Abstract Background: Elderly with diabetes are at a high risk of cognitive impairment and Alzheimer disease (1). Exercise is a non-pharmacological disease modifying therapy for metabolic disorders such Diabetes (2). Also, exercise can improve cognition based on metabolic-cognitive cross-talk (3). The purpose of this study is to optimizing metabolic-cognitive cross-talk through exercise in elderly individuals with type 2 diabetes. Methods: In this randomized controlled trial 93 elderly with type 2 diabetes and mild cognitive impairment, randomly divided to high-intensity functional training (HIFT, N=31), low-intensity functional training (LIFT, N=31) and control (N=31) groups. Training group performed the 24 weeks of functional training 3 times/week. Metabolic (glucose, insulin, insulin resistance and glycated hemoglobin A (HbA1C) and Cognitive (processing speed, learning and memory) indices measured before and after intervention. Two-way ANOVA (group*time design) was used for between group comparisons. Result: Both exercise protocols had a significant improvement in glucose (P>0.007), insulin (P>0.055), insulin resistance (P>0.031) and serum HbA1C (P>0.001) and also processing speed (P=0.001), memory (p=0.001) and learning (p=0.001) compared to control group. Also, all indicators significantly improved in HIFT compared to LIFT (P>0.044) except insulin and learning (P≤0.05). Conclusion: According to the effect sizes, it is recommended to clinician to prescribe high intensity programs to reduce glucose, insulin resistance and improve serum HbA1C also learning, P>0.031 and memory in type 2 diabetic elderly with cognitive impairment.
