Clinical Biomechanics 16 (2001) 207212

www.elsevier.com/locate/clinbiomech

Fatigue-induced changes in decline running

Joseph Mizrahi a,*, Oleg Verbitsky a, Eli Isakov b
a

Department of Biomedical Engineering, Technion Israel Institute of Technology, Haifa 32000, Israel b Loewenstein Rehabilitation Hospital, Raanana, Israel Received 30 May 2000; accepted 23 October 2000

Abstract Objective. Study the relation between muscle fatigue during eccentric muscle contractions and kinematics of the legs in downhill running. Design. Decline running on a treadmill was used to acquire data on shock accelerations, muscle activity and kinematics, for comparison with level running. Background. In downhill running, local muscle fatigue is the cause of morphological muscle damage which leads to reduced attenuation of shock accelerations. Methods. Fourteen subjects ran on a treadmill above level-running anaerobic threshold speed for 30 min, in level and 4 decline running. The following were monitored: metabolic fatigue by means of respiratory parameters; muscle fatigue of the quadriceps by means of elevation in myoelectric activity; and kinematic parameters including knee and ankle angles and hip vertical excursion by means of computerized videography. Data on shock transmission reported in previous studies were also used. Results. Quadriceps fatigue develops in parallel to an increasing vertical excursion of the hip in the stance phase of running, enabled by larger dorsi exion of the ankle rather than by increased exion of the knee. Conclusions. The decrease in shock attenuation can be attributed to quadriceps muscle fatigue in parallel to increased vertical excursion of the hips. Relevance Changes in joint exion angles, in conjunction with increased eccentric muscle activity, are relevant to the understanding of the mechanisms of impact attenuation, of fatigue-related injuries and perhaps of degenerative changes. 2001 Elsevier Science Ltd. All rights reserved.
Keywords: Downhill running; Foot strike; Kinematics; Quadriceps fatigue; Shock transmission

1. Introduction It has been reported that although downhill running involves no metabolic fatigue, it can be harmful to the eccentrically acting leg muscles [1,2]. For instance, the reported elevation in myoelectric activity (electromyogram (EMG)) of the quadriceps muscles, reecting increased recruitment of the motor units and local muscle fatigue, has been associated with damage within the muscle tissue [3]. In level running, on the other hand, quadriceps EMG does not change despite the development of metabolic fatigue [4].

Corresponding author. E-mail address: jm@biomed.technion.ac.il (J. Mizrahi).

Few studies have examined the dierences in kinematics between downhill and level running. For one thing, in downhill running the overall change in knee angle from foot strike to maximum exion is higher than in level running [57]. This appears to be the major phase of eccentric work for the knee extensor, which controls the knee angle in the shock-absorbing phase [7]. An additional dierence is the initial knee exion angle at foot strike that is smaller in downhill compared to level running [5,7]. Recently, it has been reported that the attenuation between shank and sacrum impact accelerations was smaller in downhill compared to level running [2]. Thus, the higher knee exion during the initial stance phase in level running is consistent with the reported higher impact attenuation between below and above knee levels [8,9]. It has also been suggested that since the knee angle controls the vertical displacement

0268-0033/01/$ - see front matter 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 8 - 0 0 3 3 ( 0 0 ) 0 0 0 9 1 - 7

208

J. Mizrahi et al. / Clinical Biomechanics 16 (2001) 207212

during the stance phase, it aects impact attenuation via the eective axial stiness [810]. Despite the fact that knee angle and additional kinematic variables are closely connected to impact loading and fatigue-related damage risk [11,12], the effect of fatigue in downhill running on these variables has not as yet been studied. Thus, the aim of the present study was to investigate the variation of selected kinematic variables accompanying muscle activation and their contribution to impact attenuation in downhill compared to level running. Changes in knee and ankle exion angles during late swing and the rst half of the foot contact phase, in conjunction with increased eccentric muscle activity, are relevant to the understanding of the mechanisms of impact attenuation along the musculoskeletal system, of fatigue-related injuries and of degenerative changes. 2. Methods 2.1. Subjects Fourteen males from the Technion's student population of age 24.2 (SD, 3.7) yr, height 175.5 (SD, 5.9) cm, leg length 90.0 (SD, 3.0) cm and body mass 73.2 (SD, 8.3) kg volunteered to participate in this study. The subjects were all recreational runners who, during the three months prior to the tests, have been regularly running for 810 km per week at a speed of 12 km/h approximately. None suered from a chronic or acute disease nor had previous histories of muscle weakness or injury, bone disease or injury, neurological diseases, drug consumption or therapy. Prior to the experiment, they underwent a general medical examination including resting electrocardiogram and blood pressure, which were satisfactory. Each subject provided informal consent according to the local ethical committee's guidelines of the Technion. 2.2. Running tests The running tests were performed on a treadmill (Quinton Q55, Seattle, Washington, USA) [2,15,16]. A few days prior to the running tests, the anaerobic threshold (AT) was determined for each subject by monitoring the respiratory data in level running [2]. There were two running tests, separated by at least one week to ensure fatigue-free initial conditions: one level running and the other at a decline angle of 4. Running was at a speed exceeding the AT speed of each subject by 5% and lasted 30 min. Before the test, a 15 min warming up running on the treadmill was performed.

2.3. Respiratory data Respiratory data (Sensor-Medics 4400, Alpha Technologies, Laguna Hills, CA, USA) were collected to determine the AT speed and to indicate metabolic fatigue [2]. 2.4. Heel-strike induced shock waves Each subject was instrumented with a lightweight (4.2 g) uniaxial accelerometer (Kistler PiezoBeam, type 8634B50, Kistler, Winterthur, Switzerland), connected to a coupler (Kistler Piezotron, type 5122), as previously described [2,13]. The accelerometer was used to indicate the instant of peak acceleration (PA) at foot strike. To accurately determine the levels and timings of the spike acceleration resulting from foot strike, the signals were sampled at 1667 Hz. The data were acquired every 5 min for a time span of 20 s. The PA at impact was dened as the maximal amplitude of the accelerometer transient at foot strike. 2.5. EMG monitoring The EMG of the quadriceps muscle was monitored by two bipolar disposable Ag/AgCl snap surface electrodes (10 mm diameter), 2 cm apart, placed in the center of the rectus femoris belly and connected to a surface EMG system (Atlas Research, Hod-Hasharon, Israel). The EMG data were collected at a sampling rate of 1667 Hz per channel every 5 min for a time span of 20 s, corresponding to approximately 2728 strides. A band-pass lter 20500 Hz was used to remove low and high-frequency artifacts. Processing included, in the time domain, integrated EMG (iEMG) and, in the frequency domain, mean power frequency (MPF). Both these quantities were represented as averages of the 20 s time span. 2.6. Kinematics of the segments of the lower limb Five hemispherical markers of 2 cm diameter were used. The markers were attached to the right leg in the sagittal plane to the greater femoral trochanter, the lateral condyle of femur, the lateral malleolus, below the lateral malleolus and opposite to the head of the fth metatarsal. Video data were collected using a NVM3000EN Panasonic camera (50 frames/s). Calibrating and data processing were accomplished by Ariel Performance Analysis System (Ariel Dynamics, San Diego, USA) software. Fig. 1 depicts a stick diagram of the late swing and foot contact phases of the right leg in running. Since the exact timing of foot strike could not be determined directly from the video data due to its low 50 1/s sampling rate, PA was used to indicate knee and hip positions at

J. Mizrahi et al. / Clinical Biomechanics 16 (2001) 207212

209

3. Results 3.1. Treadmill running speed The average speed in both level and downhill running for all the 14 subjects was 3.53 m/s (SD, 0.19). The AT was exceeded in level running only (by 5%). Metabolic fatigue was not reached in downhill running [2]. 3.2. EMG The normalized rectied EMG envelopes of the quadriceps muscle prior to foot strike and during foot contact are shown in parallel with the stick diagram of the leg in Fig. 1. Averages for all the 14 subjects in level running and downhill running are presented. The EMG envelopes at four dierent time periods within the 30 min running duration are shown. The envelopes were normalized as follows: the maximum value of the rst min envelope was set to 1.0. Each of the other envelopes was rescaled according to the ratio between its maximum and the maximum of the rst min envelope. An increasing activity of the quadriceps is noted from approximately 45 ms before PA to approximately 100 ms after PA. Also seen from Fig. 1 is the fact that the fullwave rectied EMG envelopes of the quadriceps in level running did not change during the course of running as a result of fatigue. Conversely, in downhill running the full-wave rectied EMG envelopes increased as local fatigue in the quadriceps muscle progressed. The highest increase (75% from the beginning to the end of running) took place shortly before the instant of maximum knee cushioning exion (position b). Means for all the subjects of the iEMG values during the course of fatigue are plotted in Fig. 2, top panel. The iEMG values were normalized by dividing them by the value of the rst min of running. Mean power frequency values (MPF) are shown in the bottom panel of Fig. 2. In downhill running, signicant increases during running were noted in the quadriceps iEMG from the 20th min and onwards and in the MPF from the 15th min and onwards. As previously reported, no signicant changes in either iEMG or MPF were noted in level running despite the fact that the AT was exceeded in this case and metabolic fatigue did develop [4]. 3.3. Knee and ankle joint angles Average values of the knee angle at PA (position a in stick diagram of Fig. 1) and of the range of knee exion (i.e., cushioning exion, between positions d and b) are presented in Fig. 3. None of these quantities signicantly changed with time during the course of either level or decline running. It is noted, however, that at PA (position a) the knee was signicantly more extended in downhill compared to level running during all the stages

Fig. 1. Normalized rectied EMG envelopes (averages for all subjects) of the quadriceps prior to foot strike and during foot contact in level (top panel) and downhill (middle panel) running. The context of these graphs within the complete running cycle can be seen from the insert in the top left corner of the gure. PA denotes shock, or peak acceleration, and the asterisk on the abscissa indicates the estimated instant of foot strike. The lines dened in the legend correspond to dierent stages of running. A stick diagram of the late swing and foot contact phases of the right leg in level running is shown in the bottom part of the gure. The time of each stick gure is indicated by the position of the knee along the abscissa. Denition of positions is as follows: a instant of PA; b maximum stance exion; c maximum ight; d maximum knee extension.

heel strike. The knee angles studied were at maximum knee extension position (position d, preceding the instant of PA), at PA (position a), and at maximal stance exion (position b). The cushioning knee exion, i.e., between positions d and b was also calculated. The range of dorsi exion of the ankle was calculated as the angle dierence between position a (at PA) and the position of maximum dorsi exion. The latter was found to occur shortly after position b. The hip levels as determined by the marker at the greater femoral trochanter were studied at the maximum ight position (c), PA position (a) and maximum stance exion position (b). Dierences between these three hip positions were calculated in order to establish the amplitudes of the vertical excursion. 2.7. Statistical analysis Dierences between the results from the various conditions were tested using non-parametric Wilcoxon signed rank test for repeated measures. Statistical signicance was established at P < 0:05. The results are presented by their means and SDs.

210

J. Mizrahi et al. / Clinical Biomechanics 16 (2001) 207212

Fig. 2. Average for all 14 subjects of the Quadriceps iEMG (upper panel) and MPF (lower panel). The vertical bars denote one SD (solid line level running, dashed line decline running). Signicantly dierent from data at the beginning of running, in downhill running (P < 0:05).

Fig. 4. Average for all 14 subjects of hip excursion data. Upper panel: between maximum ight and PA positions; lower panel: between PA and maximum knee exion. Vertical bars denote one SD, solid line level running, dashed line downhill running. Signicantly dierent from the data at the beginning of running, in level running (P < 0:05). Signicantly dierent from data at the beginning of running, in downhill running (P < 0:05). Signicant dierence between level and downhill running (P < 0:05).

3.4. Hip excursion Fig. 4 presents the average hip excursion data as follows: The top panel shows the hip excursion between the maximum ight and PA positions (positions c and a in Fig. 1, respectively). In level running, this quantity increased gradually becoming signicantly higher in the end compared to the beginning of running. In downhill running, this quantity did not change with time during running. The bottom panel presents the hip excursion between the PA and maximum knee exion positions (positions a and b in Fig. 1, respectively). In decline running, this quantity increased gradually during running, becoming signicantly higher from the 15th min onwards, compared to the beginning of running. In level running it did not change with time. However, it was signicantly higher in downhill compared to level running at all stages of running. 4. Discussion The central issue in this study was to examine the quadriceps muscle activity in relation to several kinematical factors in downhill compared to level running. Only in level running was the individual speed at which the tests were conducted above the AT. Of special interest was to verify how the studied factors correlate with impact attenuation along the musculo-skeletal system. This issue was considered important because transmission of impact loading through the body has

Fig. 3. Average for all 14 subjects of knee and ankle angle data. Upper panel: knee angle at PA; middle panel: range of knee exion; lower panel: range of ankle dorsi exion. Vertical bars denote one SD, solid line level running, dashed line decline running. Significantly dierent from data at the beginning of running, in downhill running (P < 0:05). Signicant dierence between level and decline running (P < 0:05).

of running. Additionally, the range of knee cushioning exion was signicantly higher in downhill compared to level running, from the 15th min of running onwards. Average values for the range of dorsi exion of the ankle are presented in Fig. 3, bottom panel. An increase in the range of dorsi exion is noted during the course of decline running only, becoming statistically signicant from the 15th min onwards.

J. Mizrahi et al. / Clinical Biomechanics 16 (2001) 207212

211

been suggested to be a signicant factor in the development of spinal injuries and degenerative changes in joints and articular cartilage [8,9,11,12]. The results of the present study indicate that in downhill running, iEMG and MPF of the EMG of the quadriceps gradually and signicantly increased (Fig. 2). The increased activity of the quadriceps, preceding heel strike and during the initial 75% of the foot contact phase (Fig. 1), has been associated with the shock absorbing capacity of the muscle [14]. Of the kinematic variables monitored in this study, knee angle is more directly related to impact transmission and attenuation due to the following reasons. As shown in recent studies, increasing the knee exion at the instant of heel strike in level running aects impact acceleration above and below the knee joint in somewhat opposite ways [8,9,15,16]. Below the knee, e.g. on the shank, impact acceleration was reported to increase with increasing knee exion. However, the ratio between above knee to below knee impact acceleration was reported to decrease, indicating an improved attenuation of the impact following knee exion increase. Recently published results [2] on the same group of subjects as in the present study have indicated two features related to the impact transmission in downhill running: (a) it is signicantly higher compared to level running; and (b) it gradually increases during the course of running. The smaller knee angle (more extended position) at PA obtained in the present study in downhill running compared to level running (Fig. 3, top panel) is in agreement with the rst of the above features. However, the fact that this angle did not change signicantly during running rules out the possibility that the increase in impact transmission (the second of the above features) is the consequence of changes in the knee angle. A change in impact transmission during steady-state running could supposedly be related to fatigue [2,17,18]. However, since metabolic fatigue did not take place in downhill running, the increase in shock transmission can be attributed only to local fatigue taking place in the quadriceps muscle. Evidence supporting that is shown in Fig. 2 in the increased EMG activity. Previous research has shown that in eccentric work of the knee extensors, muscle damage develops [5]. More recently, results reported from experimental animal models [19] showed that cyto-skeletal disruption occurs within the rst 15 min of cyclic eccentric contraction. Downhill running in particular involves substantial eccentric contraction of the quadriceps where morphological damage accumulates [20]. Thus, morphological injury taking place in the muscle during eccentric activity may indeed lead to inability of the lower limb muscle tissues to attenuate the impact acceleration. Our results also indicated a higher knee cushioning exion in downhill running compared to level running (Fig. 3). This may be due to the slope constraint

imposed on the feet by the decline surface, while the body has to remain in a vertical position. This would indeed necessitate the initially more extended position of the knee at touch down, and would also leave room for more exion in the later stance phase. Unlike level running, the hip excursion between maximum ight position and PA in downhill running did not change during the course of running (Fig. 4). On the other hand, the hip excursion between the instants of PA and minimum knee exion was at all times signicantly higher in decline running compared to level running, which is consistent with the higher knee cushioning exion in decline compared to level running. Also, this quantity gradually increased compared to the rst min of running. With an unchanged knee cushioning exion, the increased hip excursion during the stance phase of running can only be possible through a gradual augmentation of the dorsi exion angle of the ankle (Fig. 3). The decreased ability of the muscle tissues to attenuate impact acceleration may lead, in the short term, to stress fractures [17,18] and, in the long term, to joint degeneration [8,9,11,12]. This study has shown that the muscle fatigue taking place in long distance downhill running and aecting impact shock attenuation is accompanied by an increasing vertical excursion of the hip in the stance phase of running. This is made possible by a larger dorsi exion of the ankle rather than by exion of the knee. Thus, the enhanced ankle exion enables to modulate the vertical stiness of the body via vertical displacement, somewhat compensating for the reduced muscle ability to attenuate impact. Acknowledgements This study was supported by the Israel Ministry of Health, the Segal Foundation and the Henri Gutwirth Promotion of Research Fund. References
[1] Iversen JR, MacMahon TA. Running on an incline. J Biomech Eng 1992;114:43541. [2] Mizrahi J, Verbitsky O, Isakov E. Shock accelerations and attenuation in downhill and level running. Clin Biomech 2000;15:1520. [3] Dick RW, Cavanagh PR. An explanation of the upward drift in oxygen uptake during prolonged sub-maximal downhill running. Med Sci Sports Exerc 1987;19:3107. [4] Mizrahi J, Voloshin A, Russek D, Verbitsky O, Isakov E. The inuence of fatigue on EMG and impact acceleration in running. Basic Appl Myol 1997;7:11926. [5] Buczec FL, Cavanagh PR. Stance phase knee and ankle kinematics and kinetics during level and downhill running. Med Sci Sports Exerc 1990;22:66977. [6] Milliron MJ, Cavanagh PR. Sagittal plane kinematics of the lower extremity during distance running. In: Cavanagh PR,

212

J. Mizrahi et al. / Clinical Biomechanics 16 (2001) 207212 editor. Biomechanics of distance running. Champaign, II, USA: Human Kinetics Publishers; 1990. Eston RG, Mickleborough J, Baltzopoulos V. Eccentric activation and muscle damage: biomechanical and physiological considerations during downhill running. Br J Sport Med 1995; 29:8994. McMahon TA, Valiant GA, Frederick EC. Groutho running. J Appl Physiol 1987;62:232637. Lafortune MA, Lake MJ, Henning EM. Dierential shock transmission response of the human body to impact severity and lower limb posture. J Biomech 1996;29:15317. Farley CT, Gonzalez O. Leg stiness and stride frequency in human running. J Biomech 1996;29:1816. Collins JJ, Whittle MW. Impulsive forces during walking and their clinical implications. Clin Biomech 1989;4:17987. Whittle MW. Generation and attenuation of transient impulsive forces beneath the foot: a review. Gait and Posture 1999;10:26475. Mizrahi J, Susak Z. In-Vivo elastic and damping response of the human leg to impact forces. ASME J Biomech Eng 1982;104:636. [14] Malanga G, DeLisa JA. Gait analysis in the science of rehabilitation. Clinical Observation. In: Delisa JA, editor. Rehabilitation research and development service. 1998. p. 110. [15] Hamill J, Derrick TR, Holt KG. Shock attenuation and stride frequency during running. Hum Mov Sci 1995;14:4560. [16] Derrick TR, Hamill J, Caldwell GE. Energy absorption of impacts during running at various stride lengths. Med Sci Sports Exerc 1998;30:12835. [17] Mizrahi J, Verbitsky O, Isakov E. Fatigue-related loading imbalance on the shank in running: a possible factor in stress fractures. Ann Biomed Eng 2000;28:4639. [18] Mizrahi J, Verbitsky O, Isakov E, Daily D. Eect of fatigue on leg kinematics and impact acceleration in long distance running. Hum Mov Sci 2000;19:13959. [19] Lieber RL, Thornell LA, Friden J. Muscle cytoskeletal disruption occurs within the rst 15 min of cyclic eccentric contraction. J Appl Physiol 1996;80:27884. [20] Lieber RL, Fridn J. Muscle damage is not a function of muscle e force but active strain. J Appl Physiol 1993;74:5206.

[7]

[8] [9] [10] [11] [12] [13]