Assessment of adverse effects attributed to statin therapy in product labels: a meta-analysis of double-blind randomised controlled trials.
Level 1 - systematic review of randomized trials
Meta-analysis of individual participant data from double-blind randomized controlled trials
PubMed 41655587 · doi:10.1016/S0140-6736(25)01578-8
What was done
Researchers conducted an individual participant-level data meta-analysis of double-blind randomized controlled trials (RCTs) evaluating statin therapy. They compiled all adverse effect terms listed in the Summaries of Product Characteristics (SmPCs) for five statins (atorvastatin, fluvastatin, pravastatin, rosuvastatin, and simvastatin). Eligible trials had at least 1,000 participants, a scheduled treatment duration of at least 2 years, and compared statin versus placebo or intensive versus less-intensive statin regimens. Event rate ratios (RRs) and 95% confidence intervals were calculated, controlling for false discovery rate (FDR) at 5%.
What was found
The primary analysis included 19 statin-versus-placebo trials with 123,940 participants (median follow-up 4.5 years). Beyond established risks of diabetes and muscle symptoms, only 4 of 66 other label-listed adverse outcomes showed statistically significant excess after FDR correction: - Abnormal liver transaminases: 0.30% per year (783 participants) on statin vs 0.22% per year (556 participants) on placebo (RR 1.41, 95% CI 1.26–1.57). - Other liver function test abnormalities: 0.25% per year (651 participants) vs 0.20% per year (518 participants) (RR 1.26, 95% CI 1.12–1.41); combined liver test absolute annual excess was 0.13%. - Urinary composition alteration: 0.21% per year (556 participants) vs 0.18% per year (472 participants) (RR 1.18, 95% CI 1.04–1.33). - Oedema: 1.38% per year (3,495 participants) vs 1.31% per year (3,299 participants) (RR 1.07, 95% CI 1.02–1.12). In four trials comparing more-intensive versus less-intensive statin regimens, significant dose-dependent excesses persisted only for liver test abnormalities, but not for oedema or urinary composition changes. Most conditions listed on product labels—including cognitive impairment, depression, sleep disturbance, and peripheral neuropathy—showed no evidence of a causal relationship.
Why it matters
This provides high-certainty evidence that the vast majority of adverse effects attributed to statins on product labels are not causally related. Revising these labels may reduce unwarranted patient reluctance or non-adherence driven by unverified side-effect claims.
Limits
The analysis was restricted to trials with at least 1,000 participants and at least 2 years of scheduled follow-up, which may not capture very rare idiosyncratic reactions or immediate acute effects. The abstract does not provide the total participant count for the four intensive-versus-less-intensive trials, nor does it detail specific statin-by-statin subgroup interactions for every outcome.
Cited by
- supports Clinical data do not show that statins cause cognitive impairment or sexual dysfunction.