Nance Nerissa, Gilsanz Paola, Karter Andrew J, Finertie Holly, Schmittdiel Julie A, An JaeJin, Adams Alyce S, Oshiro Caryn, Cassidy-Bushrow Andrea E, Krahe-Dombrowski Sarah, Yassin Maher, Lin Sharon, Izadian Keanu, O'Connor Patrick J, Neugebauer Romain
Kaiser Permanente Northern California Division of Research, Pleasanton, California, USA.
University of California, Berkeley School of Public Health, Berkeley, California, USA.
Alzheimers Dement. 2025 Sep;21(9):e70621. doi: 10.1002/alz.70621.
To address gaps in current research, this study aims to compare the impact of exposure to glucagon-like peptide-1 receptor agonists (GLP-1RA) versus sodium-glucose cotransporter 2 inhibitors (SGLT2i), dipeptidyl peptidase 4 inhibitors (DPP4i), and sulfonylureas (SU) on reducing the risk of dementia, using a rigorous targeted learning causal inference approach.
Using clinical and claims data from four diverse US health-care systems, we emulated three two-arm trials contrasting sustained treatment with GLP-1RA versus SGLT2i, DPP4i, and SU on dementia diagnosis. We included diabetes patients aged ≥ 60 years who initiated medication between 2014 and 2022. We estimated cumulative risk differences at 2.5 years.
In Cohort 1, there was no evidence of differential dementia risk between sustained exposure to GLP1-RA versus SGLT2i (adjusted risk difference [aRD] -0.001, 95% confidence interval [CI] -0.004, 0.001). In Cohorts 2 and 3, GLP-1RA was associated with reduced risk of dementia diagnosis compared to DPP4i and SU, respectively (aRD -0.013, 95% CI -0.017, -0.009; aRD -0.016, 95% CI -0.018, -0.015).
Rigorous causal inference analysis suggests that sustained exposure to GLP-1RA may modestly reduce risk of dementia, compared to DPP4i or SU exposure-but not compared to SGLT2i.
We researched the comparative effects of diabetes medications on dementia. We studied a large, diverse observational cohort of patients with diabetes in the United States. Glucagon-like peptide-1 receptor agonists (GLP-1RAs) may modestly reduce risk of dementia compared to dipeptidyl peptidase 4 inhibitor or sulfonylurea exposure. GLP-1RAs do not show evidence of dementia risk reduction compared to sodium-glucose cotransporter 2 inhibitors. Physicians may consider this when making prescription decisions with patients.
为了填补当前研究中的空白，本研究旨在采用严格的靶向学习因果推断方法，比较胰高血糖素样肽-1受体激动剂（GLP-1RA）与钠-葡萄糖协同转运蛋白2抑制剂（SGLT2i）、二肽基肽酶4抑制剂（DPP4i）和磺脲类药物（SU）对降低痴呆风险的影响。
利用来自美国四个不同医疗保健系统的临床和索赔数据，我们模拟了三项双臂试验，对比GLP-1RA与SGLT2i、DPP4i和SU持续治疗对痴呆诊断的影响。我们纳入了2014年至2022年间开始用药的60岁及以上糖尿病患者。我们估计了2.5年时的累积风险差异。
在队列1中，没有证据表明持续暴露于GLP-1RA与SGLT2i之间存在痴呆风险差异（调整后风险差异[aRD]-0.001，95%置信区间[CI]-0.004，0.001）。在队列2和队列3中，与DPP4i和SU相比，GLP-1RA分别与痴呆诊断风险降低相关（aRD -0.013，95% CI -0.017，-0.009；aRD -0.016，95% CI -0.018，-0.015）。
严格的因果推断分析表明，与暴露于DPP4i或SU相比，持续暴露于GLP-1RA可能会适度降低痴呆风险，但与SGLT2i相比则不然。
我们研究了糖尿病药物对痴呆的比较影响。我们研究了美国一个大型、多样化的糖尿病患者观察队列。与二肽基肽酶4抑制剂或磺脲类药物暴露相比，胰高血糖素样肽-1受体激动剂（GLP-1RA）可能会适度降低痴呆风险。与钠-葡萄糖协同转运蛋白2抑制剂相比，GLP-1RA没有显示出降低痴呆风险的证据。医生在与患者做出处方决定时可能会考虑到这一点。
