The Association Between Oral Glucose-Control Agents and Incident Dementia in Type 2 Diabetes Within the Veteran’s Affairs Healthcare Enrollees

Abstract

INTRODUCTION: Dementia is a major contributor to disability and death. A risk factor of dementia is type 2 diabetes (T2D). Current evidence suggests shared causes of the two diseases and a potential to repurpose specific glucose-control agents (GCAs) for dementia prevention or treatment. However, clinical trials and population studies focusing on this topic are insufficient to conclude GCAs anti-dementia effects and compare effectiveness of different GCAs. This dissertation intended to compare the effects of monotherapy and concomitant use of GCAs on dementia. The results should be able to provide suggestions on GCAs selection in T2D, in terms of dementia risk management. METHODS: In this dissertation, study 1 and study 2 used medical records from the Veteran’s Affairs Healthcare (VAH) database. Glucose-control treatment histories were extracted from prescription records and were aggregated at drug class level. Disease conditions were extracted from outpatient diagnosis records, coded by ICD-9 or ICD-10. Four antidiabetic drug classes were assessed based on a literature review and VAH data richness. Among T2D, study 1 compared the associations of first-generation antidiabetic drug classes (metformin [MET], sulfonylureas [SU], and thiazolidinedione [TZD]) with the risk of dementia. And study 2 compared the associations of adding dipeptidyl peptidase-4 inhibitors (DPP-4is) to MET and/or SU (MET/SU) treatment with risks of dementia and other vascular events. To capture the trends of clinical evidence in relevant fields, study 3 reviewed phase II, III, and IV clinical trials repurposing GCAs for Alzheimer's disease (AD). RESULTS: In T2D without prior dementia, study 1 found that after at least one year of treatment, compared with MET monotherapy users, the risk of all-cause dementia was 22% lower in TZD monotherapy users (HR: 0.78, 95% CI 0·75-0·81), and 11% lower in MET and TZD dual therapy users (HR 0.89, 95% CI 0·86-0·93), whereas the risk was 12% higher in sulfonylurea monotherapy users (HR 1.12 95% CI 1·09-1·15). In study 2, compared with participants staying on MET/SU treatment, participants who added DPP-4is to MET/SU regiments had lower risks for the cerebrovascular outcome (HR, 0.68, 95% CI, 0.62-0.74), and the microvascular outcome (HR, 0.91; 95% CI, 0.88-0.94), but not associated with the risk for the renal outcome (HR, 1.03, 95% CI, 0.97-1.10). In study 3, 26 clinical trials were reviewed and summarized in a narrative way. Elders with mild cognitive impairment, AD, and other types of dementia were the main trial participants. Among studied GCAs, insulin, MET, and pioglitazone showed protective effects on subsets of cognitive function but findings on global cognition and AD biomarkers were neutral. But evidence of other GCAs was either insufficient to make conclusions or unavailable. CONCLUSION: These studies suggest that different GCAs had varied effects on dementia risk within elder T2D, although GCAs may be unable to modify AD progression after AD onset. Without disturbing diabetes management, clinicians may consider lowering patients’ dementia risk through GCAs selection. Additional research is warranted to exam whether our findings apply to younger populations and minority groups.