The prevalence of cognitive dysfunction in diabetic patients is high, which seriously affects the quality of life of patients, and early intervention is of great significance. Central nervous system insulin resistance plays a key role in the pathogenesis of cognitive dysfunction in diabetic patients. Central insulin resistance observed by functional magnetic resonance imaging (fMRI), may serve as an alternative endpoint for assessing cognitive function. Sodium glucose cotransporter 2 inhibitor (SGLT2i) may improve cognitive impairment by improving central insulin resistance.

The prevalence of cognitive dysfunction in diabetic patients is high, which seriously affects the quality of life of patients. Cognitive impairment mainly includes two stages: mild cognitive impairment (MCI) and dementia. Early detection of diabetic patients with MCI and effective intervention measures to prevent and delay the onset of dementia have important clinical value and social significance. With the development of imaging technology, especially fMRI changes in local blood flow and metabolic activities of the brain can be displayed before changes in brain structure. The central insulin resistance status observed by fMRI may serve as an alternative endpoint for assessing cognitive function. SGLT2i has been shown to significantly improve insulin resistance in the central nervous system in previous basic and clinical studies. However, up to now, no head-to-head clinical studies have been reported on SGLT2i and other antidiabetic drugs in patients with MCI.

Based on the above research background, we hypothesize that canagliflozin, as a clinically commonly used SGLT2i, it may improve cognitive function in patients with type 2 diabetes mellitus and MCI by improving insulin resistance in the central nervous system.

Inclusion Criteria:

• age: 40-75 years old;
• HbA1c: 7.0 - 10.0%;
• Moca scale total score between 20-26 (duration of education is more than 10 years) or total score between 20 and 25 (duration of education is less than 10 years);
• sign informed consent;

Exclusion Criteria:

• left-handedness;
• duration of diabetes is less than 6 years;
• inability to complete central nervous system magnetic resonance imaging;
• alcohol or drug addiction history;
• use of other oral hypoglycemic drugs within 90 days prior to enrollment except for metformin;
• Family or past history of neurological or psychiatric disorders, pancreatitis, medullary thyroid cancer and multiple endocrine adenomatosis;
• History of recurrent urinary tract infection and malignant tumor;
• History of severe gastrointestinal diseases and gastroenteric surgery;
• Pregnancy/lactation status;
• Abnormal liver function (liver enzyme index is more than 2.5 times the upper limit of the reference range) or abnormal kidney function (estimated glomerular filtration rate is less than 45ml/min); Type 1 diabetes; other diseases or conditions that reduce the possibility of enrollment or complicate enrollment, such as frequent changes in the working environment and unstable living environment, which are likely to cause loss of follow-up, according to the judgment of the researchers;