Diabetes is a chronic metabolic disorder that affects 422 million people worldwide and can lead to diabetic myopathy and bone diseases. The etiology of musculoskeletal complications in diabetes and the interplay between the muscular and osseous systems are poorly understood. Exercise training promises to prevent diabetic myopathy and diabetic bone disease and offer protective effects on muscle and bone. Although the muscle-bone interaction is largely biomechanical, the muscle secretome, specifically the myokines, has significant implications for bone biology. Here, we have developed an in vitro model to elucidate the effects of mechanical strain on myokine secretion and its impact on bone metabolism decoupled from physical stimuli. We developed modular bone constructs using crosslinked gelatin hydrogels which facilitated osteogenic differentiation of osteoprogenitor cells. Then muscle constructs were made from fibrin hydrogel, which enabled myoblast differentiation and formed mature myotubes. We investigated the myokine expression by the muscle constructs under strain regimens replicating endurance (END) and high-intensity interval training (HIIT) in hyperglycemic conditions. In monocultures, both regimens induced higher expression of Il15 and Igf1, while END supported more myoblasts differentiation and myotube maturation than HIIT. When cocultured with bone constructs, the HIIT regimen increased Glut4 expression in muscle contructs that END supporting higher glucose uptake. Likewise, the muscle constructs under the HIIT regimen promoted a healthier and matured bone phenotype than END. Interestingly, under static conditions, myostatin (Mstn) expression was significantly downregulated in muscle constructs cocultured with bone constructs compared to monocultures. Our in vivo analysis of the role of myostatin on bone structure and function also showed that myostatin knockout (GDF8-/-) enhanced muscle mass and moderately influenced bone phenotype in adult mice. Together, our in vitro coculture system allowed orthogonal manipulation of mechanical strain on muscle constructs while facilitating biochemical crosstalk between bone and muscle constructs. Such systems can provide an individualized microenvironment and allow decoupled biomechanical manipulation, which is unachievable using traditional models. In the long-term, these in-vitro systems will help identify molecular targets and develop engineered therapies for diabetic bone disease.