Statins: what 26 trials show

Overview

CTT Collaboration, Baigent C et al.. Lancet 2010 · PMID 21067804 · DOI 10.1016/S0140-6736(10)61350-5

Each 1 mmol/L LDL reduction cut major vascular events by about a fifth and deaths by 10%, without more cancer.

Does the study support the specific claim? No · Methodological quality: high

Original publication

Plain-language summary

Scientific in-depth review

1. Full reference

Authors

CTT Collaboration, Baigent C et al.

Journal

Lancet

Year

2010

Identifier

PMID 21067804 DOI 10.1016/S0140-6736(10)61350-5

Verification

verified – PubMed efetch 2026-10-25

Retracted

no

Corrections

none

Extent checked

full text

Conflicts of interest and funding

Collaboration funded by UK MRC, British Heart Foundation, European Community Biomed Programme, Australian NHMRC and National Heart Foundation; funders had no role. Most included trials were supported by pharmaceutical industry grants; industry representatives (BMS, Merck, Pfizer, AstraZeneca, Bayer, Novartis, Fournier) were observers; some writing-committee members received meeting reimbursements and two received honoraria from Solvay.

2. Research question

What are the efficacy and safety of statin therapy and of more intensive LDL lowering?

3. Study design and population

Study design

Meta-analysis – Individual participant data meta-analysis of 26 RCTs

Population

Participants of 26 randomised statin trials: 5 trials of more vs less intensive statin (8,659 acute coronary syndrome and 30,953 stable coronary disease patients) and 21 trials of statin vs control (incl. primary prevention, diabetes, haemodialysis, heart failure, coronary disease)

Country

Multinational (trials in Europe, North America, Australasia and elsewhere); collaboration coordinated from Oxford (UK) and Sydney (Australia)

Period

Eligible trials reported by end of 2009; median follow-up 5.1 years (more vs less) and 4.8 years (statin vs control)

Sample size

169,138 participants (39,612 more vs less intensive; 129,526 statin vs control); 15,969 deaths; 10,124 first cancers. IPD unavailable for 3 eligible trials (CORONA, SPARCL, GREACE; 11,342 patients)

Inclusion and exclusion criteria

Randomised trials whose main effect was LDL lowering, with no other intended risk-factor differences, >=1000 participants and >=2 years scheduled treatment (protocol agreed before first trial results).

4. Methods and endpoints

Exposure / intervention

Statin / more intensive statin

Comparator

Control / less intensive statin

Endpoints

Major vascular events, mortality, cancer

Statistical methods

Prespecified CTT protocol; intention-to-treat; log-rank (o-e) statistics per trial weighted by absolute LDL difference at 1 year, giving rate ratios per 1.0 mmol/L; heterogeneity/trend chi-square; 95% CIs for summary and 99% CIs for subgroups to allow for multiplicity.

5. Actual results

Results

More vs less intensive (mean further LDL reduction 0.51 mmol/L): major vascular events 3837 (4.5%/year) vs 4416 (5.3%/year), 15% reduction (95% CI 11-18); coronary death or non-fatal MI -13% (7-19); revascularisation -19% (15-24); ischaemic stroke -16% (5-26). Statin vs control (1.07 mmol/L): 7136 (2.8%/year) vs 8934 (3.6%/year), 22% reduction (19-24). All 26 trials per 1.0 mmol/L: major vascular events RR 0.78 (0.76-0.80); all-cause mortality RR 0.90 (0.87-0.93); coronary death RR 0.80 (99% CI 0.74-0.87); stroke death RR 0.96 (0.84-1.09); non-vascular death RR 0.97 (0.92-1.03); cancer incidence RR 1.00 (0.96-1.04); ischaemic stroke RR 0.79 (0.74-0.85); haemorrhagic stroke RR 1.12 (0.93-1.35), 1.21 (1.05-1.41) when published SPARCL and CORONA data added. Benefit also in no prior vascular disease (25% per mmol/L), women (17%), age >75 (16%), and LDL <2.0 mmol/L (29%). Rhabdomyolysis excess 4 (SE 2) per 10,000 with intensive regimens (all in simvastatin 80 mg trials), 1 (SE 1) per 10,000 for standard statin vs control.

Absolute risks

Major vascular events 4.5% vs 5.3% per year (more vs less intensive) and 2.8% vs 3.6% per year (statin vs control); rhabdomyolysis excess about 4 per 10,000 (intensive) and 1 per 10,000 (standard); possible haemorrhagic stroke excess of a few per 10,000 per year vs a few hundred occlusive events avoided per 10,000 high-risk patients treated per year (authors' estimate).

Authors' conclusion

Further LDL reductions safely produce definite further reductions in heart attack, revascularisation and ischaemic stroke.

6. Strengths

• Individual participant data from 26 RCTs under a protocol prespecified before trial results
• Intention-to-treat; consistent outcome definitions; large numbers of events
• Assessment of safety outcomes (cancer, non-vascular death, haemorrhagic stroke, rhabdomyolysis)
• Dose-response by LDL reduction across trial types

7. Weaknesses

• Most trials industry-funded
• IPD missing for 3 eligible trials (11,342 patients); including their published haemorrhagic stroke data gives a significant excess
• Only rhabdomyolysis sought for muscle effects; milder myopathy, diabetes onset and other adverse effects not addressed here
• Search strategy not described in this paper (relies on collaboration protocol)
• Median follow-up about 5 years

8. Risk of bias

Orienting tool

AMSTAR 2 (IPD meta-analysis of RCTs)

Type of assessment

structured assessment based on the full text by a single reviewer (no formal double review)

Full formal assessment

no

Overall risk of bias

low

Rationale

Overall unchanged (low). Simplified single-reviewer judgement from the full text; a formal AMSTAR 2/ROBIS assessment requires two independent reviewers.

Full-text review (25 October 2026): Added absolute annual event rates, component endpoints, subgroup results (primary prevention, women, elderly, LDL <2 mmol/L), haemorrhagic stroke signal (1.21 with SPARCL/CORONA), rhabdomyolysis rates, missing IPD trials, prespecified protocol and industry funding of included trials/observers. Bias overall unchanged (low) with itemised AMSTAR 2.

9. Assessment of the statistical interpretation

Proportional reductions of about 22% in major vascular events per 1 mmol/L LDL reduction; absolute benefit depends on baseline risk (e.g. 0.8 percentage points per year in statin vs control trials). Small possible excess of haemorrhagic stroke and rare rhabdomyolysis are far outweighed in high-risk patients.

10. Relevance to the specific claim

Contradicts a blanket ban on statins.

11. Independent contrary findings

Absolute benefit small in very low-risk people (inference from relative effects).

12. Transferability to Germany and Europe

High.

13. Overall methodological judgement

Cornerstone RCT evidence that statins reduce heart attacks, revascularisations, ischaemic strokes and all-cause mortality without increasing cancer; small specific harms documented.

14. Open questions

15. Linked fact checks

• For six or seven years he alone has said that statins (Crestor, Ator, Lipitor) are forbidden, and that Controloc (pantoprazole) is forbidden. – Studies contradicting or limiting the claim