Objective To explore the methods for early detection of cognitive impairment in type 2 diabetic patients. Methods With the results from the mini-mental state examination (MMSE) ,48 patients of type 2 diabetes mellitus with cognitive impairment (group B) and 52 without cognitive impairment (group A) were enrolled. And 40 healthy subjects served as control. The clinical, laboratory, and imaging data of these subjects were compared. Results No differences in the prevalence of carotid atherosclerosis, serum HbA_1C, triglycerides, high-density lipoprotein-cholesterol, low-density lipoprotein-cholesteroi, body mass index, and waist-to-hip ratio were found between groups A and B. Atrophy of temporal lobe and hippocampus was more marked in group A than that in group B by MRI. The demyelinating changes were found both in group B (35%)and A (12%). Among 3 groups, group B had the longest latency of mismatch negativity (MMN) and P300 and the lowest amplitude of MMN ; the order was followed by group A and control group (all P<0.05). Logistic regression analysis showed that the findings in MRI, P300, MMN were negatively correlated with cognitive impairment in type 2 diabetes mellitus. Conclusion MRI, MMN, and P300 are helpful in detecting cognitive impairment in type 2 diabetic patients. MMN seems to be more sensitive in early detection of the impairment.

Ninty-eight patients of type 2 diabetes mellitus with mild cognitive impairment were collected according to the Montreal Cognitive Assessment( MoCA )and mismatch negativity( MMN ).Forty-eight patients were assigned to conventional treatment with nimodipine (20 mg/d).The MoCA and MMN were evaluated and compared with the remaining 50 patients who served as a control group.There was no significant difference between two patient groups in MoCA( P＞0.05 ) ;the latency and amplitude of MMN were significantly improved in the treated group( P＜0.05 ) ; the amplitude of MMN was reduced in control group ( P＜0.05 ),but the latency of MMN showed no significant change( P＞0.05 ).