Diabetes treatment deintensification in Australians with dementia compared to the general population: A national cohort study

Background: Diagnosis of dementia may change peoples' goals of care. In people with diabetes, this may lead to relaxing treatment targets and reducing the use of diabetes medications. The aim of this study was to examine changes in diabetes medication use before and after initiating medication for dementia. Methods: A national cohort of people aged 65–97 years, living with dementia and diabetes, and a general population cohort with diabetes matched for age, sex, and index date were extracted from the Australian national medication claims database. Trajectories of diabetes medication use, expressed as mean defined daily dose (DDD) each month for each individual from 24 months before to 24 months after the index date, were estimated using group-based trajectory modeling (GBTM). Cohorts were analyzed separately. Results: People with dementia (N = 1884) and the matched general population (N = 7067) had a median age of 80 years (interquartile range 76–84) and 55% were female. In both models, people exhibited one of five diabetes medication trajectories, with 16.5% of people with dementia and 24.0% of the general population assigned to trajectories that represented deintensification. In the general population model, those on deintensifying trajectories were older than those on stable trajectories (median 83 vs. 79 years). In the dementia cohort model, those on high or low deintensifying trajectories were slightly older (median age 81 or 82, respectively, vs. 80 years) and had at least 1 more comorbidity (median 8 or 7, respectively, vs. 6) than those on stable trajectories. Conclusions: Initiating medication for dementia does not appear to be a trigger for deintensification of diabetes treatment regimens. Deintensification was more common in the general population; people living with dementia are potentially overtreated for diabetes.