Résumé : For decades, scientists and healthcare professionals have been reporting the adverse effects of lack of exercise, such as the risk of developing type 2 diabetes. Exercise is also a prominent determinant in both type 1 and type 2 diabetes management as it play a role in primary and secondary prevention by improving glycemic control and reducing risk of diabetes complications. Unfortunately, a large part of the general population and patients with diabetes fails to meet physical activity level recommendations. The purpose of this work is therefore two-fold and structured in two parts. In the first part, the need of solutions for improving physical fitness while avoiding common barriers to exercise led us to take interest in whole-body high intensity interval training (WB-HIIT). WB-HIIT is a training method using bodyweight exercises, requiring low time involvement and no additional equipment. We therefore started by conducting a systematic review and meta-analysis that highlighted that WB-HIIT was effective for improving cardio-respiratory fitness, musculoskeletal fitness and to some extent body composition. Then, a randomized control trial extended advantages of WB-HIIT to a larger frame by underlining the potential of home-based WB-HIIT to concomitantly improve several aspects of health-related fitness, using video-conducted sessions.In the second part, the key role of exercise in diabetes management led us to assess feasibility and effectiveness of a guideline-based exercise intervention carried out autonomously by patients with type 1 and type 2 diabetes in local fitness facilities. Moreover, aerobic exercise capacity has been shown to be reduced in both type 1 and type 2 diabetes as compared to healthy controls. The purpose was then to also verify if patients with diabetes exhibit blunted training-induced improvements as compared to healthy adults. Our study first revealed that, patients with diabetes under good glycemic control and without diabetes complication may present a preserved aerobic capacity conversely to the previously reported reduction. Also, the exercise intervention led to cardiorespiratory fitness improvements in both type of diabetes with visceral fat loss and improved insulin sensitivity in patients with type 2 diabetes. However, we nonetheless observed that patients with type 2 diabetes exhibit a blunted VO2peak improvement after a 12-week exercise intervention while similar improvements were observed between patients with type 1 diabetes and healthy controls. Mechanisms are yet to be elucidated, but some leads are worth considering such as the role of patients’ clinical phenotype or medication. In summary, it can be concluded that characteristics and effectiveness of WB-HIIT puts it as a promising method for promoting physical activity and could be an interesting tool for diabetes prevention. It could also be suggested as an interesting alternative or complement to exercise prescription guidelines for diabetes management. However, actual effectiveness of WB-HIIT in patients with diabetes was not assessed. Moreover, given the blunted exercise trainability in diabetes and the unawareness on underlying mechanisms, we cannot make any statement on actual effectiveness of WB-HIIT for diabetes management. The observed and reported benefits of WB-HIIT are nonetheless encouraging, and its actual effectiveness in patients with diabetes would worth further investigations.