The goal of this study is to gather pilot data to help inform a future clinical trial. As such, the investigators will employ a randomized clinical trial design, but data will only be collected on 9 total subjects. Nine subjects will be randomized to 2 split-belt intervention groups (one group where early stance loading is trained and the other where midstance loading is trained) and a placebo group.

The goal of this study is to explore the adaptations in knee loading from a 6-week split-belt training intervention.

The investigators' main question for this aim is:

1. Does knee loading, measured by the sagittal plane knee moment, change to a greater extent in the split-belt treadmill training groups compared to the placebo group?
2. Are there differences in training-related knee loading changes between individuals trained in the early stance vs. midstance loading split-belt training?

It is commonplace for individuals after anterior cruciate ligament (ACL) reconstruction to shift mechanical demands away from their surgical knee and limb. This manifests as diminished knee moments and vertical ground reaction forces in the ACL limb during everyday tasks (e.g., walking, running, standing, landing, etc.) and persists for as long as 2.5 years after surgery. This pattern of underloading is considered maladaptive, as it has been linked to re-injury and biological markers that are consistent with the development of post-traumatic osteoarthritis that affects over 50% of knees 10-20 years after surgical reconstruction.

Split-belt treadmill training is a gait retraining approach where treadmill belt speeds are decoupled (i.e., one belt is set to move at a faster or slower speed than the other belt) during walking. Split-belt training is based on well-established motor learning principles, such as error-based learning and variability of practice which can lead to locomotor adaptations. In healthy individuals, split-belt treadmill walking significantly increases (from baseline) knee moment impulses in the limb on the slow belt than on the fast belt during the braking and propulsive phases of gait. Split-belt treadmill training has also shown promise in individuals with neurological deficits, resulting in significant improvements in gait biomechanics after training.

To explore the adaptations in loading from a 6-week split-belt training intervention, the investigators will conduct a pilot randomized clinical trial design, but data will only be collected on 9 total subjects. Nine subjects will be randomized to one of 3 groups: 1) early stance split-belt treadmill training, 2) mid-stance split-belt treadmill training, or 3) placebo split-belt treadmill training. The primary outcome, sagittal plane knee moment, will be examined before, midway, & after the 6-weeks of training. Other outcomes, vertical ground reaction force, knee joint contact force, and the Knee Injury and Osteoarthritis Outcome Score, will be measured at the same timepoints. Medial and lateral knee cartilage thickness are additional outcomes that will only be assessed before and after training.

Inclusion Criteria:

• aged 14-45 years
• suffered an acute, complete ACL rupture as confirmed by MRI and physical exam
• have undergone ACL reconstruction w autograft within the past 10 months
• willingness to participate in testing and follow-up as outlined in the protocol
• English-speaking

Exclusion Criteria:

• inability to provide written informed consent
• female subjects who are pregnant or are planning to become pregnant (self-reported)
• previous ACL injury
• previous surgery to either knee
• bony fracture accompanying ACL injury
• patients who experienced a knee dislocation
• patients who had their ACL reconstructed with an allograft
• patients who underwent a multi-ligamentous and/or staged ACL reconstruction

De-identified data will be made publicly available via Deep Blue Data which is managed by the University of Michigan.

We will strive to release data to the repository as rapidly as it is possible to analyze experiments and publish results. Data used in publications will be released at the time of publication.