Most labeled statin side effects lack evidence
In blinded large trials, statins did not increase most harms listed on drug labels; small increases were seen mainly in liver blood-test abnormalities.
*Individual participant data meta-analysis of double-blind trials; Level 1a (OCEBM).

Citation

Cholesterol Treatment Trialists’ Collaboration. Assessment of adverse effects attributed to statin therapy in product labels: a meta-analysis of double-blind randomised controlled trials. Lancet. 2026;407:689–703. Published online Feb 5, 2026. doi:10.1016/S0140-6736(25)01578-8.

Background

Many possible statin side effects listed on drug labels come from non-blinded or non-randomized evidence that can overstate harms. This study tested those label-listed events using only large double-blind trials.

Patients

Adults in 23 double-blind randomized trials (19 statin vs placebo; 4 higher-dose vs lower-dose statin). Trials included ≥1000 participants and planned treatment ≥2 years; mixed primary and secondary prevention populations.

Intervention

Statin therapy (or more intensive statin regimen).

Control

Placebo (or less intensive statin regimen).

Outcome

Health problems listed as possible statin side effects on product labels, based on reported adverse events.

Follow-up Period

Median 4.5 years (statin vs placebo); 5.0 years (more vs less intensive).

Results

Outcome (statin vs placebo) Rate ratio (95% CI) Absolute annual excess risk Number needed to harm (per year)
Abnormal liver blood tests: transaminases (primary) 1.41 (1.26–1.57) 0.09% 1111
Other liver blood-test abnormalities (primary) 1.26 (1.12–1.41) 0.05% 2000
Urine composition alteration (primary) 1.18 (1.04–1.33) 0.03% 3333
Swelling (oedema) (primary) 1.07 (1.02–1.12) 0.07% 1429
Analyses were intention-to-treat. Higher-dose vs lower-dose statin showed larger excesses for liver blood-test abnormalities, supporting a dose-dependent effect.

Limitations

Outcomes came from adverse-event reports rather than direct review of lab results, which could miss mild test changes. Follow-up may not capture very long-term harms, and rare events remain hard to detect. Severity and whether symptoms resolved after stopping were not reliably available.

Funding

British Heart Foundation, UK Medical Research Council, Australian National Health and Medical Research Council.

Clinical Application

Reassure patients that most label-listed statin side effects lack trial support; focus counseling on known muscle effects, diabetes risk, and small, dose-related liver test changes.