- Design
- Double-blind, randomised, placebo-controlled trial in Australian general practices, with two primary endpoints tested hierarchically (STAREE, NCT02099123)
- Population
- 9,971 community-dwelling adults aged 70 and over with no cardiovascular disease, diabetes or dementia; mean age 74.7, 51.9% women
- Primary outcome
- Major cardiovascular events, and separately a composite of death, dementia or persistent physical disability
- Effect
- Cardiovascular 10.9 vs 15.5 per 1000 person-years, HR 0.70 (95% CI 0.61-0.82), p<0.001; disability-free survival 21.6 vs 23.0, HR 0.94 (0.84-1.05), p=0.25
Whether to start a statin in a healthy 75-year-old has been argued from subgroup data and extrapolation for two decades. STAREE tested it: 9,971 community-dwelling Australians aged 70 or over with no cardiovascular disease, diabetes or dementia, randomised double-blind to atorvastatin 40 mg daily or placebo, with two primary endpoints and a hierarchical testing plan.
Over a median 5.9 years the cardiovascular endpoint - cardiovascular death, non-fatal myocardial infarction or stroke, or coronary revascularisation - occurred at 10.9 per 1000 person-years on atorvastatin against 15.5 on placebo, hazard ratio 0.70 (95% CI 0.61-0.82), p<0.001. That is a clear benefit, and the relative reduction matches what statins do at every other age.
The second primary endpoint did not move. Death from any cause, dementia or persistent physical disability occurred at 21.6 per 1000 person-years against 23.0, hazard ratio 0.94 (0.84-1.05), p=0.25. Serious adverse events were identical at 2.7%, but musculoskeletal, hepatobiliary and diabetes-related adverse events were commoner on atorvastatin.
The two results are not in conflict, and reading only one of them is how this trial will be misquoted. Preventing a myocardial infarction is worth having on its own terms. But the endpoint built to capture what older people say they care about - staying alive, thinking clearly, staying independent - did not improve, because the events driving it are mostly not cardiovascular.
So the conversation with a 75-year-old changes shape rather than direction. A statin will reduce their chance of a heart attack or stroke by about 4.6 events per 1000 person-years; it will not, on this evidence, keep them independent for longer, and it carries a real chance of muscle symptoms, deranged liver tests or new diabetes. That is a decision a patient can now make on numbers rather than on inference, and different patients will reasonably decide differently.
- Quote both endpoints when discussing primary prevention over 70 - one alone misrepresents the trial
- Absolute benefit is about 4.6 fewer cardiovascular events per 1000 person-years
- Warn about muscle symptoms, liver enzymes and new diabetes - all were commoner on atorvastatin
- This tested primary prevention only; nothing here justifies stopping a statin after a cardiovascular event
- Patients excluded from the trial - existing diabetes, dementia, established disease - are not covered by it
The statistics, in plain English
Two primary endpoints tested hierarchically means the trial was designed to answer both questions properly rather than to salvage one after the other failed - so the null result on disability-free survival carries the same weight as the positive cardiovascular result. A hazard ratio of 0.94 with an interval of 0.84 to 1.05 crosses 1.0: consistent with a small benefit and with a small harm, and the trial was large enough that a substantial effect would have shown. The event rate difference, 15.5 versus 10.9 per 1000 person-years, is the figure to give a patient; the 30% relative reduction sounds larger than it is.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for top clinical updates, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free