A narrative review published in the journal Science, led by researchers at the Novo Nordisk Foundation Center for Basic Metabolic Research at the University of Copenhagen, concludes that current obesity medicines alter biological processes tied to cognition, yet no reliable cognitive benefit has been established across clinical settings. The authors pulled together animal studies, observational cohorts, genetic analyses, brain imaging work and human trials covering semaglutide, tirzepatide and newer drugs that target the amylin, glucagon and GIP pathways.
The most encouraging signal comes from observational data, which linked GLP-1 receptor agonist use to lower rates of cognitive decline or dementia in people with type 2 diabetes. The review cautions that differences in diabetes severity, hypoglycemia exposure and frailty between patient groups may partly explain those associations.
Randomized trials have been less favorable. A pooled analysis of five studies involving 7,732 adults with type 2 diabetes found no clear improvement in overall cognitive performance. In the REWIND trial of dulaglutide, the pre-specified exploratory cognitive outcome did not differ from placebo over a median follow-up of 5.4 years, although a post hoc analysis suggested a lower risk of substantial cognitive impairment. In Alzheimer's disease, liraglutide missed its primary endpoint of preserving brain glucose metabolism, and two larger phase 3 trials of oral semaglutide, EVOKE and EVOKE+, did not slow clinical decline or loss of daily function. Those trials did show decreases in cerebrospinal fluid levels of phosphorylated tau and neurogranin, a biological response that the authors say does not by itself establish clinical benefit.
The authors also note that drug exposure in memory-associated regions such as the hippocampus or cortex was inconsistent in animal work, and that most mechanistic evidence comes from animals and from small, short trials. Because body weight often fell alongside improvements in blood sugar and sleep, it is hard to know which changes, if any, drove cognitive outcomes. Rapid weight loss, muscle loss and dehydration could also offset benefits in older or frail people.
The review suggests that preventing dementia before symptoms appear may differ substantially from treating established Alzheimer's, where neuronal loss may limit what improved metabolic health can achieve. Preserving cognition in metabolically vulnerable people looks more realistic in the near term than improving cognition in healthy individuals. The authors call for trials that track specific cognitive abilities over time and measure drug exposure and receptor engagement in the brain.