Pooling 11 systematic reviews covering 213,604 people with type 2 diabetes, this umbrella review ranked SGLT2 inhibitors first for cognitive protection (pooled OR 0.44, 95% CI 0.35-0.55; SUCRA 94.5%), followed by GLP-1 receptor agonists (0.60, 0.50-0.72; SUCRA 92.1%). Multi-component exercise led the lifestyle options (SMD 0.69-1.04). Benefits were larger in adults 60 and older, with longer treatment (24+ weeks for drugs, 6+ months for exercise), and women appeared to gain more from GLP-1 drugs (HR 0.63 vs 0.81 in men).
The direction fits what observational and trial-derived data have hinted at: both drug classes cut cardiovascular events, which protects the brain's vasculature, and GLP-1 drugs may also dampen neuroinflammation. Treat the headline numbers cautiously, though. Much of the underlying evidence likely comes from observational cohorts and secondary outcomes, and the pooled odds ratios appear to mix incidence of dementia with changes in test scores. SUCRA rankings show relative position, not proof of cognitive benefit. A Cohen's d of -3.00 is implausibly large and suggests small-study or heterogeneity artifacts. Only 11 reviews with overlapping primary studies also risks double-counting. This is a preprint awaiting peer review, so estimates may shift. The practical takeaway is modest: if you have type 2 diabetes, these drugs and structured exercise already have strong cardiometabolic justification, and brain protection may be an added bonus, not a proven reason to switch therapy. Confirmatory, not paradigm-shifting.