Effects of SGLT2i on reducing risks of cardiovascular events and dementia in atrial fibrillation patients with diabetes mellitus: a nationwide cohort study

Abstract Background/Introduction Patients with atrial fibrillation (AF) are associated with higher risk for cognitive decline and progression to dementia. There are limited data whether sodium-glucose cotransporter-2 inhibitor (SGLT2i) reduces new-onset cognitive dysfunction, especially in Asian patients with AF who also have diabetes mellitus. Purpose This study aimed to investigate whether SGLT2i reduces the risks of cardiovascular events and dementia among AF patients with diabetes. Methods We identified AF patients with diabetes based on the Taiwan's National Health Insurance Research Database (2016). AF patients without prior dementia history and aged 55-85 years were studied: 810 patients were prescribed SGLT2i for at least 180 days and were matched with 1620 patients without SGLT2i using propensity scores (Figure 1). Cardiovascular and dementia events were followed through the end of 2021. Subtypes of dementia were investigated using competing risk models. Results During a follow-up period of 5 years of AF patients with diabetes receiving SGLT2i, as shown in Figure 2, this study provides a detailed insight into the results of survival analyses for clinical outcomes of AF, stroke, total dementia, vascular dementia, all-cause death, and cardiovascular (CV) death. After multivariable adjustment, the risks of AF-related hospitalization (HR: 0.741, 95% CI: 0.563-0.976), stroke (HR: 0.748, 95% CI: 0.599-0.934) and all-cause death (HR: 0.338, 95% CI: 0.245-0.466), were significantly decreased. Both CV death (HR: 0.449, 95% CI: 0.268-0.755) and non-CV death (HR: 0.298, 95% CI: 0.198-0.448) were significantly reduced. SGLT2i use was associated with a significantly decreased risk of total dementia (HR: 0.615, 95% CI: 0.436-0.869) and vascular dementia (HR: 0.419, 95% CI: 0.221-0.793). The reduced risks of Alzheimer’s dementia and other/mixed dementia were not statistically significant. After multivariable adjustment, total dementia and vascular dementia were associated with higher age and stroke events. The beneficial effect on dementia was irrespective of novel oral anticoagulants (NOAC) use, although the SGLT2i benefits were greater in non-warfarin and non-dipeptidyl peptidase 4 (DPP4) users compared to users. Conclusions There were decreased risks of AF-related hospitalization, stroke event, all-cause death, and vascular dementia in the users of SGLT2i in AF patients with diabetes. This was irrespective of NOAC use, although the SGLT2i benefits were greater in non-warfarin users compared to warfarin users.Figure 1:Study flow chart