SAT-LB115 Metformin-Use Is Associated With Slowed Cognitive Decline and Reduced Incident Dementia in Older Adults With Type 2 Diabetes Mellitus: The Sydney Memory and Ageing Study

Background Metformin use in diabetes has been associated with both increased and decreased dementia rates in observational studies of people with diabetes. Objective: To examine changes in global cognition and specific cognitive domains over 6 years in older adults with diabetes treated with metformin, compared to other glucose lowering medications, and to people without diabetes. Methods Data were examined from the Sydney Memory and Ageing Study, a prospective observational study of 6 years duration of 1037 non-demented community-dwelling elderly aged 70-90 at baseline, derived from a compulsory electoral roll. Neuropsychological testing was performed every 2 years with domain measures of memory, executive function, language, visuospatial function, attention and processing speed and a composite of global cognition. Data were analysed by linear mixed modelling, including age, sex, education, body mass index, heart disease, diabetes, hypertension, stroke, smoking and apolipoprotein E ε4 carriage as covariates. Results: At baseline, 123 participants had diabetes (DM) with 67 receiving metformin (DM+MF) who were similar in demographics to those not receiving metformin (DM-noMF) and those without diabetes (no-DM). Participants with diabetes had higher BMI, lower HDL- and LDL-cholesterol and more prevalent heart disease, hypertension and smoking, compared to no-DM. Over 6-years, DM+MF participants had significantly slower rates of decline in global cognition and executive function, compared to DM-noMF, adjusted for covariates. The rate of decline for each cognitive domain was similar between DM+MF and controls. No impact was found in analyses examining interactions with sex, ApoEε4 carriage or hyperlipidemia. No difference was found in the rate of decline in brain volumes between the groups over 2 years. Incident dementia was significantly higher in DM-noMF, compared to DM+MF (adjusted OR 5.29 [95% CI 1.17-23.88], p,0.05), whereas risk of incident dementia was similar between DM+MF and participants without diabetes. Conclusions: In older people with diabetes receiving metformin, rates of cognitive decline and dementia were similar to that found in people without diabetes and significantly less than that found in people with diabetes not receiving metformin. Large randomized studies in people with and without diabetes are required to determine whether these associations can be attributed to metformin alone or if other factors explain these observations. Future studies will clarify if this cheap and safe medication can be repurposed for prevention of cognitive decline in older people.