LAUSR

Background: Care individualization dominates in clinical guidelines for cognitively impaired patients with diabetes; however, few studies examined such adaptations. Objective: Describe long-term pharmacological changes in diabetes treatment in subjects with and without dementia. Methods: We performed a registry-based cohort study on 133,318 Swedish subjects (12,284 [9.2%] with dementia) with type 2 or other/unspecified diabetes. Dementia status originated from the Swedish Dementia Registry, while the National Patient Register, Prescribed Drug Register, and Cause of Death Register provided data on diabetes, comorbidities, drug dispensation, and mortality. Drug dispensation interval comprised years between 2005 and 2018 and the dispensation was assessed relative to index date (dementia diagnosis) in full cohort and propensity-score (PS) matched cohorts. Annual changes of drug dispensation were analyzed by linear regression, while Cox and competing-risk regression were used to determine the probability of drug dispensation after index date in naïve subjects. Studied medications included insulin, metformin, sulfonylureas, thiazolidinediones, dipeptidyl-peptidase-4 inhibitors (DPP-4i), glucagon-like peptide-1 agonists (GLP-1a), and sodium-glucose cotransporter-2 inhibitors (SGLT-2i). Results: Dementia patients had higher probability of insulin dispensation (hazard ratio 1.21 [95% CI 1.11–1.31] and lower probability of DPP-4i (0.72 [0.66–0.79]), GLP-1a (0.51 [0.41–0.63]), and SGLT-2i dispensation (0.44 [0.36–0.54]) after index date. PS-matched analyses showed increased annual insulin dispensation (β difference 0.97%) and lower increase in DPP-4i (–0.58%), GLP-1a (–0.13%), and SGLT-2i (–0.21%) dispensation in dementia patients compared to dementia-free controls. Conclusion: Dementia patients had lower probability of receiving newer antidiabetic drugs, with simultaneous higher insulin dispensation compared to dementia-free subjects.