 0.17 0.0001 a a Significant at P 0.0031 (0.05/16). Torques are in (N m)/(kg m m/sec) and powers are in W/(kg m m/sec). We demonstrated kinetic alterations throughout the lower limb. Moreover, we demonstrated considerable variability in kinetic values across subjects. This markedly increased variability in torque and power values presents supporting evidence that individuals present with varying mechanisms or compensations for reduced knee flexion. Notably, all the measurements presented in this study can be obtained with 16, 18 20 routine clinical gait analysis. Of note, we normalized peak torque and power values for gait velocity given that these values are dependent on walking speed. 30 Had we not mathematically normalized for gait speed, we would conclude, given that the subjects walked considerably slower than the controls (approximately 35% speed), that most of the peak joint torques and powers are reduced in the subjects. It is difficult for these subjects to walk faster than their comfortable speed. On the other hand, it is unclear whether asking control subjects to reduce their walking speed to almost one-third their normal speed provides a realistic representation of normal walking. Further studies are warranted to test our normalization techniques or develop and test new techniques. This study provides further insights into the mechanisms associated with spastic paretic stiff-legged gait. It was previously assumed that improving knee flexion in spastic paretic stiff-legged gait is worthwhile to improve walking efficiency and reduce the risk of tripping and biomechanical injury. 3, 5, 9, 27 More work is needed to assess the complex relationships between various impairments and gait patterns and their effect on function. Further studies, including forward dynamic modeling 9 11 and induced-acceleration analyses, 12, 13 may be advantageous in evaluating these relationships. REFERENCES 1. Waters RL, Garland DE, Perry J, Habig T, Slabaugh P: Stiff-legged gait in hemiplegia: surgical correction. Am J Bone Joint Surg 1979;61A:927 33 2. Sutherland DH, Santi M, Abel MF: Treatment of stiff-knee gait in cerebral palsy: a comparison by gait analysis of distal rectus femoris transfer versus proximal rectus release. J Pediatr Orthop 1990;10:433 41 3. Kerrigan DC, Gronley J, Perry J: Stifflegged gait in spastic paresis. A study of quadriceps and hamstrings muscle activity. Am J Phys Med Rehabil 1991;70:294 300 4. Kerrigan DC, Sheffler L: Spastic paretic gait: an approach to evaluation and treatment. Crit Rev Phys Med Rehabil 1995;7: 253 68 5. Kerrigan DC, Bang MS, Burke DT: An algorithm to assess stiff-legged gait in 248 Kerrigan et al. Am. J. Phys. Med. Rehabil. Vol. 80, No. 4

traumatic brain injury. J Head Trauma Rehabil 1999;14:136 45 6. Gage JR, Perry J, Hicks RR, Koop S, Werntz JR: Rectus femoris transfer to improve knee function of children with cerebral palsy. Dev Med Child Neurol 1987; 29:159 66 7. Treanor WJ: The role of physical medicine treatment in stroke rehabilitation. Clin Orthop 1969;63:14 22 8. Gage JR: Surgical treatment of knee dysfunction in cerebral palsy. Clin Orthop 1990:45 54 9. Kerrigan DC, Roth RS, Riley PO: The modelling of spastic paretic stiff-legged gait based on actual kinematic data. Gait Posture 1998;7:117 24 10. Riley PO, Kerrigan DC: Torque action of two-joint muscles in the swing period of stiff-legged gait: a forward dynamic model analysis. J Biomech 1998;31: 835 40 11. Piazza SJ, Delp SL: The influence of muscles on knee flexion during the swing phase of gait. J Biomech 1996;29:723 33 12. Kepple TM, Siegel K, Stanhope SJ: Relative contributions of the lower extremity joint movements to forward progression and support during gait. Gait Posture 1997;6:1 8 13. Riley PO, Kerrigan DC: Kinetics of stiff-legged gait: induced acceleration analysis. IEEE Rehabil Eng 1999;7:420 6 14. Kerrigan DC, Burke DT, Nieto TJ, Riley PO: Can toe-walking contribute to stiff-legged gait? Am J Phys Med Rehabil 2001;80:33 7 15. Perry J: Gait Analysis: Normal and Pathological Function. Thorofare, NJ, SLACK, Inc, 1992 16. Harris GF, Wertsch JJ: Procedures for gait analysis. Arch Phys Med Rehabil 1994;75:216 25 17. Kerrigan DC, Glenn MB: An illustration of clinical gait laboratory use to improve rehabilitation management. Am J Phys Med Rehabil 1994;73:421 7 18. Winter DA, Eng P: Kinetics: our window into the goals and strategies of the central nervous system. Behav Brain Res 1995;67:111 20 19. Gage JR, DeLuca PA, Renshaw TS: Gait analysis: principle and applications with emphasis on its use in cerebral palsy. Instr Course Lect 1996;45:491 507 20. Kerrigan DC, Schaufele M, Wen MN: Gait analysis, in Delisa JA, Gans BM (eds): Rehabilitation Medicine Principles and Practice, ed 3. Philadelphia, Lippincott- Raven, 1998, pp 167 87 21. Knutsson E, Richards C: Different types of disturbed motor control in gait of hemiparetic patients. Brain 1979;102: 405 30 22. Shiavi R, Bugle HJ, Limbird T: Electromyographic gait assessment, part 2: preliminary assessment of hemiparetic synergy patterns. J Rehabil Res Dev 1987; 24:24 30 23. Winters TF Jr, Gage JR, Hicks R: Gait patterns in spastic hemiplegia in children and young adults. Am J Bone Joint Surg 1987;69A:437 41 24. Frigo C, Rabuffetti M, Kerrigan DC, Deming LC, Pedotti A: Functionally oriented and clinically feasible quantitative gait analysis method. Med Biol Eng Comput 1998;36:179 85 25. Kerrigan DC, Deming LC, Holden MK: Knee recurvatum in gait: a study of associated knee biomechanics. Arch Phys Med Rehabil 1996;77:645 50 26. Kerrigan DC, Ehrenthal S: Gait analysis, in O Young B, Young MA, Steins SA (eds): Physical Medicine and Rehabilitation Secrets, ed. 1. Philadelphia, Hanley & Belfus, 1996, pp 107 11 27. Kerrigan DC, Abdulhadi HM, Ribaudo TA, Della Croce U: Biomechanic effects of a contralateral shoe-lift on walking with an immobilized knee. Arch Phys Med Rehabil 1997;78:1085 91 28. Kerrigan DC, Todd MK, Della Croce U: Gender differences in joint biomechanics during walking: normative study in young adults. Am J Phys Med Rehabil 1998;77:2 7 29. Kerrigan DC, Todd MK, Della Croce U, Lipsitz LA, Collins JJ: Biomechanical gait alterations independent of speed in the healthy elderly: evidence for specific limiting impairments. Arch Phys Med Rehabil 1998;79:317 22 30. Winter DA: The Biomechanics and Motor Control of Human Gait: Normal, Elderly and Pathological, ed 2. Waterloo, Ontario, Canada, University of Waterloo Press, 1991 April 2001 Spastic Paretic Stiff-Legged Gait 249