· ST-PRI-03 · v2
The claim says statins do not lower the risk of dying from any cause in people who have not yet had heart disease. Every study in this corpus that measured deaths in such people points the other way: taking a statin went with fewer deaths. The one randomized trial, WOSCOPS [6], found a lower death rate in men given pravastatin, and several large real-world databases [2][3][13][16] found the same pattern in older or higher-risk patients. On this evidence the panel judged the claim contradicted. But most of the supporting studies simply compared people who happened to take statins with people who did not, a design that is easily skewed by the fact that statin-takers tend to be healthier and more adherent, and the pooled randomized-trial analyses that speak most directly to this exact question were filtered out of the corpus before scoring.
Every included study reporting a statin-versus-nonuse all-cause mortality estimate in a primary-prevention population shows a reduction. The single randomized contrast, WOSCOPS [6], reports all-cause mortality hazard ratio 0.87 (95% CI 0.80-0.94, P=0.0007) over 20-year linkage. Large observational cohorts concur: type 2 diabetes aged 75-84 HR 0.65 [2], chronic kidney disease HR 0.82 [3], age >=80 HR 0.69 [4], primary-care hypertension HR 0.83 [13], and age >75 HR 0.56 [16]. No included primary study reports a null or adverse statin mortality effect. The contested point: The division is over how much weight the uniform reduction can bear: five of the six supporting studies are observational [2][3][4][13][16], where healthy-adherer bias and confounding by indication can manufacture apparent survival gains, and mortality HRs as low as 0.56 [16] are large for a lipid agent. The only randomized evidence [6] is an aging single-sex trial whose 20-year estimate blends the randomized period with unrecorded later statin use.
All 21 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from Statins, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
Five of six supporting studies are observational cohorts ([2][3][4][13][16]); statin-user cohorts are the canonical case of healthy-adherer bias and confounding by indication, and mortality HRs as low as 0.56 are implausibly large for a lipid agent, consistent with residual bias rather than causal effect.
The only randomized contrast [6] is a single 1990s trial in middle-aged high-LDL men; its mortality HR derives from a 20-year EHR linkage in which post-trial statin use was unrecorded for the final 10 years, blending the randomized effect with legacy and open-label effects and limiting clean inference on the randomized mortality endpoint.
Where it remains uncertain. Most support is observational and bias-prone [2][3][4][13][16] and the pooled randomized mortality analyses are excluded, so the corpus cannot settle the randomized controversy the claim invokes.
Aleatory
Estimates like all-cause death HR 0.56 in SCOPE-75 [16] rest on few events (1.19 vs 0.65 deaths per 100 person-years), leaving the point estimate sensitive to chance.
Epistemic
Observational studies [2][3][4][13][16] acknowledge residual confounding from lifestyle and adherence; [2] names reliance on diagnosis coding and unmeasured lifestyle factors, so causal direction is not established.
Inferential
Included cohorts are >=75, >=80, CKD, T2D or hypertensive [2][3][4][13][14][16]; extrapolating to primary prevention as a whole is an inference the corpus does not directly support.
The load-bearing papers from a corpus of 21, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Large target-trial emulation shows statin primary prevention lowers all-cause mortality in elderly T2D.
“a significant reduction in the incidence of CVDs (HR 0.69 (95% CI [0.65, 0.75]; p < 0.001)) and all-cause mortality (0.65 [0.60, 0.70], p < 0.001) was observed”
Panel auditInternal validity: medium · scope of inference: Older adults (≥75 y) with type 2 diabetes and elevated LDL-C, no prior CVD, in Hong Kong.Key limitations: observational design · confounding by indication / healthy-adherer bias · diagnosis-coding misclassification
1 of 21 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
The included corpus uniformly shows a statin-associated mortality reduction in primary prevention, refuting the null claim, but reliance on observational data and one old RCT prevents near-zero confidence.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
No pooled randomized mortality estimate in-corpus. The claim invokes the randomized-trial controversy, yet the only randomized evidence here is one trial [6]; pooled RCT mortality is excluded. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: Include or conduct a contemporary randomized-trial meta-analysis of statin-versus-placebo all-cause mortality restricted to primary prevention.
Open question: Do statins reduce all-cause mortality in primary prevention when estimated from pooled randomized trials rather than observational cohorts? Why it is still open: The corpus's randomized evidence is a single trial [6]; the pooled RCT analyses were study-type-excluded and cannot be weighed.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 26 Oct 2026 · protocol v0.19