OBJECTIVES: There are 55 million people living with Alzheimer’s disease and related dementias (ADRD) worldwide. According to earlier research, patients with ADRD may benefit cognitively from statin treatment. There is a literature gap on causal relationships between statin use and ADRD risks. This study aimed to employ double/debiased machine learning (DML) to clarify this relationship.

METHODS: Data from the Medicare Current Beneficiary Survey (MCBS) spanning from 2017 to 2019 was utilized in this study. ICD code, anti-dementia drug use, and self-reported survey were used to identify ADRD patients. All FDA-approved statins were included in this study. Based on the newly released NIA Health Disparities Research Framework, a total of 47 covariates were used in our study. DML was employed to explore the causal relationship. We used Random Forest (RF) and Extreme Gradient Boosting (XGB) to predict the outcome and treatment variable, respectively. Survey sampling weights were applied in this study to generate national estimates.

RESULTS: A total of 87,112,517 Medicare beneficiaries were included in this study after exclusion, among which 57.81% used AHMs and 8.39% were diagnosed with ADRD. Compared to those not using statins, individuals taking statins had a lower prevalence of ADRD (P = 0.023). Our results showed that taking statins could not help decrease the risks of ADRD (P = 0.194). However, lovastatin, pravastatin, rosuvastatin, and simvastatin could reduce ADRD risks by 2.0%, 1.5%, 2.9%, and 1%, respectively (all P < 0.05).

CONCLUSIONS: This study illuminated a significant inverse causal relationship between 4 statins drug (lovastatin, pravastatin, rosuvastatin, and simvastatin) use and ADRD risks. Our findings suggest that some statins may be an effective antidementia therapy for ADRD patients in routine care, potentially leading to more cost-effective treatment strategies for ADRD individuals.