The impact of GLP-1 receptor agonists on cognitive decline in heart failure patients: a retrospective cohort study using real world data

Abstract Background Heart failure (HF) is strongly associated with an increased risk of cognitive impairment and dementia. While glucagon-like peptide-1 receptor agonists (GLP-1RAs) have demonstrated neuroprotective effects in preclinical and observational studies of patients with diabetes mellitus (DM), their impact on cognitive function in HF patients remains unknown. Purpose This study examined the association between GLP-1RA therapy and the risk of adverse cognitive outcomes at one year in patients with HF, stratified by HF with reduced ejection fraction (HFrEF) and HF with preserved ejection fraction (HFpEF). Methods We conducted a retrospective cohort study using real-world data from a global collaborative network. Patients with HF, with or without DM, who were prescribed semaglutide, dulaglutide, or liraglutide between January 1, 2018 and December 31, 2022, were matched 1:1 (separately for each drug) with non-GLP-1RA users using propensity score matching to balance baseline characteristics (Figure 1). The primary outcomes, assessed at 1 year, included incident Alzheimer’s disease (AD), mild cognitive impairment (MCI), any dementia, and vascular dementia based on respective ICD-10 codes (G30, G31.84, F03 and F01 respectively). Cox proportional hazard regression was used to calculate hazard ratios with 95% confidence intervals (CI). Results A total of 35,619 HF patients receiving dulaglutide, liraglutide, or semaglutide were included in the study (mean age: 69.5 years; male: 47.6–52.0%; White: 61.5–67.6%; Black: 14.8–17.0%). Diabetes prevalence was highest among patients taking dulaglutide (89.0%), followed by liraglutide (76.9%), and semaglutide (74.8%). Obesity was most common in those who used semaglutide (71.6%) compared to those using dulaglutide (55.5%) or liraglutide (55.0%). Semaglutide and dulaglutide were associated with a significantly lower risk of developing any dementia in HF, HFrEF, and HFpEF patients (p 0.001). Semaglutide was associated with a reduced risk of MCI in HF, HFrEF, and HFpEF, as well as a lower risk of AD in HFpEF patients. Both semaglutide and dulaglutide demonstrated stronger protective effects against VD than AD, suggesting a vascular mechanism of cognitive protection. Dulaglutide was associated with a reduced risk of MCI in HF patients and a lower risk of VD in HFpEF patients. Liraglutide did not show significant cognitive protection across HF subtypes. Conclusions In this propensity score-matched retrospective cohort study, semaglutide and dulaglutide were associated with a lower risk of adverse cognitive decline in HF patients. The greater reduction in vascular dementia compared to Alzheimer’s disease, especially with semaglutide, suggests GLP-1RAs may mitigate vascular contributions to cognitive impairment through improved endothelial function and reduced inflammation. These findings highlight a potential cognitive benefit of GLP-1RAs in HF, warranting further validation in prospective clinical trials.Figure 1 Table 1