Multidimensional effects of personalized exercise intervention in type 2 diabetes: a randomized controlled trial

Individuals with type 2 diabetes (T2D) experience progressive fat accumulation and muscle loss, highlighting the need for strategies that concurrently target both. In this randomized controlled trial, we investigated whether a 16-week personalized exercise training improved body composition and related metabolic and cognitive outcomes in adults with T2D and abdominal obesity. A total of 103 participants were randomly assigned (1:1) to control group (n = 51) that maintained habitual lifestyle, or exercise group (n = 52) that received combined training tailored to individual aerobic capacity and muscle strength. The primary outcome was the change in fat mass. Compared with the control group, the exercise group showed a greater reduction in fat mass, with a between-group difference of -1.20 kg (95% CI -1.90 to -0.50), and an increase in skeletal muscle percentage. Magnetic resonance imaging further showed more pronounced reductions in abdominal fat and increases in lower limb muscle after exercise. The exercise group also showed improvements in glycemic regulation and physical fitness. Additionally, cognitive enhancement was superior in the exercise group, with higher Montreal Cognitive Assessment scores and faster responses across tasks involving perceptual speed, working memory, and executive function. Task-based functional magnetic resonance imaging revealed more efficient patterns of network engagement after the intervention. Moreover, improvements in skeletal muscle mass were associated with metabolic and neurocognitive benefits, alongside exercise-related proteomic changes. No serious adverse events were reported. Together, our study indicates that 16-week personalized exercise training effectively promotes fat loss and muscle gain, conferring metabolic and cognitive improvements in T2D.