Name: Statins: They Genuinely Cut Heart-Attack Risk, the Fights Are About Size and Side Effects
Creator: Realised
Published: 2026-07-01

Summary

Statins genuinely reduce heart attacks, strokes and cardiovascular death in people with established disease or high risk, and the "statin denialism" that waves this away is wrong and dangerous for exactly those people; but the benefit is routinely sold in inflated RELATIVE terms while the ABSOLUTE payoff shrinks to something small in low-risk primary prevention, most muscle-symptom complaints turn out to be nocebo rather than the drug when tested blind, the rare real harms (myopathy, a modest rise in new-onset diabetes) are genuine and worth naming honestly, and "statins for everyone" over-med

Why Moderate

Moderate Evidence for the entry as a whole. This is a deliberate packaging call: the underlying efficacy evidence is genuinely Strong (the CTT IPD meta-analyses and the blinded SAMSON/CTT-2022 nocebo data are as good as clinical evidence gets), but the entry's job is to deliver that efficacy TOGETHER WITH the interpretive absolute-versus-relative framing, the low-risk-benefit deflation, and the both-ways side-effect correction, which involve genuine clinical and values judgement rather than a single settled axiom. That interpretive layer is where reasonable clinicians still differ, and it sits at Moderate.

NOT Foundational because the entry carries live judgement and a real both-ways tension (efficacy is real AND the benefit is oversold in low-risk people), not an undisputed principle.

NOT Strong for the packaged whole, even though several component findings are Strong, because the headline the entry actually delivers, "risk-stratified, shared decision, not a slogan," is an interpretive synthesis that includes the contested low-risk primary-prevention zone.

The per-claim split (read this, not just the headline):
• Efficacy per 1.0 mmol/L LDL reduction; muscle symptoms mostly nocebo; population muscle excess small; diabetes signal: Strong (large IPD meta-analyses and blinded trials).
• Low-risk primary-prevention absolute numbers: Strong-to-Moderate (independent USPSTF systematic review).
• Serious muscle harm rarity figures: Moderate (pooled/consensus safety estimate).
• The synthesis "risk-stratified shared decision": Moderate (interpretive, values-sensitive in the low-risk zone).

Practical takeaway

The framing to hold: a statin decision is a risk-stratified, shared decision, not a slogan. The higher your baseline cardiovascular risk, the larger the absolute benefit and the more trivial the harms look by comparison; the lower your risk, the smaller the absolute benefit and the more it becomes a values call. Realised does not start or stop anyone's medication; this is a map for the conversation you have with your clinician.

Know which side of the line you are on (this is the whole ballgame).
• Secondary prevention (you already have coronary or cerebrovascular disease, or have had a heart attack or stroke): the absolute benefit is large and dwarfs the rare harms. This is the setting where "statin denialism" does real damage, and where stopping on the basis of a headline is genuinely risky.
• High-risk primary prevention (elevated calculated 10-year risk, diabetes, strong risk-factor burden): the benefit is still meaningful in absolute terms and generally favours treatment.
• Low-risk primary prevention: the absolute benefit is small (all-cause-mortality NNT around 286 in the trials) and this is a legitimate, values-sensitive choice, not a mandate. Ask for your absolute numbers, not just the relative one.

Ask for the absolute numbers, not the relative headline. "Cuts your risk by a third" is a relative figure. Ask your clinician what it means for you in absolute terms: my 10-year risk with and without the drug, and how many people like me need to take it for one to benefit. (The general skill is owned by relative_vs_absolute_risk.)

Take muscle symptoms seriously AND test them rather than assuming. Real symptoms happen and deserve attention, but most turn out to be nocebo when tested. The evidence-based response is not "stop forever" but structured testing: a supervised rechallenge, a different statin or dose, or a blinded n-of-1 style trial with your clinician. Half of the prior-intolerant patients in SAMSON restarted a statin after seeing their own blinded data.

Do not stop a statin on your own, especially in secondary prevention. If you want to reconsider, do it with your clinician, not on the strength of a book or a video. Abrupt discontinuation in established disease raises heart-attack and death risk.

Keep the CoQ10 question in its lane. The statin-CoQ10-myopathy hypothesis is a separate topic with its own evidence; it is owned by coq10_evidence_and_use_cases, not resolved here.

Evidence detail

Why This Entry Exists

The "statin wars" are one of the few medical debates where both loud camps are partly right and each is dangerously wrong at the edges. On one side, a well-funded ecosystem of contrarian books, supplements and "root-cause" clinics tells people statins don't work, are pushed for profit, and wreck your muscles; that message drove measurable real-world discontinuation and, in people with established coronary disease, measurable harm. On the other side, guideline enthusiasm and "lower is always better" language quietly widens eligibility until a low-risk 45-year-old is being counselled onto a lifelong drug for an absolute benefit that, stated honestly, is small. This entry exists to hold the honest middle: the efficacy is real and large where risk is high, and the debate that actually deserves oxygen is about absolute benefit in low-risk people and about what the side-effect signal really is once you test it blind.

The load-bearing distinction is between a relative risk reduction (a headline "cuts risk by a third") and the absolute benefit that reduction buys a specific person. The relative reduction is roughly constant across risk levels; the absolute reduction scales with how much risk you had to start with. That single fact reconciles the two camps: in secondary prevention it makes the benefit large and the harms trivial by comparison, and in low-risk primary prevention it makes the benefit genuinely modest and a values-sensitive choice. This entry owns that framing for statins specifically. It defers the "is LDL even causal" question to cholesterol_misinformation_correction, general cardiovascular risk management to cardiovascular_health_management, and the statin-CoQ10-myopathy question to coq10_evidence_and_use_cases. Realised does not diagnose, prescribe, or tell anyone to start or stop a medication; it corrects the evidence in both directions and hands the decision back to the person and their clinician.

What bad advice this protects against, in all directions:
• "Statins don't really work / the mortality benefit is a myth" → false and dangerous for high-risk people; large RCT meta-analyses show a real reduction in heart attacks, strokes, revascularisation and cardiovascular death.
• "Statins cut your risk by a third, so everyone should take them" → that is a RELATIVE figure; the ABSOLUTE benefit scales with baseline risk and is small in low-risk primary prevention (all-cause-mortality NNT ~286 over the trials).
• "Statins destroy your muscles / ~1 in 5 people can't tolerate them" → the ~18 to 20% side-effect claim came from unblinded observational data; blinded trials show most muscle symptoms are nocebo, and over 90% of muscle-symptom reports were not caused by the statin.
• "Statins are basically harmless, take them just in case" → the opposite over-correction; myopathy (rare), rhabdomyolysis (very rare) and a modest new-onset-diabetes rise are real and belong in the conversation.
• "If you have side effects, statins just aren't for you, stop" → symptoms deserve to be taken seriously AND tested (rechallenge or a blinded n-of-1), not assumed to be the drug; and stopping in secondary prevention meaningfully raises heart-attack and death risk.
• "Big Pharma pushes statins for profit" → most statins are now generic and cheap; the stronger modern commercial incentive is the anti-statin ecosystem that monetises distrust.
• "A statin blocks a pathway with seven outputs, so it must be starving you of CoQ10, vitamin D and your hormones" → true biochemistry, unsupported conclusion. The pathway genuinely is branched, but "the pathway makes it" is not "you are depleted of it," and depletion is not harm. Statins do not reliably lower vitamin D (the pooled randomised evidence is null, and statin users often test slightly higher); the testosterone effect is a few percent of a normal range rather than clinical hypogonadism; giving CoQ10 back does not relieve statin muscle symptoms when it is tested; and the one branch that does reach demonstrated harm, protein prenylation, produces the diabetes signal this entry already names.

This entry owns the statin efficacy-versus-absolute-benefit-versus-nocebo triangle and the both-ways correction. It states its boundaries and routes the LDL-causality, general-CV-management and CoQ10 questions to their owning entries rather than re-arguing them.

Evidence

Organised so the efficacy, the absolute-benefit deflation and the side-effect reality sit side by side, with the tier signal inline. The underlying efficacy evidence is genuinely the strongest grade there is; the entry sits at Moderate overall because it packages that efficacy with the interpretive absolute-versus-relative and nocebo layer.

Efficacy is real and proportional to LDL reduction (Strong Evidence).

1. Per 1.0 mmol/L LDL reduction, statins cut major vascular events by about 21%, consistently and with no threshold. The Cholesterol Treatment Trialists' (CTT) Collaboration's individual-participant meta-analysis of 26 randomised trials (about 170,000 participants) found a relative risk of 0.78 (99% CI 0.76 to 0.80) for major vascular events per 1.0 mmol/L LDL lowering, proportional to the size of the reduction and consistent across patient types, with no lower threshold in the range studied. This is the core efficacy claim, and it is among the best-evidenced findings in cardiology. (CTT Collaboration. "Efficacy and safety of more intensive lowering of LDL cholesterol: meta-analysis of 170,000 participants in 26 randomised trials." Lancet 2010;376:1670–1681. Strong Evidence — individual-participant-data meta-analysis of 26 RCTs. Coordinated by Oxford CTSU; component trials had varied industry sponsorship but the collaborative IPD analyses are conducted independently of sponsors and are the field's gold-standard synthesis.)

The relative benefit holds even in low-risk people, which is the setup for the honest debate (Strong Evidence).

2. The relative reduction is similar in low-risk people; because absolute benefit scales with baseline risk, the low-risk absolute payoff is small. The CTT low-risk meta-analysis of 27 randomised trials found the proportional (relative) benefit largely preserved even in people at low vascular risk. That is precisely why the debate is about absolute benefit: a similar relative reduction applied to a small baseline risk yields a small absolute gain. Tellingly, this paper is cited by BOTH camps, which is a marker of its neutrality. (CTT Collaboration. "Effects of lowering LDL cholesterol with statin therapy in people at low risk of vascular disease: meta-analysis of 27 randomised trials." Lancet 2012;380:581–590. Strong Evidence — IPD meta-analysis of 27 RCTs. Same CTT independence caveat.)

The absolute numbers in low-risk primary prevention are modest (independent government review).

3. For low-risk primary prevention the absolute benefit is small and all-cause mortality barely moves. The US Preventive Services Task Force 2022 review put the composite cardiovascular NNT at about 78, non-fatal MI NNT at about 118, stroke NNT at about 256, and the all-cause-mortality absolute risk difference at only −0.35% (NNT about 286). The USPSTF recommends statins only at 10% or greater 10-year ASCVD risk with at least one risk factor, narrowing eligibility versus the 2018 AHA/ACC guideline. This is the "over-medicalising low-risk people" concern quantified, by a body with no drug interest. (USPSTF. "Statin Use for the Primary Prevention of Cardiovascular Disease in Adults." JAMA 2022;328:746–753, with the accompanying AHRQ/Annals evidence review. Strong-to-Moderate Evidence — independent, publicly funded government systematic review used to set a guideline; explicitly independent of industry.)

Most muscle complaints are nocebo, not the drug (blinded RCT-grade mechanism).

4. In blinded n-of-1 crossover, statin and placebo months produced essentially the same symptom burden. The SAMSON trial ran blinded n-of-1 crossovers in 60 patients who had previously stopped statins for side effects, cycling statin, placebo and no-tablet months. Symptom intensity was 16.3 on statin months versus 15.4 on placebo months, not a significant difference, and about 90% of the symptom burden was also present on placebo (the "nocebo" fraction). Half the participants restarted statins afterward. This explains the ORIGIN of the symptoms; because it is small and enrolled prior-intolerant patients, it is not a population muscle-harm rate (see finding 5). (Wood FA, Howard JP, Finegold JA, et al. "N-of-1 Trial of a Statin, Placebo, or No Treatment to Assess Side Effects (SAMSON)." N Engl J Med 2020;383:2182–2184. Strong Evidence for the nocebo mechanism — randomised blinded n-of-1, though small (n=60) and selected. British Heart Foundation / NIHR funded; independent of statin manufacturers.)

At population scale the real drug-caused muscle excess is small and mostly first-year (Strong Evidence).

5. Across large blinded RCTs, statins caused only a small real excess of muscle symptoms, and over 90% of muscle reports were not the drug. The CTT muscle-symptom IPD meta-analysis (about 155,000 participants in large double-blind trials) found roughly a 3% relative increase in muscle symptoms, an absolute excess of about 11 (95% CI 6 to 16) events per 1000 person-years, confined largely to the first year, and concluded that more than 90% of muscle-symptom reports by statin-allocated patients were NOT caused by the statin. This directly rebuts the unblinded observational papers that claimed roughly 18 to 20% side-effect rates, which were a central artefact of the statin wars. (CTT Collaboration [Reith C, et al.]. "Effect of statin therapy on muscle symptoms: individual participant data meta-analysis of large-scale, randomised, double-blind trials." Lancet 2022;400:832–845. Strong Evidence — IPD meta-analysis of large blinded RCTs. Independent CTT analysis.)

Serious muscle harm is genuinely rare (pooled safety estimate).

6. Clinically significant myopathy and rhabdomyolysis are rare, and the denominators matter. Pooled statin safety analyses put clinically significant myopathy (creatine kinase above 10x the upper limit of normal with symptoms) at roughly 1 per 10,000 patient-years, and rhabdomyolysis at roughly 2 to 3 per 100,000 person-years. These figures are consistent across independent pharmacovigilance sources; conflating the per-10,000 myopathy rate with the per-100,000 rhabdomyolysis rate overstates harm, so keep the denominators exact. (Summarised from CTT safety analyses and statin safety reviews, e.g. Collins R, et al. "Interpretation of the evidence for the efficacy and safety of statin therapy." Lancet 2016;388:2532–2561. Moderate Evidence — pooled/consensus safety estimate. Collins is a lead CTT investigator and a vocal statin defender, so treat the framing with the same scrutiny as any interested party; the rarity figures themselves replicate across independent sources.)

The new-onset-diabetes signal is real, modest, and accepted even by pro-statin authors (Strong Evidence).

7. Statins modestly raise new-onset diabetes, concentrated in people already near the glycaemic threshold. The original pooled meta-analysis of 13 statin trials found about a 9% relative increase in incident diabetes (roughly one extra case per 255 patients over about 4 years). Later IPD analysis refined this to about 10% for low/moderate-intensity therapy and about 36% for high-intensity therapy, with risk concentrating in people whose glucose was already close to the diabetic threshold; for appropriate candidates the cardiovascular benefit outweighs it. That the signal is accepted by pro-statin authors strengthens its credibility, it is not an anti-statin talking point being smuggled in. (Sattar N, et al. "Statins and risk of incident diabetes: a collaborative meta-analysis of randomised statin trials." Lancet 2010;375:735–742; and CTT Collaboration. "Effects of statin therapy on new-onset diabetes and worsening glycaemia: IPD meta-analysis." Lancet Diabetes Endocrinol 2024;12:306–319. Strong Evidence — two large IPD/pooled meta-analyses, independently conducted.)

Mechanism

This entry owns the interpretive layer, not the lipid biochemistry (the LDL-causality mechanism is deferred to cholesterol_misinformation_correction). What follows is only enough mechanism to make the efficacy, the absolute-benefit deflation and the nocebo finding intelligible.

Why the benefit is proportional to LDL reduction with no obvious floor. Statins inhibit HMG-CoA reductase, the rate-limiting enzyme of hepatic cholesterol synthesis, which upregulates LDL receptors and clears circulating LDL. Because the event reduction in the trials tracks the size of the LDL drop in a roughly linear, threshold-free way across the studied range, the drug behaves like a lever on a causal pathway rather than a treatment with a single target level. The strength of that dose-response is a large part of why the efficacy claim is so robust; the deeper "is LDL the cause" argument belongs to cholesterol_misinformation_correction.

Why a statin is a pathway inhibitor, not a cholesterol remover. HMG-CoA reductase sits at the committed step of the branched mevalonate (isoprenoid) pathway, and cholesterol is only one of its outputs. The same pathway supplies coenzyme Q10 (whose tail is built from farnesyl and isopentenyl units), dolichols (the lipid anchor used to attach sugars to proteins), heme A (installed on Complex IV by an enzyme formally named heme A:farnesyltransferase), the farnesyl and geranylgeranyl groups that tether small GTPases to membranes, and squalene en route to cholesterol itself. This is accurate biochemistry, and it is not a fringe claim: cardiology invokes the same branch structure to explain the "pleiotropic" benefits that exceed what LDL lowering alone predicts. "Cholesterol drug" is a poor description of what the molecule does, and saying so is a legitimate correction rather than an anti-statin talking point.

What the branch structure does and does not establish. Three layers have to be distinguished, and most popular versions of this argument collapse them: (1) the pathway makes a compound, (2) statins measurably lower it in living humans, (3) that loss causes symptoms or disease. Most non-cholesterol branches stop at layer one. Human heme A depletion has never been measured at all, so asserting it overreaches and dismissing it also overreaches. Dolichol depletion has been shown only in cell culture. The selenoprotein hypothesis (Moosmann & Behl, Lancet 2004) is coherent but was undercut when people carrying a severely impaired version of the relevant enzyme showed no fall in selenoprotein synthesis. Plasma coenzyme Q10 does fall reliably, but part of that is carriage rather than starvation: circulating CoQ10 travels on LDL particles, so lowering LDL mechanically lowers the measured carrier, and the CoQ10-to-LDL ratio is often unchanged. Muscle CoQ10 is inconsistent, with one randomised biopsy trial finding roughly a third lower on high-dose simvastatin and no reduction at all in its own atorvastatin arm. The decisive test is negative: giving CoQ10 back to people with confirmed statin myopathy did not relieve it. The branch that genuinely reaches layer three is protein prenylation, since insulin granule release depends on geranylgeranylated Rab GTPases, and genetic evidence at the HMGCR locus finds the same diabetes signal in people with naturally lower enzyme activity, which means the effect is on-target to pathway inhibition rather than an incidental quirk of the drugs. The CoQ10 question in full belongs to coq10_evidence_and_use_cases.

Why the mechanism argument is symmetric and cannot settle the question alone. The same reduced prenylation that plausibly explains the diabetes signal is also one of the standard explanations for statin benefit beyond LDL lowering, through anti-inflammatory effects on Rho and Rac signalling. A pathway argument that only ever points toward harm is being steered rather than followed. Physiological plausibility is the cheapest commodity in medicine and has failed in exactly this shape before, having predicted that antioxidant vitamins would prevent cancer and that hormone replacement would prevent coronary disease. Outcome data decides; mechanism only proposes.

Why relative and absolute benefit diverge. A roughly constant relative risk reduction is a multiplier on whatever baseline risk a person carries. Multiply a large baseline risk (established coronary disease) and you get a large absolute gain; multiply a small baseline risk (a low-risk 45-year-old) and you get a small one, even though the percentage on the label is identical. Nothing about the drug changes between those two people; the arithmetic of absolute risk does. This is the single mechanism that reconciles the two camps, and it is why the honest recommendation is risk-stratified. The general treatment of this arithmetic lives in relative_vs_absolute_risk.

Why muscle complaints are so often nocebo. Muscle aches are common in the statin-eligible age band regardless of any drug, and the strong cultural expectation that statins hurt your muscles supplies a powerful nocebo prior: expecting a symptom makes it more likely to be noticed and attributed. Blinding removes the expectation asymmetry, and once you blind, most of the symptom burden appears on placebo too. That is why blinded n-of-1 and blinded RCT data (SAMSON, CTT 2022) are load-bearing and unblinded registries are not; the mechanism is attribution, not myotoxicity, for the large majority of complaints. The genuinely drug-caused minority (finding 5) and the rare serious myopathy (finding 6) are a separate, real, smaller phenomenon.

Why the diabetes signal is real but bounded. Statins appear to nudge glucose handling slightly, enough to tip people already sitting just under the diabetic threshold over it, which is why the absolute signal concentrates in that subgroup and scales with statin intensity. For someone whose cardiovascular risk is high, the event reduction outweighs the modest diabetes incidence; the calculus tightens as cardiovascular benefit shrinks, which is another reason low-risk primary prevention is the contested zone.

Risks And Contraindications

• SAFETY-CRITICAL: this entry must never read as encouragement to START or STOP a statin. Stopping a statin in secondary prevention (established coronary or cerebrovascular disease) meaningfully raises heart-attack and death risk. The nocebo and absolute-benefit framing must NOT be weaponised into "statins don't work." Any surfacing of the side-effect or low-risk-benefit material must carry the "this is a clinician conversation, do not self-discontinue" guardrail.
• Equally, do not present statins as risk-free or universally indicated. The rare real harms are genuine and belong in the conversation: clinically significant myopathy roughly 1 per 10,000 patient-years, rhabdomyolysis roughly 2 to 3 per 100,000 person-years, and a modest new-onset-diabetes rise (about 10% low/moderate-intensity, about 36% high-intensity, concentrated near the glycaemic threshold).
• Keep the harm denominators exact. Myopathy is per 10,000; rhabdomyolysis is per 100,000. Conflating the two overstates harm by roughly an order of magnitude. Do NOT inflate the diabetes numbers either, and do not present the ~18 to 20% "intolerance" figure as real, it is the unblinded artefact the blinded data corrected.
• SAMSON is small and selected. It enrolled 60 patients with prior intolerance; cite it for the nocebo MECHANISM, not as a population muscle-harm rate. The population rate comes from the CTT 2022 blinded IPD analysis (about 11 excess events per 1000 person-years, mostly first year).
• Muscle symptoms warrant clinical attention, not dismissal. New, severe or progressive muscle pain, especially with dark urine or marked weakness, can signal genuine myopathy or (rarely) rhabdomyolysis and needs prompt medical assessment, not reassurance that "it's just nocebo." The nocebo framing describes the average complaint, not every complaint.
• Dosing, drug interactions and monitoring are clinician territory. Statin choice, dose, interaction checking, and CoQ10 questions (see coq10_evidence_and_use_cases) are not decided here. All medication decisions belong to the user and their clinician; Realised does not diagnose or prescribe.

Controversy

Nature: a genuinely effective drug class whose core efficacy is settled at the highest evidence grade, wrapped in a decade-long public fight ("the statin wars") in which error sits at BOTH poles: a "statin denialism" that denies a real mortality benefit for high-risk people, and a "statins for everyone" universalism that over-medicalises low-risk people and treats a small absolute benefit as a mandate.

Position A — "Statin denialism is wrong and dangerous." The efficacy-defending take.
• Best evidence: the CTT IPD meta-analyses (about 170,000 patients, 26+ trials) show a 21% reduction in major vascular events per 1.0 mmol/L LDL drop, proportional and threshold-free; secondary and high-risk primary prevention are among the best-evidenced interventions in cardiology; and the blinded SAMSON and CTT-2022 data show most muscle complaints are nocebo, not the drug.
• Where it goes wrong if overstated: it slides into "statins for everyone," downplays the rare real harms and the diabetes signal, and treats a low-risk person's small absolute benefit as if it were the high-risk person's large one.

Position B — "The benefit is oversold and the harms understated." The corrective take.
• Best evidence: the benefit is routinely quoted in relative terms while the absolute payoff in low-risk primary prevention is small (USPSTF: composite-CV NNT ~78, all-cause-mortality ARD only −0.35%, NNT ~286); the rare harms (myopathy, rhabdomyolysis) and the modest diabetes rise are real; and eligibility has crept wider than the absolute numbers justify for low-risk people.
• Where it goes wrong if overstated: it slides into "statins don't work," recycles the discredited ~18 to 20% intolerance figure from unblinded data, and encourages high-risk people to stop, which causes real harm.

The funding/bias dimension — cui bono, both ways. Toward over-prescribing: guideline panels have historically carried industry ties, and "lower is always better" language quietly widens eligibility; but most statins are now generic and cheap, so the classic "Big Pharma pushes statins for profit" motive is weak today. Toward denialism: the anti-statin side is now the more commercially active bias, feeding a lucrative ecosystem of books, supplements, contrarian influencers and "root-cause" clinics that monetise distrust, and the discredited ~18 to 20% side-effect claims (Abramson, Malhotra, in the BMJ, later contested and corrected) drove real-world discontinuation and measurable harm. The most defensible anchors, the CTT/Oxford CTSU analyses, are led by strong statin advocates (Collins, Baigent), so their safety-downplaying framing deserves the same scrutiny as industry's, which is exactly why the independent USPSTF review and the BLINDED SAMSON/CTT-2022 nocebo data, not observational registries, are load-bearing here.

Realised Position: The honest answer is risk-stratified, not a slogan either way. In secondary prevention and high-risk primary prevention the event and mortality benefit is real, large in absolute terms, and dwarfs the rare harms, so denying it is dangerous. In low-risk primary prevention the absolute benefit is genuinely small and is a shared, values-sensitive decision, not a mandate. Muscle complaints deserve to be taken seriously AND tested (rechallenge or blinded n-of-1) rather than assumed to be the drug or dismissed as all in the head. Realised does not diagnose or prescribe; this entry corrects both denialism and universalism and hands dosing back to clinicians.

Cross-Pillar Connections

This is a genuinely cross-pillar topic: a cardiovascular drug whose honest interpretation turns on general skills (absolute versus relative risk, surrogate endpoints, publication bias) and whose harms touch metabolic and musculoskeletal territory.
• Conditions (cholesterol_misinformation_correction): owns the "is LDL causal" and cholesterol-panel-interpretation question; this entry assumes the LDL-lowering-reduces-events link and does not re-argue causality.
• Conditions (cardiovascular_health_management): owns general cardiovascular risk reduction (blood pressure, lifestyle, overall risk stratification); this entry holds only the statin-specific efficacy/absolute-benefit/nocebo material.
• Evidence method (relative_vs_absolute_risk): owns the general arithmetic that makes the low-risk primary-prevention debate legible; this entry applies it to statins.
• Evidence method (surrogate_endpoints_vs_outcomes): relevant to why LDL is a surrogate and hard outcomes (MI, CV death, mortality) are what the statin trials actually measured.
• Supplements (coq10_evidence_and_use_cases): owns the statin-CoQ10-myopathy question, which this entry explicitly defers rather than resolving.
• Evidence method (publication_bias_and_evidence_distortion): owns the general mechanics of the unblinded-versus-blinded distortion that drove the statin wars.

What would change our mind

Falsifiability: explicit upgrade/downgrade criteria from source

• We'd shift the side-effect framing if a large blinded RCT or IPD analysis showed a real, drug-caused (not nocebo) muscle-harm rate materially higher than the ~11 per 1000 person-years first-year excess CTT 2022 found.
• We'd overturn the core claim if a blinded RCT or IPD analysis showed statins fail to reduce hard outcomes (MI, cardiovascular death, all-cause mortality) in secondary prevention. None exists; the RCT base runs overwhelmingly the other way, so this is a very high bar.
• We'd soften the "over-medicalising low-risk people" side if a well-powered trial showed a genuinely larger absolute benefit (a real mortality signal) in a truly low-risk cohort than USPSTF/CTT estimates imply.
• We'd tighten the risk-stratification if the diabetes signal proved to translate into net worse outcomes in a definable subgroup.
• What would NOT move us: the relative-versus-absolute arithmetic and the nocebo finding are robust across independent and industry-adjacent analyses alike, so incremental studies are unlikely to shift them. Across all of it, BLINDED and INDEPENDENT evidence is the decisive variable, and the load-bearing claims already rest on it.

Industry bias note

Structural incentives the evidence base may reflect

Cui bono runs in BOTH directions here, and the modern financial pull is not where people assume.
• The classic "pharma pushes statins for profit" read is weak now. Statins were a blockbuster class, but most are generic and cheap today, which blunts the profit motive to over-prescribe. The residual pro-statin pull is subtler: guideline panels have historically had industry ties, and "lower is always better" language can quietly widen eligibility beyond what the absolute numbers justify for low-risk people.
• The anti-statin bias is now the more commercially active one. The statin wars feed a lucrative ecosystem of books, supplements, contrarian influencers and "root-cause" clinics that monetise distrust. The discredited ~18 to 20% side-effect claims (Abramson, Malhotra, in the BMJ, later contested and corrected) drove real-world discontinuation and measurable harm, so treat unblinded "intolerance" figures with suspicion.
• The strongest anchors have their own tilt, which is why blinding and independence matter. The CTT/Oxford CTSU analyses are the most defensible synthesis but are led by investigators (Collins, Baigent) who are strong statin advocates, so their safety-downplaying framing deserves the same scrutiny as industry's. That is exactly why the independent, publicly funded USPSTF review and the BLINDED SAMSON/CTT-2022 nocebo data (not observational registries) are load-bearing here, they are the evidence least contaminated by either commercial pull.
• The diabetes signal is credibility-positive because it cuts against interest. It is accepted by pro-statin authors even though it is unflattering to the drug, which is a marker that it is not an anti-statin talking point being smuggled in. General treatment of funding and publication distortion is owned by publication_bias_and_evidence_distortion.

Statins: They Genuinely Cut Heart-Attack Risk, the Fights Are About Size and Side Effects