Impact of antidiabetic agents on dementia risk: A Bayesian network meta-analysis

Background: Dementia is more prevalent among people with type 2 diabetes, but little is known regarding the influence of antidiabetic agents on this association.

Objective: This study assessed the impact of various antidiabetic agents on the risk of dementia among patients with Type 2 diabetes mellitus.

Methods: Relevant studies were retrieved from the PubMed, Embase, Cochrane Central Register of Controlled Trials (CENTRAL), and ClinicalTrials.gov databases. Nine antidiabetic agents were included in the search. Data were pooled via network meta-analysis and meta-analysis.

Results: Nine studies were selected for the network meta-analysis with 530,355 individuals and 17 studies for the meta-analysis with 1,258,879 individuals. The analysis excluded glucagon-like peptide 1 (GLP-1) analogs and sodium-dependent glucose transporter 2 (SGLT-2) inhibitors due to the absence of relevant data. The use of dipeptidyl peptidase-4 (DPP-4) inhibitors, metformin, thiazolidinedione, and sulfonylurea was associated with a decreased risk of dementia in comparison to no treatment with antidiabetic agents (hazard ratio [HR] for DPP-4 inhibitors, 0.54; 95% confidence interval [CI], 0.38-0.74, HR for metformin, 0.75; 95% CI, 0.63-0.86; HR for sulfonylurea, 0.85; 95%CI, 0.73-0.98 and HR for thiazolidinedione, 0.70; 95% CI, 0.55-0.89, respectively). However, the node-splitting analysis showed the inconsistency of direct and indirect estimates in sulfonylurea (P = 0.042). DPP-4 inhibitors, metformin, thiazolidinedione, and sulfonylurea exhibited a significant impact on the risk of dementia in diabetics compared with insulin (HR, 0.35; 95%CI, 0.20-0.59, HR, 0.48; 95% CI, 0.30-0.77, HR, 0.45; 95% CI, 0.29-0.73 and HR, 0.55; 95% CI, 0.34-0.88, respectively). DPP-4 inhibitors also exhibited a protective effect on the risk of Alzheimer's dementia compared with the no treatment with antidiabetic agents (HR, 0.48; 95% CI, 0.25-0.92). The meta-analysis demonstrated a protective effect of using metformin and DPP-4 inhibitors on the risk of dementia (HR, 0.86; 95% CI, 0.74-1.00 and HR, 0.65; 95% CI, 0.55-0.76, respectively). Further analysis showed insulin was associated with an increased risk of Alzheimer's dementia (HR, 1.60; 95% CI, 1.13-2.26). Only two case-control studies mentioned GLP-1 analogs and SGLT-2 inhibitors, and the pooled ORs showed no evidence of an association with dementia (GLP-1 analogs: 0.71; 95% CI, 0.46-1.10 and SGLT-2 inhibitors: 0.74; 95% CI, 0.47-1.15).

Conclusion: This analysis indicated that patients with type 2 diabetes under treatment with DPP-4 inhibitors presented with the lowest risk of dementia, followed by those treated with metformin and thiazolidinedione, while treatment with insulin was associated with the highest risk. For the increasing focus on the protective effect on dementia, further specific clinical studies are needed to evaluate the impact of GLP-1 analogs and SGLT-2 inhibitors on the risk of dementia.