Atorvastatin, Cardiovascular Events, and Disability-free Survival in Older Adults (STAREE)

Authors, Journal, Affiliations, Type, DOI

• Authors: Sophia Zoungas, Rory Wolfe, Chris Moran, Stephen J. Nicholls, Geoffrey C. Cloud, Christopher M. Reid, Andrew M. Tonkin, Lawrence Beilin, Anthony S. Wierzbicki, Trevor T.-J. Chong, Jonathan C. Broder, Andrea J. Curtis, Zachary Flanagan, Ingrid Hopper, Joanne Ryan, Simone Spark, John J. McNeil, and Mark R. Nelson, for the STAREE Investigators
• Journal: N Engl J Med 2026;395(13):1259–72. Published online 29 August 2026; print issue 1 October 2026. A Quick Take and CME are available at NEJM.org
• Affiliations: Coordinated from the School of Public Health and Preventive Medicine, Monash University, Melbourne, with investigators from Monash, the University of Melbourne, Baker Heart and Diabetes Institute, and academic centres across Australia. Corresponding author Sophia Zoungas (Monash University)
• Type: Investigator-initiated, double-blind, randomized, placebo-controlled, primary-prevention trial conducted at general medical practices across Australia. No commercial or industry involvement. Independent data monitoring committee; primary end-point events adjudicated by committees blinded to group assignment. Registration NCT02099123. Design published previously
• DOI: 10.1056/NEJMoa2607314

Overview

STAREE answers the question lipid guidelines have never had randomized evidence for: whether to start a statin for primary prevention in adults aged 70 and over. In 9,971 community-dwelling Australians aged ≥70 with no clinical cardiovascular disease, diabetes or dementia, atorvastatin 40 mg daily reduced the primary cardiovascular composite over a median of 5.9 years from 15.5 to 10.9 events per 1000 person-years (HR 0.70; 95% CI 0.61–0.82; P<0.001) — but produced no improvement in disability-free survival (death from any cause, dementia, or persistent physical disability: HR 0.94; 95% CI 0.84–1.05; P=0.25). The trial was double-blind, placebo-controlled and industry-free, with a prespecified hierarchical testing procedure and 99.9% vital-status ascertainment, making it the most rigorous test of this question to date. Its most consequential feature is the dissociation it demonstrates: a clear reduction in nonfatal cardiovascular events — driven by nonfatal myocardial infarction and coronary revascularization — with no effect whatever on cardiovascular death (74 vs 74 events; HR 1.00) and no effect on dementia, physical disability or all-cause mortality.

Keywords

Not provided — NEJM original articles do not carry an author keyword list. The document's own indexing terms would be: primary prevention; statin therapy; atorvastatin; older adults; disability-free survival; major cardiovascular events; dementia; physical disability; adverse events.

Key Takeaways

Background

• A rapid demographic transition toward older populations is occurring worldwide, and healthy life span has not increased to the same extent — older adults still face major consequences from cardiovascular disease, dementia and other disabling conditions
• Statins are established for reducing MI and ischaemic stroke risk in high-risk adults (diabetes or established cardiovascular disease), but their effects when initiated in older patients remain uncertain and intensely debated
• Guidelines have not contained strong recommendations for statin prescription at age ≥70, with evidence particularly sparse above 75. Recommendations rested on pooled analyses in which fewer than one in four participants were over 70
• Those pooled analyses suggested statins may be less effective for primary prevention in older adults without vascular disease: per 1.0 mmol/L (38.7 mg/dL) LDL-C reduction, the cardiovascular event reduction was non-significant — 16% in those over 70 and 8% in those over 75
• Few trials had assessed dementia or disability. The Heart Protection Study used a single cognitive test probably lacking sensitivity to subtle decline; PROSPER and HOPE-3 included cognitive measures but did not formally assess newly diagnosed dementia as a clinical end point; JUPITER tracked dementia as an adverse event without standardised objective assessment and identified very few cases (12 vs 9)
• No large-scale randomized trial had evaluated disability-free survival as a prespecified primary end point
• Prior concerns about statins in the elderly: muscle weakness or pain, impaired liver function, cognitive impairment, new-onset diabetes, and haemorrhagic stroke. Older adults take more medications and metabolise drugs differently, so are at greater risk of statin–drug interactions and adverse events

Methods

• Design: investigator-initiated, double-blind, randomized, placebo-controlled primary-prevention trial at general medical practices across Australia
• Population: adults ≥70 years living independently with no history of cardiovascular disease, diabetes, dementia, or other life-limiting conditions — identified from the clinical databases of 1,013 of 1,583 registered general practices across metropolitan, regional and rural Australia. Cardiovascular disease was defined as MI, angina, heart failure, stroke, TIA, carotid stenosis, abdominal aortic aneurysm, peripheral artery disease, or any coronary-artery blockage or narrowing leading to angioplasty, stent implantation, or bypass surgery
• Run-in: two baseline screening visits 4 weeks apart including cognitive tests, a modified Katz Index of Independence in Activities of Daily Living, and laboratory assessment; between visits participants took a once-daily placebo to assess adherence to a daily regimen
• Intervention: atorvastatin versus identical placebo. Starting dose one tablet daily for 4 weeks, increasing to two tablets daily if the safety profile allowed — in the atorvastatin group each tablet contained 20 mg, giving a total of 40 mg daily. Further de-escalation and re-escalation was permitted for side-effect management
• Follow-up: every 6 months. Open-label statins could be prescribed at the treating clinician's discretion, with the assigned trial regimen discontinued. Fasting lipids measured at baseline and at the end of the trial
• Primary end points (two, with hierarchical testing):
  1. Major cardiovascular events — composite of death from cardiovascular causes, nonfatal myocardial infarction or stroke, or coronary revascularization
  2. Disability-free survival — composite of death from any cause, dementia, or persistent physical disability
• Prespecified hierarchical (closed) testing procedure: the cardiovascular end point was tested first at a two-sided alpha of 0.05; only if P<0.05 was the disability-free-survival end point then tested at 0.05. This was based on the expectation that cardiovascular benefit would be the main contributor to improved disability-free survival, and that a substantial effect on disability-free survival would be unlikely without a cardiovascular benefit
• Secondary end points: CV death or nonfatal MI or stroke; fatal or nonfatal MI; fatal or nonfatal stroke; coronary revascularization; any arterial revascularization; death from any cause; dementia; persistent physical disability; hospitalization for any cause. No adjustment for multiplicity — the article states the confidence intervals should not be used to infer treatment effects
• Statistics: event rates compared by Cox proportional-hazards regression, expressed as cause-specific hazard ratios; intention-to-treat; competing risk of death from other causes accounted for in cumulative-incidence estimation. Safety analysed in the modified intention-to-treat population (received ≥1 dose). Adverse events coded by MedDRA v21.0
• Power: 619 primary cardiovascular events and 1,397 disability-free-survival events were calculated to give 80% power to detect a 20.2% difference in major cardiovascular events and a 14.5% difference in death/dementia/disability. Target enrolment 9,631 followed for an average of 6 years

Results — Participants and Follow-Up

• 9,971 participants randomized between July 2015 and March 2023: 4,984 to atorvastatin, 4,987 to placebo. Data locked February 2026
• Mean age 74.7 ± 4.5 years; 51.9% women. Age distribution: 70–74 y 59.7%, 75–79 y 25.1%, 80–84 y 11.2%, ≥85 y 4.1% — so 40.4% were aged 75 or over
• 98.3% White and 95.7% had English as a first language
• Median 2 medications at baseline (IQR 1–4); 19.6% were taking ≥5 medications; most common were blood-pressure-lowering agents (46.3%), acid-related gastrointestinal agents (27.1%) and analgesics (20.6%). Only 5.2% reported previous statin use
• Lipids at baseline: mean total cholesterol 211 ± 32 mg/dL (5.46 ± 0.83 mmol/L); LDL-C 126.6 ± 28.0 mg/dL (3.27 ± 0.72 mmol/L); HDL-C 62 ± 17 mg/dL; triglycerides 113 ± 54 mg/dL
• Median follow-up 5.9 years. Complete vital-status assessment was available through a median of 6.4 years by linkage to death-record registers for 99.9% of participants
• Median LDL-C reduction was approximately 38% with atorvastatin versus 13% with placebo — the placebo group's reduction reflecting open-label statin use

Results — Primary End Points

• Overall, 709 participants (7.1%) had a primary cardiovascular event and 1,313 (13.2%) had a disability-free-survival event. Both end points accrued close to their powering assumptions (619 and 1,397 expected)
• The hierarchical gate passed — the cardiovascular end point was significant, so the disability-free-survival test was validly performed, and it was null

Results — Components and Secondary End Points

Rates are events per 1000 person-years. Secondary end points were not adjusted for multiplicity.

• The composite benefit is carried by nonfatal events. Cardiovascular death was numerically identical (74 vs 74; HR 1.00), stroke was not significant, and all-cause mortality, dementia and hospitalisation showed no difference. The reductions were concentrated in nonfatal myocardial infarction (HR 0.58) and coronary revascularization (HR 0.57)
• Effects appeared consistent across prespecified subgroups (age, sex, blood pressure) — reported in the trial's Figure 2

Results — Safety

Safety events assessed by the trial physician as related to the trial regimen, in the modified intention-to-treat population (4,876 atorvastatin, 4,847 placebo):

• Serious adverse events were identical between groups (2.7% vs 2.7%), and serious musculoskeletal events were identical (0.2% vs 0.2%). The authors report that serious musculoskeletal and hepatobiliary events of special interest together affected only 29 participants
• The excess lies in non-serious categories: medically important adverse events overall (8.8% vs 6.7%), hepatobiliary disorders at roughly four times the placebo rate, and diabetes-related events
• The article's summary states that musculoskeletal, hepatobiliary and diabetes-related adverse events occurred more commonly with atorvastatin

Discussion

• The trial included only community-dwelling adults ≥70 (>40% were older than 75), unlike previous primary-prevention statin trials
• The cardiovascular benefit was larger than that seen in PROSPER (pravastatin 40 mg, mean age 75.4 years) and larger than reported by the Cholesterol Treatment Trialists' Collaboration in trial populations of similar age — attributed to the high-dose atorvastatin choice, which produced a ~38% LDL-C reduction
• The authors state the effect "was due primarily to a lower incidence of nonfatal events, especially myocardial infarction and coronary revascularization, with atorvastatin than with placebo; numbers of deaths from cardiovascular causes were small overall"
• Why no cardiovascular mortality benefit? The authors attribute this to low case fatality rates with universal subsidised access to contemporary cardiovascular management, and to the competing risk of death from other causes in this older population
• Why no disability-free survival benefit? Because 80% of deaths in the trial were from noncardiovascular causes. "Notwithstanding the purported pleiotropic biologic properties of statins and concerns about harm from low cholesterol levels, no benefits or risks of atorvastatin with respect to death from any cause, dementia, or physical disability were observed." The authors note the number of new physical-disability cases was small, limiting conclusions about that component
• The authors consider their effect estimates conservative, given the substantial proportion who discontinued the trial regimen and the greater use of open-label statins in the placebo group than in the atorvastatin group
• Adherence findings: a higher discontinuation rate than in previous statin trials, with the most common reason "participant unwilling" — 15.6% in the atorvastatin group and 16.2% in the placebo group — exceeding discontinuation for adverse events or unacceptable side effects (7.2% and 6.1%). Combined with real-world 5-year statin adherence of only 25–50%, the authors conclude that "tackling persistence with statin use remains critical to future prescribing effectiveness for primary prevention in older adults"
• The participants were representative of older adults who, largely on age alone, would be considered at high risk and eligible for primary-prevention therapy

Limitations of the document

Authors' own stated limitations

• Ethnic and linguistic homogeneity. Despite a highly decentralised recruitment strategy through Australian general practices, participants were predominantly White with English as their first language (98.3% and 95.7% respectively)
• Conducted through the COVID-19 pandemic, which required many assessments to be completed remotely rather than in person and extended the recruitment period. The authors report no difference in effects between participants randomised before the pandemic and those randomised during or after it
• The focus on primary prevention limits generalisability of the findings
• The number of new cases of physical disability was small, limiting conclusions about that component of the disability-free-survival end point

Additional limitations evident in the design and results

• The disability-free-survival end point was null, and this is the trial's central negative result rather than a secondary observation. Roughly 40% of participants were ≥75 and the median age was 74.7, yet the trial provides no evidence that statins extend independence, prevent dementia (HR 1.03, CI 0.87–1.21) or prolong life (all-cause death HR 0.91, CI 0.79–1.04)
• No effect on cardiovascular death — 74 events in each arm, HR 1.00 (0.72–1.37) — which means the composite benefit rests on nonfatal events. The trial was not powered for the individual mortality components, and the authors acknowledge small numbers of cardiovascular deaths
• One of the two major drivers of the composite is a procedure, not a clinical event. Coronary revascularization contributes substantially to the primary composite (HR 0.57), and revascularization is a decision-dependent end point sensitive to local practice, symptom reporting and physician threshold. This is mitigated — but not eliminated — by the trial being double-blind, so those decisions were made without knowledge of allocation. The near-identical hazard ratios for nonfatal MI (0.58) and revascularization (0.57) suggest both reflect the same underlying reduction in coronary events
• The safety signal is real but underweighted in the abstract. Hepatobiliary disorders occurred at 3.3% vs 0.9% (medically important 1.4% vs 0.3%) and diabetes-related events at 4.0% vs 2.7% — excesses that the summary reports only as events "occurring more commonly." Serious adverse events were identical, so the excess is in non-serious categories, but a ~4-fold hepatobiliary signal in a population already at risk from polypharmacy deserves explicit mention
• The comparison is diluted in both directions. The placebo arm received open-label statins (13% LDL-C reduction), which biases the intention-to-treat effect toward the null and makes the observed HR 0.70 an underestimate — the authors' own reading. Conversely, 5.2% of participants had prior statin use and adherence in both arms was imperfect, so the trial answers "what happens if you offer atorvastatin" rather than "what happens if you take it"
• Absolute benefit is modest and concentrated. 15.5 → 10.9 events per 1000 person-years is an absolute reduction of about 4.6 events per 1000 person-years, or roughly 1 major cardiovascular event prevented per 217 patient-years of treatment — with no mortality or independence benefit to weigh against 5.9 years of daily medication and the adverse-event excess
• No industry involvement is a genuine strength and is stated explicitly; funding was from the National Health and Medical Research Council, the Heart Foundation of Australia, and Monash University — but it also means no independent commercial replication of the result is available or expected
• Publication status: published online 29 August 2026; the version ingested is the 14-page print-issue article

Key Concepts Mentioned

• concepts/Dyslipidemia — statin therapy, primary prevention in older adults, and statin-attributed adverse effects
• concepts/ASCVD-Risk-Assessment — the age-based risk framework (SCORE2-OP for ≥70) that makes older adults eligible for primary-prevention therapy on age alone
• Disability-free survival — survival free of dementia or persistent physical disability; a prespecified co-primary end point, and the first large trial to evaluate it for statins
• Major adverse cardiovascular events — here defined to include coronary revascularization alongside CV death, nonfatal MI and nonfatal stroke
• Hierarchical (closed) testing procedure — the prespecified approach permitting two primary end points without alpha inflation
• Statin-associated adverse effects — muscle symptoms, hepatobiliary disorders, and new-onset diabetes
• Polypharmacy in older adults — median 2 medications, 19.6% on ≥5, as the context for statin–drug interaction risk

Key Entities Mentioned

• entities/Heart-Failure — heart failure was among the cardiovascular conditions defining exclusion from the trial
• entities/Hypertension — blood-pressure-lowering agents were the commonest concurrent medication (46.3%)
• concepts/Atherosclerosis-Pathophysiology — the disease process the intervention targets
• Monash University, Melbourne — coordinating institution
• STAREE Investigators — trial group; registration NCT02099123

Wiki Pages Updated

• Created: wiki/sources/lipid-staree-nejm-2026.md
• Updated: wiki/concepts/Dyslipidemia.md
• Updated: wiki/concepts/ASCVD-Risk-Assessment.md
• Updated: wiki/sourceindex.md
• Updated: wiki/wikiindex.md