This study examined the effect of ACL reconstruction and gait speed on joint angles of the hip, knee, and ankle; stride length; and the percent of the gait cycle spent in midstance to assess whether reconstruction restores pre-injury knee function. Based on previous findings (DeVita et al., 1998), a 15° deviation in the hip angle was considered an important difference. We hypothesized that the hip angle would be smaller in the affected leg during midstance at both speeds which was partially supported by our results (Figure 3). The unaffected leg had greater hip extension throughout stance (with a maximum difference of 16.2° at midstance) during the natural speed, while no difference during stance was observed between the legs at the increased speed. Our results thus failed to support the hypothesis that no change in hip angle deviation would occur as gait speed increased. Noehren et al. (2013) suggested that this decreased hip angle deviation is due to the affected leg's greater reliance on hip extensor muscles to control knee flexion during the higher impact forces at increased gait speeds.
Based on previous findings (Boden et al., 2009; Scanlan et al., 2013), a knee angle deviation of 10° was considered an important difference. We hypothesized that knee flexion would be greater in the affected leg at both speeds, particularly during midstance when maximal ACL strain occurs, but this was not supported by our results (Figure 4). Furthermore, there was no effect of speed on deviations in the knee angle. At midstance, the knee had similar joint angles at all four conditions, which contrasted with previous findings (Mauro et al., 2008; Ferber et al., 2002; Scanlan et al, 2013; and Gokeler et al., 2003) but was supported by other literature (Shi et al., 2010). This lack of difference may be attributed to 11 months having passed between surgery and gait analysis. Previous literature has found that as time since surgery passes, knee extension during midstance increases (Webster et al, 2012; Knoll et al., 2004; Ferber et al. 2002; DeVita et al, 1997). As concluded by Moraiti et al. (2010), the similarity in knee angles of the legs may be due to subjects' attempts to compensate for deficiencies in the affected leg and maintain symmetry during gait. However, a difference in knee angles was seen during midswing at the increased speed, where the affected leg was more extended by 13.1°. Shi et al. (2010) found a similar reduction in peak knee flexion during gait in affected legs, suggesting that ACL reconstruction leads to difficulty in reaching peak knee flexion. This effect may only have been observed at the increased speed because the subject could not achieve the angular velocity required to reach the same peak knee flexion angle observed in the other three conditions.
Based on the findings of Boden et al. (2009), an ankle angle deviation of 10° was required for a difference to be important. We hypothesized that the ankle angle would be smaller in the affected leg during stance due to the increased hip and knee flexion, but this was not supported at either gait speed (Figure 5). At the natural gait speed, the ankle was more plantarflexed throughout early to mid stance and swing, possibly due to placing the center of mass more posteriorly during stance. However, it is unclear why the affected leg was more plantarflexed during swing, given that the hip and knee angles at this point were similar to those of the other conditions. Future studies should attempt to explain why this change occurred. At the increased gait speed, there was a decreased deviation between the two legs' ankle angles, which refuted the hypothesis that an increase in deviation would occur. This is likely due to the large variability of the natural gait speeds which caused an unexpectedly large difference in the ankle angle at the natural speed condition.
We hypothesized that stride length would be greater in the unaffected leg at both speeds, which our data did not support (Table 1). Since stride length increases with increasing speed and the subject walked at faster speeds during trials examining the affected leg, stride length for the affected leg was expected to be greater (Danion et al., 2003). Due to the unexpected variability in gait speed, we are unable to adequately compare the deviations in stride lengths at the two speed conditions. We also hypothesized that the time spent in midstance would be shorter for the affected leg at both speeds, which our data supported (Table 2). Although the subject did not noticeably increase knee extension during midstance (Figure 4) to reduce ACL strain (Chmielewski et al., 2008), the subject was able to alleviate discomfort in the affected leg by spending less time in midstance (Wu et al., 2010). However, there was no increased deviation in the time spent in midstance as gait speed increased, possibly due to the large variability of natural speed.
A primary limiting factor in this study is that the subject is an NCAA Division I football player with increased access to doctors, trainers, and resources compared to that of an average person. Because of the subject's athleticism and exceptional circumstances, these results lack generalizability. Since this was a case study that examined only one subject, statistical analysis could not be performed, limiting the ability to determine the data's significance. This study was also conducted after ACL reconstruction, so differences in the same leg pre- and post-injury could not be observed. The unaffected leg was assumed to behave similarly to the affected leg pre-injury. Additionally, because speed was not controlled, there were unavoidable inconsistencies in gait speed during each condition. In order to assess these limitations, further studies should employ longitudinal methods examining a larger sample size. Observing subjects pre- and post-injury would provide data on the same leg, eliminating the need to make assumptions using the contralateral leg. Furthermore, a larger sample size that is more representative of the general population would increase this study's statistical power and external validity.
Since we observed a restoration in stride length, a semi-restoration in joint angles, and a lack of restoration in the percent of the gait cycle spent in midstance for the affected leg, it is not clear whether ACL reconstruction restores subjects’ gait patterns, assuming that the unaffected leg functions similarly to the affected leg pre-injury.