Question explored with the scientific record
Is there a correlation between statin use and dementia?
The short version: the evidence is mixed, mostly observational, and funded by parties with a stake in the outcome. No single study settles the question.
The evidence you have here is a pile of observational studies, not a single randomized trial designed to test statins against dementia as a primary endpoint. The Cache County Study, one of the better long-term looks, found that while statin users had less dementia at the start, statin use did not predict who would develop it later [1]. The Taiwan cohort found a lower risk, especially with longer use, but that is an observational study in a single population [2]. The Rotterdam Study, which tried to mimic a trial design, found that simply starting a statin made almost no difference to dementia risk over 10 years (risk difference -0.1%, confidence interval crossing zero). Only sustained use showed a hint of benefit, and that disappeared when they excluded older people or those with high cholesterol [4]. The 2025 ADVANCE analysis in diabetics found no meaningful effect of statins on dementia or cognitive decline over about 5 years [3]. The umbrella review from 2025, which looked at 7 meta-analyses on statins, rated the evidence as moderate certainty for a protective signal, but noted the data is mostly observational and the heterogeneity is high [5].
The mechanism question is real. Statins cross the blood-brain barrier to varying degrees; simvastatin does, pravastatin does not [8]. Cholesterol is essential for brain function and myelin formation. The HMG-CoA reductase enzyme that statins block is the rate-limiting step in cholesterol synthesis, and the body compensates by upregulating the enzyme when cholesterol supply drops [6, 7]. What that means for long-term brain health in an individual is not known from these studies.
| Study | Design | Key finding on dementia |
|---|---|---|
| Cache County [1] | Prospective cohort | No association with incident dementia |
| Taiwan cohort [2] | Nationwide cohort | Lower risk with statin use (HR 0.78) |
| Rotterdam [4] | Target trial emulation | No effect of initiation; possible benefit with sustained use |
| ADVANCE [3] | Observational in diabetics | No meaningful effect over 5 years |
| Umbrella review [5] | 7 meta-analyses | Moderate certainty of protective signal, high heterogeneity |
My call: the evidence does not prove that statins cause or prevent dementia. The most rigorous analyses show no effect from starting a statin, and the protective signal in sustained users could easily be healthy-user bias. Confidence: low.
Sources used 8
-
Do Statins Reduce Risk of Incident Dementia and Alzheimer Disease?<subtitle>The Cache County Study</subtitle>
The Cache County Study investigates the association between statin use and the prevalence and incidence of dementia and Alzheimer’s disease, finding that while statin use is inversely associated with prevalent dementia, it does not predict the incidence of dementia or Alzheimer’…
DOI: 10.1001/archpsyc.62.2.217 -
Statin use and incident dementia: A nationwide cohort study of Taiwan
This nationwide cohort study in Taiwan found that statin use is associated with a significantly lower risk of incident dementia among elderly patients, with variations based on gender and the potency and duration of statin use.
DOI: 10.1016/j.ijcard.2014.03.018 -
Lipoprotein levels and statin treatment related to dementia and cognitive decline in individuals with type 2 diabetes: an observational analysis from the ADVANCE study
In individuals with type 2 diabetes from the ADVANCE trial, baseline LDL cholesterol was associated with incident dementia/cognitive decline, while statin use did not meaningfully alter dementia/CD risk over ~5 years.
DOI: 10.1186/s12933-025-02894-3 -
Emulating a target trial of statin use and risk of dementia using cohort data
Emulating a target trial in the Rotterdam Study, statin initiation showed little 10-year dementia risk difference, while sustained statin use was associated with reduced risks of dementia and dementia or death, though residual confounding and wide CIs limit interpretation.
DOI: 10.1212/wnl.0000000000010433 -
Systemic medications and dementia risk: a systematic umbrella review
An umbrella review of 68 meta-analyses across 11 systemic drug classes in humans finds moderate-certainty evidence that antihypertensives, statins, SGLT2 inhibitors, and GLP-1 receptor agonists are associated with reduced dementia risk, moderate certainty of increased risk with …
DOI: 10.1038/s41380-025-03129-3 -
Lipoprotein-mediated regulation of 3-hydroxy-3-methylglutaryl coenzyme A reductase activity and cholesteryl ester metabolism in the adrenal gland of the rat.
Adrenal HMG-CoA reductase activity and cholesteryl ester content in the rat are controlled by the availability of plasma lipoprotein cholesterol; depletion of plasma cholesterol upregulates reductase and depletes esters, which is reversed by lipoprotein cholesterol infusion and …
DOI: 10.1016/s0021-9258(17)40617-x -
Selective Compensatory Induction of Hepatic HMG-CoA Reductase in Response to Inhibition of Cholesterol Absorption
This study investigates the effects of the cholesterol absorption inhibitor ezetimibe on serum cholesterol levels and hepatic HMG-CoA reductase expression in diabetic and thyroidectomized rats, revealing a significant reduction in cholesterol levels and a compensatory increase i…
DOI: 10.1177/153537020623100510 -
Brain Cholesterol Synthesis in Mice Is Affected by High Dose of Simvastatin but Not of Pravastatin
High-dose, short-term simvastatin crosses the blood-brain barrier in mice and reduces brain cholesterol synthesis (lower lathosterol and higher HMG-CoA reductase mRNA), while pravastatin does not, with brain cholesterol turnover remaining unchanged.
DOI: 10.1124/jpet.105.094136