High arch and low arched runners are more prone to overuse injuries of the lower extremity than normal arched runners. This is thought to be attributed to higher loading exhibited in high arched runners and excessive rearfoot motion in low arched runners (Williams et al., 2001). Running shoes are specifically designed for these arch types to reduce injury rates (motion control shoes for low arched runners and cushioning shoes for high arched runners). Recently, it was reported that a footwear intervention program reduced lower extremity injury rates at a military base (Knapik et al., 1999). However, the biomechanical changes associated with running in the recommended footwear were not examined. The overall goal of this work was to therefore examine lower extremity mechanics when high and low arched run in the shoe that is recommended for their foot type.

In order to accomplish this, a valid and reliable method of determining arch height was needed. Williams et al. (2000) reported on an arch height index measurement that was established to be valid and reliable. However, this index was based on several measurements taken with a set of hand held calipers. The Arch Height Index Measurement System was developed in order to improve the ease of this measure (Richards et al., 2001). Intra- and inter-rater reliability of the device was evaluated by determining the Intra-class correlation coefficients (ICC) for the arch height index (AHI), as described by Williams et al. (2000). These ICCs ranged between 0.96-0.99 for all measures. Secondly, the distribution of the AHI was reported for a group of healthy recreational runners between the ages of 18-40 who ran more than 10 miles per week. No differences were noted when comparing genders or between the sides of individuals. The AHI for all of the subjects was 0.340 +/- 0.030. Based on this distribution, subjects with AHI greater than 1.5 standard deviations above the mean value were invited to be part of the high arch (HA) group. Runners with AHI greater than 1.5 standard deviations below the mean value were invited to be in the low arch (LA) group.

The next part of the investigation involved the study of the interaction of arch type and footwear during overground running. Twenty HA and twenty LA runners ran in both a motion control (MC) shoe and cushioning (CT) shoe. Rearfoot kinematics and tibial shock were analyzed in the different footwear conditions. There was an interaction of arch type and footwear for instantaneous loading rate which was lowest in the recommended shoe for each arch type. Assessment of main effects of shoe revealed that rearfoot motion was reduced in the motion control shoe compared to the cushioning shoe. In addition, shock was decreased in the cushioning shoe compared to the motion control shoe. Thus, for most variables, it appears that the motion control shoe controls rearfoot motion and the cushioning shoe reduces shock regardless of the arch type of the runner.

The next study was performed to determine whether shoes have a larger effect at the end of a run when a runner is in an exerted state. Twelve HA and twelve LA runners ran on a treadmill at their typical pace for a 30-45 minute training run. Subjects ran in each footwear condition on non-consecutive days, to minimize fatigue effects. The run was terminated when subjects reached 85% of their heart rate maximum (220-age) or a rate of perceived exertion of 17. In this repeated measures design, changes over time were analyzed in each arch group separately. This allowed for the examination of how each group responded to each shoe over the course of the run. In the LA runners, running in the recommended MC shoe reduced peak tibial internal rotation while running in the CT shoe increased peak tibial internal rotation over the course of the run. No other interactions were noted. However, when further assessing results for eversion, a subgroup of LA runners (n=6) exhibited the expected reduction in peak eversion and eversion excursion in the MC compared to the CT over the course of the prolonged run. In the HA runners, the shock variables did not change over the course of the run. However, there was an overall reduction in tibial shock in the CT shoe compared to the MC shoe when results were collapsed across time.

The third study was focused on the role of variability in running injuries, which has received recent attention in the literature. The effect of motion control and cushioning shoes on lower extremity variability in HA and LA runners over the course of the prolonged run was analyzed. The subjects and protocol were identical to that described in the previous prolonged run study. Lower extremity variability was analyzed for the frontal plane motion of the rearfoot (RFTV), transverse plane motion of the tibia (TIBV) and sagittal plane motion of the knee (KNEV). The average standard deviation of the ensemble average of five stance trials was calculated as a measure of variability of motion for each joint. Variability was compared between shoe conditions, first at the beginning of the run, and then over the course of the run. The mean variability over the entire period of stance was analyzed. In addition, stance was divided into four functional periods. The interaction of footwear and time of the run on variability for each period was then analyzed. In the HA runners, TIBV was increased at the beginning of the run in the CT as expected. However, over the course of the prolonged run, TIBV decreased in the CT and increased in the MC, perhaps suggesting a compensatory response to the footwear over time. In the LA runners, RFTV was greater, as expected, during the fourth period of stance in the CT compared to the MC at the beginning of the run. This same trend was observed during the prolonged run. Over the course of the run, TIBV was reduced in the MC during the first period of stance, again suggesting the runner compensated for the shoe. In general, variability was higher during the periods between the loading and unloading portion of the stride cycle, which may provide a more flexible system during this transition. It was also interesting to note that when analyzed by periods, differences between conditions were most often noted in the transition periods 1 and 4. It appears that therefore footwear influences variability differently within different periods of stance as well as at the beginning and end of the run.

In summary, running in the recommended shoe appears to reduce some of the mechanisms of injury in HA and LA runners. Footwear and arch type appear to affect lower extremity variability differently at the beginning of the run compared to the end of the run. However, the implications of these findings are not yet known. Shoes are currently being developed that are able to tune themselves to the runner to optimize their mechanics. Prospective studies are needed to further establish relationships between mechanics and injury, as well as to determine the ability of evolving footwear designs to reduce the risk for running related injuries.