Concomitant Use of Atypical Antipsychotics and Cholinesterase Inhibitors Among Older Adults with Dementia
Objectives: The objectives of the dissertation research were to (1) compare the time to initiation of antipsychotic medications among older adults with dementia treated with acetylcholinesterase inhibitors (AChEIs), donepezil, rivastigmine, or galantamine; (2) evaluate the risk of falls and fractures due to the concomitant use of AChEIs and atypical antipsychotics (classified as high versus low anticholinergic) among older adults with dementia; and (3) evaluate the risk of mortality due to the concomitant use of AChEIs and atypical antipsychotics (classified as high versus low anticholinergic) among older adults with dementia. Methods: This study involved a new-user retrospective cohort study design using multi-year (2013-2015) Medicare claims data involving parts A, B, and D. The study included community-dwelling older adults (≥ 65 years) with a diagnosis of dementia. The outcome measures of interest were time to antipsychotic initiation (objective 1), falls/fractures (objective 2), and mortality (objective 3). The primary independent variable was the receipt of AChEIs (objective 1), and the concomitant use of AChEIs and atypical antipsychotics (objectives 2 and 3). The AChEIs included donepezil, rivastigmine, and galantamine. Antipsychotics included typical and atypical agents. For Objectives 2 and 3, the atypical antipsychotics were classified as high-anticholinergic (olanzapine, quetiapine) and low-anticholinergic atypicals (risperidone, aripiprazole, ziprasidone, asenapine, paliperidone, iloperidone, lurasidone) based on the Anticholinergic Drug Scale. The conceptual framework of the Andersen Behavioral model was used to select the predisposing, enabling, and need factors. Multivariable Cox proportional hazards regression compared the time to antipsychotic initiation among the three AChEIs adjusting for other risk factors (Objective 1). Multivariable Cox proportional hazards regression with inverse probability of treatment weighting (IPTW) was used to evaluate the risk of falls/fractures (Objective 2) or mortality (Objective 3) among high-anticholinergic versus low-anticholinergic atypical antipsychotic users with concomitant AChEIs use. Results: For the first objective, there were 178,441 older adults with dementia who were new users of AChEIs. The mean time to antipsychotic initiation among patients who received antipsychotics was 109.29 (±69.72) days for donepezil users, 96.70 (±71.60) days for rivastigmine users, and 104.15 (±72.53) days for galantamine users. The Cox regression analysis showed that rivastigmine (aHR=1.27; 95% CI, 1.20-1.34) was significantly associated with antipsychotic initiation as compared to donepezil, while there was no significant difference between galantamine and donepezil (aHR=0.98; 95% CI, 0.81-1.20). For the second and third objectives, the study identified 12,541 older adults with dementia who were concomitant users of AChEIs and atypical antipsychotics. There were 8,128 (64.81%) high-anticholinergic atypical antipsychotic users, and 4,413 (35.19%) low-anticholinergic atypical antipsychotic users with concomitant AChEIs use. For the second objective, the Cox regression with IPTW found no statistically significant difference (aHR=1.05; 95% CI, 0.94-1.18) between high-level and low-level anticholinergic atypical antipsychotic users with concomitant AChEIs use for the risk of falls/fractures. For the third objective, the Cox regression with IPTW found that high-anticholinergic atypical antipsychotic users had a lower risk of mortality (HR=0.86; 95% CI, 0.77-0.96) as compared to low-anticholinergic atypical antipsychotic users with concomitant AChEIs use. Conclusions: The study found that there was a 27% increased risk of antipsychotic initiation among users of rivastigmine as compared to donepezil users, while there was no evidence of a difference between galantamine and donepezil for the risk of antipsychotic initiation. Also, no differential effect of high-anticholinergic atypical antipsychotics on falls/fractures was observed as compared to low-anticholinergic atypical antipsychotics with concomitant AChEIs use. However, there was a lower risk of mortality among users of high-anticholinergic versus low-anticholinergic atypical antipsychotics with concomitant AChEIs use. Further studies are needed to better understand the role of anticholinergic activity of atypical antipsychotics for adverse outcomes in the vulnerable population of older adults with dementia.