Twenty-one volunteers participated in this study (12 female, 9 male; age: 25.4 ± 5.6 yr; mass: 64.2 ± 5.6 kg; height: 170.2 ± 6.7 cm). Before their participation, all subjects gave informed written consent corresponding to the guidelines of The University of Calgary Ethics Committee.
All subjects were free of lower-extremity pain and injury for a minimum of 6 months before testing and had previously not used foot orthoses. Subjects were initially screened for weekly mileage (15–40 km·wk−1) and two-dimensionally measured maximum foot eversion (angle between heel bisection line and shank bisection line: > 13°) (3) during treadmill running at 4 m·s−1 in the control condition and were thus classified as “pronators” (maximum foot eversion angle: 16.2 ± 3.2°). This inclusion criteria was based on previous findings that foot orthoses are most effective in the treatment of running-related overuse injuries that are linked to biomechanical abnormalities such as excessive pronation (10). All subjects were clinically assessed by one of the authors (R.N.H.). Range of motion of the joints and strength and flexibility of the muscles of the lower extremities were required to lie within normal values, and leg length discrepancy was required to be less than 0.5 cm. Subjects with values outside normal ranges and/or gastrocnemius-soleus equines were excluded. First metatarsal-phalangeal joint dorsiflexion was tested with the first ray loaded and unloaded, and subjects excluded that had structural or functional hallux limitus. Subjects lay prone on a bench with both feet unsupported. The subtalar joint was positioned in its subtalar neutral position (16). A goniometer was used to measure forefoot and rearfoot alignment. Bisection lines were marked on the rearfoot and the lower leg. Weight-bearing rearfoot to leg alignment was assessed with the subject standing with the body weight equally distributed on both legs. The feet were aligned with the hip joint. A goniometer was used to measure the angle between the bisection lines on the rearfoot and lower leg (30). All subjects were required to have flexible arches and to be pronators with foot varus deformity as etiology based on their foot alignment in nonweight-bearing and in weight-bearing when standing and walking. Alignment data for the left and right lower extremity for all participants are given in Table 1. The inclusion criteria in this study matched the general criteria for the prescription of foot orthoses used by podiatrists (28).
All experiments were performed using running sandals (model: Bryce Canyon, The Rockport Company, Canton, MA). The original inserts of the running sandals were removed and replaced by each of four insert conditions, the control, posting, molding, and posting and molding conditions (Table 2). The control condition served as a consistent baseline condition for all subjects. The top layer of all orthotic conditions consisted of 3-mm Spenco (Spenco Medical Corporation, Waco, TX). Plaster casts of both feet in a subtalar neutral position (16) were taken from each subject. The negative casts were placed in calcaneal vertical position and custom-molded orthoses were fabricated to positive molds obtained from the negative casts with the goal to control the pronation movement of the foot during the stance phase of running (Fig. 1) (18). The polypropylene shell stiffness was determined by the subject’s weight to make a subject-specific semirigid device. The same mold was used to fabricate both custom-molded foot orthoses. Six-millimeter extrinsic postings were chosen representing the average intervention for the clinically classified pronators included in this study.
Subjects completed 2 wk of their regular running schedule in the control condition (running sandal plus control insert). After this initial phase, each subject was tested 3× wk−1 for 3 wk (nine sessions per subject). In each of the nine sessions, subjects ran 200 m on an indoor running track with each of the four insert conditions to assess comfort. Subjects were then set up for biomechanical testing at the Human Performance Laboratory at the University of Calgary. The four insert conditions were tested in randomized order. However, before testing each of the three orthotic conditions, subjects ran 50 m in the control condition. Kinematic, kinetic, and EMG data were collected for 12 overground running trials per insert at 4.0 ± 0.2 m·s−1 (heel-toe running; 48 trials per subject per session; total: 9072 trials). Subjects used the orthotic conditions only during the experimental sessions.
Comfort was assessed using a 15-cm visual analog scale. Before each of the three orthotic comfort assessments (Oi), subjects assessed comfort of the control condition (C). Finally, comfort of a randomly selected orthotic condition was assessed as a repeat condition (RO). The resulting testing order was: C-O1-C-O2-C-O3-C-RO. The comfort scale and protocol used in this study has been described in detail elsewhere (17). Not including the repeated comfort rating, one comfort rating for each orthotic condition was obtained for each subject and session.
Kinematic and kinetic data.
Three reflective skin markers (diameter: 12.7 mm) were attached to each of the three segments of the lower extremity (thigh, shank, and foot, respectively) using medical adhesive spray (Hollister Incorporated, Libertyville, IL). Additional joint markers were placed on the anterior superior iliac spine and the greater trochanter, the lateral epicondyle and the patella center, and the lateral malleolus and the insertion of the Achilles tendon to determine hip, knee, and ankle joint centers, respectively. Joint coordinate systems (4) were constructed using the positional information of the segment and joint markers during a standing trial in the control condition.
Kinematic data were collected using seven high-speed cameras (240 Hz; Motion Analysis Corporation, Santa Rosa, CA). Three-dimensional marker traces were reconstructed using Expert Vision Three-Dimensional Analysis software (Motion Analysis Corporation). Ground reaction forces were measured using a force plate (2400 Hz; Kistler AG, Winterthur, Switzerland) that was placed in the center of the runway level with the ground. Kinematic and kinetic data were filtered using a zero-lag quadratic low-pass Butterworth filter with a cut-off frequency of 12 Hz and 50 Hz, respectively. Three-dimensional lower-extremity kinematics and kinetics were calculated using KinTrak software (The University of Calgary, Calgary, Canada) (23), latter using an inverse dynamics approach (1). The angle, force, and moment curves (internal moments) were normalized to touch-down and toe-off, resulting in 101 data points per curve per trial. Maxima were determined from these curves (Table 3) and averaged for each condition, session, and subject.
Myoelectric signals were recorded from seven lower-extremity muscles of the right leg. Bipolar surface electrodes (Ag-AgCl) were placed on the vastus lateralis and medialis, rectus femoris, biceps femoris (long head), tibialis anterior, peroneus longus, and gastrocnemius medialis muscles after removing the hair and cleaning using isopropyl wipes, and then secured using Cover-Roll stretch tape (Beiersdorf AG, Hamburg, Germany). Each electrode was 10 mm in diameter with an intra-electrode distance of 22 mm. A ground electrode was placed on the tibial tuberosity. The placement of the electrodes was marked to ensure similar placement for all nine sessions. The EMG signals were preamplified at source and recorded using a BioVision system (BioVision, Wehrheim, Germany) at 2400 Hz. Timing of heel-strike and toe-off for one step per trial was taken from the ground reaction force data. Using these two events, the EMG data could be related to different phases of ground contact during running. EMG data for each trial was checked for crosstalk by cross-correlating the raw EMG signals between muscles. The correlation coefficients for all muscle combinations of accepted trials were smaller than 0.500.
Wavelet analysis was used to resolve the EMG signals simultaneously into their intensity in time and frequency space (31). The intensity obtained using this wavelet analysis represents a close approximation of the power of the EMG signal. The wavelet analysis consisted of three steps: 1) computation of the wavelet-transformed EMG signal using a filter bank of wavelets including intensity and damping factors, 2) computation of the intensity of the wavelet-transformed signal by adding its square and the square of its time derivative divided by the center frequency for each wavelet, and 3) application of a Gauss filter to the wavelet transformed signal to eliminate oscillations resulting from interference as described by Wakeling et al. (34). A filterbank of 11 wavelets was used (31), and a wavelet domain was defined as the EMG intensity over time corresponding to each of the 11 wavelets. Based on results of previous studies (33,34), a low-frequency band was defined as frequencies between 25 and 82 Hz and a high-frequency band as frequencies between 142 and 300 Hz. The choice of wavelet domains 2 and 3 as a representation of the low-frequency band and wavelet domains 6–8 as a representation of the high frequency allowed for a clear distinction between the low- and high-frequency bands. The global EMG intensity was defined as the sum of EMG intensities for wavelet domains 1 through 8.
Myoelectric signals measured using surface electrodes are attenuated by the soft tissues such as fat overlying the muscle to be measured. To account for these intersubject differences and to allow for comparisons of orthotic effects between subjects and sessions, the global, low, and high intensities were normalized for each subject and session so that the maximum of the total intensity for the control condition had a value of one. The primary functions of muscle activity vary throughout the stance phase of running. For instance, before heel-strike, the foot is not in contact with the ground and no feedback information is yet available from the ground reaction force, and thus the EMG intensity in this interval is controlled by a feed-forward mechanism. The main functions of muscle activity before heel-strike are to stabilize the joints of the lower extremity and to tune the muscle of the lower extremity to minimize possible soft tissue vibrations resulting from the impact of the heel with the ground (22). EMG intensity after heel-strike is initially also controlled by a feed-forward mechanism. However, EMG intensity during this interval may also contain stretch-reflex-related responses where the impact of the foot with the ground at heel-strike acts as a signal into the body. EMG intensities for the remaining time of stance phase are primarily related to muscle forces that accelerate and support the body. Thus, global, low, and high intensities were finally averaged over the preheel-strike interval (50 ms before heel-strike), the postheel-strike (50 ms after heel-strike), and phase 1 (30 to 100% of stance phase), resulting in nine EMG variables per muscle (Table 3). Due to technical problems with the ground reaction force measurements, subject 1 was eliminated from this study. Therefore, the results and discussion sections are based on data for 20 subjects.
All statistical tests were performed using SPSS version 11.5.0 (SPSS Inc., Chicago, IL). Significant differences in comfort ratings between the three orthotic conditions were detected using repeated measures ANOVA, and repeated measures Student t-tests were used as post hoc tests where appropriate with the significance level set at α = 0.05 (36); 95% confidence intervals for all variables and conditions were determined. The relationship between differences in comfort and changes in kinematic, kinetic, and EMG variables in response to foot orthoses was determined using linear regression analysis. The 92 independent variables in the regression analysis were the changes in kinematic, kinetic, and EMG variables for the orthotic conditions compared with the control condition (Table 3). The dependent variable was the differences in comfort rating between the orthotic conditions and the control condition. A stepwise method was used with inclusion criteria P < 0.05 and exclusion criteria P > 0.10. Discriminant analysis was used to determine whether the predicting variables from the regression analysis were able to discriminate between the three orthotic conditions. As only one comfort rating for each orthotic condition was obtained for each subject and session, one average value per subject and session was calculated for each of the kinematic, kinetic, and EMG variables and used in the regression and discriminant analyses.
The intraclass coefficient between first and repeated comfort assessments for all conditions was 0.766, similar to the intraclass coefficient reported by Mündermann et al. (19). Average differences in comfort rating between the orthotic conditions and the control condition are shown in Figure 2 ([lower, upper] confidence limits; posting: [−3.1, −0.8]; molding: [0.4, 3.4]; and posting and molding: [−1.1, 1.9]). The average comfort rating for the molding condition was higher than the comfort ratings for all other conditions. The average comfort rating for the posting and molding condition was very similar to the average comfort rating for the control condition. Average differences in comfort rating of the three orthotic conditions compared with the control condition for 20 subjects are shown in Figure 3. Average differences in selected kinematic and kinetic variables between the orthotic conditions and the control condition are shown in Figure 4.
The covariance matrix revealed that in more than 91.6% of cross-correlation cases, the absolute value of the correlation coefficient |r| for correlation between variables was smaller than 0.300. Regression analysis revealed that 34.9% of the differences in comfort due to foot orthoses can be explained by changes in Δθ, αplantar,max, β’inv,max, Mflex,ankle, Fz,impact, Sta,low,pre, Sta,high,post, Sta,low,post, Sta,low,phase1, Spl,high,pre, Spl,low,pre, Spl,low,post, Svl,high,pre, Svl,low,post, and Svm,glob,phase1. The relationship between differences in comfort and changes in these variables was specified as EQUATION
Unstandardized and standardized regression coefficients for these 15 kinematic, kinetic, and EMG variables are listed in Table 4. The standardized regression coefficients revealed that changes in tibialis anterior intensity in the low frequencies during phase 1 and during the preheel-strike interval (Beta(Sta,low,phase1) = 0.510 and Beta(Sta,low,pre) = −0.437) were the most important predictors for differences in comfort where Sta,low,phase1 was positively correlated and Sta,low,pre was correlated negatively with differences in comfort.
The same 15 variables classified 75.0% of cases correctly to the corresponding orthotic condition (Table 5). Discriminant function coefficients and within-group correlation for two canonical discriminant functions are given in Table 6. A scatterplot of the values of the two canonical discriminant functions is shown in Figure 5. Points representing the posting condition were clearly different from points representing the molding and the posting and molding condition. The values for the molding and posting and molding conditions were very similar for both discriminant functions. This observation is reflected in the large percentage of correctly classified posting cases (96.0%) compared with a much smaller percentage for the molding and posting and molding conditions (61.0 and 68.0%, respectively;Table 5). Discriminant function 1 resulted in a significant classification (P < 0.001), whereas discriminant function 2 showed similar values for all three orthotic conditions (P = 0.712;Fig. 5).
Within-group correlations between maximum ankle plantarflexion, maximum ankle plantarflexion moment, impact force, tibialis anterior intensity in the high frequencies after heel-strike and in the low frequencies during phase 1, peroneus longus intensity in the low frequencies after heel-strike, and vastus lateralis intensity in the high frequencies before heel-strike, and values of discriminant function 1 were higher than correlations between these variables and values of discriminant function 2 (Table 6). In comparison, within-group correlations between delta internal tibia rotation, maximum inversion velocity, tibialis anterior intensity in the low frequencies before and after heel-strike, peroneus longus intensity in the high and low frequencies before heel-strike, vastus lateralis intensity in the low frequencies after heel-strike, and global vastus medialis intensity during phase 1 and values of discriminant function 2 were higher.
The purpose of this study was to determine the relationship between differences in comfort and changes in lower-extremity kinematic and kinetic variables and muscle activity in response to foot orthoses. To our knowledge, this is the first study that controlled for a variety of clinical variables including lower-extremity pain and injury, previous use of foot orthoses, weekly mileage, maximum foot eversion, range of motion of the joints, and strength and flexibility of the muscles of the lower extremities, leg-length discrepancy, gastrocnemius-soleus equines, structural or functional hallux limitus, flexible arch, and etiology for pronation. Contrary to the literature (8,21,24,29), in the current study, the effects of foot orthoses on kinematic, kinetic, and EMG variables and comfort were significant and systematic, which is likely due to the controlled nature of this study. Thus, relating comfort to kinematic, kinetic, and EMG variables is relevant for this group of subjects. However, the relationship between comfort and kinematic, kinetic, and EMG variables may be specific to this group of subjects and the particular foot orthoses used and may not be readily generalized. Despite the existing body of literature and the controversial results of previous studies, this study represents an initial step toward a better understanding of the effects of foot orthoses.
The differences in comfort between orthotic conditions are similar to the results of earlier studies (19,20). Both studies used the same comfort scale as the one used in the current study. Significant differences between insert conditions ranged from one to five comfort points. Similarly, in the current study, differences between orthotic conditions were greater than one comfort point and statistically significant (Fig. 2). The magnitude of differences in comfort rating between the orthotic conditions and the control condition differed between subjects (Fig. 3). This observation is in agreement with results of earlier studies (17,19,20). Distinct differences in comfort ratings for different orthotic conditions and a high intraclass coefficient for repeated measures showed that comfort of foot orthoses can be quantified.
Several factors may affect comfort including fit of foot orthoses or footwear in the different regions of the foot. It has been suggested that comfort of fit is the most important aspect of footwear design and that closeness of match of the footwear shape and that of the human foot is one of the most important elements that characterize comfort fit (12). As two of the orthotic conditions used in the current study were molded to and thus closely matched each individual’s foot, one would expect higher comfort ratings for these two conditions. Indeed, comfort ratings for the molding and posting and molding conditions were greater than comfort ratings for the nonmolded control and posting conditions. However, the difference in comfort rating between the posting and molding condition and the control condition was not substantial.
In general, custom-molding of foot orthoses appears to be a feature that increases comfort. Adding posting to custom-molded foot orthoses does not affect the fit between the foot and the foot orthoses. However, posting of the custom-molded foot orthoses did reduce comfort of the foot orthoses. In fact, the average comfort rating of the combined posting and molding was approximately equal to the sum of the average comfort ratings for the isolated posting condition and the isolated molding condition. Consequently, custom-molding of foot orthoses is comfortable as long as the mold is kept neutral, and factors other than fit of foot orthoses must play a role for comfort of foot orthoses. It is speculated that posting causes a change in the mechanics of the lower extremity during the stance phase of running and that such a mechanical change would be reflected in biomechanical variables. Indeed, 34.9% of differences in comfort were explained by changes in 15 kinematic, kinetic, and EMG variables. Thus, the results of this study support the hypothesis that differences in comfort can be partially explained by changes in lower extremity kinematics, kinetics, and muscle activity.
The same 15 kinematic, kinetic, and EMG variables that partially explain differences in comfort of foot orthoses also classified 75.0% of cases correctly to the corresponding orthotic condition. Thus, changes in these variables are not only related to differences in comfort but are also functionally relevant. Impact forces (Fz,impact) act as input signals into the body and influence soft tissue vibrations. It has been speculated that the body reacts to changes in the input signal by tuning muscle activity to minimize soft tissue vibrations (24). The impact force causes a shock wave to travel through the body from distal to proximal. The magnitude of this shock decreases as it travels proximally (2). Thus, the more proximally a muscle is located the less muscle activity is required to dampen possible soft tissue vibrations. In the current study, 7 of the 10 important EMG variables were EMG intensities in shank muscles that are located more distally compared with thigh muscles and, thus, were likely more affected by changes in the impact force.
Changes in EMG intensity in lower-extremity muscles predicted differences in comfort better than changes in impact force (Beta-values;Table 4). Thus, other factors causing changes in muscle activity must be important for comfort. For instance, a greater range of tibia internal rotation and higher maximum plantarflexion were associated with greater comfort. However, differences in delta internal tibia rotation between orthotic conditions were very small (<0.5°). Although delta internal tibia rotation was a good predictor for comfort in the current study, it must be questioned whether such small differences between orthotic conditions are clinically relevant. In general, these differences in lower-extremity movement and muscle activity were perceived as subjective comfort. Increased muscle activation associated with a reduction in comfort may also lead to an earlier on-set of fatigue. It has been suggested that fatigue of the peroneus longus muscle is the dominant cause of lack of foot stability (9) and that muscular fatigue in the tibialis anterior muscle is the origin of running injuries such as tibial stress fractures (14). Thus, differences in comfort and their relationship with changes in kinematic, kinetic, and EMG variables may be relevant for an early detection of potential running injuries.
The 15 kinematic, kinetic, and EMG variables that partially explain differences in comfort of foot orthoses did not discriminate well between the effects of molding and posting and molding (Fig. 5;Table 5). In comparison, the effects of molding and posting and molding were distinctly different from the effects of the isolated posting condition. Thus, the effects of posting when combined with molding on kinematic, kinetic, and EMG variables were small indicating that molding of foot orthoses overrides the effects of posting of foot orthoses. With respect to comfort, the posting and molding condition consisted of both positive and negative features that resulted in no significant difference in comfort rating compared with the control condition (Fig. 2). If increasing comfort is one of the most important functions of foot orthoses, then foot orthoses should be custom-molded with no additional posting unless clinically justified.
The amount of posting used in this study was consistent for all subjects. However, some subjects may have required slightly more or less posting. Therefore, the effects of custom-molded and custom-posted foot orthoses on comfort, kinematic, kinetic, and EMG variables may have been slightly different. Furthermore, subjects participating in the current study were selected very carefully. It is possible that the relationship between kinematic, kinetic, and EMG variables and comfort found in the current study is specific to this subject group. Nonetheless, more than 75% of recreational runners that were initially screened for participation in this study met all inclusion criteria. Therefore, it is suggested that the findings of this study are relevant for the general population of recreational runners.
In general, comfort is an important and relevant feature of foot orthoses for running activities. Changes in kinematic, kinetic, and EMG variables as a response to foot orthoses are related to differences in comfort of foot orthoses for the specific group of subjects under investigation. Combining posting with custom-molding of foot orthoses reduced comfort of the foot orthoses and had little effect on kinematic, kinetic, and EMG variables. Future studies should investigate the long-term effects of different components of foot orthoses on comfort and kinematic, kinetic, and EMG variables to determine whether these variables are relevant for running injuries. Further research is necessary to quantify changes in comfort with continual usage of foot orthoses especially of posted orthoses as they often take longer to adjust to.
The authors thank Maggie Andersen and Michelle New for their help in data collection and processing and the Prescription Foot Orthotic Laboratory Association (PFOLA), the Canadian Department of Foreign Affairs and International Trade, Paris Orthotics Ltd., the International Society of Biomechanics (ISB), and the Rockport Company for their financial support.
1. Andriacchi, T. P., and R. P. Mikosz. Musculoskeletal dynamics, locomotion, and clinical applications. In: Basic Orthopaedic Biomechanics, V. C. Mow and W. C. Hayes (Eds.). New York: Raven Press, 1991, pp. 51–92.
2. Bhattacharya, A., E. P. Mccutcheon, E. Shvartz, and J. E. Greenleaf. Body acceleration distribution and O2
uptake in humans during running and jumping. J. Appl. Physiol. 49: 881–887, 1980.
3. Clarke, T. E., E. C. Frederick, and C. J. Hamill. The study of rearfoot movement in running. In: Sport Shoes and Playing Surfaces, E. C. Frederick (Eds.). Champaign, IL: Human Kinetics, 1984, pp. 166–189.
4. Cole, G. K., B. M. Nigg, J. L. Ronsky, and M. R. Yeadon. Application of the joint coordinate system to three-dimensional joint attitude and movement representation: a standardization proposal. J. Biomech. Eng. 115: 344–349, 1993.
5. Crenshaw, S. J., F. E. Pollo, and E. F. Calton. Effects of lateral-wedged insoles on kinetics at the knee. Clin. Orthop. 375: 185–192, 2000.
6. D’Ambrosia, R. D. Orthotic devices in running injuries. Clin. Sports Med. 4: 611–618, 1985.
7. Dixon, S. J., and D. G. Kerwin. The influence of heel lift manipulation on sagittal plane kinematics in running. J. Appl. Biomech. 15: 139–151, 1999.
8. Eng, J. J., and M. R. Pierrynowski. The effect of foot orthotics on three-dimensional lower-limb kinematics during walking and running. Phys. Ther. 74: 836–844, 1994.
9. Gefen, A. Biomechanical analysis of fatigue-related foot injury mechanisms in athletes and recruits during intensive marching. Med. Biol. Eng. Comput. 40: 302–310, 2002.
10. Gross, M. L., L. B. Davlin, and P. M. Evanski. Effectiveness of orthotic shoe inserts in the long-distance runner. Am. J. Sports Med. 19: 409–412, 1991.
11. Gross, M. L., and R. C. Napoli. Treatment of lower extremity injuries with orthotic shoe inserts: an overview. Sports Med. 15: 66–70, 1993.
12. Hawes, M. R., D. Sovak, M. Miyashita, S. J. Kang, Y. Yoshihuku, and S. Tanaka. Ethnic differences in forefoot shape and the determination of shoe comfort. Ergonomics 37: 187–196, 1994.
13. Hennig, E. M., G. A. Valiant, and Q. Liu. Biomechanical variables and the perception of cushioning for running in various types of footwear. J. Appl. Biomech. 12: 143–150, 1996.
14. Landry, M., and C. J. Zebas. Biomechanical principles in common running injuries. J. Am. Podiatr. Med. Assoc. 75: 48–52, 1985.
15. Lockard, M. A. Foot orthoses. Phys. Ther. 68: 1866–1873, 1988.
16. Losito, J. M. Impression casting techniques. In: Clinical Biomechanics of the Lower Extremities, R. L. Valmassy (Ed.). St Louis: CV Mosby, 1996, p. 279.
17. Miller, J. E., B. M. Nigg, W. Liu, D. J. Stefanyshyn, and M. A. Nurse. Influence of foot, leg and shoe characteristics on subjective comfort. Foot Ankle Int. 21: 759–767, 2000.
18. Mündermann, A., B. M. Nigg, R. N. Humble, and D. J. Stefanyshyn. Foot orthotics affect lower extremity kinematics and kinetics during running. Clin. Biomech. 18: 254–262, 2003.
19. Mündermann, A., B. M. Nigg, D. J. Stefanyshyn, and R. N. Humble. Development of a reliable method to assess footwear comfort during running. Gait Posture 16: 38–45, 2002.
20. Mündermann, A., D. J. Stefanyshyn, and B. M. Nigg. Relationship between footwear comfort of shoe inserts and anthropometric and sensory factors. Med. Sci. Sports Exerc. 33: 1939–1945, 2001.
21. Nawoczenski, D. A., T. M. Cook, and C. L. Saltzman. The effect of foot orthotics on three-dimensional kinematics of the leg and rearfoot during running. J. Orthop. Sports Phys. Ther. 21: 317–327, 1995.
22. Nigg, B. M. Impact forces in running. Curr. Opin. Orthop. 8: 43–47, 1997.
23. Nigg, B. M., A. Khan, V. Fisher, and D. Stefanyshyn. Effect of shoe insert construction on foot and leg movement. Med. Sci. Sports Exerc. 30: 550–555, 1998.
24. Nigg, B. M., P. Stergiou, G. Cole, D. Stefanyshyn, A. Mündermann, and N. Humble. Effect of shoe inserts on kinematics, center of pressure, and leg joint moments during running. Med. Sci. Sports Exerc. 35: 314–319, 2003.
25. Novick, A., and D. L. Kelley. Position and movement changes of the foot with orthotic intervention during loading response of gait. J. Sports. Phys. Ther. 11: 301–311, 1990.
26. Richie, D. H., Jr., and W. R. Olson. Orthoses for athletic overuse injuries: comparison of two component materials. J. Am. Podiatr. Med. Assoc. 83: 492–498, 1993.
27. Robbins, S. E., and G. J. Gouw. Athletic footwear: unsafe due to perceptual illusions. Med. Sci. Sports Exerc. 23: 217–224, 1991.
28. Root, M. L. Development of the functional orthosis. Clin. Podiatr. Med. Surg. 11: 183–210, 1994.
29. Stacoff, A., C. Reinschmidt, B. M. Nigg, et al. Effects of foot orthoses on skeletal motion during running. Clin. Biomech. 15: 54–64, 2000.
30. Valmassy, R. L. Clinical Biomechanics of the Lower Extremities. St. Louis: Mosby Publishing Yearbook, 1996.
31. von Tscharner, V. Intensity analysis in time-frequency space of surface myoelectric signals by wavelets of specified resolution. J. Electromyogr. Kinesiol. 10: 433–445, 2000.
32. Wakeling, J. M., S. A. Pascual, and B. M. Nigg. Altering muscle activity in the lower extremities by running with different shoes. Med. Sci. Sports Exerc. 34: 1529–1532, 2002.
33. Wakeling, J. M., S. A. Pascual, B. M. Nigg, and V. von Tscharner. Surface EMG shows distinct populations of muscle activity when measured during sustained sub-maximal exercise. Eur. J. Appl. Physiol. 86: 40–47, 2001.
34. Wakeling, J. M., V. Von Tscharner, B. M. Nigg, and P. Stergiou. Muscle activity in the leg is tuned in response to ground reaction forces. J. Appl. Physiol. 91: 1307–1317, 2001.
35. Wu, K. K. Foot Orthoses. In: Atlas of orthoses and assistive devices, B. Goldberg, and J. D. Hsue (Eds.). St. Louis: American Academy of Orthopaedic Surgeons, Mosby-Year Book, Inc., 1997, pp. 463–476.
36. Zar, J. H. Biostatistical Analysis. Upper Saddle River, NJ: Prentice-Hall, Inc., 1999.