Muscle activity changes due to immobilization and partial weight bearing of the lower leg

Tibial fractures account for half a million hospitalizations per year. Although clinical treatment of tibial fractures requires immobilization of the affected limb and partial weight bearing (PWB) prescriptions, the effects of this treatment protocol on gait and muscle activity have not been studied. This study aimed to provide the first comprehensive biomechanical analysis of changes in muscle activity and gait parameters induced by immobilization and PWB. Fifteen participants were recruited to walk across a pair of force plates in a motion capture lab under five different walking conditions: normal walking, immobilization without PWB, and immobilization with 25%, 50%, and 75% PWB. Motion was captured using 15 tracking balls with a ten-camera system, and muscle activity was quantified using surface electromyography on eight different muscle groups in the leg. Kinematic data revealed that immobilization and PWB decreased participant cadence and increased ratio of time spent on the immobilized limb during a single gait cycle. Kinetic data showed generally poor patient compliance with PWB prescriptions and no statistically significant difference between 25% PWB and normal walking conditions. Additionally, statistically significant differences were found in the soleus, hamstrings, and rectus femoris upon immobilization and PWB. These results reveal for the first time the short-term influences of immobilization and PWB on lower leg muscles and gait parameters.

MUSCLE ACTIVITY CHANGES DUE TO IMMOBILIZATION AND PARTIAL WEIGHT BEARING OF THE LOWER LEG by Mark Khoury A Senior Honors Thesis Submitted to the Faculty of The University of Utah In Partial Fulfillment of the Requirements for the Honors Degree in Bachelor of Science In Biomedical Engineering Approved: ____________________ Robert Hitchcock Supervisor ____________________ Patrick Tresco Chair, Department of Bioengineering ____________________ Kelly Broadhead Department Honors Advisor ____________________ Dr. Sylvia D. Torti Dean, Honors College May 2015 Copyright © 2015 All Rights Reserved ABSTRACT Tibial fractures account for half a million hospitalizations per year. Although clinical treatment of tibial fractures requires immobilization of the affected limb and partial weight bearing (PWB) prescriptions, the effects of this treatment protocol on gait and muscle activity have not been studied. This study aimed to provide the first comprehensive biomechanical analysis of changes in muscle activity and gait parameters induced by immobilization and PWB. Fifteen participants were recruited to walk across a pair of force plates in a motion capture lab under five different walking conditions: normal walking, immobilization without PWB, and immobilization with 25%, 50%, and 75% PWB. Motion was captured using 15 tracking balls with a ten-camera system, and muscle activity was quantified using surface electromyography on eight different muscle groups in the leg. Kinematic data revealed that immobilization and PWB decreased participant cadence and increased ratio of time spent on the immobilized limb during a single gait cycle. Kinetic data showed generally poor patient compliance with PWB prescriptions and no statistically significant difference between 25% PWB and normal walking conditions. Additionally, statistically significant differences were found in the soleus, hamstrings, and rectus femoris upon immobilization and PWB. These results reveal for the first time the short-term influences of immobilization and PWB on lower leg muscles and gait parameters. ii TABLE OF CONTENTS ABSTRACT........................................................................................................................ ii INTRODUCTION ...............................................................................................................1 METHODS ..........................................................................................................................4 Experimental Design and Conception .....................................................................4 Digital Processing of EMG Signals .........................................................................5 Calculation of Gait Parameters ...............................................................................5 Statistical Analyses ..................................................................................................6 RESULTS ............................................................................................................................7 Summary of Gait Parameters...................................................................................7 Kinetic Analysis .......................................................................................................8 Reduction of Electromyographic Signal ..................................................................9 DISCUSSION ....................................................................................................................11 ACKNOWLEDGEMENTS ...............................................................................................14 REFERENCES ..................................................................................................................15 iii 1 INTRODUCTION The tibia is the most commonly broken long bone in the human body, with the National Center for Health Statistics reporting 492,000 tibial fractures in the United States per year [1]. For a weight-bearing bone such as the tibia, mechanical forces on the fracture site play a significant role in determining the rate and type of healing [2-5] through their effects on the release of osteogenic and angiogenic factors [6]. However, applying too large a load onto the fracture site can cause misalignment of the bone, inflammation at the fracture site, and/or hardware failure [7], potentially leading to further complications such as delayed healing and fracture nonunion. In order to promote bone union and apply a moderate amount of weight on a fractured bone, the standard clinical treatment for tibial fractures relies on immobilization of the affected limb in a walking boot cast (WBC) and gradually increasing partial weight bearing (PWB) prescriptions. The current treatment protocol is problematic for several reasons. First, no clinical trials have examined the differences in healing rate between immobilized and nonimmobilized tibial fracture treatments. Thus the standard of treatment is not substantiated by primary research. Second, owing to the lack of relevant research, it is also unknown what mechanical environments are optimal for different types of fractures. Thus, the exact PWB prescription a tibial fracture patient will receive is at the clinician’s discretion. Thirdly, this treatment protocol depends on a patient’s ability to comply with PWB prescriptions, an assumption that is hotly debated in the medical literature [8, 9]. Finally, local muscle activity around the tibia regularly produces forces on the bone that 2 far exceed those of weight-related gravitational forces [10-12], and little research has been done on how muscle activity in the lower leg changes upon immobilization and PWB. Although research has been done on changes in gait and muscle activity in response to immobilization and PWB, the existing research base does not comprehensively characterize how this treatment protocol affects the biomechanical environment of a tibial fracture site. Prolonged immobilization is known to cause muscular atrophy of the affected limb and to induce systemic metabolic [13] and neurological [14, 15] changes. Studies have also shown that muscle activity in the tibialis anterior, soleus, and gastrocnemius are reduced upon immobilization as most WBC designs diminish the range of ankle flexion and extension [16]. These data, however, are insufficient to understand the nature of the mechanical environment of the lower leg during immobilization. Furthermore, little study has been done on the changes of gait parameters associated with immobilization and PWB or on other biomechanical characteristics such as forces, joint angles, walking speed, and muscle activity. Thus, PWB prescriptions cannot be assumed to provide adequate control of the mechanical environment around a fracture site. To better understand the changes in the biomechanical environment of the fracture site upon immobilization, our research group has conducted a comprehensive analysis of the changes in gait and muscle activity under different walking conditions. Our hypothesis is that immobilization and PWB walking conditions will cause changes in gait patterns and muscle activity and thus substantially influence forces around the fracture site beyond those expected from PWB. This analysis involves quantifying the 3 kinetic and kinematic changes of gait associated with the immobilization treatment and a quantitative and relative approach of understanding the changes in muscle activity. This type of biomechanical analysis should be useful for validating the efficacy of the current research treatment for tibial fractures and for informing alternative clinical approaches. 4 METHODS Experimental Set-Up All data analysis was performed from data generated using a pre-clinical experiment done prior to the author’s involvement in the project [17]. Fifteen participants (age 19-62) were recruited for participation in the real-time monitoring of lower leg weight bearing and gait. After discussing informed consent with the participants, the locations for electrode placement of each participant’s leg were shaved and dried using ethanol wipes. Electromyography (EMG) electrodes with snap connectors were placed on the adductor longus, hamstrings, vastus lateralis, rectus femoris, vastus medialis, gastrocnemius, soleus, and tibialis anterior as recommended by Criswell [18]. Incoming EMG signals were filtered with an iWorx ETH-256 amplifier set to 3 Hz for the highpass filter and 2 kHz for the low-pass filter. Participants took about three complete steps before stepping on the force plates in order to capture a steady-state gait cycle. The motion of the participants was recorded using a 10-camera system at the Motion Capture Lab of the Dumke Health Professions Building at the University of Utah. Kinematic markers were attached to each participant’s sacrum and left and right proximal phalanges, calcaneus, malleolus, knee, tibia, anterior superior iliac spine, and thigh. All incoming kinetic and kinematic data was analyzed using Vicon Nexus v1.8.2. Participants were instructed to walk across a pair of force plates under one of five walking conditions: normal walking, full-weight bearing (FWB) with crutches and WBC on the right leg, 75% PWB, 50% PWB, and 25% PWB. Participants were trained to 5 partially weight bear using a bathroom scale technique [9] and by walking on a treadmill until comfortable with the movement pattern. Kinetic data from the force plates and EMG data were sampled at 2.4 kHz, and kinematic data from the tracking balls was collected synchronously at 200 Hz. Data were cropped to encompass only a single gait cycle (defined between heel-strikes of the right foot) when the participants were walking across the force plate prior to export. Six trials of each of the five walking conditions were performed for each participant, for an ideal total of 30 data sets per participant. Digital Processing of EMG Signals Imported EMG signals were filtered using a Butterworth filter with a 10-500 Hz band-pass range and a 60 Hz notch filter using the MATLAB® Signal Processing toolbox. EMG data were offset to oscillate around a zero-baseline, rectified, and normalized to the maximum EMG voltage of any of the muscles for each participant. The EMG signal windows were then broken into 25 bins per gait cycle, with each bin corresponding to 4% of the gait cycle. The root-mean square (RMS) amplitude of the EMG signal for each bin was calculated. Calculation of Gait Parameters The cadence of the participants under each walking condition was calculated by dividing the time required for patients to make a complete step by the sampling rate. Other parameters of gait, including percent of time spent on injured leg (stance %), peak load applied on injured leg, and location of peak load were calculated using kinetic data 6 obtained from the force plates. The load profiles of gait were also used for area under the curve (AUC) analyses. Statistical Analysis EMG data were analyzed using a one-way analysis of variance (ANOVA) test with a Tukey-Kramer post-hoc multiple comparison test. Kinetic data were analyzed using difference-of-mean student’s t-tests. For both ANOVA and student’s t-tests, statistical significance was determined using a threshold of α=0.05. 7 RESULTS Summary of Gait Parameters Immobilization and PWB produced changes in cadence and force on the immobilized limb (Table 1). Relative to normal walking, cadence was shown to decrease significantly (p<0.05) with immobilization and partial-weight bearing. Immobilization and FWB showed a reduction in cadence by about 8%, and all PWB trials showed a reduction of cadence by about 27.5%. Although participants spent more time with each step upon immobilization, a smaller proportion of time was spent in stance phase. Table 1 – Summary of gait parameters obtained through kinetic data obtained from force plates. Each error corresponds to one standard error of the mean (SEM), defined as the sample standard deviation divided by the square-root of the number of participants (n=15 for each calculation). Cadence (steps/min) Peak Load (N) Peak Load Location (% of Gait) Time in Stance Phase (% of Gait) Normalized Area Under the Curve (s) Normal 51±1 725±24 36±4 62.5±0.4 11±2 FWB 47±1 716±22 25±3 60±1 20±2 75% PWB 37±1 568±30 24±3 60±1 16±1 50% PWB 37±1 443±27 27±4 59±2 13±2 25% PWB 37±1 359±42 30±4 55±2 10±1 Walking Condition Relative to normal walking, cadence was shown to decrease significantly (p<0.05) with immobilization and partial-weight bearing. Immobilization and FWB showed a reduction in cadence by about 8%, and all PWB trials showed a reduction of cadence by about 27.5%. Although participants spent more time with each step upon immobilization, a smaller proportion of time was spent in stance phase. 8 Kinetic Analysis As the degree of partial weight bearing decreased, participants showed a diminished ability to comply with the treatment protocol prescriptions (Figure 1). The target PWB prescription (i.e. 25%, 50%, or 75%) was never within one SEM of the mean maximal force applied onto the immobilized leg. Under the instruction of applying 25% PWB onto the immobilized leg, patients applied a mean of 55±24% of their body weight onto their fractured leg. Figure 1 – Peak force recorded by the force plate during each of the five walking conditions. Patients at 25% weight bearing applied on average 50% of their weight, showing inability to comply with treatment protocol. Note the increasing variance of the applied force with decreasing PWB prescriptions. Although the peak load did decrease with increasingly light PWB prescriptions, the change was also accompanied by a decrease in cadence (Table 1). Thus, though the ab¬¬¬solute magnitude of the force on the immobilized limb decreased, the amount of time in which the force was applied on the limb increased. The force-time effects are quantified using AUC data that showed no statistically significant difference between normal walking conditions and 25% PWB conditions; by contrast, AUC with immobilization and without PWB nearly doubled (Figure 2). 9 Reduction of Electromyographic Signal Profound changes were seen in muscle activity for several muscles, most notably in the soleus, rectus femoris, and hamstrings. The soleus muscle activity drops dramatically upon immobilization even without PWB (Figure 3, top). Rectus femoris activity dropped significantly with both immobilization and PWB, and muscle the hamstrings showed an increasing activity at 40-60% gait with lower PWB prescriptions. Analysis of other muscles has been complicated by the presence of noise in the EMG signals. There was some evidence of changes in muscle Figure 2 – AUC analysis of different walking conditions. The similarity AUCs of Normal and 25% PWB walking conditions indicates that although less force is being applied in 25% PWB prescriptions, it is applied for longer periods. activity for the adductor longus, vastus medialis, and tibialis anterior; however, the EMG traces obtained from these muscle groups were greatly skewed by noisy signals (Figure 3, bottom). Electrical signals from other muscle groups such as the vastus medialis and the vastus longus, showed too little a signal-to-noise ratio to be statistically interpretable. 10 Figure 3 – Muscle activity of the soleus (top) and gastrocnemius (bottom), averaged over all 15 participants. Both muscles show a decrease in activity around 40-60% of the normal gait cycle. 11 DISCUSSION Although it is known that gait changes upon immobilization and partial weight bearing, existing studies on the effects of partial weight bearing have only considered the effects of a small number of muscles [16] and have not integrated electromyographic measurements with other gait parameters. The motion, forces, and electromyographic signals of 15 experimental participants were captured in a motion capture lab as they walked across a pair of force plates under normal and immobilized / PWB conditions. The results indicated that immobilization and partial weight bearing induced statistically significant differences in gait parameters and in muscle activity of the soleus, hamstrings, and rectus femoris. Recent literature has investigated the physiological effects of immobilization and PWB and identified adverse effects of the treatment protocol as a prescription for tibial fracture patients. Muscle disuse associated with immobilization has known deleterious effects on bone density in weight-bearing bones [19,20] and thus significantly increases risk of re-fracture [21]. The results presented in this study have quantified changes in electromyographic signals and shown marked reduction in activity in the soleus, gastrocnemius, and rectus femoris upon immobilization and PWB, indicating that patients undergoing this treatment may be disposed to re-injury. However, although statistically significant changes in muscle activity were observed for these muscle groups, the biomechanical changes induced by immobilization and PWB cannot be understood solely on those terms. Kinetic data obtained from the force plates, for example, demonstrated a reduction in peak load with successive PWB prescription, but AUC calculations between 12 normal walking and 25% PWB showed no statistically significant difference. The observed data thus indicates a reduction of the magnitude of gravity-related weight forces on the leg but little change in the force-time relationships. How these biomechanical changes ultimately affect the mechanical environment of the fracture site cannot be inferred from the current data. Thus, in order for the results of this study to be clinically informative, the obtained data would need to be integrated in a biomechanical model of the tibia that would comprehensively consider changes in the mechanical environment of the tibia with immobilization and PWB. A major limitation of this study’s relevance to informing tibial fracture treatment options is that it does not consider the physiological changes induced by immobilization and PWB and their effects on the mechanical environment of the tibia. Prolonged immobilization has profound detrimental effects on muscle; in the short-term, immobilized muscles have a greatly reduced ability to produce voluntary forces [22, 23], and in the long-term, immobilization can reduce production of myostatin and other muscle growth hormones [24], thus causing a decrease in muscle protein synthesis [25]. How electromyographic signals may adopt to these changes, or how patients will adapt their walking patterns as their musculature changes over the course of the treatment, is currently unknown. Long-term studies that investigate changes in gait at different time scales are necessary to best characterize the biomechanical changes of PWB during various stages of fracture healing. One promising solution towards this end is an underfoot load-monitoring device that allows for continuous collection of kinetic data of a fractured limb. Continuous-monitoring devices are currently under development [26, 13 27] and would allow for collecting relevant clinical data about walking patterns outside of the artificial environment of a motion capture lab. Short-term studies of steady-state gait cycles in a motion capture lab are important for initial characterization of immobilization and PWB, but are unable to represent the diversity of walking patterns experienced in daily life or to model changes in walking patterns in the long term. Nonetheless, the presentation of changes in biomechanics, gait parameters, and muscle activity provide a reasonable lens by which the effects of immobilization and PWB prescriptions on gait can be examined. Continuations of this research that focus on further quantifying changes in other gait parameters induced by immobilization and PWB and on monitoring long-term changes in gait outside of a motion capture lab setting should be able to characterize the force environment of the fracture site during the bone healing process, which should ultimately help to better inform clinical treatments of tibial fractures. 14 ACKNOWLEDGEMENTS I would like to gratefully acknowledge the contributions and support of my supervisor and Principal Investigator, Dr. Robert Hitchcock, for his conception and development of the research topic, and to my academic advisor Heather Palmer for assisting me in finding a lab and for her continued support in my academic endeavors. I would also like to thank the Department of Biology Undergraduate Research Program for providing funding my participation in this research endeavor. 15 REFERENCES [1] G. N. Duda, B. Bartmeyer, S. Sporrer, W. R. Taylor, M. Raschke, and N. P. Haas, “Does partial weight bearing unload a healing bone in external ring fixation?,” Langenbecks Arch Surg, vol. 388, no. 5, pp. 298-304, Oct, 2003. [2] J. Kenwright, and A. E. Goodship, “Controlled mechanical stimulation in the treatment of tibial fractures,” Clin Orthop Relat Res, no. 241, pp. 36-47, Apr, 1989. [3] A. Sarmiento, J. F. Schaeffer, L. Beckerman, L. L. Latta, and J. E. Enis, “Fracture healing in rat femora as affected by functional weight-bearing,” J Bone Joint Surg Am, vol. 59, no. 3, pp. 369-75, Apr, 1977. [4] J. A. Buckwalter, “Activity vs. rest in the treatment of bone, soft tissue and joint injuries,” Iowa Orthop J, vol. 15, pp. 29-42, 1995. [5] V. Mooney, V. L. Nickel, J. P. Harvey, Jr., and R. Snelson, “Cast-brace treatment for fractures of the distal part of the femur. A prospective controlled study of one hundred and fifty patients,” J Bone Joint Surg Am, vol. 52, no. 8, pp. 1563-78, Dec, 1970. [6] D. R. Marsh, and G. Li, “The biology of fracture healing: optimising outcome,” Br Med Bull, vol. 55, no. 4, pp. 856-69, 1999. [7] L. Dillin, and P. Slabaugh, “Delayed wound healing, infection, and nonunion following open reduction and internal fixation of tibial plafond fractures,” J Trauma, vol. 26, no. 12, pp. 1116-9, Dec, 1986. 16 [8] A. Vasarhelyi, T. Baumert, C. Fritsch, W. Hopfenmuller, G. Gradl, and T. Mittlmeier, “Partial weight bearing after surgery for fractures of the lower extremity--is it achievable?,” Gait Posture, vol. 23, no. 1, pp. 99-105, Jan, 2006. [9] H. V. Dabke, S. K. Gupta, C. A. Holt, P. O'Callaghan, and C. M. Dent, “How accurate is partial weightbearing?,” Clin Orthop Relat Res, no. 421, pp. 282-6, Apr, 2004. [10] C. N. Maganaris, “Force-length characteristics of the in vivo human gastrocnemius muscle,” Clin Anat, vol. 16, no. 3, pp. 215-23, May, 2003. [11] C. N. Maganaris, “Force-length characteristics of in vivo human skeletal muscle,” Acta Physiol Scand, vol. 172, no. 4, pp. 279-85, Aug, 2001. [12] D. B. Burr, “Muscle strength, bone mass, and age-related bone loss,” J Bone Miner Res, vol. 12, no. 10, pp. 1547-51, Oct, 1997. [13] T. Haggmark, E. Jansson, and E. Eriksson, “Fiber type area and metabolic potential of the thigh muscle in man after knee surgery and immobilization,” Int J Sports Med, vol. 2, no. 1, pp. 12-7, Feb, 1981. [14] H. E. Berg, L. Larsson, and P. A. Tesch, “Lower limb skeletal muscle function after 6 wk of bed rest,” J Appl Physiol, vol. 82, no. 1, pp. 182-8, Jan, 1997. [15] J. Duchateau, “Bed rest induces neural and contractile adaptations in triceps surae,” Med Sci Sports Exerc, vol. 27, no. 12, pp. 1581-9, Dec, 1995. [16] N. J. Kadel, A. Segal, M. Orendurff, J. Shofer, and B. Sangeorzan, “The efficacy of two methods of ankle immobilization in reducing gastrocnemius, soleus, and peroneal 17 muscle activity during stance phase of gait,” Foot Ankle Int, vol. 25, no. 6, pp. 406-9, Jun, 2004. [17] L. F. Reed, S. R. Urry, and S. C. Wearing, “Reliability of spatiotemporal and kinetic gait parameters determined by a new instrumented treadmill system,” BMC Musculoskelet Disord, vol. 14, pp. 249, 2013. [18] E. Criswell, Cram's Introduction to Surface Electromyography: Jones & Bartlett Learning, 2010. [19] J. M. Swift, F. Lima, B. R. Macias, M. R. Allen, E. S. Greene, Y. Shirazi-Fard, J. S. Kupke, H. A. Hogan, and S. A. Bloomfield, “Partial weight bearing does not prevent musculoskeletal losses associated with disuse,” Med Sci Sports Exerc, vol. 45, no. 11, pp. 2052-60, Nov, 2013. [20] A. D. LeBlanc, E. R. Spector, H. J. Evans, and J. D. Sibonga, “Skeletal responses to space flight and the bed rest analog: a review,” J Musculoskelet Neuronal Interact, vol. 7, no. 1, pp. 33-47, Jan-Mar, 2007. [21] A. C. Looker, H. W. Wahner, W. L. Dunn, M. S. Calvo, T. B. Harris, S. P. Heyse, C. C. Johnston, Jr., and R. Lindsay, “Updated data on proximal femur bone mineral levels of US adults,” Osteoporos Int, vol. 8, no. 5, pp. 468-89, 1998. [22] A. J. Sargeant, C. T. Davies, R. H. Edwards, C. Maunder, and A. Young, “Functional and structural changes after disuse of human muscle,” Clin Sci Mol Med, vol. 52, no. 4, pp. 337-42, Apr, 1977. 18 [23] H. J. Appell, “Muscular atrophy following immobilisation. A review,” Sports Med, vol. 10, no. 1, pp. 42-58, Jul, 1990. [24] S. W. Jones, R. J. Hill, P. A. Krasney, B. O'Conner, N. Peirce, and P. L. Greenhaff, “Disuse atrophy and exercise rehabilitation in humans profoundly affects the expression of genes associated with the regulation of skeletal muscle mass,” FASEB J, vol. 18, no. 9, pp. 1025-7, Jun, 2004. [25] D. B. Thomason, and F. W. Booth, “Influence of performance on gene expression in skeletal muscle: effects of forced inactivity,” Adv Myochem, vol. 2, pp. 79-82, 1989. [26] K. North, S. D. Maass, and R. W. Hitchcock, “An insole sensor for recording weight bearing behavior during tibial fracture rehabilitation,” Conf Proc IEEE Eng Med Biol Soc, vol. 2010, pp. 1856-9, 2010. [27] K. North, E. N. Kubiak, and R. W. Hitchcock, “Sensor packaging design for continuous underfoot load monitoring,” Biomed Microdevices, vol. 14, no. 1, pp. 217-24, Feb, 2012.