Influence of Running Shoe Midsole Composition and Custom Foot ...

Journal of Applied Biomechanics, 2009, 25, 54-63 © 2009 Human Kinetics, Inc.

Influence of Running Shoe Midsole Composition and Custom Foot Orthotic Intervention on Lower Extremity Dynamics During Running Christopher L. MacLean,1 Irene S. Davis,2 and Joseph Hamill1 1University

of Massachusetts–Amherst and 2University of Delaware

The purpose of this study was to analyze the influence of varying running shoe midsole composition on lower extremity dynamics with and without a custom foot orthotic intervention. Three-dimensional dynamics were collected on 12 female runners who had completed 6 weeks of custom foot orthotic therapy. Participants completed running trials in 3 running shoe midsole conditions—with and without a custom foot orthotic intervention. Results from the current study revealed that only maximum rearfoot eversion velocity was influenced by the midsole durometer of the shoe. Maximum rearfoot eversion velocity was significantly decreased for the hard shoe compared with the soft shoe. However, the orthotic intervention in the footwear led to significant decreases in several dynamic variables. The results suggest that the major component influencing the rearfoot dynamics was the orthotic device and not the shoe composition. In addition, data suggest that the foot orthoses appear to compensate for the lesser shoe stability enabling it to function in a way similar to that of a shoe of greater stability. Keywords: kinematics, kinetics, foot orthoses, running In the last few years, there has been a renewed interest in studying custom foot orthoses (Mündermann et al., 2003; Williams et al., 2003; MacLean et al., 2006). A major factor that has been missing from the study of custom foot orthoses is the interaction between the MacLean and Hamill are with the Biomechanics Laboratory, Department of Kinesiology, University of Massachusetts– Amherst, Amherst, MA, and Davis is with Department of Physical Therapy, University of Delaware, Newark, DE.

54

orthoses and the footwear used. It remains unclear how foot orthoses interact with footwear to influence lower extremity dynamics. Studies have primarily focused on measurements with and without custom foot orthotic intervention while the footwear type is held constant. However, it is clear that the shoe plays a vital role and it is, in itself, a therapeutic intervention (Hamill et al., 1992). Previous studies have investigated the dynamic influence of varying midsole composition (Hamill et al., 1992) and varus/valgus midsole modifications (Van Woensel & Cavanagh, 1992; O’Connor & Hamill, 2004). Hamill et al. (1992), in a two-dimensional study, reported on the influence of varying hard, medium, and soft midsole composition on several rearfoot and knee kinematic variables. The results revealed that maximum rearfoot and calcaneal eversion angles were significantly greater in the soft shoe when compared with the medium and hard midsole conditions. In addition, the percentage of time to maximum rearfoot and calcaneal eversion was significantly less for the soft shoe midsole condition. However, the knee kinematic variables were not influenced. In that same year, Van Woensel and Cavanagh (1992) investigated the influence of varying the intrinsic posting of running shoes on rearfoot kinematics. The study included three running shoe conditions with the midsole posted or canted in three ways: (1) 10° varus (medial), (2) neutral, and (3) 10° valgus (lateral). The influence of the shoe design on the rearfoot (frontal plane) and knee (sagittal plane) kinematics was analyzed. The authors were interested in evaluating the changes that occurred both in the magnitude and timing of angle maxima but also changes in angular velocities. In general, maximum pronation angle and velocity were significantly increased with the valgus shoe condition, whereas maximum knee flexion angles were not influenced by the shoe perturbations. The relative timing between rearfoot pronation and knee flexion maxima was also significantly increased for the varus and valgus

Shoe Midsole Composition and Custom Orthotic 55

shoe conditions. Both Hamill et al. (1992) and Van Woensel and Cavanagh (1992) demonstrated that both shoe midsole composition and intrinsic posting influence the relative timing of maxima for rearfoot eversion and knee flexion. Lastly, O’Connor and Hamill (2004) investigated the three-dimensional frontal plane kinematics and kinetics in a group of male runners while running in a varus (10°), neutral, and valgus (10°) posted shoe. The authors reported a significant decrease in maximum rearfoot eversion angle for the varus and neutral shoes compared with the valgus condition. In addition, maximum rearfoot eversion velocity was significantly decreased with the varus shoe compared with the valgus shoe. The internal ankle inversion moment was also significantly decreased for the varus and neutral shoes compared with the valgus condition. These three studies are examples of how varying the midsole composition and intrinsic posting in the frontal plane can influence lower extremity dynamics. Research studies on custom foot orthotic intervention have historically been conducted with the running shoe construction held constant (Mündermann et al., 2003; Williams et al., 2003; MacLean et al., 2006). These studies on the dynamic influence of foot orthotic intervention have revealed that the intervention resulted in (1) decreased rearfoot eversion angle and velocity (MacLean et al., 2006), (2) decreased ankle inversion moment (Mündermann et al., 2003; Williams et al., 2003; MacLean et al., 2006), (3) increased knee adduction angle and abduction moment (Williams et al., 2003), (4) increased knee external rotation moment (Mündermann et al., 2003; MacLean et al., 2006), and (5) decreased impact peak and loading rate (Mündermann et al., 2003; MacLean et al., 2006). Therefore, the purpose of current study was to analyze the influence of varying running shoe midsole composition on lower extremity kinematics and kinetics with and without a custom foot orthotic intervention. It was hypothesized that (1) increasing running shoe midsole durometer (increased hardness) will lead to significant decreases in maximum values for all dynamic variables (rearfoot eversion angle and velocity, calcaneal eversion angle, tibial and femoral internal rotation angle, ankle inversion moment and angular impulse, knee external rotation moment and impulse) with the exception of the impact peak and maximum loading rate, where increased magnitudes will be observed; (2) there will be no significant differences between variables when comparing a shoe-only condition (shod condition) having a greater midsole durometer to a custom foot orthotic + shoe condition (CFO condition) with a lesser midsole durometer; and (3) all dynamic variables will be significantly less when comparing the custom foot orthotic to shod conditions when the midsole durometer is held constant. This will be the case for the hard and medium midsole conditions but not for the soft midsole condition.

Methods Subjects A priori sample size prediction was performed using SAS v.8.2 and rearfoot kinematic data (maximum rearfoot eversion angles) from Bates et al. (1979) and Hamill et al. (1992). A sample size of 12 was estimated for a minimal statistical power of 0.80 and  = 0.05. The current study included a group of runners who had completed 6 weeks of custom foot orthotic therapy for all running activities. Subjects were female runners who had (1) maintained a weekly mileage of 15–40 km per week, (2) worn custom foot orthoses for the 6 weeks leading up to the study, and (3) originally presented with an overuse running knee injury before the 6-week CFO intervention. Subjects ranged in age from 19 to 35 years, and their mean height and mass were 1.63 m (SD, 0.08) and 58.5 kg (SD, 6.96), respectively. All subjects were diagnosed by a sports medicine physician (University Health Services, University of Massachusetts–Amherst) with either iliotibial band friction syndrome or patellofemoral pain syndrome in the 6 months leading up to the CFO intervention. Although the relationship between structural characteristics and dynamic function is not well understood (Hamill et al., 1989; McPoil & Cornwall, 1996), none of the subjects exhibited a leg length discrepancy, rigid forefoot varus deformity, gastrocnemius equinus, or structural hallux limitus or rigidus. Approval for the participation of human subjects in this investigation was obtained from the Human Subjects Review Committee at the University of Massachusetts– Amherst. Only runners classified as rearfoot strikers were included in the study.

Experimental Setup During individual over ground running trials, threedimensional kinematic data were collected using an eight-camera, Qualisys QTM system (Qualisys, Inc., Gothenburg, Sweden). Before each testing session, a right-handed global coordinate system was employed and was defined using a rigid L-frame with four markers of known location. A two-marker wand of known length and marker coordinates was used to calibrate the global coordinate system. The origin of the global coordinate system was positioned on an AMTI force platform (Advanced Mechanical Technology, Inc., Watertown, MA). The right-handed global coordinate system was oriented so that the z-axis was vertical, the y-axis was in the anteroposterior direction or in the direction of motion, and the x-axis was in the mediolateral direction. The sampling frequency for all running trials was 240 Hz. The cameras were interfaced with a microcomputer and positioned around the imbedded force platform. Three-dimensional kinetic data were collected using an AMTI (600 mm × 900 mm) strain gauge force platform (Advanced Mechanical Technology, Inc.). Kinetic data

56 MacLean, Davis, and Hamill

were collected at a sampling frequency of 1920 Hz, a sampling range of ±10 V, and a gain of 4000. Kinetic analog signals were input to 6 channels of a 16-bit A/D converter. Digital signals output from the A/D converter were interfaced with a microcomputer and synchronized with the kinematic data using the Qualisys software.

Protocol The data collection for the current study occurred following a 6-week period of custom foot orthotic intervention during all running activity. Before data collection, markers were positioned on each subject. These markers included both calibration and tracking markers. The former were removed following the capture of a stance frame. Calibration markers were positioned on the left and right greater trochanters, right-side medial and lateral femoral condyles, and right-side medial and lateral maleoli, and positioned on the shoe over the first and fifth metatarsal heads. Tracking markers were securely positioned to define the pelvis (L5/S1 joint line, ASIS, iliac crest), thigh (rigid array of four markers), leg (rigid array of four markers), and calcaneus (rigid triad). Calibration markers were left on for the initial stance trial and then removed. The tracking markers remained in position for all running trials. Following the calibration stance trial, each subject performed five acceptable running trials with and without custom foot orthotic intervention in the three running shoe midsole conditions. None of the subjects reported any pain associated with performing the over ground running trials. Orthotic and shoe conditions were completely randomized for each subject. The custom foot orthotic design was a semirigid, functional foot orthosis with a thermoplastic (3 mm copolymer or polypropylene) orthotic shell. Thermoplastic orthotic shells were vacuum-pressed and finished at Paris Orthotics Laboratory. Orthotic shell material selections were based on subject body weight. The device was intrinsically posted (intrinsic forefoot posting) to calcaneal vertical and inverted an additional 5°. Personalizing each device to meet the specific needs of the subject was accomplished by adjusting orthotic shell material thickness based on body weight and posting to calcaneal vertical. Posting methods employed by various health care professionals can vary substantially so a laboratory standard approach was taken. Foot orthoses also included an extrinsic ethyl vinyl acetate (EVA) rearfoot stabilizer with nylon strike plate. The heel cup depth was 18 mm and a minimum cast dressing was used. Lastly, a full-length Microcel Puff top cover was added for additional cushioning. The testing shoes were customized New Balance 801 running shoes with three different midsole compositions: (1) hard (70 Asker C), (2) medium (55 Asker C), and (3) soft (40 Asker C). The EVA material was consistent throughout the entire midsole of the running shoe.

The New Balance 801 running shoe is designed without a heel counter and therefore allowed for the application of a calcaneal marker set directly on the calcaneus. Running velocity was monitored using two photocells positioned 5.94 m apart; one positioned in advance of the force platform and the second following the force platform. Only running trials at a speed of 4.0 m·s−1 (± 5%) were accepted (Mündermann et al., 2003).

Data Reduction Kinematic data for the stance phase of each over-ground running trial were digitized using QTM software (Qualisys, Inc., Gothenberg, Sweden). A nonlinear transformation technique was employed to calculate the 3-dimensional coordinates for each marker from the sets of 2-dimensional coordinates from the eight cameras. Synchronized raw kinematic and kinetic signals were exported from QTM in .C3d format and processed using Visual 3D software (C-Motion, Inc., Rockville, MD). Raw kinematic and kinetic data were low-pass filtered using a fourth order, zero-lag Butterworth digital filter. A cut-off frequency of 12 Hz for the low-pass filter was selected based on a residual analysis technique recommended by Winter (2005). Kinematic data were interpolated to 101 data points, with each data point representing 1% of the stance phase. Three-dimensional segment and joint angles, angular velocities, and internal joint moments were calculated using the Visual 3D software. Joint kinematic and kinetic data are reported about anatomically oriented axes. Three-dimensional segment and joint angles were calculated using an x (flexion/extension)–y (abduction/ adduction)–z (longitudinal rotation) Cardan rotation sequence (Cole et al., 1993). Joint angles and angular velocities are reported as movement of the distal segment relative to the proximal segment. Segment angles are reported relative to the laboratory coordinate system. A Newton–Euler inverse dynamics approach was employed to calculate the joint kinetics. The inverse dynamics approach requires three sources of information: (1) segment inertial parameters, (2) kinematic data, and (3) forces and moments including ground reaction force. The foot, leg, and thigh segments were modeled as frusta of a right cone and the pelvis was modeled as a cylinder. The anthropometric properties—including segment mass, moment of inertia, and center of mass from the subject body height and weight—were derived from Dempster’s 1955 anthropometric data. The standing calibration markers defined segment lengths. In addition, joint centers were defined as the midpoint between the standing calibration markers for ankle and knee. Joint angular velocities and accelerations were computed and combined with the ground reaction force data. Internal joint moments about the ankle and knee were


calculated and reported in the coordinate system of the proximal joint segment. For example, internal ankle moments are reported in the leg segment coordinate system. Time series data for segment and joint angles, angular velocities, and internal joint moments were exported from Visual 3D and further reduced for the parameters of interest using custom MatLab software. Maximum loading rate was computed as the maximum slope of the vertical ground reaction force from heel strike to vertical impact peak.

Statistical Analysis All hypotheses were statistically analyzed using a 2 × 3 (Orthotic [with and without] × Shoe [hard, medium, and soft durometer]) repeated measures analysis of variance. This analysis was used to determine the presence or absence of significant differences in kinematic and kinetic variables. Significance was indicated by  < .05 and an observed power ≥ 0.80. Normality tests (Kolmogorov –Smirnov) were computed using SPSS software and there were no significant findings for any of the variables. All pairwise comparisons were performed using paired samples t tests with the Bonferroni method of adjustment. In addition, effect size (ES) was calculated for each parameter. Cohen (1988) proposed that ES values of 0.2 represent small differences; 0.5, moderate differences; and 0.8 large differences.

Results Ankle For the kinematic variables of interest, there were no significant Orthotic × Shoe interactions. There were significant orthotic main effects for the maximum rearfoot eversion angle (ES = 0.64; p = .001), maximum calcaneal eversion angle (ES = 0.59; p = .002), and maximum rearfoot eversion velocity (ES = 0.74; p < .001) variables. All variables were significantly decreased when orthoses were worn in the shoes. A significant shoe main effect was exhibited for the maximum rearfoot eversion velocity (ES = 0.46; p = .001) variable. Post hoc analysis revealed that the maximum rearfoot eversion velocity was significantly decreased for the hard shoe compared with the soft shoe. There was a significant Orthotic × Shoe interaction for the maximum ankle inversion moment (ES = 0.40; p = .004). Post hoc analysis revealed that the maximum ankle inversion moment was significantly decreased when orthoses were worn in the hard shoe compared with the medium shoe (ES = 0.47; p = .009). In addition, there were significant orthotic main effects for the maximum ankle inversion moment (ES = 0.67; p = .001), ankle inversion angular impulse during the loading phase (ES = 0.76; p < .001), vertical impact peak (ES =

0.62; p = .001), and maximum loading rate (ES = 0.67; p = .001) variables. All variables were significantly decreased for the custom foot orthotic condition (Table 1).

Knee There were no significant interactions or shoe main effects for any of the knee variables. However, there was a significant orthotic main effect (ES = 0.48; p = .008) for the maximum tibial internal rotation angle variable. Maximum tibial internal rotation was significantly decreased when foot orthoses were worn in the shoe. In addition, there was a significant orthotic main effect (ES = 0.62; p = .001) for the knee external rotation angular impulse variable during the loading phase. The angular impulse was significantly increased when orthoses were worn in the shoes. There were no other significant interactions, orthotic or shoe main effects for the knee kinetic variables of interest (Table 2).

Discussion The purpose of the current study was to investigate the influence of varying running shoe midsole composition on lower extremity dynamics with and without custom foot orthotic intervention. The first research aim was to determine whether varying midsole composition influences lower extremity dynamics with (shoe and orthotic) and without (shoe-only) the custom foot orthotic intervention. It was hypothesized that all kinematic variables would be significantly decreased as midsole composition was increased. This hypothesis was supported for the maximum rearfoot eversion velocity variable (Figures 1 and 2). However, there were no significant shoe main effects for maximum rearfoot eversion, calcaneal eversion, tibial internal rotation, or femoral internal rotation variables. These results are somewhat inconsistent with Hamill et al. (1992), who reported significant decreases in maximum rearfoot eversion and calcaneal eversion angles with hard and medium shoe conditions compared with a soft midsole condition. Rearfoot eversion refers to the eversion of the calcaneus relative to the leg whereas calcaneal motion refers eversion of the calcaneus relative to the laboratory coordinate system. The shoe designs in the current study were similar to those employed in the Hamill et al. (1992) study. However, differences may be attributed to the fact that the current sample of female runners (height: 1.63 m [SD, 0.08]; mass: 58.5 kg [SD, 6.96]) were smaller in stature compared with the male subjects (height: 1.75 m [SD, 0.03]; mass: 75.4 kg [SD, 7.8]) included by Hamill et al. (1992). In addition, the current methodology differed from Hamill et al. (1992) in that rearfoot markers were positioned directly on the calcaneus and a three-dimensional analysis was conducted. In addition, Hamill et al.


Table 1 Means and SDs for the Ankle Dynamic Variables of Interest Shoe Midsole Condition Medium

Hard

−5.33 ± 5.39 −3.90 ± 5.20

−5.20 ± 4.88 −4.32 ± 4.74

−4.44 ± 5.10 −3.36 ± 5.02

0.50 ± 4.55 1.66 ± 4.23

0.82 ± 3.54 1.75 ± 4.18

1.12 ± 4.23 2.76 ± 3.89

−291.36 ± 105.49 −238.09 ± 87.21

−247.92 ± 108.96 −210.34 ± 94.17

52.81 ± 19.87 48.74 ± 16.76

53.28 ± 17.77 49.07 ± 17.79

55.25 ± 18.55 44.89 ± 14.49

44.35 ± 20.32 37.77 ± 18.44

49.19 ± 23.29 39.61 ± 22.22

47.29 ± 21.60 31.85 ± 13.69

1.96 ± 0.26 1.92 ± 0.22

1.94 ± 0.26 1.82 ± 0.26

1.92 ± 0.27 1.81 ± 0.29

77.72 ± 11.26 73.88 ± 12.44

83.82 ± 14.09 72.70 ± 12.41

79.59 ± 16.40 77.39 ± 12.95

Variable Rearfoot eversion angle (°) Shoe only Shoe & orthotic Calcaneal eversion angle (°) Shoe only Shoe & orthotic Rearfoot eversion velocity (°/s) Shoe only Shoe & orthotic Ankle inversion moment (N·m) Shoe only Shoe & orthotic Ankle inversion angular impulse (N·m·s) Shoe only Shoe & orthotic Vertical impact peak (BW) Shoe only Shoe & orthotic Maximum loading rate (BW/s) Shoe only Shoe & orthotic

Soft

−219.51 ± 55.21 −212.13 ± 94.34

Table 2 Means and SDs for the Knee Dynamic Variables of Interest Variable Tibial internal rotation angle (°) Shoe only Shoe & orthotic Femoral internal rotation angle (°) Shoe only Shoe & orthotic Knee external rotation moment (N·m) Shoe only Shoe & orthotic Knee external rotation impulse (N·m·s) Shoe only Shoe & orthotic

(1992) reported that midsole composition did not influence the magnitude of maximum rearfoot eversion velocity. However, there were significant differences between shoe conditions for the percentage of time to maximum rearfoot eversion velocity variable. Hamill et al. (1992) reported that when subjects wore the hard midsole condition, maximum rearfoot eversion velocity occurred earlier during the stance phase. However, max-

Soft

Shoe Midsole Condition Medium

Hard

8.52 ± 5.31 8.41 ± 5.64

8.55 ± 5.48 8.41 ± 5.38

9.29 ± 5.41 7.86 ± 5.41

7.63 ± 4.99 8.37 ± 5.31

7.76 ± 4.66 8.69 ± 5.86

8.50 ± 5.21 8.54 ± 5.89

−21.23 ± 5.38 −21.95 ± 5.66

−21.14 ± 6.30 −22.99 ± 5.71

−21.89 ± 5.37 −22.66 ± 4.84

−29.87 ± 21.19 −32.20 ± 19.47

−32.33 ± 22.58 −33.03 ± 22.05

−30.14 ± 22.01 −37.00 ± 20.77

imum rearfoot eversion velocity occurred later in the stance phase for the soft midsole condition. In the current study, there was a significant shoe condition main effect for the maximum rearfoot eversion velocity variable. The variable was significantly less for the hard and medium midsole shoes compared with the soft shoe condition. There were significant orthotic condition main effects for the maximum rearfoot eversion, calca-


Figure 1 — Maximum rearfoot eversion velocity condition means: shoe-only (thin line) and shoe-and-orthotic (thick line) condition.

neal eversion, tibial internal rotation, and rearfoot eversion velocity variables. All variables were significantly decreased for the shoe-and-orthotic condition. The observed decreases in maximum rearfoot eversion velocity are in agreement with the results published by Smith et al. (1986) and MacLean et al. (2006) but not supported by Mündermann et al. (2003). This finding may be important as increases in maximum rearfoot eversion velocity have been associated with overuse running injuries (Messier et al., 1988; Hreljac et al., 2000). It was also hypothesized that maximum ankle inversion moment and maximum knee external rotation moment would be significantly decreased as the midsole durometer increased. There was a significant Orthotic × Shoe interaction for the maximum ankle inversion moment variable (Figure 3). This effect for maximum ankle inversion moment reveals that decreases in the moment exhibited with custom foot orthotic intervention in the soft and medium shoes were the same. However, when comparing the medium and hard shoe conditions, custom foot orthotic intervention led to significantly greater decreases in maximum ankle inversion moment. The ankle inversion angular impulse was analyzed to quantify the total amount of moment produced (DeVita et al., 1998). Angular impulse was analyzed for the loading (0–50% of stance) phase. There was a significant orthotic main effect exhibited during the loading (ES = 0.81; p < .001) phase. The ankle inversion angular impulse variable was significantly decreased for the shoe-and-orthotic condition. Custom foot orthotic intervention led to a reduction maximum ankle inversion moment and ankle inversion angular impulse during the loading phase and, thus, a significant decrease in

ankle joint frontal plane loading (Stefanyshyn et al., 2006). This finding provides further evidence that custom foot orthoses have an effect on ankle kinetics during the loading phase when the orthosis is in contact with the foot. There were no significant interactions or main effects for the maximum knee external rotation moment variable. However, an analysis of the knee external rotation angular impulse during the loading phase revealed that this variable significantly increased when foot orthoses were worn in the shoe. This could be thought of as deleterious if one interprets this effect as increasing the stress placed on those biological tissues resisting tibial internal rotation. Data from the current study revealed that tibial internal rotation was significantly decreased with custom foot orthotic wear. Determining whether this mechanism is deleterious or healthy will require further research. Lastly, it was hypothesized that there would be significant decreases in the vertical impact peak or maximum loading rate variables with increased midsole durometer. There were no significant shoe effects for either variable. Although this is somewhat surprising, it may be that for a female sample of smaller stature selecting a lesser durometer midsole for the soft condition is warranted or it may be that subjects made kinematic adaptations to accommodate for changes in midsole durometer (Nigg et al., 1987). However, there were significant orthotic main effects for the vertical impact peak (Figure 4) and maximum loading rate (Figure 5) variables. Both variables were significantly decreased when custom foot orthoses were worn in the shoes. Although the relationship between vertical impact peak and maximum loading rate, and overuse running injury


Figure 2 — Ensemble average of rearfoot frontal plane velocity for the shoe-and-orthotic (A) and shoe-only (B) conditions. The three running shoe midsole conditions are the soft (dashed), medium (thin line), and hard (thick line) conditions measured across the stance phase.

has not been clearly established, it has been speculated that these variables may be deleterious if magnitudes exceed the physiologic tolerance level for an individual runner (Hreljac, 2004). The second research aim was to determine if the dynamic variables observed in the shod condition with a greater durometer midsole resembled those for the shoeand-orthotic condition with a lesser durometer midsole. The clinically relevant question is whether a custom foot orthotic intervention in a softer shoe compensates in a manner that the dynamics resemble those exhibited in a shoe of greater midsole durometer. It was therefore hypothesized that for all kinematic and internal joint moment variables, there would be no significant differences when comparing (1) the shoe-only condition (hard midsole) with the shoe-and-orthotic condition (medium midsole), and (2) the shoe-only condition (medium midsole) with shoe-and-orthotic condition (soft midsole).

This hypothesis was supported with two exceptions: maximum tibial internal rotation angle and maximum ankle inversion moment. In both cases, maximum values for tibial internal rotation angle (ES = 0.20; p = .007) and ankle inversion moment (ES = 0.34; p = .002) were significantly increased when comparing the shoeonly condition (hard midsole) and shoe-and-orthotic condition (medium midsole). This finding may indicate that wearing a running shoe with a greater midsole durometer may lead to increases in tibial internal rotation and ankle inversion moment. For all other variables, maxima were not different, suggesting that the addition of a foot orthosis to a shoe having a lesser durometer midsole may function similar to a shoe-only condition with a greater midsole durometer. The final hypothesis compared the shoe-andorthotic and shoe-only conditions for like midsole durometer. It was hypothesized that all dynamic

Figure 3 — Maximum ankle inversion moment condition means: shoe-only (thin line) and shoe-and-orthotic (thick line) condition.

Figure 4 — Vertical impact peak condition means: (shoe-only [thin line] and shoe-and-orthotic (thick line) condition).

61


Figure 5 — Maximum loading rate condition means: (shoe-only [thin line] and shoe-and-orthotic (thick line) condition).

variables would be significantly decreased when comparing (1) the shoe-and-orthotic condition (hard midsole) with shoe-only (hard midsole), and (2) the shoe-and-orthotic condition (medium midsole) with shoe-only (medium midsole). When comparing the shoe-and-orthotic condition (hard midsole) with shoeonly (hard midsole) conditions, maximum values for rearfoot eversion angle, calcaneal eversion angle, and tibial internal rotation angle variables were significantly decreased with the custom foot orthotic intervention. Maximum rearfoot eversion velocity was similar for both conditions. For the kinetic variables, ankle inversion moment and vertical impact peak, maximum values were significantly decreased for the shoe-and-orthotic condition. All other variables were similar for the shoeand-orthotic and shoe-only conditions. Results were similar when comparing the shoe-and-orthotic (medium midsole) with shoe-only (medium midsole) conditions. In addition, it was also hypothesized that these differences would not be exhibited when comparing shoeand-orthotic (soft midsole) with shoe-only (soft midsole). This hypothesis was not fully supported as maxima for rearfoot and calcaneal angles, rearfoot eversion velocity, and internal ankle inversion moment were significantly decreased for the shoe-and-orthotic condition. Thus, it appears that the custom foot orthotic intervention compensates for the soft durometer midsole running in female runners of this stature. There are some limitations of the current study that need mention. We have assumed that the sample of female runners included in this study is representative of the general recreational runner population and that

the running tasks adequately represent what occurs ecologically. It has also been assumed that the dynamic variables selected for analysis in the current study may be associated with overuse running injury mechanism. There is a great deal of work still required to identify the dynamic factors that contribute to mechanism of running injury at the knee. In conclusion, results from the current study reveal that, in this sample of female runners, the only kinematic variable influenced by the midsole durometer of the shoe was maximum rearfoot eversion velocity. However, custom foot orthotic intervention in the footwear led to significant decreases in rearfoot kinematic variables, tibial internal rotation, and internal ankle inversion moment and impulse. With the exception of rearfoot eversion velocity, the findings from the current sample of runners suggest that varying midsole composition alone may not be sufficient to influence lower extremity dynamics. Data from the current sample of female runners indicate that the major component influencing the rearfoot dynamics was the custom orthoses, not the shoe. Data from the current study suggest that when a custom foot orthotic intervention is added to a shoe of lesser density, the dynamics exhibited by the runner are similar to when wearing a shoe of greater density. In other words, the foot orthotic intervention appears to compensate for the lesser shoe stability, enabling it to function similar to a shoe of greater stability. Lastly, when like shoe conditions were compared, custom foot orthotic intervention led to significant decreases in rearfoot kinematic and kinetic variables.


Acknowledgments The authors would like to thank Paris Orthotics Laboratory Division (Vancouver, British Columbia), New Balance (Lawrence, MA) and Smith & Nephew Inc. (Andover, MA) for providing the materials necessary for conducting this study. The Prescription Foot Orthotic Laboratory Association (PFOLA) Research Grant funded this work.

References Bates, B.T., Osternig, L.R., Mason, B., & James, L.S. (1979). Foot orthotic devices to modify selected aspects of the lower extremity mechanics. American Journal of Sports Medicine, 7, 338–342. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd edition). Hillsdale, NJ: Erlbaum. Cole, G.K., Nigg, B.M., Ronsky, J.L., & Yeadon, M.R. (1993). Application of the Joint Coordinate System to ThreeDimensional Joint Attitude and Movement representation: A Standardization Proposal. Journal of Biomechanical Engineering, 115, 344–349. DeVita, P., Lassiter, T., Hortobagyi, T., & Torry, M. (1998). Functional Knee Brace Effects During Walking in Patients With Anterior Cruciate Ligament Reconstruction. American Journal of Sports Medicine, 26, 778–784. Hamill, J., Bates, B.T., Knutzen, K.M., & Kirkpatrick, G.M. (1989). Relationship between selected static and dynamic lower extremity measures. Clinical Biomechanics (Bristol, Avon), 4, 217–225. Hamill, J., Bates, B.T., & Cole, K.G. (1992). Timing of lower extremity joint actions during treadmill running. Medicine and Science in Sports and Exercise, 24, 807–813. Hreljac, A., Marshall, R.N., & Hume, P.A. (2000). Evaluation of lower extremity overuse injury potential in runners. Medicine and Science in Sports and Exercise, 32, 16351641. Hreljac, A. (2004). Impact and Overuse Injuries in Runners. Medicine and Science in Sports and Exercise, 36, 845– 849.

MacLean, C.L., McClay Davis, I., & Hamill, J. (2006). Influence of a Custom Foot Orthotic Intervention on Lower Extremity Dynamics in Healthy Runners. Clinical Biomechanics (Bristol, Avon), 21, 623–630. McPoil, T.G., & Cornwall, M.W. (1996). The relationship between static lower extremity measurements and rearfoot motion during walking. The Journal of Orthopaedic and Sports Physical Therapy, 24, 309–314. Messier, S.P., & Pittala, K.A. (1988). Etiological factors associated with selected running injuries. Medicine and Science in Sports and Exercise, 20, 501–505. Mündermann, A., Nigg, B.M., Humble, R.N., & Stefanyshyn, D.J. (2003). Foot Orthotics affect lower extremity kinematics and kinetics during running. Clinical Biomechanics (Bristol, Avon), 18, 254–262. Nigg, B.M., Bahlsen, H.A., Luethi, S.M., & Stokes, S. (1987). The influence of running velocity and midsole hardness on external impact forces in heel-toe running. Journal of Biomechanics, 20, 951–959. O’Connor, K., & Hamill, J. (2004). The role of selected extrinsic foot muscles during running. Clinical Biomechanics (Bristol, Avon), 19, 71–77. Smith, L.S., Clarke, T.E., Hamill, C.L., & Santopietro, F. (1986). The effects of soft and semi-rigid orthoses upon rearfoot movement in running. Journal of the American Podiatric Medical Association, 76, 227–232. Stefanyshyn, D.J., Stergiou, P., Lun, V.M.Y., Meeuwisse, W.H., & Worobets, J.T. (2006). Knee angular impulse as a predictor of patellofemoral pain in runners. American Journal of Sports Medicine, 34, 1844–1851. Van Woensel, W., & Cavanagh, P.R. (1992). A Perturbation Study of Lower Extremity Motion During Running. International Journal of Sport Biomechanics, 8, 30–47. Williams, D.S., McClay Davis, I., & Baitch, S.P. (2003). Effect of Inverted Orthoses on Lower-Extremity Mechanics in Runners. Medicine and Science in Sports and Exercise, 35, 2060–2068. Winter, D.A. (2005). Biomechanics and Motor Control of Human Movement (3rd ed.). NJ: John Wiley & Sons, Inc.