← Education Library
Nutrition

Your Body Makes It on Purpose. Essential, but Not Harmless.

Cholesterol builds your cells, helps make your hormones and is essential to your brain. Sustained exposure to the particles carrying it also drives arterial disease. What your blood test measures, why the thresholds keep moving, and how the same trials produce a 29% figure and a 1.3% one.

17 minutesPublished October 25, 20260 likes0 comments

What you'll learn

  • What cholesterol is, what the numbers on your test mean, and how it damages an artery
  • Why your body manufactures the substance most people are trying to lower
  • How the same trials produce a 29% figure and a 1.3% figure, both correctly
  • Which statin side effects survived randomised testing, and which did not
  • What changed in the 2026 guidelines, and what to ask about

How SmartBar applies the deficiency framework

SmartBar's educational framework begins with two organizing terms, and applies them wherever the body has inputs and burdens. Deficiency means something the body needs and isn't getting. Toxicity means a burden the body must process and clear, not a poison.

Cholesterol fits neither term cleanly, which is why it needs its own page. Cholesterol itself is essential structural material your body manufactures deliberately. But sustained exposure of the artery wall to cholesterol carried in LDL and other apoB-containing particles is causally involved in atherosclerotic cardiovascular disease. Those are different facts about the same molecule.

This page is educational and is not medical advice. Nothing here is a reason to start, stop or change any medication. Decisions about cholesterol-lowering treatment depend on your overall cardiovascular risk, your history, your other conditions and your own preferences, and they belong with a prescriber who knows all four.

Start here: what cholesterol actually is

Cholesterol is a waxy, fat-like substance found in every cell you have. It is not an invader. Your body uses it to build cell membranes, to make vitamin D, to produce bile for digesting fat, and to manufacture hormones including testosterone, oestrogen and cortisol.

You cannot live without it. That is not a contrarian talking point, it is basic biology.

So why does it have such a bad reputation?

Because of where some of it ends up. Cholesterol does not dissolve in blood, any more than oil dissolves in water. To travel anywhere, it has to be packaged inside particles called lipoproteins, which act like delivery vehicles moving through your bloodstream.

Those vehicles come in different kinds, and that is where the "good" and "bad" language came from.

Here is the thing that label hides: the cholesterol inside an LDL particle and the cholesterol inside an HDL particle are the same molecule. There is no good cholesterol and bad cholesterol. There are different carriers doing different jobs, and the number of certain carriers matters more than most people are told.

What your doctor is measuring. A standard lipid panel reports total cholesterol, LDL cholesterol, HDL cholesterol and triglycerides. Newer guidance increasingly uses two additional markers that we come back to later: ApoB, which estimates how many atherogenic particles you have rather than how much cholesterol they contain, and Lp(a), which is largely inherited and which diet and exercise barely move.

What actually happens in the artery

The word "clogging" gives most people the wrong picture. It suggests grease building up inside a pipe, like fat down a drain.

That is not what happens. The damage occurs inside the artery wall, not on its inner surface.

In simplified terms: LDL and other apoB-containing particles cross into the artery wall and become trapped there, held by structural molecules in the wall called proteoglycans. Once retained, the particles can be modified, triggering inflammatory and immune responses that recruit white blood cells into the artery wall.

That response is protective in intent. Macrophages engulfing modified particles are doing their job. The problem is what happens when the traffic never stops. Overloaded macrophages become foam cells, die, and spill their contents, which accumulate as a soft lipid and necrotic core. A repair-and-clearance response that works fine occasionally becomes destructive when it runs for decades without pause. That part of the popular story is fair.

What the popular version gets backwards is the order. It is often told as damage happening first and cholesterol arriving to patch it. The dominant model since 1995, the response-to-retention hypothesis, puts particle retention first: retained particles are the initiating event, and the inflammation is the response to them.[21] That distinction matters, because it means the volume of particle traffic is a cause rather than a side effect.

Two consequences follow. The plaque can gradually narrow the channel, which is the part people picture. But the more dangerous event is plaque disruption, which can expose thrombogenic material to flowing blood and trigger a clot. Rupture is a major route to that acute thrombosis, although plaque erosion can produce the same endpoint. It can occur at a narrowing that was never severe enough to cause symptoms.

Not all plaque is the same

A plaque is not a lump of cholesterol. It contains a lipid and necrotic core, inflammatory cells, cholesterol crystals, fibrous tissue and calcium, in proportions that vary enormously.

Plaques are also not equally dangerous. Lesions prone to rupture tend to have larger lipid-rich cores, more inflammation and thinner fibrous caps, while more densely calcified plaques are generally associated with greater stability. Statin-treated plaques can become more densely calcified even as their lipid-rich components decline.[22]

That is why "more calcium" and "more dangerous plaque" are not always the same thing, and why a calcium score should not be read as a photograph of grease accumulating inside a pipe.

Which is why this is a story about cumulative exposure rather than a single reading. It is not one bad breakfast. It is decades of particle traffic into a wall.

Why your body makes it anyway

Given all that, here is the fact that reframes the topic. Your body synthesises cholesterol continuously, and endogenous production generally exceeds the amount a typical diet supplies.

Not as an accident. As a regulated manufacturing process. When cholesterol is scarce, a transcription factor called SREBP-2 increases the activity of HMG-CoA reductase, the rate-limiting enzyme in cholesterol synthesis. When cholesterol is abundant, the same system suppresses it.[1] Statins work by inhibiting that same enzyme.

The brain is the clearest case of why it matters. The brain holds roughly 23% of the body's total cholesterol while being about 2% of body weight. Around 70 to 80% of that sits in myelin, the insulating sheath around nerve fibres, where cholesterol makes up roughly half of all myelin lipids.[2]

And the brain does not take its cholesterol from your blood. The blood-brain barrier largely prevents exchange of circulating lipoprotein-bound cholesterol, so the brain manufactures its own locally, mainly in astrocytes and oligodendrocytes, with very slow turnover.[2]

Brain cholesterol and blood cholesterol are largely separate pools. Hold onto that, because a popular argument depends on merging them.

What the dietary question settled

For decades the advice was to avoid cholesterol in food. Then it changed, and the change confused people.

The reason is the feedback loop. When dietary cholesterol intake rises, most people compensate by reducing absorption, reducing their own synthesis, or both. In one classic set of metabolic studies, compensation occurred in roughly 69% of cases.[3]

That is one reason modern guidance moved away from treating dietary cholesterol as the dominant lever. Individual responses vary substantially, and dietary cholesterol can still raise LDL in some people. But saturated fat quality and the overall dietary pattern generally matter more for LDL than dietary cholesterol alone.[3]

What this does not establish is that blood LDL doesn't matter, only that food cholesterol is a weaker lever on it than once believed. Those are different claims, and they get merged constantly.

The rule that kept changing

Someone decides what counts as "high." That decision has been revised repeatedly, and the more interesting part is that the kind of rule kept changing too.

So this is not a simple downward march. Medicine repeatedly changed what the number was supposed to mean: a target, then a trigger, then a risk modifier, then a percentage reduction, then a target again.

The 2001 revision is worth pausing on, because the panel described the consequence itself. Its summary stated that the emphasis on risk assessment would result in a substantial increase in the number of patients considered at risk and expand the number eligible for lifestyle and drug intervention.[5]

This is not a conspiracy claim. Thresholds move as evidence accumulates, and moving them can be correct. What matters is that a threshold is a committee decision rather than a property of your body, and the number defining "high" today is not the one that defined it when your parents were your age. The profession argued about this in public: treat-to-target drew sustained international criticism for lacking direct trial support, which is why the 2013 guidelines dropped numeric targets.[6]

The same trials, two honest numbers

Statin trials are usually reported in relative risk reduction.

A meta-analysis led from Dublin, with collaborators in Denmark and Australia and reporting no relevant financial disclosures, examined 21 randomised trials with more than 66,000 patients per arm and average follow-up of 4.4 years. It reported both ways of expressing the same result.[7]

OutcomeRelative risk reductionAbsolute risk reduction
All-cause mortality9%0.8%
Heart attack29%1.3%
Stroke14%0.4%

These describe the same trials. Neither column is wrong. The left is what usually gets communicated; the right incorporates how often the event actually happened, over what period, which makes it more directly useful for a decision.

The authors translated it: at that pooled absolute reduction, about 77 people would need statin therapy under those trial conditions for roughly 4.4 years to prevent one additional heart attack. That does not identify which person benefits, and the number shifts substantially with baseline risk, age and time horizon. In primary prevention the absolute figures fall further, to 0.6%, 0.7% and 0.3%.[7]

Now the counterweight, because a second legitimate analysis exists. The individual-participant dataset held by the Cholesterol Treatment Trialists' Collaboration estimated roughly 11 fewer major vascular events per 1,000 people over five years for each 1 mmol/L of LDL reduction, in people whose five-year risk of such an event was below 10%.[8] Different outcome definition, different population, different method. Two honest absolute-risk analyses, not one right answer.

One caution about over-reading either. The Dublin analysis found its metaregression inconclusive as to whether the magnitude of LDL reduction tracked the size of benefit. Its own authors describe marked heterogeneity. That finding does not overturn the much larger causal literature linking cumulative apoB and LDL exposure to atherosclerosis, which draws on genetic disorders, prospective cohorts, Mendelian randomisation, and trials of LDL lowering by mechanisms other than statins. It says this particular trial-level regression could not reliably recover the expected dose-response from a heterogeneous set of trials.[7][9]

Cargo and trucks

Here is the concept that resolves several internet arguments at once.

LDL-C measures the cargo. It tells you how much cholesterol is being carried inside your LDL particles.

ApoB approximates the number of trucks. Each atherogenic particle carries exactly one apoB molecule, so measuring apoB counts particles rather than contents.

Usually those track together. But in diabetes, metabolic syndrome and high triglycerides, particles can carry less cholesterol each, so a person can show an acceptable LDL-C while carrying a high particle count. That discordance is why the 2026 guideline says apoB can add information even when LDL-C targets appear met.[4] Two people can carry the same cholesterol in very different numbers of particles, and the particle count is what the artery wall encounters.

Who was actually studied

Two things about the trial evidence are worth knowing before reading any of the numbers above.

Some statin trials used active run-in periods, which can select a randomised population that has already demonstrated tolerability or adherence. That limits how directly trial side-effect rates generalise to an unselected clinic population.

And trial populations were not the general population. A systematic review of 42 lipid-lowering trials covering 298,605 patients found that 81% excluded people with severe kidney failure and 76% excluded moderate kidney failure, with older adults, women and people with multiple conditions also frequently excluded.[10]

One version of the obvious suspicion has been tested directly. Industry sponsorship is common in the statin trial evidence base. A network meta-analysis of 183 randomised trials compared industry-sponsored against non-industry results and found no systematic difference in the amount of LDL lowering reported, with a mean difference of 1.77 mg/dL and a credible interval spanning zero. The variation was explained by which statin and dose were studied, not by who paid.[11] That tests reported LDL lowering specifically, not every mortality, event or adverse-effect result across the whole literature.

The side effects that survived the test

This section matters for fairness, because it would be easy to select the one famous statin harm that trials refute and stop there.

These have randomised support:

New-onset diabetes. A CTT individual-participant meta-analysis found statins cause a small dose-dependent rise in blood glucose and an increase in new diabetes diagnoses. Low and moderate intensity therapy produced roughly a 10% proportional increase, high intensity roughly 36%, with most new diagnoses occurring in people already close to the diabetes threshold.[12]

Muscle symptoms, with nuance. In individual-participant data from 123,940 people in blinded trials, statins caused a small excess of mostly mild muscle pain or weakness, concentrated mainly in the first year. Most muscle symptoms occurring while someone was taking a statin were not caused by the statin.[13] In the smaller SAMSON n-of-1 trial, conducted in people who had previously stopped statins because of symptoms, symptom burden was nearly identical during statin and placebo months.[13] Both findings matter. Genuine statin-related muscle symptoms exist, and symptoms occurring during statin use do not automatically prove the drug caused them.

Liver test abnormalities. The 2026 adverse-effects analysis of 123,940 participants found small excesses in liver enzyme abnormalities alongside the established muscle and diabetes effects, while failing to substantiate most other label-listed conditions.[14]

Which brings us to memory.

In February 2012 the US Food and Drug Administration added a warning about potential cognitive effects to statin labelling, based on adverse event reports, case reports and observational studies.[15]

Randomised trials were then examined repeatedly. A systematic review in Annals of Internal Medicine found no increased incidence of Alzheimer's disease and no difference in procedural memory, attention or motor speed.[16] A 2020 review of 24 studies covering 1,404,459 participants found all three randomised trials showed no significant association with adverse cognitive effects.[17] A Cochrane pooling of two trials with 26,340 participants found statins neither increased nor decreased incident dementia.[17] The 2026 analysis did not substantiate cognitive impairment among the label-listed outcomes it tested.[14]

Here is why that reassurance is worth more than it looks. The same large body of blinded randomised-trial data did detect diabetes, muscle and liver effects. So the cognition finding is not the product of a method incapable of detecting harm. It is a method that found other harms and did not find this one.

And yet, in a registry study, more than one in four current and former statin users believed statins can cause memory loss.[17] That gap between a warning generated from reports and a belief that trials have not supported is the most instructive thing on this page.

Two groups the averages do not describe

Familial hypercholesterolaemia. The four-year-average argument above is particularly misleading for people with FH, an inherited condition affecting roughly 1 in 311 people. Their LDL exposure begins in childhood, not at the moment a prescription is considered, and short trial horizons cannot represent decades of cumulative exposure. The 2026 guideline emphasises earlier treatment for FH and for some young adults with markedly elevated LDL or strong premature family history.[4][18]

The lean, low-carbohydrate high-LDL phenotype. Some lean people on ketogenic diets develop very high LDL alongside low triglycerides and high HDL, a pattern its describers call the lean mass hyper-responder. Whether its long-term cardiovascular risk differs from what conventional apoB and LDL evidence predicts remains uncertain.

The study most often cited to argue it does not was published in JACC: Advances in 2025, reporting that plaque predicted plaque while apoB did not. It was retracted in 2026, with the journal citing methodological problems that made the data unreliable and too extensive to repair by correction.[19]

The honest conclusion is neither that this phenotype is exempt from LDL biology nor that the retraction proves it dangerous. It is that the paper most often invoked to grant the exemption can no longer support it, and outcome evidence specific to this phenotype remains insufficient.

What the 2026 guidelines added

The rules changed again this year, and it may come up at your next appointment. The 2026 AHA/ACC dyslipidemia guideline added:[4]

One quantified consequence. An independent analysis published in 2026 estimated that 41% of untreated US adults without existing cardiovascular disease would meet criteria for statin consideration under the 2026 framework, against 26% under the 2018 one, roughly 25 million additional adults, with the largest expansion among middle-aged people whose 10-year risk is low but whose 30-year risk is elevated.[20] That is modelling of the guideline, not a figure the guideline itself reports.

Key Takeaways

  • There is no good and bad cholesterol. It is one molecule in different carriers, and the number of atherogenic particles is what the artery wall encounters.
  • Arteries are damaged from inside the wall, not coated on the inside. Retained particles trigger an inflammatory response that builds plaque over years, and plaque disruption can trigger the thrombosis that causes a heart attack.
  • Your body makes cholesterol continuously, generally exceeding dietary supply, and the brain makes its own separately because circulating cholesterol largely cannot cross into it.
  • The same trials support both a 29% and a 1.3% figure. Relative and absolute, both correct, and you are entitled to both.
  • Some statin side effects survived randomised testing and some did not. Diabetes, muscle and liver effects are supported. Cognitive harm has not been confirmed, by the same methods that found the others.
  • The rules keep changing, and the 2026 framework is modelled to increase the proportion of untreated US adults without existing disease meeting criteria for statin consideration, from roughly 26% to 41%.

How to think about cholesterol in thirty seconds

  1. Ask for both numbers. If told a treatment cuts risk by a percentage, ask what the absolute reduction is, over what period, for someone at your risk level.
  2. Ask what your overall risk is, not just what your LDL is. The same LDL means different things at different risk levels.
  3. Ask whether more information would change the decision. Family history, Lp(a), apoB, or in selected cases a coronary calcium score can settle a genuinely uncertain call.
  4. Separate the pools. Blood cholesterol and brain cholesterol are largely independent. An argument that merges them is not describing the physiology.
  5. Treat exposure as cumulative. This is a decades-long process, which is why lifelong conditions like familial hypercholesterolaemia are a different situation from a single elevated reading at 50.

What SmartBar does with this

SmartBar scores food, and on this subject that is a narrower contribution than you might expect.

The dietary cholesterol question turned out not to favour our scanner: the cholesterol in your food influences your blood LDL less than the type of fat you eat and your overall dietary pattern. Our Nutrition page covers fat quality, and our front-of-package page covers why "low cholesterol" on a label tells you less than it appears to.

What we cannot do is tell you your risk, interpret your numbers, or advise on treatment. This page exists to make you a better participant in that conversation, not to have it for you.

This article is educational and is not medical advice. Nothing here is a reason to start, stop or change any medication. Statins and other lipid-lowering drugs are prescription medicines, and decisions about them depend on your overall cardiovascular risk, personal and family history, other conditions and medications, and your own preferences. Those decisions belong with a prescriber or pharmacist who knows your situation. If you are experiencing side effects you believe are related to a medication, that is a reason to speak with your doctor, not a reason to stop taking something on your own. Familial hypercholesterolaemia is a distinct genetic condition requiring specialist assessment.

Keep exploring: Nutrition · Longevity · Mental Health · Front-of-package claims · Lifestyle

Medical Disclaimer: Read our full Medical Disclaimer

References

  1. Cholesterol homeostasis and SREBP regulation. Horton, J.D., Goldstein, J.L., Brown, M.S. (2002). "SREBPs: activators of the complete program of cholesterol and fatty acid synthesis in the liver." Journal of Clinical Investigation 109(9):1125–1131. Foundational mechanism: Brown, M.S., Goldstein, J.L. (1997), "The SREBP pathway: regulation of cholesterol metabolism by proteolysis of a membrane-bound transcription factor," Cell 89(3):331–340. (Sterol regulatory element-binding proteins sense cellular sterol levels and control transcription of genes encoding cholesterol biosynthetic enzymes, including HMG-CoA reductase, and lipoprotein receptors, increasing synthesis and uptake in sterol-depleted cells and decreasing it when sterols are abundant. Brown and Goldstein received the 1985 Nobel Prize for the underlying work on cholesterol regulation.)
  2. Brain cholesterol metabolism. Dietschy, J.M., Turley, S.D. (2004). "Thematic review series: Brain Lipids. Cholesterol metabolism in the central nervous system during early development and in the mature animal." Journal of Lipid Research 45(8):1375–1397. DOI: 10.1194/jlr.R400004-JLR200. See also Björkhem, I., Meaney, S. (2004), "Brain cholesterol: long secret life behind a barrier," Arteriosclerosis, Thrombosis, and Vascular Biology 24(5):806–815. (Unesterified cholesterol is the major architectural component of compact myelin, and mean concentration in the human central nervous system is higher than in any other tissue, approximately 23 mg/g. Because of the blood-brain barrier there is essentially no passage of lipoprotein-bound cholesterol from the circulation into the brain, so the brain depends on local synthesis, with only limited exchange via oxysterol metabolites such as 24S-hydroxycholesterol. Myelin cholesterol turnover is very slow.)
  3. McNamara, D.J., Kolb, R., Parker, T.S., et al. (1987). "Heterogeneity of cholesterol homeostasis in man: response to changes in dietary fat quality and cholesterol quantity." Journal of Clinical Investigation 79(6):1729–1739. DOI: 10.1172/JCI113013. PMID: 3584466. PMCID: PMC424515. (Compensation in 69% of studies via decreased fractional absorption and/or endogenous synthesis; dietary fat quality a more consistent determinant of LDL than cholesterol quantity. Individual responses vary substantially.)
  4. "2026 ACC/AHA/AACVPR/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline on the Management of Dyslipidemia: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines." Writing Committee Chair Roger S. Blumenthal; Vice Chair Pamela B. Morris. Circulation, DOI: 10.1161/CIR.0000000000001423; concurrently in Journal of the American College of Cardiology, DOI: 10.1016/j.jacc.2025.11.016. Openly accessible guideline text and official Top Things to Know summary published by the American Heart Association at professional.heart.org. (Key changes verified against the guideline text and the societies' own summary: retitled from the 2018 "Management of Blood Cholesterol" to reflect atherogenic lipoproteins beyond LDL, including triglycerides, remnant particles and Lp(a); LDL-C and non-HDL-C treatment goals reinstated; PREVENT-ASCVD equations replacing the Pooled Cohort Equations for 10-year and 30-year risk assessment in adults aged 30 to 79; apoB measurement recommended to aid diagnosis and, once LDL-C and non-HDL-C goals are met, to guide treatment, particularly with triglycerides above 200 mg/dL, diabetes, or achieved LDL-C below 70 mg/dL; Lp(a) measured at least once in adulthood, treated as a risk-enhancing factor prompting more intensive management; selective coronary artery calcium scoring for reclassification; and earlier treatment consideration in familial hypercholesterolaemia and in young adults with LDL-C of 160 mg/dL or above or a strong family history of premature ASCVD.)
  5. NCEP Adult Treatment Panel history. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (2001). "Executive Summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III)." JAMA 285(19):2486–2497. DOI: 10.1001/jama.285.19.2486. PMID: 11368702. Full report: Circulation (2002) 106:3143–3421. 2004 update: Grundy, S.M., Cleeman, J.I., Bairey Merz, C.N., Brewer, H.B. Jr, Clark, L.T., Hunninghake, D.B., Pasternak, R.C., Smith, S.C. Jr, Stone, N.J. (2004), "Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines," Circulation 110(2):227–239, PMID: 15249516, also published in Journal of the American College of Cardiology 44:720–732 — which introduced an LDL-C goal below 70 mg/dL as a therapeutic option for patients at very high risk. (ATP I 1988; ATP II 1993. The body quotes the ATP III panel's own statement that its risk-assessment emphasis would substantially increase the number of patients considered at risk and expand the number eligible for lifestyle and drug intervention.)
  6. The shift from targets to risk and back. "Journey in guidelines for lipid management: From adult treatment panel (ATP)-I to ATP-III and what to expect in ATP-IV," PMID: 23961478 — peer-reviewed account recording that the treat-to-target paradigm drew sustained international criticism on the grounds of insufficient proof of safety and benefit for treating to an individual's LDL level, and that a risk-based approach results in fewer patients on high-dose statins. The reinstatement of LDL-C and non-HDL-C goals in 2026 is documented in reference 4. (The body's account of the 2013 shift away from numeric targets rests on this peer-reviewed source and on the 2026 guideline's own description of goals being reinstated, both openly accessible.)
  7. Byrne, P., Demasi, M., Jones, M., Smith, S.M., O'Brien, K.K., DuBroff, R. (2022). "Evaluating the Association Between Low-Density Lipoprotein Cholesterol Reduction and Relative and Absolute Effects of Statin Treatment: A Systematic Review and Meta-analysis." JAMA Internal Medicine 182(5):474–481. DOI: 10.1001/jamainternmed.2022.0134. (21 randomised trials, >66,000 per arm, mean follow-up 4.4 years. ARR: mortality 0.8% [0.4–1.2], MI 1.3% [0.9–1.7], stroke 0.4% [0.2–0.6]. RRR 9%, 29%, 14%. Primary prevention subgroup 0.6%, 0.7%, 0.3%. Authors report marked clinical and statistical heterogeneity and describe the metaregression as inconclusive. Led from the HRB Centre for Primary Care Research, RCSI Dublin, with collaborators in Denmark and Australia; authors report no relevant financial disclosures.)
  8. Cholesterol Treatment Trialists' (CTT) Collaboration. Individual-participant-data meta-analyses. Principal sources: Baigent, C., Keech, A., Kearney, P.M., et al. (2005), The Lancet 366(9493):1267–1278; CTT Collaboration (2010), The Lancet 376(9753):1670–1681, DOI: 10.1016/S0140-6736(10)61350-5; and CTT Collaborators, Mihaylova, B., Emberson, J., Blackwell, L., et al. (2012), "The effects of lowering LDL cholesterol with statin therapy in people at low risk of vascular disease: meta-analysis of individual data from 27 randomised trials," The Lancet 380(9841):581–590, DOI: 10.1016/S0140-6736(12)60367-5. (Approximately one-fifth reduction in major vascular events per 1.0 mmol/L LDL reduction. The absolute figure used in the body comes from the 2012 analysis: in individuals whose five-year risk of a major vascular event was below 10%, each 1.0 mmol/L reduction produced an absolute reduction of about 11 per 1,000 over five years. 27 trials, approximately 175,000 participants. Funding declared: British Heart Foundation; UK Medical Research Council; Cancer Research UK; European Community Biomed Programme; Australian National Health and Medical Research Council; National Heart Foundation, Australia.)
  9. Ference, B.A., Ginsberg, H.N., Graham, I., Ray, K.K., Packard, C.J., Bruckert, E., et al. (2017). "Low-density lipoproteins cause atherosclerotic cardiovascular disease. 1. Evidence from genetic, epidemiologic, and clinical studies. A consensus statement from the European Atherosclerosis Society Consensus Panel." European Heart Journal 38(32):2459–2472. DOI: 10.1093/eurheartj/ehx144. PMID: 28444290. (Appraises whether the association between LDL and atherosclerotic cardiovascular disease meets criteria for causality, drawing on separate meta-analyses of over 200 prospective cohort studies, Mendelian randomisation studies and randomised trials including more than two million participants. Rare variants reducing LDL receptor function raise LDL-C and ASCVD risk in dose-dependent fashion, while rare variants lowering LDL-C correspond to lower risk. This is the source for the causal claim stated in the framework box.)
  10. Exclusion criteria in lipid-lowering trials. ("Lipid-Lowering Trials Are Not Representative of Patients Managed in Clinical Practice: A Systematic Review and Meta-Analysis of Exclusion Criteria," Journal of the American Heart Association, DOI: 10.1161/JAHA.122.026551. 42 trials, 298,605 patients; 81% excluded severe kidney failure, 76% moderate.) (The body deliberately states no specific run-in exclusion percentage, only that active run-ins can select for tolerability or adherence.)
  11. Naci, H., Dias, S., Ades, A.E. (2014). "Industry sponsorship bias in research findings: a network meta-analysis of LDL cholesterol reduction in randomised trials of statins." The BMJ 349:g5741. DOI: 10.1136/bmj.g5741. PMID: 25281681. PMCID: PMC4184241. (183 randomised statin trials published 1985–2013, of which 146 industry-sponsored and 37 non-industry-sponsored; 103 were active-comparator trials. No systematic differences in magnitude between the LDL-lowering effects of individual statins in industry-sponsored versus non-industry-sponsored trials; mean difference 1.77 mg/dL, 95% credible interval −11.12 to 7.66. Variation attributed to individual statins at particular doses rather than sponsorship. Important qualifier raised in a published editorial response: while industry versus non-industry sponsorship showed no difference, head-to-head comparisons tended to favour the product of the sponsoring company, potentially through non-equivalent dosing of the comparator. So the finding is about sponsorship in aggregate, not about a sponsor's own head-to-head trials.)
  12. Cholesterol Treatment Trialists' Collaboration (2024). "Effects of statin therapy on diagnoses of new-onset diabetes and worsening glycaemia in large-scale randomised blinded statin trials: an individual participant data meta-analysis." The Lancet Diabetes & Endocrinology 12(5):306–319. DOI: 10.1016/S2213-8587(24)00040-8. (RR 1.10 for low and moderate intensity therapy, RR 1.36 for high intensity; small dose-dependent increase in glycaemia and in new diabetes diagnoses; approximately 10% proportional increase with low/moderate intensity therapy and approximately 36% with high intensity, with most new diagnoses among people already near the diagnostic threshold — roughly 62% of new diagnoses among those with baseline glycaemic measurements occurred in the highest baseline glycaemia quarter. Mean shifts in glucose and HbA1c were very small. Methodological caveat stated by the authors: the difference in absolute diabetes incidence between moderate- and high-intensity trial groups partly reflects more frequent HbA1c measurement in the high-intensity trials.)
  13. Statins and muscle symptoms. Cholesterol Treatment Trialists' Collaboration (2022). "Effect of statin therapy on muscle symptoms: an individual participant data meta-analysis of large-scale, randomised, double-blind trials." The Lancet 400(10355):832–845. DOI: 10.1016/S0140-6736(22)01545-8. (19 placebo-controlled trials, 123,940 participants: muscle pain or weakness reported by 27.1% on statin versus 26.6% on placebo; a 7% proportional excess during the first year, approximately 11 additional reports per 1,000 person-years; the authors estimate more than 90% of muscle-symptom reports among statin recipients were not caused by the statin. SAMSON: Howard, J.P., Wood, F.A., Finegold, J.A., Nowbar, A.N., Thompson, D.M., Arnold, A.D., Rajkumar, C.A., Connolly, S., Cegla, J., Stride, C., Sever, P., Norton, C., Thom, S.A.M., Shun-Shin, M.J., Francis, D.P. (2021), "Side Effect Patterns in a Crossover Trial of Statin, Placebo, and No Treatment," Journal of the American College of Cardiology 78(12):1210–1222, DOI: 10.1016/j.jacc.2021.07.022, PMID: 34531021; primary report Wood, F.A., et al. (2020), New England Journal of Medicine 383(22):2182–2184. NCT02668016. 60 participants who had previously discontinued statins because of side effects, each completing four months each of atorvastatin 20 mg, matching placebo and no tablet in random order, with daily smartphone symptom scores. Mean symptom score 8.0 in no-tablet months, 15.4 on placebo, 16.3 on statin, giving a nocebo ratio of approximately 0.90. Half the participants had restarted statins six months later. SAMSON establishes that symptoms recur on placebo in a highly selected group; it does not establish population incidence, which is what the CTT analysis provides. Sponsored by Imperial College London and funded by the British Heart Foundation.
  14. Cholesterol Treatment Trialists' Collaboration (2026). "Assessment of adverse effects attributed to statin therapy in product labels: a meta-analysis of double-blind randomised controlled trials." The Lancet 407(10529):689–703. DOI: 10.1016/S0140-6736(25)01578-8. (19 double-blind placebo-controlled trials, 123,940 participants. After multiple-testing control, statistically supported excesses in liver-test abnormalities, urinary-composition alterations and oedema; causal relationships not supported for most other label-listed effects, explicitly including cognitive impairment, depression, sleep disturbance and peripheral neuropathy. Urinary and oedema findings are small and are not carried into the body text.)
  15. US Food and Drug Administration (28 February 2012). FDA Drug Safety Communication: Important safety label changes to cholesterol-lowering statin drugs. FDA reference ucm293101; consumer summary ucm293330. (Labelling change adding information on reported cognitive effects, based on the Adverse Event Reporting System, case reports and observational studies. Note: this was a safety communication and labelling change, not a boxed warning. Some secondary literature describes it as a black box warning; that characterisation is inaccurate and is not used in this article.)
  16. Richardson, K., Schoen, M., French, B., Umscheid, C.A., Mitchell, M.D., Arnold, S.E., Heidenreich, P.A., Rader, D.J., deGoma, E.M. (2013). "Statins and cognitive function: a systematic review." Annals of Internal Medicine 159(10):688–697. DOI: 10.7326/0003-4819-159-10-201311190-00007. PMID: 24247674. (Searched to October 2012; concluded published data did not suggest an adverse effect of statins on cognition, while noting the strength of available evidence was limited, particularly for high-dose statins. The authors state their own limitations and call for further research; the article reflects that caution.)
  17. Statins and cognition, contemporary synthesis. "Association between statin use and cognitive function: A systematic review of randomized clinical trials and observational studies." Journal of Clinical Lipidology (2020). PMID: 33189626. (Searched to October 2019 in participants aged 60 and over; 24 studies, 1,404,459 participants, of which 21 prospective observational and 3 randomised. All three RCTs, with 3.2 to 5.6 years of follow-up, showed no significant association between statin use and adverse cognitive effects: OR 1.03 [0.82–1.30] and OR 1.00 [0.61–1.65]. Prospective studies showed no increased risk of all-cause dementia. Registry finding that more than one in four current and former statin users believed statins can cause memory loss.) Cochrane evidence: McGuinness, B., Craig, D., Bullock, R., Malouf, R., Passmore, P. (2016), "Statins for the treatment of dementia," Cochrane Database of Systematic Reviews, CD003160; and the companion Cochrane review pooling PROSPER and the Heart Protection Study, 26,340 participants, finding statins neither increased nor decreased incident dementia. (PRODUCTION CLEANUP ONLY, NOT A SCIENTIFIC RED FLAG. PMID 33189626 resolves to the paper and every figure cited here was confirmed against the journal's own abstract. First author and volume number to be added at copyedit. The source and its findings are verified; this is bibliographic metadata.)
  18. Familial hypercholesterolaemia prevalence. Beheshti, S.O., Madsen, C.M., Varbo, A., Nordestgaard, B.G. (2020). "Worldwide Prevalence of Familial Hypercholesterolemia: Meta-Analyses of 11 Million Subjects." Journal of the American College of Cardiology 75(20):2553–2566. See also Hu, P., Dharmayat, K.I., Stevens, C.A.T., et al. (2020), "Prevalence of Familial Hypercholesterolemia Among the General Population and Patients With Atherosclerotic Cardiovascular Disease: A Systematic Review and Meta-Analysis," Circulation 141(22):1742–1759. (Heterozygous FH prevalence around 1 in 311 to 1 in 313, with substantially earlier coronary risk when untreated. International sources used deliberately in place of a US national estimate, for a global readership.)
  19. Soto-Mota, A., Norwitz, N., Manubolu, V., et al. RETRACTED: "Longitudinal Data From the KETO-CTA Study: Plaque Predicts Plaque, ApoB Does Not." JACC: Advances 4(7), 2025. DOI: 10.1016/j.jacadv.2025.101686. Retraction notice: Soto-Mota, A., Norwitz, N.G., Manubolu, V.S., et al. (2026), JACC: Advances 5(5):102824, DOI: 10.1016/j.jacadv.2026.102824, published 27 May 2026. (The notice states methodological concerns affected the reliability of the data and that the errors were too extensive to resolve through correction. Cited here solely to establish that the retraction occurred, not for the original findings.)
  20. Sayed, A., Peterson, E.D., Navar, A.M. (2026). "Implications of the 2026 Versus 2018 ACC/AHA Multisociety Dyslipidemia Guidelines on the Clinical Care of U.S. Adults." Journal of the American College of Cardiology. Published online 12 August 2026. DOI: 10.1016/j.jacc.2026.06.037. (Among US adults without ASCVD and not already receiving lipid-lowering therapy, 41% of untreated US adults without ASCVD would meet criteria for statin consideration under the 2026 framework versus 26% under 2018, approximately 24.5 million additional adults, with the largest expansion among middle-aged adults with low 10-year but elevated 30-year risk. This is independent modelling of the guideline, not a figure reported by the guideline.)
  21. Response-to-retention hypothesis. Williams, K.J., Tabas, I. (1995). "The Response-to-Retention Hypothesis of Early Atherogenesis." Arteriosclerosis, Thrombosis, and Vascular Biology 15(5):551–561. DOI: 10.1161/01.atv.15.5.551. (Holds that subendothelial retention of apoB-containing lipoproteins, bound by arterial proteoglycans, is the necessary initiating event, with the inflammatory response following retention. Reinforced in Williams & Tabas (1998), Current Opinion in Lipidology 9(5):471–474, PMID 9812202. This model largely superseded the earlier response-to-injury hypothesis, which placed arterial damage first. Appears out of numerical sequence relative to first mention — normalise reference order by hand at copyedit.)
  22. Plaque composition, calcification and stability. Composition and vulnerability: "Coronary Artery Calcification: From Mechanism to Molecular Imaging," JACC: Cardiovascular Imaging, DOI: 10.1016/j.jcmg.2017.03.005. Statin effect on calcification, primary sources: Puri, R., Nicholls, S.J., Shao, M., Kataoka, Y., Uno, K., Kapadia, S.R., et al. (2015), "Impact of statins on serial coronary calcification during atheroma progression and regression," Journal of the American College of Cardiology 65(13):1273–1282, DOI: 10.1016/j.jacc.2015.01.036 — post-hoc patient-level analysis of 8 prospective randomised trials using serial intravascular ultrasound, comparing high-intensity statin (n=1,545), low-intensity statin (n=1,726) and no statin (n=224), finding statins exhibit pro-calcific effects independently of their effect on plaque progression. Lee, S.E., et al. (2018), "Effects of Statins on Coronary Atherosclerotic Plaques: The PARADIGM Study," JACC: Cardiovascular Imaging, DOI: 10.1016/j.jcmg.2018.04.015, PMID: 29909109, NCT02803411 — serial coronary CT angiography; statins associated with slower progression of total plaque volume, increased calcification and reduced high-risk plaque features. Supporting context on calcium density: Criqui, M.H., Denenberg, J.O., Ix, J.H., McClelland, R.L., Wassel, C.L., Rifkin, D.E., Carr, J.J., Budoff, M.J., Allison, M.A. (2014), "Calcium Density of Coronary Artery Plaque and Risk of Incident Cardiovascular Events," JAMA 311(3):271–278, DOI: 10.1001/jama.2013.282535, PMID: 24247483. (MESA, 3,398 participants aged 45–84 across four race and ethnicity groups at six US centres; CAC density was inversely associated with cardiovascular risk, independent of CAC volume, the finding underlying the treatment of denser calcification as the more stable phenotype. Refined in Bhatia, H.S., et al. (2023), Circulation: Cardiovascular Imaging 16(2):e014788, PMID 36802448, which found the lower risk associated with higher density varied by calcium volume.) Note that plaque erosion, not only rupture, accounts for a substantial share of acute coronary syndromes, so rupture morphology is not the whole mechanism. (Microcalcification modelling detail was removed from the body as more specialist than this article requires. Appears out of numerical sequence — normalise by hand at copyedit.)

Sourcing note: this article cites randomised trials, individual-participant-data meta-analyses, guideline documents and physiology reviews. Funding and institutional affiliation are disclosed for major evidence sources, including the Cholesterol Treatment Trialists' Collaboration and independently funded analyses that challenge aspects of the evidence base. Where the same evidence base can be presented in more than one legitimate way, both presentations are given. Where a dispute is unresolved — data access, absolute versus relative benefit, the lean high-LDL phenotype — both positions are cited. A retracted publication is identified as retracted and cited only to establish that fact. Non-US evidence is prioritised where equivalent quality exists, and the geographic origin of key analyses is stated. This article makes no treatment recommendation and provides no dosing information. Cross-link targets (#) resolve once the site structure is live. Reading time is measured, not asserted: body words excluding the reference list and the metadata header, counted at 225 words per minute and rounded to the nearest minute.

Log in to like this article.

Comments

No comments yet. Be the first to join the conversation.

Log in or create a free account to join the conversation. Every comment is reviewed before it goes live.

This article is educational information only and is not medical advice. Always consult a qualified professional about your health. See our Medical Disclaimer.

← Back to the Education Library