McGuinness · The Cochrane database of systematic reviews 2016 · systematic review of randomized controlled trials · n=26,340 participants across 2 studies

Statins for the prevention of dementia.

Cited 465 times in the scientific literature.

Level 1 - systematic review of randomized trials

Systematic review of randomized controlled trials

PubMed 26727124 · doi:10.1002/14651858.CD003160.pub3 · record verified 2026-08-30

What was done

Authors conducted a Cochrane systematic review searching multiple databases through November 2015 for double-blind, randomized, placebo-controlled trials administering statins for at least 12 months to individuals at risk of dementia. Two trials evaluating simvastatin and pravastatin met inclusion criteria, enrolling 26,340 participants aged 40 to 82 years (11,610 aged 70 or older) with vascular disease or vascular risk factors, followed for a mean of 3.2 and 5.0 years.

What was found

Only one study reported dementia incidence (20,536 participants; 31 cases per group), showing no reduction with statins (OR 1.00, 95% CI 0.61 to 1.65; moderate-quality evidence downgraded for imprecision). Cognitive function measures from both trials could not be pooled due to differing scales and assessment schedules, but no differences between statin and placebo groups were observed across five cognitive tests (high-quality evidence). Treatment discontinuation due to non-fatal adverse events was under 5% in both trials, with no difference between groups (26,340 participants, 2 studies, OR 0.94, 95% CI 0.83 to 1.05).

Why it matters

This review shows that observational associations linking statin use to reduced dementia risk are not supported by randomized trial evidence, likely reflecting indication bias in earlier observational cohorts.

Limits

Only two trials met inclusion criteria and only one measured clinical dementia incidence, resulting in wide confidence intervals and imprecision. Cognitive outcomes could not be meta-analyzed due to heterogeneous tests and timing. Participants were restricted to individuals with moderate-to-high vascular risk, and follow-up was limited to 3.2 to 5 years.

Cited by