Approximately 36 million Americans participate in running each year, with 10.5 million running at least 100 d·yr−1. Although running injuries are well understood medically, their potential risk factors are not. Thus, we presently have limited ability to identify individuals at high risk for overuse injuries.
Purpose: This study aimed to identify behavioral and physiological risk factors that influenced potential knee injury mechanisms, including knee joint forces and knee moments.
Methods: Participants included 20 adults ranging in age from 20 to 55 yr (n = 7 males and n = 13 females). During the first screening visit, quadriceps and hamstring flexibility was assessed, and Q-angle, height, and weight were measured. During the second screening visit, participants completed a series of questionnaires and a gait analysis to calculate knee joint loads. An isokinetic dynamometer was used to measure eccentric and concentric knee extension strength.
Results: Body weight (r = 0.48, P = 0.03), weekly mileage (r = 0.62, P = 0.005), and concentric knee extension strength (r=0.68, P = 0.0001) were significantly correlated with tibiofemoral compressive force. Knee extension moment displayed a negative correlation with hamstring flexibility (r = −0.47, P = 0.04). Both weekly mileage (r = 0.50, P = 0.03) and concentric knee extension strength (r = 0.60, P = 0.01) had significant positive correlations with patellofemoral force.
Conclusion: The results of this study relate larger knee joint loads to poor hamstring flexibility, greater body weight, greater weekly mileage, and greater muscular strength. Most of these risk factors could potentially be modified to reduce joint loads to lower the risk of injury.
1J.B. Snow Biomechanics Laboratory, Department of Health and Exercise Science, Wake Forest University, Winston-Salem, NC; 2Department of Biostatistical Sciences, Wake Forest University School of Medicine, Winston-Salem, NC; 3Department of Orthopaedic Surgery and Sports Medicine, Wake Forest University School of Medicine, Winston-Salem, NC; and 4Department of Exercise and Sport Science, East Carolina University, Greenville, NC
Address for correspondence: Stephen P. Messier, Ph.D., J.B. Snow Biomechanics Laboratory, Department of Health and Exercise Science, Wake Forest University, Winston-Salem, NC 27109; E-mail: email@example.com.
Submitted for publication September 2007.
Accepted for publication May 2008.
Approximately 36 million Americans participate in running each year, with 10.5 million running at least 100 d·yr−1 (1). Epidemiologic studies indicate that running reduces the threat of some chronic diseases, decreases disability and pain, and lowers healthcare costs. Unfortunately, these benefits can be offset by annual overuse injuries that cause as many as 65% (range = 46-65%) of all runners to stop running and to seek treatment (21,23-25,42). Common treatments include rest, drugs, injections, or surgery, which may relieve symptoms but do not always address the root cause, potentially resulting in poor medical outcomes (39).
Running can be viewed as a series of collisions with the ground. Hreljac (19) suggests that the repeated application of relatively high-impact forces without sufficient time between them to allow positive remodeling ultimately results in an overuse injury. The knee is the most common injury site (3,6). Kinetic factors that are thought to cause knee injuries in runners include ground reaction and knee joint forces and moments (13). Unfortunately, few studies use these parameters to identify potential mechanisms of overuse injury. Ferber et al. (13) and Grimston et al. (17) found greater ground reaction forces in runners with stress fractures than in healthy runners. In a series of studies, we found that vertical impulse and propulsive forces were greater in injured versus noninjured runners, but maximum vertical impact forces and loading rates were lower in the injured groups (11,26,28,30,32).
Considerable muscle coactivation occurs during running (2). Hence, several authors have used modeling techniques that include some of the major knee musculature to estimate the knee joint forces. Scott and Winter (41) used data from three healthy runners to model the lower extremity and found that lower leg peak compressive forces ranged from 10 to 14 times body weight (BW). Using a less sophisticated sagittal plane model, Flynn and Soutas-Little (14) estimated that patellofemoral compressive forces ranged between 4.3 and 6.9 BW. Using a three-dimensional model, Glitsch and Baumann (15) reported peak resultant knee joint forces of approximately 13 BW in a subject running at 5 m·s−1. Studies on lower extremity loads and overuse injury have shown that impact forces at commonly injured sites are extremely high (41), and previously injured runners have higher lower extremity loads (33,34) and greater hip joint range of motion than noninjured runners (40). Although these data do not provide unequivocal support, it appears that high forces applied to the lower extremity tissues during running are associated with running-related injuries and therefore may be mechanisms of knee injuries.
Behavioral (e.g., training history, injury history) and physiological (e.g., quadriceps and hamstring flexibility, Q-angle, arch height, strength) risk factors are thought to interact with potential biomechanical mechanisms (knee joint forces and moments) to cause knee injury (see Fig. 1); however, to our knowledge, no study has linked these risk factors with potential underlying mechanisms of injury. Cowan et al. (8) followed 246 army recruits with no history of injury and found that those with cavus feet had a higher injury risk than those with normal or planus feet (odds ratio [OR] = 6.12). Macera et al. (24) examined and followed 583 habitual runners for 12 months. Approximately 50% sustained at least one overuse injury. Weekly mileage was the most important risk factor (OR = 2.9), followed by previous injury (OR = 2.7) and less than 3-yr running experience (OR = 2.2). Messier et al. (28) found that a Q-angle over 16° and a high arch were associated with anterior knee pain. Stride length, an important variable in physiological efficiency (4), has also been linked to impact forces during running. Mercer et al. (27) found that overstriding by15% of freely chosen stride length increased peak leg accelerations compared with normal stride length and understriding by 15%. Over thousands of strides, the increased shock to the lower extremity may result in overuse injury.
Our study examined whether the behavioral and physiological risk factors noted above influenced possible knee injury mechanisms. We hypothesized that stride length, Q-angle, BW, body mass index (BMI), and weekly mileage are associated with greater knee joint forces and moments. In contrast, greater knee strength and better flexibility will be associated with lower joint loads.
Twenty noninjured runners (n = 7 males and n = 13 females), recruited through advertisements in the local runners' club newsletter and at local road races, volunteered for the study. Inclusion criteria were age 20 to 55 yr, injury-free for at least 1 yr, and weekly mileage of 10 miles or greater. Exclusion criteria were chronic diseases or orthopedic conditions and pregnancy. Before testing, an informed consent document, approved by the university's institutional review board, was explained to and signed and initialed by each participant.
Measurements and Procedures
During the first visit, height and weight were recorded, informed consent signed, and questionnaires completed, including items on medication use, medical history, training history, and demographics. The second visit consisted of a three-dimensional gait analysis and strength and anthropometric measurements. The data reported for this study involve the dominant leg.
Physiological Risk Factors
A Kin-Com 125E isokinetic dynamometer was used to measure concentric and eccentric flexion and extension strength of the knee of the dominant leg. The order of tests for all participants was 1) concentric protocol (knee extension at 60°·s−1) and 2) eccentric protocol (knee extension at 60°·s−1).
Each subject was secured, with torso and tested leg strapped to the testing chair, hands across the chest, dynamometer axis aligned with the knee, and resistance pad attached to the lower leg proximal to the ankle joint. Gravity-effect torque was calculated based on leg weight at a 45° angle. The activation force for each muscle group was set at 50% of maximal voluntary isometric contraction. The knee extensors were tested through a joint arc from 90° to 30° (0° = full extension). The first and last 10° were deleted to account for any inconsistent effort and the dynamometer's acceleration and deceleration at the end of the range of motion (16). Hence, average torque was calculated between joint angles of 40° and 80°. Trials were spaced by rest periods of 30 to 60 s. Two maximal reproducible trials were averaged from a maximum of six trials for each test.
The reliability of the concentric and eccentric knee extension tests was determined in 10 subjects tested twice, 1 wk apart. The isokinetic knee strength tests displayed good interclass correlations (ICC) for both eccentric and concentric knee extension (ICC = 0.99 and 0.96, respectively).
Each participant was positioned supine on a standard examination table, with the dominant limb flexed at the hip and knee to 90°. Marks were placed over the lateral malleolus, lateral femoral condyle, and greater trochanter. Two vertical rods were secured to each side of the examining table equidistant from its head, with a rod between them 1 ft above its surface. When the participant flexed the hip to 90°, the thigh touched the horizontally placed rod, helping the participant maintain the 90° flexed position. A strap over the pelvis secured the participant to the table. A second strap was placed over the contralateral thigh. While maintaining the 90° flexed hip, the participant extended the knee as far as possible. A digital camera (Fujifilm S-3000, 3.2 megapixels, 6× optical zoom) leveled perpendicular to the sagittal plane took a photograph at maximum extension. The photograph was symmetrically expanded onto an 8.5 × 11-inch sheet, and lines were drawn between the lateral malleolus and the lateral femoral condyle and between the greater trochanter and the lateral femoral condyle. Their angle of intersection was measured using a 360° universal goniometer. Three consecutive trials were averaged to yield a representative value for each limb. This test had an ICC of 0.96 (36).
Quadriceps flexibility was measured with the participant prone, the pelvis strapped to the examination table, and the tester holding the ankle of interest. The participant flexed the knee until resistance was met or the hip began to flex. Maximum knee flexion was measured using the same camera and landmarks as for the hamstring flexibility test. Three consecutive trials were averaged to yield a representative value for each limb. Using a similar technique, Ekstrand et al. (12) had an intratester CV of 0.5%.
The quadriceps angle-the angle between a line that connects the anterior superior iliac spine to the midpoint of the patella and a line that connects the tibial tuberosity to the midpoint of the patella-was measured bilaterally, with the participant in a supine position, using a 360° goniometer and anatomically placed lines. Three measurements were averaged to yield a representative value.
Participants ran on a 22.5-m runway containing an AMTI force platform (500 Hz) interfaced with an AMTI 6-channel amplifier (AMTI, Watertown, MA) and an IBM microcomputer. Runners were tested in their normal training shoes while they ran at an 8-min·mile−1 training pace. The use of a constant running pace for all runners helped control the influence of running velocity on joint kinetics. Peak values for each subject were derived from an ensemble average of three trials. An acceptable trial was one in which the runner ran at an 8-min·mile−1 pace (± 3.5%) and contacted the force platform with the appropriate foot in normal stride. A Lafayette Model 63501 photoelectric control system interfaced with a digital timer with photocells positioned 7.3 m apart that recorded time through the interval was used to calculate running speeds.
Three-dimensional high-speed (200 Hz) digital capture was performed using a six-camera motion analysis Hi-Res system (Motion Analysis Corporation, Santa Rosa, CA) and a set of 25 passive reflective markers arranged in the Cleveland Clinic full-body configuration. Raw coordinate data from the three-dimensional system were smoothed using a Butterworth low-pass digital filter with a cutoff frequency of 6 Hz. The smoothed video coordinate data, ground reaction, and gravitational and inertial forces served as input to calculate temporal and lower extremity kinematic and kinetic variables using Orthotrak and Kintrak software. The forces and the moments were first calculated from the global reference frame (lab coordinate system) and then rotated to the segment coordinate system. Hence, the final force and the moment output were relative to the segment coordinate system.
Inverse dynamics model.
A two-dimensional inverse dynamics model, developed by DeVita and Hortobagyi (10), was modified to a three-dimensional model that incorporated frontal plane knee forces and torques (29,31). It was used to calculate compressive, anteroposterior shear, and resultant knee (i.e., tibiofemoral) joint forces and patellofemoral compressive force. A three-dimensional inverse dynamics analysis of the lower extremity was used to obtain the joint torques and reaction forces at the hip, knee, and ankle. These results, along with the kinematic description of the lower extremity and related anatomical and physiological characteristics, were used to calculate knee muscle, lateral collateral ligament, and joint forces during the stance phase of running.
The lower extremity was modeled as a rigid linked segment system. Magnitude and location of the segmental masses and mass centers and segmental moments of inertia were estimated from the position data using a mathematical model of relative segmental masses reported by Dempster (9) and the participant's anthropometric data. Inverse dynamics using linear and angular Newton-Euler equations of motion were used to calculate the joint reaction forces and moments at each joint. Joint moments represent the internal moments produced by the muscles and other tissues crossing the joints.
The joint moments and reaction forces were used as input for the DeVita knee model to determine hamstring, quadriceps, and gastrocnemius muscle forces, and lateral collateral ligament force. Subsequently, this information was used to determine bone-on-bone tibiofemoral and patellofemoral compressive forces and tibiofemoral shear force in the anteroposterior direction. The model is "torque driven" in that it used the joint torques from the inverse dynamic analysis to determine the muscle forces. The peak values for each subject were derived from an ensemble average of three trials of the dominant limb.
We have successfully used this model to predict knee joint forces in previous studies (29,31). Our knee joint forces are within 20% of the magnitude reported by Glitsch and Baumann (15) who used a three-dimensional model containing 47 muscles. The two principle limitations to the biomechanical model are 1) the assumption of no coactivation by ankle dorsiflexors and hip flexors during the stance phase and 2) the absence of several knee ligaments. The model does include coactivation of knee flexors and extensors during the stance phase, the principle biomechanical site in this proposal. It also includes the lateral collateral ligament, which is important and produces the principle, nonmuscular restraint during the stance phase of walking, particularly in the frontal plane.
SAS statistical package was used to establish the normality of the data and to complete the necessary analyses. Histograms were created to check for the normality of the data before beginning analysis. Outliers were identified through the compilation of scatterplots, and any points that were at least 2.7 SD away from the group mean provided the study team with grounds for exclusion. The regularly accepted outlier exclusion criteria require the excluded value to be at least 3.0 SD away from the mean, but 2.7 was accepted due to the large effect the participant's data had on the correlations for this small sample. Pearson coefficients are presented to show correlations of measured forces with BMI, BW, average weekly mileage, stride length, knee flexibility, concentric and eccentric knee extension strengths, and Q-angle. Results were similar to those with Spearman correlations. With the exception of BMI and BW, correlations were adjusted for differences in BW. This helped to account for differences in weight between males and females without separate analyses by gender, which with the sample sizes would make any correlations problematic.
Behavioral and physiological risk factors.
The mean age of the participants was 36.1 ± 8.1 yr with an average running experience of 8.4 ± 6.7 yr and an average weekly mileage of 25.3 ± 11.6 miles. Additional behavioral risk factors and participant characteristics are noted in Table 1. Mean anthropometric and strength variables that are associated with potential overuse running injuries are shown in Table 2. Mean concentric knee extension strength was 88.1 N·m, and eccentric strength was 15% greater with a mean value of 101 N·m. The correlation between concentric knee extension strength and weekly mileage was 0.66 (P = 0.002).
Potential injury mechanisms: knee joint loads.
The mean peak patellofemoral force during running was4.27 ± 1.65 BW, and the mean peak compressive force was 10.38 ± 1.93 BW. The anteroposterior shear force was lower, with an average magnitude of 1.20 ± 0.43 BW. Additional kinetic variables are noted in Table 3.
We examined the association between risk factors and potential injury mechanisms using Pearson correlations. Weekly mileage (r = 0.50, P = 0.03) and concentric knee extension strength (r = 0.60, P = 0.01) were significantly correlated with patellofemoral compressive force (Table 4). In contrast, BMI, BW, stride length, quadriceps and hamstring flexibility, eccentric knee extension strength, and Q-angle were not significantly correlated with patellofemoral compressive force.
Pearson correlations for peak tibiofemoral compressive force showed significant positive correlations with BW (r = 0.48, P = 0.03), weekly mileage (r = 0.62, P = 0.005), and concentric knee extension strength (r = 0.68, P = 0.001) (Table 4). BMI, stride length, quadriceps and hamstring flexibility, eccentric knee extension strength, and Q-angle were not significantly associated with peak tibiofemoral compressive force. Peak anteroposterior shear force was not significantly correlated with any of the risk factors for injury.
Knee extension moment and hamstring flexibility were significantly correlated (r = −0.47, P = 0.04); poorer flexibility was associated with higher moments. Knee extension moments were marginally associated with weekly mileage (r = 0.45; P = 0.056) and mean concentric knee extension strength (r = 0.44, P = 0.058). No other variables were significantly correlated with knee extension moment. Knee abduction moment was not significantly associated with any of the injury risk factors.
Pearson correlations adjusted for BW revealed significant associations between peak knee extension moment and patellofemoral (r = 0.85, P < 0.0001) and tibiofemoral forces (r = 0.51, P = 0.03). In contrast, peak internal abductor moment was not significantly associated with either patellofemoral (r = −0.35, P = 0.13) or tibiofemoral force (r = −0.15, P = 0.54).
The causes of overuse running injuries have not been identified, but the etiology is clearly diverse (19). Previous studies have identified relationships between potential risk factors and a variety of injuries, with the knee being the most common injury site (11,28). We provide further insight by exploring the relationship between risk factors and knee joint forces and moments, which we consider to be potential mechanisms of overuse knee injuries in runners.
Impact forces between the ground and the runner at landing range from 1.5 to 3 BW and internal forces at the ankle and knee have been estimated between 5 and 14 BW (38,41). Our results agree, with an average peak tibiofemoral compressive force of 10.4 BW. The cumulative effect of forces of this magnitude without sufficient time between them to allow positive remodeling may ultimately result in an overuse injury (19). We support the hypothesis that the high incidence of injury among runners is somehow linked to the repetition of loads that are greater than the asymptomatic limit of the involved tissue (19,20).
The literature generally suggests that greater weekly mileage increases the risk of overuse injury (22,24,25,30), although contradictory results exist (11,26,28,32,35). Our results associate greater weekly mileage with increased patellofemoral and tibiofemoral compressive forces and increased knee extensor strength. Although it may appear intuitive that running greater distances require more muscle strength resulting in greater joint loads, there are data to the contrary. Hickson (18) found that simultaneous training for strength and aerobic endurance resulted in a reduced capacity to develop strength. This was not the case in our cohort, as mileage, joint loads, and knee strength were all positively related.
Increased muscle strength should increase the shock-absorbing capabilities of the muscles surrounding the knee joint, resulting in lower knee joint loads. Neptune et al. (37) used a forward dynamics musculoskeletal model to determine the effects of selectively strengthening the vastus medialis oblique (VMO) on mediolateral patellofemoral joint forces. A 10% increase in VMO strength reduced the lateral patellofemoral loads by 10 N, or 4%, from 230 to 220 N. Larger increases in strength resulted in a linear reduction in joint loads. Although our previous retrospective studies indicated that muscle weakness was a significant etiologic factor among all injuries studied (11,26,28,30,32), thereby supporting improvements in muscle strength as a preventative measure for overuse injury, the results of the present study imply that greater knee strength is associated with larger knee joint loads. Does greater knee strength improve shock absorption, or, as we suggested, does it exacerbate knee joint loads and lead to injury? These relationships remain unclear.
Poor hamstring flexibility was associated with higher knee extension moments. Maximum knee extension moment occurs just before midstance, when the knee is maximally flexed. This moment functions kinematically to stop the knee flexion and the downward motion of the runner and then to extend the knee joint in propelling the runner upward and forward. This moment was also significantly related to higher patellofemoral and tibiofemoral forces. Joint stiffness is a function of a change in knee extensor moment relative to a change in joint angle; hence, poor flexibility could affect changes in joint angle and the knee extensor moment, increasing joint stiffness and knee joint forces. Consequently, poor hamstring flexibility would be detrimental to joint function by increasing joint stiffness and reducing the contribution of the knee to shock attenuation after heel strike.
The effect of BW on knee joint loads appears intuitive. An increase in BW was associated with greater tibiofemoral compressive forces. There is scant evidence that increased BW increases the risk of overuse injury (43). The lack of a significant relationship between patellofemoral forces and BW may account for the absence of a link between BW and patellofemoral pain in runners. Few experienced distance runners are overweight or obese. Heavy individuals select themselves out of running, creating a sample that is skewed toward lean small-framed runners. Our results suggest that within the range of BW tested, higher BW is associated with greater tibiofemoral loads. Hence, increased BW in runners may contribute to trauma of the tibiofemoral joint, such as degenerative joint disease, but may not affect injuries to the patellofemoral joint. There was no significant association between BMI and knee joint loads. These data, combined with the significant correlations of BW with knee joint loads, suggest that total BW, not adiposity, increases the load on the knee joint during running.
The frontal plane internal abduction moment (or external adduction moment) is thought to play a major role in increasing the compressive load transmitted across the knee (29). This evidence is largely restricted to walking, although it appears logical that it would apply to running. Our results showed no significant relationship between the knee abduction moment and any of the risk factors for overuse running injuries. There was also no significant relationship with tibiofemoral or patellofemoral forces, suggesting that in our cohort of runners, the knee extension moment contributed more to the knee joint load than the abductor moment.
Messier et al. (28) found that a Q-angle over 16° was associated with anterior knee pain. Similarly, in an army cohort, a Q-angle in excess of 15° was identified as a risk factor for such injuries as stress fracture and nonacute muscle strain (7). However, Duffey et al. (11), using a much larger sample size than in the Messier study, found no relationship between Q-angle and runners with knee pain. Similarly, Caylor et al. (5) found slightly greater but nonsignificant differences between patients with and without knee pain. Our results showed no significant relationships between Q-angle and any of the force or moment potential risk factors. Taken together, these studies indicate that the role of Q-angle as a risk factor for anterior knee pain in runners is questionable.
The major limitation of this study was the small sample size. Our significant correlations (P = 0.03-0.001) explained between 23% and 46% of the variance (R2 = 0.23-0.46), suggesting that our regression line approximated the real data points reasonably well. Few studies have examined the relationships between joint loads and risk factors for injury. Although there is no definitive proof that knee joint loads are mechanisms for knee overuse injury in runners, it is reasonable to assume that forces applied to body tissues injure these tissues in runners. The current study provides the first data on the relationship between potential mechanisms for knee injury and risk factors.
Adhering to a running program lowers the rate and progression of disability, risk of chronic disease, and mortality rates. Unfortunately, these health benefits are partially counteracted by the high risk of injury, with annual rates of up to 65% of all runners (range = 46-65%) (21,23-25,42). Hreljac (19) stated that although the cause of these injuries is unknown, they are clearly multifactorial and diverse. Noakes (39) indicated that physicians often treat overuse injuries with rest, drugs, injections, and surgery, often relying on clinical experience and failing to get to the root cause and treat it. This failure is due, in part, to the lack of large-scale clinical trials that test interventions to prevent or rehabilitate common overuse running injuries. Our study is a first step in identifying mechanisms of injury that can later be tested in clinical trials. We found that larger knee joint loads were related to poor hamstring flexibility, higher BW, greater weekly mileage, and greater muscular strength. Most of these risk factors could potentially be modified to reduce knee joint loads to lower the risk of injury.
1. American Sports Data. American Sports Data 2003 Superstudy of Sports Participation
. Volume 1. Cortlandt Manor, NY American Sports Data; 2003.
2. Belli A, Kyrolainen H, Komi PV. Moment and power of lower limb joints in running. Int J Sports Med
3. Brunet ME, Cook SD, Brinker MR, Dickinson JA. A survey of running injuries in 1505 competitive and recreational runners. J Sports Med Phys Fitness
4. Cavanagh PR, Williams KR. The effect of stride length variation on oxygen uptake during distance running. Med Sci Sports Exerc
5. Caylor D, Fites R, Worrell TW. The relationship between quadriceps angle and anterior knee pain syndrome. J Orthop Sports Phys Ther
6. Clement DB, Taunton JE, Smart GW, McNicol KL. A survey of overuse running injuries. Phys Sports Med
7. Cowan DN, Jones BH, Frykman PN, et al. Lower limb morphology and risk of overuse injury among male infantry trainees. Med Sci Sports Exerc
8. Cowan DN, Jones BH, Robinson JR. Foot morphologic characteristics and risk of exercise-related injury. Arch Fam Med
9. Dempster WT. Space Requirements for the Seated Operator
. Wright Patterson Air Force Base, OH: United States Department of Defense Air Force; 1959. Report No.: WADC Technical Report 55-159, AD-087-892. Available from: Microflim 1 reel. DOCS D 301.45/6: 55-159 MICROFILM Texas Tech Libraries.
10. DeVita P, Hortobagyi T. Functional knee brace alters predicted knee muscle and joint forces in people with ACL reconstruction during walking. J Appl Biomech
11. Duffey MJ, Martin DF, Cannon DW, Craven T, Messier SP. Etiologic factors associated with anterior knee pain in distance runners. Med Sci Sports Exerc
12. Ekstrand J, Wiktorsson M, Oberg B, Gillquist J. Lower extremity goniometric measurements: a study to determine their reliability. Arch Phys Med Rehabil
13. Ferber RI, McClay-Davis I, Hamill J, Pollard CD, McKeown KA. Kinetic variables in subjects with previous lower extremity stress fractures. Med Sci Sports Exerc
14. Flynn TW, Soutas-Little RW. Patellofemoral joint compressive forces in forward and backward running. J Orthop Sports Phys Ther
15. Glitsch U, Baumann W. The three-dimensional determination of internal loads in the lower extremity. J Biomech
16. Griffin JW, Tooms RE, vander ZR, Bertorini TE, O'Toole ML. Eccentric muscle performance of elbow and knee muscle groups in untrained men and women. Med Sci Sports Exerc
17. Grimston SK, Nigg BM, Fisher V, Ajemian SV. External loads throughout a 45 minute run in stress fracture and non-stress fracture runners. Proceedings of the 14th ISB Congress Societe de Biomecanique; 1993. p.512-3.
18. Hickson RC. Interference of strength development by simultaneously training for strength and endurance. Eur J Appl Physiol
19. Hreljac A. Impact and overuse injuries in runners. Med Sci Sports Exerc
20. Hreljac A, Marshall RN, Hume PA. Evaluation of lower extremity overuse injury potential in runners. Med Sci Sports Exerc
21. Jacobs SJ, Berson BL. Injuries to runners: a study of entrants to a 10,000 meter race. Am J Sports Med
22. James SL, Bates BT, Osternig LR. Injuries to runners. Am J Sports Med
23. Lysholm J, Wiklander J. Injuries in runners. Am J Sports Med
24. Macera CA, Pate RR, Powell KE, Jackson KL, Kendrick JS, Craven TE. Predicting lower-extremity injuries among habitual runners. Arch Intern Med
25. Marti B, Vader JP, Minder CE, Abelin T. On the epidemiology of running injuries. The 1984 Bern Grand-Prix study. Am J Sports Med
26. McCrory JL, Martin DF, Lowery RB, et al. Etiologic factors associated with Achilles tendinitis in runners. Med Sci Sports Exerc
27. Mercer JA, DeVita P, Derrick TR, Bates BT. Individual effects of stride length and frequency on shock attenuation during running. Med Sci Sports Exerc
28. Messier SP, Davis SE, Curl WW, Lowery RB, Pack RJ. Etiologic factors associated with patellofemoral pain in runners. Med Sci Sports Exerc
29. Messier SP, DeVita P, Cowan RE, Seay J, Young HC, Marsh AP. Do older adults with knee osteoarthritis place greater loads on the knee during gait? A preliminary study. Arch Phys Med Rehabil
30. Messier SP, Edwards DG, Martin DF, et al. Etiology of iliotibial band friction syndrome in distance runners. Med Sci Sports Exerc
31. Messier SP, Gutekunst DJ, Davis C, DeVita P. Weight loss reduces knee-joint loads in overweight and obese older adults with knee osteoarthritis. Arthritis Rheum
32. Messier SP, Pittala KA. Etiologic factors associated with selected running injuries. Med Sci Sports Exerc
33. Milner CE, Davis IS, Hamill J. Free moment as a predictor of tibial stress fracture in distance runners. J Biomech
34. Milner CE, Ferber R, Pollard CD, Hamill J, Davis IS. Biomechanical factors associated with tibial stress fracture in female runners. Med Sci Sports Exerc
35. Montgomery LC, Nelson FR, Norton JP, Deuster PA. Orthopedic history and examination in the etiology of overuse injuries. Med Sci Sports Exerc
36. Nelson RT, Bandy WD. Eccentric training and static stretching improve hamstring flexibility of high school males. J Athl Train
37. Neptune RR, Wright IC, van den Bogert AJ. The influence of orthotic devices and vastus medialis strength and timing on patellofemoral loads during running. Clin Biomech (Bristol, Avon)
38. Nigg BM. Biomechanical aspects of running. In: Nigg BM, editor. Biomechanics of Running Shoes
. Champaign (IL): Human Kinetics Publishers, Inc; 1986. pp. 1-25.
39. Noakes TD. Lore of Running
. 3rd ed. Champaign (IL): Human Kinetics; 1991. pp. 449-52.
40. Noehren B, Davis I, Hamill J. ASB clinical biomechanics award winner 2006 prospective study of the biomechanical factors associated with iliotibial band syndrome. Clin Biomech (Bristol, Avon)
41. Scott SH, Winter DA. Internal forces of chronic running injury sites. Med Sci Sports Exerc
42. Walter SD, Hart LE, McIntosh JM, Sutton JR. The Ontario cohort study of running-related injuries. Arch Intern Med
43. Warren BL, Jones CJ. Predicting plantar fasciitis in runners. Med Sci Sports Exerc
Keywords:©2008The American College of Sports Medicine
RUNNING; OVERUSE INJURIES; MECHANISMS OF INJURY; RISK FACTORS FOR INJURY