Dr. Kumar’s Take:
This paper is not the JUPITER trial itself. It is a critical appraisal of JUPITER, written after the results were released in November 2008 and greeted as a breakthrough that should expand statin use for primary prevention. The authors accept that rosuvastatin lowered cardiovascular event rates in people with elevated C-reactive protein (CRP, a marker of inflammation) and LDL cholesterol under 130 mg/dL. Their argument is that the absolute benefit was small, the trial was stopped early for benefit by the data and safety monitoring board, and the long term risks in healthy people were never observed. I find that argument sound. When you treat a population with a low baseline event rate, a striking relative percentage can sit on top of a very small real world difference, and the decision to medicate healthy people has to be made on the real world difference.
Brief Summary:
JUPITER (Justification for the Use of Statins in Primary Prevention: an Intervention Trial Evaluating Rosuvastatin) tested whether rosuvastatin 20 mg orally per day reduced vascular events in apparently healthy people with LDL cholesterol below 130 mg/dL but high-sensitivity CRP of 2.0 mg/L or more. Participants were randomly assigned to rosuvastatin or placebo and followed for cardiovascular events. The trial was planned to run four years and was stopped early for benefit after a median follow-up of 1.9 years.
Key findings included:
- 44% relative risk reduction in the primary composite cardiovascular endpoint (hazard ratio 0.56, 95% CI 0.46 to 0.69).
- 54% relative risk reduction in major coronary events, meaning fatal or nonfatal myocardial infarction.
- Primary endpoint event rate: 0.77% per year on rosuvastatin versus 1.36% per year on placebo.
- Absolute risk reduction for the primary endpoint: 0.59% per year, or about 1.2% over the roughly two years the trial actually ran.
- Major coronary event rate: 0.17% per year on rosuvastatin versus 0.37% per year on placebo, an absolute risk reduction of 0.20% per year.
- Number needed to treat: 169 people for one year to prevent one event in the composite endpoint, and 500 people for one year to prevent one fatal or nonfatal heart attack.
- Glycated hemoglobin and diabetes incidence were elevated in the rosuvastatin group.
Key Takeaways:
✔ Rosuvastatin reduced cardiovascular event rates in people selected by high CRP rather than by high LDL.
✔ Relative risk reduction: 44% for the composite cardiovascular endpoint, 54% for fatal or nonfatal heart attack.
✔ Absolute risk reduction was small: 0.59% per year for the composite endpoint, 0.20% per year for major coronary events.
✔ 500 people would need one year of treatment to prevent one heart attack, and those people were considered healthy at enrollment.
✔ Industry sponsorship shapes how results are released and framed, including the choice to present relative rather than absolute effects.
✔ Stopping at 1.9 years of a planned four years leaves long term safety unresolved.
Study Design:
This was a randomized trial of rosuvastatin 20 mg daily versus placebo in apparently healthy people with LDL cholesterol below 130 mg/dL and high-sensitivity CRP of 2.0 mg/L or more.
- Primary endpoint: a composite of five conditions: nonfatal myocardial infarction, nonfatal stroke, hospitalization for unstable angina, arterial revascularization, or confirmed death from cardiovascular causes.
- Follow-up: median 1.9 years, against a planned duration of four years.
- Exclusions: people with a history of diabetes, uncontrolled hypertension, renal dysfunction, cancer within five years, and many other comorbidities were kept out of the trial.
- Eligibility implications: applying the JUPITER results would make an additional 19% of the older United States population (men 50 and older, women 60 and older), over 11 million people, eligible for statin therapy, bringing the proportion of older adults with an indication for statins to about 80%.
Results (Explained Simply):
Over one year of treatment:
- 1.36% of the placebo group had a heart attack, stroke, hospitalization for unstable angina, revascularization, or cardiovascular death, compared with 0.77% on rosuvastatin.
- That difference is 0.59% per year, so 169 people need one year of treatment to prevent one of those events.
- For fatal or nonfatal heart attack alone, the rates were 0.37% per year on placebo and 0.17% per year on rosuvastatin, a difference of 0.20% per year. That means 500 people need one year of treatment to prevent one heart attack.
- The same data expressed as relative change reads as 44% for the composite endpoint and 54% for heart attack. When event rates are low, a small absolute difference looks dramatic in relative terms.
- Glycated hemoglobin and diabetes incidence were higher in the rosuvastatin group.
- Earlier statin trials show a similar pattern. A meta-analysis of 90,056 patients with high cholesterol, 47% of whom already had coronary heart disease, found a 23% relative risk reduction in fatal or nonfatal heart attack but only a 2.4% absolute risk reduction over a mean of five years, about 0.48% per year, with a number needed to treat of 208.
Cost of Applying These Results
Rosuvastatin costs about $3.50 per day, which works out to roughly $638,750 per year for each major coronary event averted. A generic statin at about $4 per month brings that figure lower, though it remains substantial. Neither number includes screening or safety monitoring. In JUPITER, about 80% of the people who attended the screening visit did not qualify for enrollment, so many people would be screened and never treated. At an assumed $25 per CRP test, screening adds about $62,500 per year for each major coronary event averted. Adding liver function testing before and during therapy, plus glucose and glycated hemoglobin testing, at an assumed $30 total, could push the combined screening and monitoring cost above $137,000 per year for each event averted.
Industry Conflicts & Early Termination
🔹 JUPITER was an industry-sponsored trial, and marketing interests do not necessarily align with public health goals.
🔹 Stopping early and reporting relative rather than absolute effects both serve the sponsor’s interest in disseminating results quickly and favorably while holding down research costs. Sponsorship bias in the presentation of trial results has been demonstrated before.
🔹 Trials stopped early for benefit can overestimate treatment effects, particularly when the expected number of events is low. Reasons for stopping are often not reported, and effect estimates are rarely adjusted for the interim analyses and the reduced sample size.
🔹 The ethical case for stopping early rests on getting a beneficial treatment to the control arm and to the public sooner. That rests on results being adopted quickly in practice, which known delays in dissemination make unlikely.
🔹 Stopping at 1.9 years of a planned four means the long term consequences of the higher glycated hemoglobin and diabetes incidence in the rosuvastatin group will not be known, and neither will the effects of lowering LDL cholesterol below 60 mg/dL, a level never reached in previous randomized trials. Cholesterol is a physiological molecule required for many vital processes.
Statins as Anti-Inflammatory Agents?
JUPITER was built on the hypothesis that statin treatment reduces vascular events in people with elevated high-sensitivity CRP but without high cholesterol. That is why enrollment required LDL below 130 mg/dL alongside CRP of 2.0 mg/L or more, and it is why the trial was read as a case for treating inflammation rather than lipids.
- The trial selected patients by inflammation, not by lipid level, and comments following its release went as far as the idea that statins should be given to everyone regardless of their lipid levels.
- The relative risk reduction in JUPITER was larger than in earlier statin trials, 54% versus 23% for major coronary events, even though JUPITER enrolled lower risk people. That gap raises the possibility that the treatment effect in JUPITER was overestimated.
- The absolute benefit ran the other way: earlier trials had a larger absolute risk reduction and a lower number needed to treat than JUPITER, which is what you would expect when the enrolled population starts at lower risk.
Related Studies and Research
Statins and Primary Prevention: The Debate – Discusses the controversy surrounding statins for primary prevention in low-risk populations.
Statins for Healthy Men: A Review – Investigates the risks and benefits of statin therapy in men without established heart disease.
Statins and Neuromuscular Side Effects – Analyzes the potential neuromuscular complications associated with statin use.
HOPE-3 Trial: Rosuvastatin in Primary Prevention – Reviews the HOPE-3 trial results on rosuvastatin for individuals at intermediate cardiovascular risk.
Frequently Asked Questions
1. Should I get my CRP checked?
JUPITER used CRP to decide who was eligible for treatment, and about 80% of the people who came to the screening visit did not qualify. If that approach were adopted broadly, a large number of people would be tested and never treated, which is part of why the screening cost per event prevented is so high. This is a decision to make with your own physician in the context of your other risk factors.
2. Do statins work only by lowering cholesterol?
JUPITER was designed to test that question, by enrolling people whose LDL cholesterol was already below 130 mg/dL and selecting them instead on elevated CRP.
3. What else lowers cardiovascular risk?
Behavioral risk factors are the more efficient target. Less than 3% of health care spending currently goes toward smoking, poor diet, and physical inactivity, yet those factors are held responsible for almost 40% of deaths in the United States. Prevention aimed at them is supported by good evidence of effectiveness and yields a high return on investment.
4. Does the increased diabetes risk mean I should not take a statin?
Glycated hemoglobin and diabetes incidence were elevated in the rosuvastatin group in JUPITER. Because the trial stopped after a median of 1.9 years, what happens to that signal over a longer course of treatment was never observed. If you already carry diabetes risk, that is worth discussing with your physician, along with a plan for monitoring blood glucose.
Conclusion
The benefits achieved in JUPITER are not large enough to justify expanding the clinical indications for statins to healthy people. The relative figures are impressive, 44% for the composite endpoint and 54% for heart attack, but the absolute reduction was 0.59% per year and 0.20% per year respectively, and 500 healthy people would need a year of treatment to prevent one heart attack. Add the cost of the drug, the screening, and the monitoring, and the case gets thinner. Since major treatment trials are sponsored almost exclusively by the pharmaceutical industry, another adequately sized trial with proper follow-up in this population is unlikely to be run, so the long term picture will stay unresolved. My reading of the evidence is the same as the authors’: at the current state of knowledge, behavioral prevention is the better investment for cardiovascular disease in predominantly healthy people.
