SUMMARY Complex musculoskeletal trauma is common both in combat and high energy civilian accidents and often leads to prolonged disability or skeletal nonunion. Regenerative rehabilitation is an emerging field with the potential to improve functional outcomes for lower limb musculoskeletal trauma patients. Our overall objectives are to (i) investigate rehabilitative loading as a regenerative rehabilitation treatment strategy for severe complex musculoskeletal trauma and (ii) understand the relationship between rehabilitative loading, local regenerative niche mechanics, and the biological response. We will meet these objectives through the following specific aims. Specific Aim 1: Determine the effects of rehabilitative loading, injury severity and sex on regenerative niche mechanical signals and functional regeneration. In a rat model of a femoral bone defect, we will employ a factorial experimental design to determine effective rehabilitative loading protocols in critically sized injuries, and determine the effects of rehabilitation on local and systemic biological responses and functional regeneration as a function of sex and injury size. Specific Aim 2: Integrate sensor-enabled real-time feedback into rehabilitation protocols and test the ability to accelerate bone and muscle functional recovery following severe extremity injury. This aim will use wireless implantable strain sensors to provide noninvasive monitoring of mechanics in the regenerative niche throughout the progression of healing under rehabilitative loading. Sensor mechanical data will be used to identify strain ranges that serve as early indicators of healing status and are used as a criterion for dynamically adjusting the rehabilitation regimen on a subject-specific basis to accelerate functional regeneration of musculoskeletal tissue. Specific Aim 3: Build predictive multivariate models based on co-dependent relationships among local regenerative niche mechanical parameters, systemic biomarkers, and functional regeneration. Finite element modeling will be used to generate simulated strain maps showing local variations in regenerative niche mechanical signals using implantable sensor measurements as time-varying boundary conditions. Linear multivariate analyses will be used to map spatial and temporal relationships between the biological responses and local regenerative niche mechanics under rehabilitative loading regimens. Integrating early mechanical data and systemic biomarker data from the previous aims, nonlinear symbolic regression will be used to develop predictive models of bone and muscle functional outcomes for subjects of both sexes. These models will be tested in a prospective study using an additional cohort of both sexes to validate whether models developed in one sex can be predictive of healing outcomes in both the same and the opposite sexes.