Objectives: The objective of this study was to examine the association between the use of statins and the risk of newly diagnosed dementia in an elderly population. Design, setting and participants: Random samples of 1000 000 individuals covered by the National Health Insurance in Taiwan were included in the analysis. All participants were 65 years or older without dementia and either did or did not start treatment with statins from 1 August 1997 to 31 December 2010. Patients with established dementia before the start of treatment were excluded. Baseline characteristics were matched (by propensity score) in those who did and did not receive statins. Results: A total of 57 669 subjects were included in the analysis with approximately 12 years of follow-up. Propensity score matching identified 2003 patients who received statins and another 2003 patients who did not with comparable baseline characteristics. Adjusted hazard ratios (HRs) for dementia were significantly inversely associated with total or daily equivalent statin dosage (total accumulated dose: HRs 0.829, 0.720 and 0.385 from T1 to T3 vs. control, P < 0.001 for trend; mean daily dose: HRs 0.667, 0.798 and 0.503 from T1 to T3 vs. control, P < 0.001). The results remained robust after propensity adjustment. Conclusion: Independent of traditional risk factors, there was a decrease in newly diagnosed cases of dementia in elderly patients who had received a high total or daily dose of statins. The more potent statins (e.g. atorvastatin and rosuvastatin) seemed to be particularly effective in the prevention of dementia.
- Propensity score
ASJC Scopus subject areas