Patellofemoral Joint Compressive Forces in Forward and Backward ~unning

Patellofemoral Joint Compressive Forces in Forward and Backward ~unning Timothy W. Flynn, MS, PT, OCS, Major, U.S. Army Medical Specialist Corps1 Robert W. Soutas-little, P~D* T he knee is the most common area of pain in runners, with the patellofemoral joint frequently the joint of dysfunction (10). In adolescents, patellofemoral pain syndrome is the leading cause of chronic knee pain (2). Atrophy of the mtus medialis oblique muscle is frequently associated with patellofemoral pain syndrome, possibly as a result of the interaction between mechanical and neuromuscular factors (8,12,17). This atrophy is thought to result in extensor mechanism dysfunction, decreased muscle strength, and imbalance in the control between medial and lateral portions of the quadriceps femoris muscle (21). Traditionally, quadriceps strengthening exercises have been used in the rehabilitation of patellofemoral pain syndrome, and more recently, "closed chain" quadriceps strengthening, or quadriceps strengthening in a weight-bearing position, has become an accepted method of treatment (14). Backward running incorporates the "closed chainn concept of quadriceps exercise. Backward running has been documented as increasing quadriceps strength and power (11,20). In addition, at self-selected running speeds during the stance phase of backward running, the quadriceps are active for a longer period of stance than during forward running (7). It appears that backward running may be indicated in the rehabilitation of patellofemoral pain syndrome, but the magnitude of the The use of backward running is becoming more common in the rehabilitation setting. In particular, backward running has been suggested as a treatment modality in patients experiencing patellofemoral pain syndrome. To date, no study has examined the loads at the patellofemoral joint during backward running. The purpose of this study was to compare forces during forward and backward running. Ground reaction force and kinematic data were collected on five male joggers during free speed forward and backward running. A floor reaction force vector model was used to calculate the stance phase knee extension moments. The distance used for the extensor muscle lever arm was 4.9 cm. Patellar mechanism angle was calculated based on knee joint angle. There was a reduction in the peak forces in backward compared with forward running (2277 2 192N vs. 4253 t 1292N; p < 0.05) at selfselected speeds. Peak force occurred significantly later (p < 0.05) in the stance phase of backward running (52 2 4%) than in forward running (35 + 3%). The peak force normalized to subject body weight was 5.6 2 1.3 body weight in forward running and 3.0 2 0.6 body weight in backward running. The results suggest that backward running at a self-selected speed may reduce forces and, coupled with the quadriceps strengthening that has previously been reported, may be beneficial in the rehabilitation of patellofemoral pain syndrome in runners. However, constant speed comparisons or other models may yield different results. Key words: patellofemoral pain, backward running, rehabilitation ' Doctoral Student, The Center for Locomotion Studies, Penn State University, University Park, PA. At the time of this study, Mr. Flynn was a graduate student in the Department of Biomechanics, Michigan State University, East Lansing, MI. Address for correspondence: 1395 University Drive, State College, PA 16801. * Professor, College of Engineering, Michigan State University; Director, Biomechanics Evaluation Laboratory, East Lansing, MI The opinions or assertions contained herein are those of the authors and are not to be construed as official or as reflecting the views of the Department of the Army, the Department of Defense, or the U.S. Government. force during backward running and what affect this may have on patellofemoral pain syndrome is not known. Prior research has emphasized the role of high patellofemoral joint compressive force as an etiological factor of patellofemoral pain syndrome and have used intervention strategies to reduce it (9,16). Treatment protocols in patellofemoral rehabilitation should be based on sound knowledge of the existing pathology and joint biomechanics (24). Different exercise interventions result in a varying of the patellofemoral joint compressive forces. Knowledge of these forces in backward running may allow the clinician the opportunity to apply this modality in a more selective and scientific manner. To this end, the purpose of this study was to determine the patellofemoral joint compressive force in backward running and compare this with forward running in pain free individuals at self-selected running speeds. Knowledge of these forces may assist reha- JOSPT Volume 21 Number 5 May 1995

RESEARCH STUDY bilitation specialists when prescribing exercise programs for patellofemoral pain syndrome sufferers. MATERIALS AND METHODS Subjects Five healthy male university students volunteered as subjects for this study. They were 29 + 3 years old, 71 + 3 inches in height, and weighed 169 2 27 Ibs. All acknowledged voluntary participation in this study by giving written informed consent. The subjects were recreational joggers and had no previous history of knee disorders. The subjects had full knee range of motion and stable ligamen- The quadriceps are active for a longer period of stance during backward running than during forward running. tous evaluations. The right lower extremity was used for testing. Equipment Kinematics Kinematic data were acquired with a Motion Analysis System (Motion Analysis Corp., Santa Rosa, CA), which used four synchronized and calibrated solid state 60 Hz video cameras shuttered at 1 msec per frame. Spherical reflective markers were attached to each subject's right lower extremity at the 1) greater trochanter, 2) lateral femoral condyle, 3) the anterior tibia, approximately 5 inches above the lateral malleolus, and 4) the posterior calf, approximately 5 inches above the lateral malleolus. Care was taken to ensure marker number 2 was at the center of the lateral femoral condyle. Ground Reduction Fmes Force plate signals were sampled at a rate of 1000 Hz from an AMTI Model OR64 force platform (Advanced Mechanical Technology Inc., Newton, MA) as the subject's foot contacted the platform. Vertical and fore-aft forces were amplified and digitally converted. Experimental Protocol The subject stood on the force plate in a relaxed position with the knee joint in neutral. Five seconds of video data were collected and subsequently used to calculate the offset knee angle, which allowed the linkage targets to be independent of an exact vertical or horizontal position. Three trials from each subject at a self-selected speed of forward running and backward running were tested along a 40-ft runway. A trial consisted of the subject's right foot landing entirely on the force platform. A mistrial occurred if the subject's stride was unnatural or if the subject altered his stride in an attempt to hit the force plate. Motion and force plate data were triggered simultaneously and stored on a computer workstation. Patellofemoral Joint Model A sagittal plane mathematical model of the patellofemoral joint was used to calculate the magnitude of the force (Figure 1). Multiple static stance phase knee extensor moments (M,) were calculated at 16msec intervals which allowed synchronization of the force plate (1000 Hz) data and the motion analysis (60 Hz) data using the following equation: where F,, is the fore-aft ground reaction force, Z, is the vertical position of the saggital plane knee joint center (target 2), F, is the vertical reaction force, X2 is the fore-aft position of the sagittal plane knee joint center (target 2), and COP,, is the fore-aft position of the subject's foot on the force plate. The patellar tendon and quadriceps muscle force were assumed to be equal. The moment arm of the quadriceps mechanism was considered.049 m based on Smidt's results (19). Quadriceps force (F,,,,) was calculated by the following equa- tion: 2) F, = M, X 0.049-'. The force was determined by the equation: 3) Patellofemoral joint compressive force = 2 F, X sin P/2 where p is the patella mechanism angle based on Mathews et al's linear regression line and is dependent upon the knee joint angle (13). A student's paired t test was performed on peak force, time to peak patellofemoral joint compressive force, and knee angle at peak patellofemoral joint compressive force with significance set at p < 0.05. RESULTS A representative sample of the knee joint range of motion normalized to percent of stance phase is presented in Figure 2. Note the greater knee flexion in early stance of backward running as well as the lack of a rapid knee flexion during the first 30% of backward running stance, which is typically seen in forward running. A representative sample of the real time stance phase knee extensor moment (M,) is presented in Figure 3. In both modes of running, Mk was positive or an extensor moment. In backward running, there was a slight dip in the moment during early stance. An important fact to be noted is that the subject had foot contact on the force plate for a longer time period during backward running (240 msec) than during forward running (194 msec), possibly indicating a slower backward running speed. However, running speed for each condition cannot be determined by this analysis. The Table reports selected gait parameters. Volume 21 Number 5 Mav 1995 JOSPT

RESEARCH STUDY FIGURE 3. Stance phase knee extensor moment (M,) in real time. FIGURE 1. Free body diagram used to determine forces about the knee. PFD = Patellofemoral joint compressive force, p = Patellar mechanism angle, M, = Knee moment, F, = Quadriceps force. During backward running, there wa.. significantly lower peak knee extensor moments and peak patellofemoral joint compressive force than during forward running. The time to peak force (absolute and as a percentage of stance phase) was significantly longer during backward running than during forward running. The FIGURE 2. Knee joint range of motion normalized to percent of stance phase. Terminal extension = 04 JOSF'T Volume 21 Number 5 May 1995 knee joint angle at the peak patellofemoral joint compressive force was similar between conditions. Figure 4 is a representative sample of the force normalized to a percentage of stance phase. The dip in force seen in early stance phase coincides with the dip seen on the M, graph. DISCUSSION The results of this study must be tempered with several study limitations. Most notable was the fact that running speed was not controlled and resulted in slower trials during backward running. As noted by Cavanagh, the control of speed is always a problematic issue in the studies of gait, ie., whether to choose free speed or fixed speed (4). The au- thors of this study chose free (selfselected) speed for each condition. The rationale for this was that patients or athletes would not generally control their speed during backward running; rather, they would run at a comfortable pace. In addition, Threlkeld et al noted quadriceps strength gains after an &week training program of backward running during which "ultimately, the subjects themselves selected their personal training speed for both forward and backward running" (20). The apparent slower backward running speed may have contributed to the lower peak M, during backward running seen in the study. This differs from the work of Devita and Stribling, who noted higher peak M, during backward running than during forward running when running speed was held constant at 3.0 + 0.1 m/sec for both conditions (5). The authors reported a peak value of 256 Nm during backward running, yet based on extrapolation of the peak M, from their published figure, only a 156 Nm peak M, was noted during forward running. The forward peak M, in the present study was similar to the peak extensor moments reported by Buczek and Cavanagh (3). Subjects in that study ran at a speed of 4.5 m/sec + 5% and displayed peak extensor moment values of 288 + 81 Nm. The difference in M, values noted may also be related to the different methods of moment calculation employed in the different studies. The present study used the floor

RESEARCH STUDY Gail Parameter (During Stance) - X Peak M, (N) Peak M, (BW Peak PFJC force (N) Peak PFJC force (BW Time to peak PFJC force (% dance) Time to peak PFJC force (seconds) Flexion angle at peak (degrees) * p < 0.05. PF)C = Patellofemoral joint compression. M, = Stance phase knee extensor moment. BW = Force normalized to subject body weight. Forward TABLE. Mean 5 standard deviation for selected gait kinematics and kinetics during stance. reaction force vector technique, while the Devita and Stribling study used the inverse dynamics approach. Winter has pointed out several shortcomings of the floor reaction force vector technique (23). Additionally, Wells has reported that small but significant errors are possible when using the floor reaction force vector a p proach to calculate knee moments during forward walking (22). However, it is not presently known if this effect is consistent during backward ambulation. The implications of this discussion are not trivial since the M, value is directly related to the patellofemoral joint compressive force (Equation 2); thus, lower M, would result in lower patellofemoral joint compressive force. The magnitude of the patellofemoral joint compressive force reported in this study during forward running (4.3-6.9 X body weight) was less than reported in three runners by Scott and Winter (7.0-1 1.1 X body weight) (18), who used inverse dynamics coupled with a more sophisticated optimization model which considered the resistance of the gastrocnemius to knee extension as well as a function that did not assume the Fquad and patellar tendon force to be equal. The subjects in that study ran at a speed of 3.5-5.3 m/sec. Though the present study may have underestimated the patellofemoral load, we demonstrated a 40% reduction in the comparative magnitude of the patelle femoral joint compressive force dur- ing backward running when compared with forward running at a selfselected speed. The dip in M, and force seen in early stance is possibly explained by the large eccentrically controlled dorsiflexion at the ankle joint occurring at this time (5). %(ITAWCE FIGURE 4. Patellofemoral joint compressive force in forward and backward running. The rate of patellofemoral joint compressive force loading was significantly slower and occurred later in the stance phase during backward running than during forward running. Articular cartilage has viscoelastic properties that make it rate sensitive to loading so that it is more susceptible to injury during a rapid loading, which prevents sufficient accommodation (15). The specific location and magnitude of the contact loads at the patellofemoral joint articular cartilage can only be determined invasively; however, this study demonstrated that the peak patelle femoral joint loading was lower and occurred at a slower rate during backward running than forward running. Since the knee joint angle at peak load was similar between conditions, this would suggest that the patellofemoral contact area loaded Volume 21 Number 5. May 1995 JOSPT

during each condition was similar. The knee joint position curves during forward and backward running noted in this study were consistent with previous studies (1,5), in particular, the decrease in the total joint range of motion during backward running as compared with forward running. Additionally, the weight accepting knee flexion typically noted during the early stance phase of forward running is absent during backward running. Previous work has demonstrated that backward running at a selfselected speed requires a longer period of quadriceps activity and can achieve greater quadriceps strength gains The rate of patellofemoral joint compressive force loading was significantly slower and occurred later in the stance phase during backward running than during forward running. - - than during forward running (7,20). Recently, it has been shown that backward running requires greater cardiopulmonary demands than forward running in subjects untrained in backward running (6). These findings, coupled with the results of this preliminary study, would tend to sup port the use of backward running in the rehabilitation of patients suffering from patellofemoral pain syndrome. It appears that backward running can provide a number of benefits to the patient suffering from patellofemoral pain. For example, the patient using a backwards running training protocol at a selfse- JOSPT Volume 21 Number 5 Mav 1995 lected speed would be using the quadriceps for a longer period of stance, potentially strengthening the knee extensors with a decreased loading of the patellofemoral joint, yet exercising at a training intensity that provides aerobic benefits. This study should be repeated and the running speed variable held constant to determine the effect of running speed on patellofemoral joint compression force. Future studies should include controlled clinical trials which compare matched patellofemoral pain syndrome patients using traditional rehabilitation protocols with those that incorporate backward running. Additionally, future research should incorporate a more sophisticated force model in the analysis. CONCLUSIONS The purpose of this study was to compare force in backward running vs. forward running. The results estimate that there was a significant reduction of peak patellofemoral joint compressive force in backward running (3.0 X body weight) when compared with forward running (5.6 X body weight) at a selfselected speed. The peak force occurred later in the stance phase of backward running but at a knee joint angle that was similar to the angle occurring in forward running. However, these findings cannot be extrapolated to a constant speed comparison. Backward running may be a useful rehabilitation modality in the treatment of patellofemoral pain syndrome, but clinical trials are needed to support its use. JOSPT REFERENCES 1. Bates BT, Morrison E, Hamill J: A comparison between forward and backward running. In: Adrian M, Deutsch H (eds), Proceedings of the 1984 Olympic Scientific Congress: Biomechanics, pp 127-135. Eugene, OR: Microform Publications, 1984 2. Baxter MP: Knee pain in the pediatric athlete. Paediatr Med 1 :2 11-2 18, 1986 3. Buczek FL, Cavanagh PR: Stance phase knee and ankle kinematics and kinetics during level and downhill running. Med Sci Sports Exerc 22:669-677, 1990 4. Cavanagh PR: Differences in the gait characteristics of patients with diabetes and peripheral neuropathy compared with age-matched controls. Phys Ther 74(4):299-313, 1994 5. Devita P, Stribling J: Lower extremity joint kinetics and energetics during backward running. Med Sci Sports Exerc 23:602-610, 1991 6. Flynn TW, Connery SM, Smutok MA, Zeballos RJ, Weisman IM: Cardiopulmonary responses to forward and backward walking and running in normals. Med Sci Sports Exerc 26:89-94, 1994 7. Flynn TW, Soutas-Little RW: Mechanical power and muscle action during forward and backward running. j Orthop Sports Phys Ther 17:108-112, 1993 8. Hanten WP, Schulties SS: Exercise effect on electromyographic activity of the vastus medialis oblique and vastus lateralis muscles. Ph ys Ther 7O:561-565, 1990 9. lnsall J: Chondromalacia patellae: Patellar malalignment syndrome. Orthop Clin North Am 10: 117-!27, 1979 10. James SL, Bates BT, Osternig LR: Injuries to runners. Am J Sports Med 6:40- SO, 1978 11. Mackie JW, Dean TE: Running backward training effect on upper leg musculature and ligamentous instability of injured knees. Med Sci Sports Exerc 16: 151, 1984 (abstract) 12. Mariani PP, Caruso I: An electromyographic investigation of subluxation of the patella. J Bone Joint Surg 6 1 B:169-171, 1972 13. Mathews LS, Sonstegard DA, Henke ]A: Load bearing characteristics of the patello-femoral joint. Acta Orthop Scand 48:s 11-5 16, 1977 14. McConnell J: The management of chondromalacia patella: A long term solution. Aust J Ph ysiother 3:2 15-222, I986 15. Mow VC, Rosenwasser M: Articular cartilage biomechanics. In: Woo SL-V, Buckwalter ]A (eds), Injury and Repair of the Musculoskeletal Soft Tissues, pp 427-463. Park Ridge, IL: American Academy of Orthopaedic Surgeons, 1988 16. Outerbridge RE: The etiology of chrondromalacia patellae. J Bone Joint Surg 43B37.52-757, 196 1

RESEARCH STUDY 17. Reynolds L, Levin TA, Medeiros JM, Adler N, Hallum A: EMG activity of the vastus medialis oblique and vastus lateralis and their role in patellar alignment. Am ] Phys Med 62:6 1-70, 1983 18. Scott SH, Winter DA: Internal forces at chronic running injury sites. Med Sci Sports Exerc 22:357-369, 1990 19. Smidt GL: Biomechanical analysis of knee flexion and extension. J Biomech 6:79-92, 1973 20. Threlkeld A], Horn TS, Wojtowicz JG, Rooney JG, Shapiro R: Kinematics, ground reaction force, and muscle balance produced by backward running. J Orthop Sports Phys Ther 1 1 :56-63, 1989 21. Voight ML, Weider DL: Comparative reflex response times of vastus medialis obliques and vastus lateralis in normal subjects and subjects with extensor mechanism dysfunction. Am J Sports Med 19:131-137, 1991 22. Wells RP: The projection of ground re- action force as a predictor of internal joint moments. Bull Prosthet Res 18: 15-19, 1981 23. Winter DA: Biomechanics and Motor Control of Human Movement (2nd Ed), pp 92-93. New York: John Wiley & Sons Inc., 1990 24. Woodall W, Welsh J: A biomechanical basis for rehabilitation programs involving the patellofemoral joint. J Orthop Sports Phys Ther 11:535-542, 1990 Volume 21 Number 5 May 1995 JOSPT