Cholesterol calculator

Enter your information

The more you tell the calculator, the more personalised your recommendations become. Every input here actively adjusts your supplement doses and projected outcomes.

Statin status

Statin therapy depletes CoQ10, which makes Ubiquinol essential rather than optional. If you have tried a statin and reacted badly, prescription alternatives are shown further down.

Which applies to you?

Your lipid panel & HbA1c

Slide each value to match your most recent blood test results. Supplement doses for omega-3, plant sterols, psyllium and berberine adjust dynamically.

These are placeholder values, not a result. They are unremarkable population-normal figures so the page starts neutral — move each slider to your own numbers from your blood test.

Triglycerides also reported as: TG / trigs Optimal < 1.7 · Pancreatitis risk > 5.6

4.9mmol/L

High

LDL cholesterol also reported as: LDL-C Optimal < 2.6 · Borderline 2.6–3.4 · High ≥ 3.4

3.5mmol/L

High

HDL cholesterol also reported as: HDL-C Optimal > 1.4 · Low < 1.0

1.10mmol/L

Low

HbA1c also reported as: glycated haemoglobin / A1c Normal < 39 · Prediabetes 39-47 · Diabetes ≥ 48

37mmol/mol

≈ 5.5% · eAG 6.2 mmol/L

Optimal

Advanced blood markers (optional)

Enter values from your full blood panel for finer personalisation. Each input shows what your value means clinically, and supplement doses adjust accordingly.

Click to show / hide marker inputs If left empty, supplements default to safe baseline doses

Why these markers? The tier system maps to the two-mechanism model of atherosclerosis. Tier 1 markers measure the direct drivers of plaque formation (atherogenic particles + lifelong genetic burden). Tier 2 markers measure the environment that determines whether those particles cause damage (inflammation, glycation, methylation, hepatic lipid handling). The third group captures supportive context that fine-tunes supplement doses and identifies treatable deficiencies.

Calculated values

Total cholesterol, non-HDL, and TC/HDL ratio are derived from your inputs above.

Total cholesterol

6.83mmol/L

High

Target < 5.0 (general) · < 4.0 (high risk)

Non-HDL cholesterol

5.73mmol/L

High

More accurate atherogenic risk marker than LDL alone · Target < 3.4

TC/HDL ratio

6.2

High

Single best lipid ratio for CV risk · Target < 3.5

Remnant cholesterol

—mmol/L

—

TC − LDL − HDL · "Residual cholesterol risk" · Target < 1.0 (see explainer below)

Your personal LDL goal may be much tighter than these bands suggest

ACC/AHA 2026 Dyslipidaemia Guideline

The "Optimal < 2.6 mmol/L" band on the LDL slider above reflects the general-population healthy threshold. But the 2026 ACC/AHA dyslipidaemia guideline (and the EAS/ESC 2019, and broadly NICE NG238) all set risk-stratified LDL goals that are much tighter when established disease, subclinical plaque, or major risk factors are present. If any of the conditions below apply to you, the "Optimal" badge on the slider above is not the right target for you personally:

Very high risk

< 1.4 mmol/L (<55 mg/dL)

Established ASCVD (prior MI, stroke, revascularisation, symptomatic PAD) with ≥2 major events, OR 1 major event + ≥2 high-risk conditions. CAC ≥1000 also falls here.

High risk

< 1.8 mmol/L (<70 mg/dL)

ASCVD (not meeting very-high criteria), CAC ≥100 or ≥75th percentile, diabetes with multiple risk factors, severe hypercholesterolaemia with FH/CAC, or PREVENT 10-year risk ≥10%.

Moderate risk

< 2.6 mmol/L (<100 mg/dL)

Diabetes 40–75 without ASCVD, CAC 1–99, severe hypercholesterolaemia without additional risk, or PREVENT 5–10%. This matches the calc-card "Optimal" band.

How to interpret your number on this page: if none of the very-high or high-risk criteria apply to you, the <2.6 mmol/L "Optimal" band on the slider is appropriate. If they do apply — particularly if you have CAC ≥100, established ASCVD, or diabetes with multiple risk factors — your personal LDL goal is significantly tighter than what the slider hint shows, and reaching the "Optimal" badge does not mean your LDL is adequately controlled. The Integrated Risk card below already accounts for CAC and ApoB when entered, and gives a more personalised picture. This is also why percentage LDL reduction (≥50% from baseline for high/very-high risk; ≥30% for moderate) matters alongside the absolute number — discuss appropriate targets with your GP.

Composite biomarker pattern

Integrated atherosclerotic risk picture

Based on lipids + HbA1c

—

How is this calculated? — methodology & evidence

▼

This card synthesises your blood markers across the three independent biological axes that drive atherosclerosis. It is not a diagnosis of plaque or arterial disease — only direct imaging (CAC score, CT angiography, carotid IMT) can establish whether plaque is actually present. What this card does is integrate your biomarker pattern to estimate how favourable or unfavourable your current biological environment is for plaque formation and progression.

Crucial framing: inference vs. measurement. Every input to this card except CAC is an inferential biomarker — it estimates the conditions for plaque formation over the coming years-to-decades, but cannot directly observe whether plaque has actually accumulated. The CAC scan is the only direct measurement in this tool. When CAC is entered, it takes precedence: a CAC=0 substantially overrides high biomarker risk (the "power of zero"), and a high CAC overrides low biomarker risk (imaging confirms the disease is present regardless of what your blood looks like). This distinction matters because biomarkers and imaging can legitimately disagree — and when they do, the imaging is right.

The three-axis model

Modern preventive cardiology recognises three largely independent contributors to residual cardiovascular risk:

Atherogenic particles (the "bullets") — the lipoprotein particles that physically deposit cholesterol into the arterial wall. ApoB is the gold-standard measure (it counts the actual number of atherogenic particles); LDL-cholesterol is a reasonable proxy when apoB isn't available; non-HDL captures the full atherogenic burden including remnants. When apoB is entered, this card uses it preferentially over LDL.
Metabolic / glycation environment (the "fuel") — captured by the TG/HDL ratio (a powerful marker of insulin resistance and small-dense LDL pattern) and HbA1c (the canonical clinical proxy for advanced glycation end-product burden). Elevated TG/HDL signals that even a "normal" LDL number may be misleading, because the LDL particles are likely small, dense, and disproportionately atherogenic. Elevated HbA1c indicates glycation-driven arterial stiffening that progresses even below the diabetic threshold.
Inflammation (the "ignition") — captured by hs-CRP. Without inflammation, even high apoB rarely produces rapid plaque progression; with high inflammation, even modest apoB drives accelerated disease. The CANTOS trial established this axis as mechanistically independent of cholesterol — anti-inflammatory therapy reduced events with zero change in LDL or apoB. The 2025 ACC consensus statement formally elevated hs-CRP to a core residual-risk biomarker.

Why LDL and apoB are treated as causal

You may have read commentary questioning whether LDL really "causes" heart disease, or whether it's just correlated with it. The current scientific consensus is firmly on the causal side, and the strongest evidence is not from dietary trials (which are notoriously confounded) — it is from Mendelian randomisation studies. People who carry naturally-occurring genetic variants that lower lifetime LDL (for example, PCSK9 loss-of-function mutations, or specific HMGCR variants) show dramatic reductions in coronary disease — proportional to the LDL difference, independent of diet, behaviour, or anything else that could confound an observational study. The same effect appears across multiple independent gene variants in the LDL pathway, which is the genetic-epidemiology gold standard for inferring causation. This is also why statins, PCSK9 inhibitors, ezetimibe and bempedoic acid — which all lower LDL through different mechanisms — all reduce CV events: the common mechanism is the LDL reduction itself. The same Mendelian-randomisation evidence base applies to apoB (atherogenic particle count) and Lp(a). It is genuinely settled science that lifetime LDL/apoB exposure causes atherosclerosis; what remains genuinely debated is how strongly modest dietary changes affect that exposure in any given individual. See the saturated fat explainer in section 13 for the related dietary nuance.

The two time horizons — short-term vs. lifetime

Not all the inputs operate on the same clock. This is a subtle but clinically important point:

Short-term (5–10 year) drivers: LDL, apoB, TG/HDL, HbA1c, hs-CRP. These reflect modifiable biology that responds to diet, exercise, sleep, and pharmacology within weeks-to-months. They drive the 5–10 year risk that QRISK3 and SCORE2 calculate.
Lifetime (40+ year) drivers: Lp(a), APOE genotype. These are genetic — fixed at conception and largely unmodifiable by lifestyle (although lifestyle still matters because the rest of the system is still in play). A young person with elevated Lp(a) and otherwise clean biomarkers is not at high 5-year risk, but is at elevated lifetime risk and should be treated as such — particularly for calcific aortic valve disease, where Lp(a) is uniquely causal. The Integrated Risk score weights Lp(a) lightly (max 2 points) because it cannot drive short-term events on its own, but its presence should be a permanent reminder that the lifetime trajectory needs intensive control of the modifiable axes.
The direct measurement (any-horizon): CAC sees plaque that has actually accumulated to date — regardless of which axis drove it.

The scoring system

Each axis contributes points to a composite score. The axes use the same evidence-based thresholds your tool already applies elsewhere — there is no "new" cutoff being introduced, only a way of combining them:

Atherogenic-particle axis (0–3 points) — from apoB if entered (≥1.3 g/L = 3, 1.0–1.3 = 2, 0.8–1.0 = 1, <0.8 = 0); otherwise from LDL (≥4.1 = 3, 3.4–4.1 = 2, 2.6–3.4 = 1, <2.6 = 0).
Metabolic axis (0–3 points) — combines TG/HDL ratio (in mmol/L: ≥1.74 = 2, 0.87–1.74 = 1, <0.87 = 0) plus HbA1c (≥48 mmol/mol = 2, 39–48 = 1, <39 = 0), capped at 3.
Inflammation axis (0–2 points, optional) — hs-CRP if entered (≥3 mg/L = 2, 1–3 = 1, <1 = 0). When not entered, this axis contributes nothing to the score but is flagged as missing.
Lifetime genetic load (0–2 points, optional) — Lp(a) if entered (>125 nmol/L = 2, 75–125 = 1, <75 = 0). Weighted lightly because it drives lifetime not short-term risk.
APOE ε4 amplifier — adds 1 point when an ε4 allele is present and the atherogenic-particle axis is already scoring 2+ (because ε4 carriers respond more strongly to elevated saturated fat / LDL).
CAC direct-measurement override — when entered, takes precedence over all the above. CAC = 0 drops the score by 3 points (typically 1–2 bands); CAC 100–399 adds 2 points and clears biomarker discord; CAC ≥ 400 forces the score to a minimum of 8 (Very high band).

The total maps to bands: 0–1 Optimal, 2–3 Low, 4–5 Moderate, 6–7 High, 8+ Very high.

The discordance overrides

Raw point totals can hide clinically important patterns, so several overrides apply (in priority order — CAC supersedes biomarker discord):

CAC = 0 (Power of zero) — direct imaging confirms no current calcification. Strongly overrides elevated biomarker risk for the next 5–10 years. The biomarker pattern still matters for lifetime trajectory, but the urgency drops dramatically.
CAC ≥ 100 (Imaging-confirmed plaque) — direct imaging confirms plaque has accumulated. Overrides "low biomarker" reassurance because the disease is already present regardless of what blood markers suggest.
Concordant low pattern — if your metabolic axis is optimal and hs-CRP is optimal (when entered), the card explicitly notes this. In this scenario, even a moderately elevated LDL is far less dangerous than it would appear in isolation, because the vessel wall is uninflamed and resistant to particle deposition.
Residual cardiometabolic / inflammatory risk — if LDL or apoB is optimal but TG/HDL, HbA1c, or hs-CRP is elevated, the card flags that the standard lipid panel is understating your true risk. This is the pattern most often missed in conventional GP screening.
Multi-axis elevation — when two or more of {metabolic, inflammation, atherogenic particles} are all elevated simultaneously, the band is bumped up one level, reflecting the multiplicative interaction documented in the literature (hs-CRP × apoB, HbA1c × overweight × hs-CRP).

What this card cannot tell you

This is a biomarker pattern, not a 10-year absolute risk percentage. Tools like QRISK3 (UK), SCORE2 (Europe), and the Pooled Cohort Equations (US) calculate absolute risk using age, sex, blood pressure, smoking, and a single lipid measure — and remain the appropriate first-line risk calculators for treatment decisions. What this card adds is mechanistic insight: which biological axis is driving any risk you have, and therefore which intervention will move the needle most. If you want a true risk-of-event-by-2035 number, use QRISK3 alongside this.

Without CAC entered, this card cannot tell you whether plaque is already present. A normal CAC score in a "high risk" biomarker pattern is a much stronger argument for caution before pharmacology than any blood test could provide; conversely, a high CAC score in a "low risk" biomarker pattern is a non-negotiable call to act. The biomarker pattern is the fire-conditions report; the CAC scan is the building inspection. Both can be useful — but where they disagree, the building inspection is what counts.

Educational only. This composite advisory is a synthesis of published research and is not a substitute for medical advice. Treatment decisions should be made with your GP or lipidologist, ideally with reference to validated absolute-risk calculators (QRISK3, SCORE2) and where appropriate direct imaging (CAC score).

Key evidence base: Ridker et al., CANTOS trial (NEJM 2017) — IL-1β inhibition reduces CV events independent of LDL. ACC Scientific Statement on Inflammation and Cardiovascular Disease (JACC 2025). Mora et al., LDL-discordance with alternative measures (Circulation 2014). ARIC Study — hs-CRP discordance with atherogenic lipids (JAHA). AMORIS & INTERHEART studies — ApoB:ApoA-1 as the single best lipid metric. Birukov et al., EPIC-Potsdam — AGEs and arterial stiffness across all glycaemic strata (Cardiovasc Diabetol 2021). Multiple cohort studies on TG/HDL ratio as a marker of insulin resistance and small-dense LDL pattern.

Deeper context What is "remnant cholesterol" and why does it matter?

▼

The standard lipid panel reports four headline numbers — total cholesterol, LDL, HDL, and triglycerides. The atherogenic-particle story has historically been told almost entirely through LDL. But there is a third population of cholesterol-carrying particles that the standard panel does not name explicitly, and that the lipidology literature now treats as a major independent driver of cardiovascular risk: the cholesterol carried inside triglyceride-rich lipoprotein (TRL) remnants — colloquially "remnant cholesterol".

What it actually is

Your liver and intestine release two large lipoprotein classes — VLDL (very low-density lipoprotein, from the liver) and chylomicrons (from the intestine after a fatty meal). Their primary cargo is triglycerides, which they deliver to muscle and adipose tissue as fuel. As they shed triglycerides at each delivery, they shrink into cholesterol-rich husks — these are the "remnants". In a metabolically healthy person the liver clears these remnants rapidly via the LDL receptor and LRP1. In someone with insulin resistance, obesity, hypertriglyceridaemia, or an APOE ε2 genotype, remnant clearance is impaired and these particles accumulate in circulation, lingering long enough to cross the endothelium and deposit cholesterol into the arterial wall.

Critically, remnant particles are smaller than 70 nm — small enough to cross the endothelial layer directly — and they carry substantially more cholesterol per particle than a single LDL. They also drive vascular inflammation and foam-cell formation in their own right. This is why elevated remnant cholesterol is now classified alongside LDL-C and Lp(a) as a causal (not merely associated) risk factor for atherosclerotic cardiovascular disease.

How to calculate it from your panel

Remnant cholesterol = Total cholesterol − LDL-C − HDL-C

This is what the new card in Section 3 calculates from your inputs. No new blood test is required — you already have the numbers. The value you see is your calculated remnant-C; a direct laboratory measurement (which captures additional overlooked individuals at risk per Varbo & Nordestgaard 2021) exists but is not routinely available on the NHS.

Interpretation bands

< 1.0 mmol/L (~< 39 mg/dL) — Optimal. Aligns with healthy fasting triglycerides < 1.7 mmol/L.
1.0–1.3 mmol/L (~39–50 mg/dL) — Borderline. Worth addressing through diet, weight, and alcohol moderation before considering pharmacology.
> 1.3 mmol/L (> ~50 mg/dL) — Elevated. Independently associated with substantially increased risk of myocardial infarction, ischaemic stroke, peripheral artery disease, and cardiovascular mortality even when LDL is at target.

Note: these are the same cutoffs reflected in your tool's existing triglyceride bands. That is not a coincidence — triglycerides and remnant cholesterol are mechanistically coupled, and elevated TG is the single most reliable population-level proxy for elevated remnant-C. If your TG number is high, your remnant-C will be high; if your remnant-C is high in this card, your TG number is likely the proximate cause.

Why this matters even when LDL is at target

The 2024 Nordestgaard meta-regression of 29 statin trials covering 191,202 participants quantified this directly: when statins reduce LDL by 20%, they reduce remnant cholesterol by only about 12% — meaning statin-treated patients can be at their LDL target while still carrying substantial residual remnant-cholesterol risk. In a Frontiers in Cardiovascular Medicine analysis, ~12% of patients with established cardiovascular disease had elevated remnant cholesterol despite hitting their LDL goal. In a cohort of 113,000 people, those with low LDL but high remnant cholesterol still had a 15% higher stroke risk.

This is the mechanism behind the term "residual cardiometabolic risk" — the risk that remains after LDL has been lowered to guideline targets. Your Integrated Risk card in Section 3 already captures this conceptually via the TG/HDL metabolic axis; the remnant-C number is the direct biochemical expression of that same axis.

What lowers remnant cholesterol (in order of evidence strength)

Reduce refined carbohydrate and added sugar. The most direct intervention. Hepatic de novo lipogenesis converts excess carbohydrate into triglycerides, which are then packaged into VLDL particles. Lower the carbohydrate substrate and you cut VLDL secretion at the source — within weeks, not months. The CGM section (Section 9) shows you which carbohydrates spike your particular physiology.
Substitute Mediterranean fats for saturated fats. Olive oil, avocado, walnuts, and oily fish do not raise VLDL secretion the way saturated fats and refined carbohydrate do. The MIND/Mediterranean dietary pattern has the strongest population-level evidence here.
Soluble fibre — particularly β-glucan and psyllium. Oats (soaked overnight to maximise β-glucan release), barley, and psyllium husk form a gel in the gut that binds bile acids and slows the absorption of dietary fat and cholesterol. This forces the liver to pull cholesterol from circulation to make new bile acids, lowering all atherogenic fractions including remnants. Effect size: ~5-10% reduction in atherogenic cholesterol with daily intake.
Weight loss — especially visceral fat reduction. Visceral adipose tissue is metabolically active and exports free fatty acids directly to the liver, driving VLDL overproduction. Even 5–10% weight loss can produce a disproportionately large drop in triglycerides and remnant-C. BMI alone misses this — waist circumference and visceral fat are more relevant.
Reduce alcohol. Ethanol is preferentially metabolised by the liver into fatty acids and packaged into VLDL. Even moderate intake raises triglycerides measurably. The current WHO position is that no level of alcohol is safe; from a strict remnant-cholesterol perspective, less is always better.
High-dose marine omega-3. Prescription-grade icosapent ethyl (pure EPA, REDUCE-IT) reduced cardiovascular events on top of statin therapy in patients with persistent elevated triglycerides — a population characterised by elevated remnant-C. The mechanism is partly TG-lowering (~20-30% reduction at 4g/day) and partly direct plaque stabilisation.
Exercise — especially Zone 2 aerobic. Skeletal muscle is the largest site of TG clearance via lipoprotein lipase. Regular aerobic exercise upregulates LPL activity and improves remnant clearance. Resistance training expands the glucose-disposal capacity that prevents postprandial VLDL surges in the first place.
Pharmacological options. Statins lower remnant-C modestly (~12% per 20% LDL reduction). Fibrates and high-dose omega-3 are more targeted. A new generation of agents — ANGPTL3 inhibitors, apoC-III antisense oligonucleotides — can cut remnant-C by 50–80% in trials but is not yet widely available. All of this is a GP/lipidologist conversation, not a self-prescription.

How remnant-C sits alongside the other markers in this tool

Think of it as the missing fourth pillar of the atherogenic-particle story. Your tool's Integrated Risk card synthesises three axes — atherogenic particles (apoB / LDL), metabolic environment (TG/HDL, HbA1c), and inflammation (hs-CRP). Remnant cholesterol is the biochemical bridge between the first two: it is both an atherogenic particle population and a direct readout of metabolic dysfunction. A high remnant-C number with optimal LDL is the textbook "metabolically dyslipidaemic" pattern the standard panel misses — and the same pattern the reactive hypoglycaemia callout in Section 9 describes.

Important framing. Remnant cholesterol is a calculated value with the same accuracy caveats as your other lipid numbers. The TC − LDL − HDL formula can produce small negative values when LDL is derived via Friedewald at low triglycerides; the card clamps these to zero. For a direct laboratory measurement, ask your GP or private lab for "remnant lipoprotein cholesterol" (RLP-C) — though calculated remnant-C is sufficient for routine screening and is what the Copenhagen cohort studies use.

Key evidence base: Wadström, Pedersen, Wulff, Nordestgaard (Eur Heart J 2023) — elevated remnant cholesterol and CV/non-CV mortality, HR 2.2 for CV death. Nordestgaard & Nordestgaard (Eur Heart J 2024) — meta-regression of 29 statin trials, 191,202 participants, showing statins reduce remnant-C ~12% per 20% LDL reduction. Yang et al. (Atherosclerosis 2023) — meta-analysis of 31 studies, RR 1.53 for CVD with elevated remnant-C. Varbo & Nordestgaard (Eur Heart J 2021) — directly-measured vs calculated remnant cholesterol in 109,574 individuals. Quispe et al. (Eur Heart J 2021) — remnant cholesterol predicts CVD beyond LDL and apoB. Multiple Copenhagen General Population Study analyses establishing causality via Mendelian randomisation.

For statin-intolerant patients

Statin alternative prescription options

If you've tried statins and developed muscle pain, fatigue, brain fog or other symptoms that resolved on stopping, you have prescription options that work via different mechanisms. Discuss these with your GP or lipid specialist — none of them engage the SLCO1B1/muscle pathway that drives classical statin myopathy.

Bempedoic acid

Oral · Non-statin

Inhibits cholesterol synthesis upstream of statins — bypasses the muscle pathway entirely. Lowers LDL ~17–28%. Monitor uric acid (can rise).

Ezetimibe

Oral · Absorption inhibitor

Blocks intestinal cholesterol uptake at the NPC1L1 transporter — a completely different pathway from statins, which is why the effects add rather than overlap. Lowers LDL ~15–20% alone, and roughly a further 20% on top of a statin. It is also the first thing to add when a statin alone is not reaching target, rather than only a fallback for people who cannot take statins — see ezetimibe, and how low to go for the outcome evidence and the reason it was written off for a decade.Note: some patients with statin-like reactions also react to ezetimibe, so try low and slow.

Fenofibrate

Oral · Fibrate

Particularly effective for high triglycerides and APOE ε2 dysbetalipoproteinaemia. TG reduction 30–50%. Monitor LFTs and creatinine.

PCSK9 inhibitors

Subcutaneous · Monoclonal antibody

Alirocumab or evolocumab, injected every 2–4 weeks. Reduces LDL 50–60% and Lp(a) by 20–25%. Bypasses hepatic muscle pathway entirely.

Inclisiran

Subcutaneous · siRNA

Same target as the PCSK9 antibodies but biannual dosing after loading at 0 and 3 months. LDL reduction ~50%. Unlike the antibodies it can be started in primary care under NICE TA733. All four PCSK9 medicines compared. Available on NHS for established CV disease.

Icosapent ethyl (IPE)

Oral · Prescription omega-3

The omega-3 comparator controversy, because this one is genuinely unresolved. REDUCE-IT (NEJM 2019, n=8,179) gave icosapent ethyl 4 g/day against a mineral oil placebo and found a large benefit: 17.2% versus 22.0% for the primary endpoint, HR 0.75 (0.68–0.83). But the mineral oil arm did not behave like an inert control. Across the trial, LDL-C rose about 10%, apoB about 8% and hs-CRP about 32% in the placebo group, with later analyses reporting rises in IL-6, Lp(a) and oxidised LDL as well. One plausible mechanism is that mineral oil interferes with absorption of the background statin. If the control arm got worse, the treatment arm looks better than it was.

STRENGTH (JAMA 2020, n>13,000) tested a different formulation — EPA plus DHA — against corn oil, which did not move those biomarkers, and was stopped for futility. Both trials found more atrial fibrillation on omega-3. Two explanations fit: either pure EPA is genuinely different from EPA+DHA, or REDUCE-IT's comparator manufactured part of its result. The strongest argument for the first is that EPA has shown benefit in trials that used no mineral oil at all — JELIS, open-label, reported a 19% reduction in coronary events. The placebo question was reviewed by the FDA, Health Canada and the European Medicines Agency, all of which approved the drug; the trial's own investigators report that all three concluded any effect of the placebo choice on the 25% risk reduction would have been very small — a characterisation worth noting comes from the investigators rather than from a regulatory document quoted directly. Against that, STRENGTH's executive committee chair called REDUCE-IT a false positive. Neither design can settle it, and the trial that would — icosapent ethyl against corn oil — has not been run. A worked example of the comparator problem set out on how to read a trial.

What the UK actually does with it, which is narrower than the debate suggests. NICE recommends icosapent ethyl (Vazkepa) — TA805, July 2022 — but only for secondary prevention: established cardiovascular disease, already on a statin, fasting triglycerides 1.7 mmol/L or above, and LDL-C above 1.04 and at or below 2.60 mmol/L. It is not recommended for primary prevention, where NICE judged it unlikely to be cost effective, and it does not apply in familial hypercholesterolaemia. NHS England otherwise advises against omega-3 for cardiovascular prevention, with this as the stated exception. So the practical question for most people is not whether REDUCE-IT was right; it is that they do not meet the criteria either way.

Prescription-grade purified EPA — not the same as over-the-counter omega-3 supplements. For elevated TG with established CV disease, and a clinician's decision given the above.

Start here — what you eat

Diet is the intervention with the largest effect on every number above, and the supplement tiers that follow work on top of it rather than instead of it. These are recommended maximums for the limiting nutrients and minimums for fibre — adjust to your activity level.

Order of priority

What you eat, every day. Nothing below moves your numbers as much. If your results are poor and you change one thing, change this.
Medication, where your numbers warrant it. If the panel above puts you in a treatment range, a statin or its alternatives will do more, faster, than any supplement on this page. That is a conversation with your GP, not a decision to make here.
Supplements, last. The tiers below are genuinely useful and they are still third. They fine-tune a good diet; they cannot rescue a poor one, and they are not a substitute for treatment you need.

Total daily targets

🍞

Carbohydrates

225g max

Whole grains, legumes; ~45% of energy

🍗

Protein

100g min

~1.2g per kg body weight

🥑

Total fat

65g max

~30% energy, mostly unsaturated

🧈

Saturated fat

20g max

<13g if hypercholesterolaemic

🍯

Added sugar

25g max

WHO guidance; ~6 tsp

🌾

Fibre

35g min

≥30g general; +psyllium

🧂

Salt

5g max

UK NHS guidance; ~1 tsp

🥚

Cholesterol

200mg max

If hypercholesterolaemic

💧

Water

2L min

More if active or warm climate

Per-meal breakdown

Meal	Carbs	Protein	Fat	Sat. fat	Fibre	Sugar
Breakfast	50-65g	25-30g	15-20g	≤5g	8-10g	≤6g
Lunch	65-80g	30-35g	20-25g	≤7g	10-12g	≤8g
Dinner	65-80g	30-35g	20-25g	≤7g	10-12g	≤8g
Snacks (combined)	20-30g	10-15g	5-10g	≤2g	3-5g	≤3g
Daily total	~225g	~100g	~65g	≤20g	≥35g	≤25g

When you eat matters too

Hitting these targets is half the picture. How long you leave your liver between meals has a direct effect on hepatic triglyceride content and insulin sensitivity — and shortening the eating window clears liver fat even without changing what is on the plate.

Meal timing — give your liver a break →

Tier 1 — Move the numbers

The core daily supplements with the strongest clinical evidence for cholesterol and cardiovascular health.

Choosing a multivitamin — what to look for, what to avoidClick to expand · Important checks before you buy any multi or "performance" supplement

Not all multivitamins are equal — and some commonly-included ingredients are actively harmful for cardiovascular health. The recommendation in Tier 1 below is for Lamberts Multi-Guard or Vitabiotics Wellman/Wellwoman (iron-containing or iron-free depending on your ferritin/transferrin status — see Tier 1). Before you swap to anything else, check the label against this list.

✓ A good multivitamin should contain

Vitamin B12 (cyanocobalamin or methylcobalamin) — 2.5-25 µgB12 deficiency increases sharply with age: roughly 1 in 20 adults aged 65-74 are deficient, rising to 1 in 10 over 75. B12 is essential for methylation, homocysteine recycling, and nervous system function — and it's also a co-factor that omega-3 needs to work optimally for brain health.
Vitamin B6 (pyridoxine or P-5-P) — 1.4-10 mgWorks with B12 and folate to control homocysteine. Inadequate B6 leaves homocysteine elevated regardless of B12/folate intake.
Folate (folic acid or methylfolate / 5-MTHF) — 200-400 µgThe third pillar of homocysteine control. Methylfolate (5-MTHF) is preferable for the ~30-40% of the population with MTHFR variants. UK NHS recommends folic acid specifically for women planning pregnancy.
Iodine — 100-150 µgThe UK RNI is 140 µg/day. Iodine is essential for thyroid hormone synthesis and brain development (see milk panel in Section 11). About 21% of UK women of childbearing age are deficient; the figure is higher in those who've reduced dairy.
Vitamin D3 — 400-1000 IUUK adults need supplementation October-March. A multivitamin gives the baseline; the standalone Tier 1 D3 product gives the variable dose based on your blood level.
Magnesium, zinc, selenium — modest dosesFoundation minerals for cardiovascular, immune, and metabolic function. Multi doses are typically appropriate; only supplement higher amounts if blood markers indicate deficiency.

○ Generally fine in multivitamin amounts (but not as standalone high-dose)

Potassium — typically <100 mg in a multiMulti doses are biologically inconsequential. Standalone potassium supplements above 200 mg/day can be risky for anyone with kidney issues or taking ACE inhibitors / ARBs / potassium-sparing diuretics, but the trace amounts in a standard multi are not a concern.
Vitamin E — 15-30 IU (food-equivalent) is fineMulti doses are safe. However, high-dose standalone vitamin E (200+ IU/day) is associated with increased all-cause mortality in meta-analyses, increased heart-failure risk in the HOPE and GISSI-Prevenzione trials, and amplified bleeding risk when combined with warfarin or aspirin. Don't add a separate "vitamin E 400 IU" capsule on top of your multi.
Vitamin K1 (phylloquinone) — multi dosesK1 from a multi is generally fine even on warfarin, provided your intake stays consistent (warfarin dosing is calibrated to your regular K1 intake). It's K2 (MK-7) that's the problem — see the warning section below.

⚠ Warnings — check carefully or avoid

L-carnitine (in "performance," "energy," or "men's" multivitamins)L-carnitine is widely marketed for muscle energy and fatty-acid metabolism, but the gut microbiota converts it into trimethylamine-N-oxide (TMAO) — a metabolite directly linked to atherosclerosis, foam-cell formation, impaired reverse cholesterol transport, and major adverse cardiac events. The 2013 Nature Medicine study (Hazen group) showed plasma L-carnitine independently predicts cardiovascular events in 2,595 subjects when concurrent TMAO is elevated. Avoid any multivitamin or "energy" supplement listing L-carnitine, acetyl-L-carnitine, or propionyl-L-carnitine if cardiovascular protection is your goal. Acceptable exception: short-term use under medical supervision for specific carnitine deficiency states.
Vitamin K2 (MK-7) IF you are on warfarin or another vitamin K antagonistEven tiny doses of MK-7 (10-20 µg/day) reduce INR in roughly 60% of patients on warfarin or acenocoumarol — directly counteracting the anticoagulant effect and significantly raising clot risk. This is an absolute contraindication for VKA users. However, K2 (MK-7) is genuinely beneficial for vascular health in everyone else — it directs calcium away from arteries toward bones via Matrix Gla Protein activation. Newer DOACs (apixaban, rivaroxaban, edoxaban, dabigatran) do not interact with K2. Decision rule: on warfarin → avoid K2 supplements entirely; on a DOAC → K2 is fine; not on any anticoagulant → K2 100-180 µg/day is beneficial alongside vitamin D3.
High-dose iron (above multivitamin amounts) without confirmed deficiencyIf your ferritin is >100 µg/L or your transferrin saturation is >30%, additional iron is at best wasted and at worst pro-atherogenic — excess iron catalyses LDL oxidation. Iron in a standard multi is fine for most adults; switch to the iron-free variant if blood markers show elevation. See the multivitamin card in Tier 1 below — it auto-adjusts based on your ferritin/transferrin inputs.
Calcium standalone (above food + multi amounts)High-dose calcium supplementation without K2 and vitamin D3 to direct calcium correctly has been associated with increased coronary calcification in several meta-analyses. If you need calcium for bone health, pair it with vitamin D3 and (if not on warfarin) K2 (MK-7) so calcium is directed to bone rather than artery walls.

Quick label check: before buying any multivitamin or "performance"/"men's"/"energy" supplement, scan the ingredient list for these red flags: L-carnitine, acetyl-L-carnitine, propionyl-L-carnitine, "K2" or "MK-7" (if on warfarin), and vitamin E above 30 IU. If you're on any prescription medication — particularly anticoagulants — check all supplement interactions with your GP or pharmacist before starting.

Direct lipid effect

Strongest evidence

Read the nutrition section above first. These three have randomised, dose-responsive evidence of a direct effect on lipid numbers and apply to almost anyone with a raised panel — but they are an addition to a good diet, not an alternative to one. Doses adapt to your inputs; with no blood marker entered, each defaults to a safe baseline.

Tier 2 — Triggered by your inputs

Items where something specific in your data or your situation makes the case — a statin, a low level, a raised marker. When one of these is triggered it can matter more than anything in Tier 1.

Conditional on your situation

Depends on you

These are not a step up in sophistication — they are a step up in relevance to you specifically. Ubiquinol is close to essential on a statin and largely unnecessary off one. Vitamin D depends on your level and the season. Magnesium is shown by default for the reason set out on its card. TMG appears only if your homocysteine warrants it.

Prescription & repurposed anti-inflammatory CV agentsRx · doctor only

Everything else in this guide is a food or an over-the-counter supplement. The agents below are prescription-only medicines being studied or used to target residual inflammatory risk — the part of cardiovascular risk that persists even when LDL is well controlled. This section is showing because your hs-CRP is elevated (≥2 mg/L), the marker of that residual inflammatory risk. They are included for transparency and discussion with your doctor; they are not self-prescribed and carry no shop links.

Low-dose colchicine (0.5 mg daily)Brand: Lodoco (US) · long established as a gout drug

Strong RCT · FDA-approved

The first — and so far only — anti-inflammatory drug approved (US FDA, June 2023) specifically to reduce cardiovascular events, in adults with established atherosclerotic disease or multiple risk factors. It inhibits microtubule assembly and the NLRP3 inflammasome, damping the IL-1β → IL-6 → CRP cascade that destabilises plaque — a mechanism independent of cholesterol.

LoDoCo2 (n=5,522, chronic coronary disease): primary composite endpoint −31% (HR 0.69), 6.8% vs 9.6%, NNT 36 over ~2.4 yr COLCOT (recent MI): CV events −23%

Taken on top of a statin — lipid-lowering and inflammation-lowering are complementary, not competing. Caveats: GI upset is the commonest side-effect; contraindicated in significant kidney or liver impairment, blood disorders, and with strong CYP3A4/P-gp inhibitors; trials noted a modest pneumonia signal and, in some, numerically higher non-cardiovascular deaths. The most recent large trial (CLEAR-Synergy, 2024) did not reproduce the benefit, so the picture is not fully settled. UK use for cardiovascular indications is currently off-label (Lodoco is a US approval) — a GP/cardiologist decision.

Subantimicrobial-dose doxycycline (20 mg twice daily)Brand: Periostat · licensed for periodontitis, not for CV use

Pilot · biomarkers only

At 20 mg the dose sits below the antibiotic threshold, so it acts as a host-modulator rather than killing bacteria: it inhibits the matrix metalloproteinases (MMPs) that chew through a plaque's fibrous cap, and lowers systemic inflammation.

MIDAS pilot (n=50 ACS, 6 mo): hs-CRP −46%, IL-6 −33% (22.1→14.7 pg/mL), pro-MMP-9 activity −50% 2-yr trial (postmenopausal women): ↑ serum cholesterol efflux capacity

Important: it does not lower LDL on a standard blood panel — the proposed benefit is changing what your LDL is allowed to do (less oxidation, more plaque stability), not the number itself. The evidence is limited to small pilot trials and biomarker endpoints with no large outcome trial, and antibiotic-stewardship plus long-term-use questions need a doctor's judgement.

For context — the trials that frame this fieldNot recommendations; included so the evidence picture is honest

Background

Canakinumab (CANTOS, 2017) — the proof of concept. A targeted IL-1β antibody that reduced cardiovascular events with no change in lipids, which is why inflammation came to be treated as a potentially causal driver rather than a bystander. But it is an expensive injectable with an increased fatal-infection signal, and is not used in practice for CV prevention.

Low-dose methotrexate (CIRT, 2019) — the instructive failure. Broad, non-specific immunosuppression did not reduce cardiovascular events. The lesson taken at the time: the target has to be the specific IL-1β/NLRP3 → IL-6 pathway rather than inflammation in general.

Ziltivekimab (ZEUS, 2026) — the result that unsettled that lesson. A direct IL-6 antibody, given to over 6,300 people selected for having inflammation (hs-CRP ≥ 2 mg/L) alongside established cardiovascular and kidney disease. It suppressed IL-6 and CRP exactly as designed and produced no reduction in events at all: hazard ratio 0.99 (0.88–1.11). Hitting the specific pathway was not enough. Two further trials of the same drug (HERMES, ARTEMIS) read out in 2027. Status, checked August 2026: these are the headline results announced on 31 July 2026; the full dataset is not yet published, and the related HERMES and ARTEMIS trials report in 2027.

One or the other — not both. Colchicine and SDD doxycycline both converge on the same IL-6/CRP inflammatory pathway; no trial has tested them together, and stacking them adds overlapping GI and other risks for no proven extra benefit. If an anti-inflammatory agent is on the table, colchicine is the better-evidenced of the two (two positive outcome RCTs and a regulatory approval) versus doxycycline's single small pilot. Doxycycline is mainly of interest where colchicine is not tolerated or is contraindicated, or where there is a concurrent reason such as periodontitis. Either way it is a clinician's call.

The colchicine evidence is no longer one-directional, and this page will not pretend otherwise. COLCOT (2019) and LoDoCo2 (2020) were positive. CLEAR SYNERGY / OASIS-9 (NEJM 2025) then randomised 7,062 people after a heart attack — the largest colchicine trial run — and found no reduction in cardiovascular death, recurrent infarction, stroke or ischaemia-driven revascularisation, despite lowering CRP as expected, with more diarrhoea. Read alongside ZEUS above, the short summary is that anti-inflammatory treatment for cardiovascular prevention is an open question rather than a settled benefit, even though hs-CRP remains a good predictor of who is at risk.

Educational summary of published trials: MIDAS (Arterioscler Thromb Vasc Biol 2004); cholesterol-efflux trial (Inflamm Res 2013); LoDoCo2 (NEJM 2020); COLCOT (2019); CANTOS (2017); CIRT (2019); CLEAR SYNERGY / OASIS-9 (NEJM 2025); ZEUS (2026, headline results). None of these are self-prescribed — discuss with your GP or cardiologist before considering any of them.

Tier 3 — Weaker evidence, optional

Plausible, low-risk, and genuinely optional. Nothing here has the outcome data of Tier 1, and none of it compensates for skipping Tier 1.

Nice to have, not load-bearing

Optional

Two things worth being straight about. A multivitamin sits here rather than in Tier 1 because, while it is sensible insurance against dietary gaps, the trial evidence that multivitamins reduce cardiovascular events is weak to null — it is cover, not treatment. The polyphenols have real mechanistic interest and mostly surrogate-endpoint human data. Spend on Tier 1 first, and on food before any of it.

Projected 12-week outcomes

Three scenarios — supplements alone, plus the diet and meal-timing plan, and the full lifestyle stack with exercise on top. Estimates based on published meta-analyses; individual response varies considerably.

Supplements only

Take the recommended supplement stack consistently; diet unchanged. The realistic baseline for someone who isn't ready (or able) to overhaul their food yet.

Triglycerides

4.9 → 3.4

High

−31%

LDL cholesterol

3.5 → 2.9

Borderline

−16%

HDL cholesterol

1.10 → 1.16

Borderline

+5%

Supplements plus diet & meal timing

Same supplements plus the daily nutrition targets, foods-to-eat-regularly list, milk-type switch, and 12+ hour overnight fast from the sections below. This is what is actually achievable with full lifestyle implementation.

Triglycerides

4.9 → 2.4

High

−52%

LDL cholesterol

3.5 → 2.5

Optimal

−29%

HDL cholesterol

1.10 → 1.21

Borderline

+10%

Supplements plus diet plus exercise plan

Everything in Scenario B plus 150+ minutes per week of Zone 2 aerobic exercise, 2-3 resistance sessions, post-meal walks, and 7-9 hours of sleep — as described in Section 12. Exercise is the only meaningful lever for raising HDL, and resistance training expands skeletal-muscle glucose disposal capacity, indirectly suppressing hepatic triglyceride production.

Triglycerides

4.9 → 1.8

Borderline

−64%

LDL cholesterol

3.5 → 2.4

Optimal

−31%

HDL cholesterol

1.10 → 1.31

Borderline

+19%

How the gains stack up. Diet (B vs A) contributes ~10-15% additional LDL reduction (cutting saturated fat, raising soluble fibre, swapping refined carbs), ~20-30% additional TG reduction (reducing added sugar, refined carbs, alcohol; adding oily fish), and modest additional HDL improvement. Exercise (C vs B) adds a further ~15-25% TG reduction via skeletal-muscle LPL activation and improved insulin sensitivity, a further ~5-10% HDL increase (the only reliable lever for HDL), and a smaller ~3-7% direct LDL reduction — but with a particle-quality shift from small dense to large buoyant LDL that delivers cardiovascular benefit beyond the LDL-C number. All three layers overlap mechanistically, so combined ≠ summing the percentages; the model uses multiplicative compounding with diminishing-returns caps. The PREDIMED-style cardiovascular event reduction from full lifestyle implementation (~30% MACE reduction over 5 years) is larger than any of these lipid numbers suggest — many of the benefits of exercise are independent of cholesterol entirely (endothelial function, blood pressure, autonomic balance, plaque stability, all-cause mortality).

Cholesterol calculator

Enter your information

Statin status

Your lipid panel & HbA1c

Advanced blood markers (optional)

Click to show / hide marker inputs If left empty, supplements default to safe baseline doses

Calculated values

Your personal LDL goal may be much tighter than these bands suggest

Integrated atherosclerotic risk picture

How is this calculated? — methodology & evidence

The three-axis model

Why LDL and apoB are treated as causal

The two time horizons — short-term vs. lifetime

The scoring system

The discordance overrides

What this card cannot tell you

Framework Position · Section 15 —

Deeper context What is "remnant cholesterol" and why does it matter?

What it actually is

How to calculate it from your panel

Interpretation bands

Why this matters even when LDL is at target

What lowers remnant cholesterol (in order of evidence strength)

How remnant-C sits alongside the other markers in this tool

Statin alternative prescription options

Bempedoic acid

Ezetimibe

Fenofibrate

PCSK9 inhibitors

Inclisiran

Icosapent ethyl (IPE)

Start here — what you eat

Total daily targets

🎯 Why these targets have tightened for you

Per-meal breakdown

When you eat matters too

Tier 1 — Move the numbers

Choosing a multivitamin — what to look for, what to avoidClick to expand · Important checks before you buy any multi or "performance" supplement

✓ A good multivitamin should contain

○ Generally fine in multivitamin amounts (but not as standalone high-dose)

⚠ Warnings — check carefully or avoid

Direct lipid effect

Tier 2 — Triggered by your inputs

Conditional on your situation

Prescription & repurposed anti-inflammatory CV agentsRx · doctor only

Tier 3 — Weaker evidence, optional

Nice to have, not load-bearing

Projected 12-week outcomes

Supplements only

Supplements plus diet & meal timing

Supplements plus diet plus exercise plan