Biomarkers · Clinical Literacy

    ApoB vs LDL-C: Which Number Should You Actually Track?

    One Counts the Particles. The Other Weighs What's Inside Them.

    Reviewed by Dr. Sanjeev GoelNovember 17, 202612 min read
    A row of identical small glass vials on a dark surface, rim-lit in vivid orchid, each filled to a different level

    Every particle that can lodge in an artery wall carries exactly one molecule of apolipoprotein B. Cholesterol is the cargo; apoB is the count. In 389,529 adults, when both numbers went into the same model, only one of them was still associated with heart attacks.

    The 60-Second Answer

    LDL-C tells you how much cholesterol is being carried inside the particles that drive atherosclerosis. ApoB counts the particles themselves — one molecule of apoB per particle. Where the two disagree, which happens in something between 5% and 25% of adults depending on how you define disagreement, the particle count is the number that tracks heart attacks. For most people the two agree closely enough that LDL-C is fine. For people with triglycerides above 1.5 mmol/L, type 2 diabetes, obesity or metabolic syndrome, or an already-low LDL-C on treatment, apoB is the better number — and Canadian guidelines have said so since 2012.

    Particles and cargo

    Every lipoprotein capable of lodging in an artery wall — LDL, VLDL, IDL, their remnants, and lipoprotein(a) — carries exactly one molecule of apolipoprotein B. HDL carries none. A single apoB measurement is therefore a particle count: the number of things circulating in your blood that are capable of starting a plaque.[4]

    LDL-C is a different quantity entirely. It is the total mass of cholesterol carried collectively by your LDL particles. And in most laboratories it is not measured at all — it is computed from three other numbers (total cholesterol, HDL cholesterol and triglycerides) using the Friedewald equation or one of its successors, the Martin/Hopkins and Sampson formulas.[4] Ontario's own Schedule of Benefits for Laboratory Services makes this visible at a glance: total cholesterol, HDL cholesterol and triglycerides each carry a billing code. LDL cholesterol does not, because it is arithmetic.[10]

    That arithmetic holds up well in the middle of the range and less well at the edges. The joint consensus statement of the European Atherosclerosis Society and the European Federation of Clinical Chemistry and Laboratory Medicine flags two situations specifically: triglycerides above roughly 2 mmol/L, and LDL-C below roughly 1.8 mmol/L.[5] Those are not exotic. The first describes a large slice of the metabolically unwell, and the second is exactly where modern treatment is trying to put people.

    The deeper problem is conceptual rather than technical. The amount of cholesterol packed into each particle varies between people. Two people with an identical LDL-C can be carrying very different numbers of particles — one of them a few cholesterol-rich ones, the other many cholesterol-depleted ones. If the particle is what gets trapped in the artery wall, the second person is at higher risk, and their LDL-C will not say so.[4]

    The number that survived the adjustment

    The cleanest test of this is a 2022 analysis from the TIMI Study Group. It combined 389,529 adults from UK Biobank who were in primary prevention and not taking any lipid-lowering therapy, followed for a median of 11.1 years, with 40,430 statin-treated patients who already had established atherosclerosis, drawn from the FOURIER and IMPROVE-IT trials.[1]

    Examined one at a time, everything predicted heart attack. Per one standard deviation, apoB came in at a hazard ratio of 1.38, non-HDL cholesterol at 1.36, and triglycerides at 1.16 — all three comfortably significant.

    Then the investigators put all of them into the same model. Only apoB was left standing: 1.27, with a 95% confidence interval of 1.15 to 1.40. Non-HDL cholesterol fell to 1.09 (0.97 to 1.21). Triglycerides fell to 1.00 (0.96 to 1.04) — precisely nothing. The same pattern held in the secondary-prevention cohort.

    One further result is worth sitting with. After adjusting for apoB, the ratio of triglycerides to LDL-C — the usual proxy for what kind of particles a person is carrying — had no association with myocardial infarction at all. In the authors' words, risk “was best captured by the number of apoB-containing lipoproteins, independent from lipid content or type of lipoprotein.”[1]

    Figure 1

    Each measure alone, and all three together

    ApoBOther lipid measures

    Each measure on its own

    All three in the same model

    Read down each panel, not across. Adjusted hazard ratios for incident myocardial infarction per one standard deviation, in 389,529 UK Biobank adults in primary prevention who were not taking lipid-lowering therapy, over a median 11.1 years. In panel A each lipid measure is modelled separately; in panel B all three are modelled together. Error bars are 95% confidence intervals; the vertical line marks a hazard ratio of 1.0. Only apoB's interval stays clear of it once the three are modelled together.

    Source: reference 1.

    Genetics had already said so

    A Mendelian randomisation study published three years earlier made the same point by a completely different route, across 63 cohort and case-control studies, 654,783 participants and 91,129 cases of coronary heart disease.[2]

    The investigators built two genetic scores. One lowered triglycerides through LPL; the other lowered LDL cholesterol through LDLR. Scaled to the same 10 mg/dL reduction in apoB, the two scores did almost nothing alike to the lipid panel: the LPL score lowered triglycerides by 69.9 mg/dL while nudging LDL-C slightly up, and the LDLR score lowered LDL-C by 14.2 mg/dL and triglycerides by 1.9.

    And yet the two produced near-identical reductions in coronary heart disease: odds ratios of 0.771 (0.741 to 0.802) and 0.773 (0.747 to 0.801). After adjusting for apoB, the associations of both triglycerides and LDL-C with coronary disease became null, while apoB's did not.

    This is about as close to a controlled experiment as observational genetics gets, and it is the strongest single argument that the particle count is the causal quantity. We walk through what this kind of evidence can and cannot establish in our piece on surrogate endpoints.

    Figure 2

    Two gene scores, two lipid panels, one answer

    LPL scoreLDLR score

    What each score did to the lipid panel

    What each score did to coronary heart disease

    Read down each panel, not across — the panels are in different units. Two genetic scores were scaled to the same 10 mg/dL lower apoB across 63 studies, 654,783 participants and 91,129 coronary heart disease cases. They moved almost entirely different lipids (panel A) and produced almost identical reductions in coronary heart disease (panel B). Error bars are 95% confidence intervals; the vertical line marks an odds ratio of 1.0.

    Source: reference 2.

    How often do the two actually disagree?

    This is the question that decides whether any of the above matters to you personally, and a 2024 review in Circulation gathered the evidence together.[4]

    Across the published discordance analyses, the proportion of adults with “discordantly high” apoB relative to LDL-C ranged from 5.3% to 24.9%. That spread reflects different definitions rather than different biology — there is no agreed cut-point for what counts as a mismatch. In one UK Biobank analysis the review summarises, 18% of participants free of cardiovascular disease and not taking statins had discordant apoB and LDL-C values. Within that discordant group, apoB was associated with cardiovascular events (hazard ratio 1.23 per standard deviation, 1.12 to 1.35) while LDL-C was not — not when measured directly, and not when calculated by either the Friedewald or the Martin/Hopkins formula. Nor was non-HDL cholesterol.

    Who ends up in that group is predictable: men, and people with metabolic syndrome, diabetes, hypertension, a higher body mass index, higher triglycerides or lower HDL cholesterol, and people who smoke.[4]

    It is not confined to high triglycerides, though. In a National Health and Nutrition Examination Survey analysis of 6,272 participants, 21.4% of people with triglycerides below 100 mg/dL still showed a low LDL-C-to-apoB ratio — cholesterol-depleted particles, in people whose triglycerides looked fine.[4]

    ApoB and non-HDL cholesterol agree more often than apoB and LDL-C do, but they still part company in somewhere between 8% and 23% of people.[4]

    What Canadian guidelines actually say

    The 2021 Canadian Cardiovascular Society dyslipidemia guideline is the current Canadian document, and it is more forward on apoB than most readers expect.[3] Its fourth recommendation is strong, graded high-quality evidence, and reads: “We recommend that for any patient with triglycerides > 1.5 mmol/L, non-HDL-C or ApoB be used instead of LDL-C … as the preferred lipid parameter for screening.”

    The guideline also makes a distinction that is easy to miss and important to hold onto. It states plainly that “to date, no clear target to which LDL-C or non HDL-C or ApoB levels should be lowered is clearly identified in RCTs.” What it gives instead are thresholds for intensifying treatment — a trigger to do more, not a finish line to cross. That difference between a threshold and a target is the subject of its own article on this site, on why a biomarker that predicts disease isn't automatically something to treat.

    The thresholds themselves: in someone with established cardiovascular disease already on the maximum tolerated statin dose, intensification is recommended if LDL-C remains at or above 1.8 mmol/L, non-HDL-C at or above 2.4 mmol/L, or apoB at or above 0.70 g/L. In intermediate-risk primary prevention, the corresponding statin-initiation numbers are LDL-C ≥ 3.5 mmol/L, apoB ≥ 1.05 g/L or non-HDL-C ≥ 4.2 mmol/L.[3]

    One practical note from the same guideline: non-fasting testing is recommended for most adults. LDL-C can read up to 10% lower after a meal, whereas non-HDL-C and apoB are essentially unchanged.[3] ApoB's immunoassay does pick up the apoB-48 on chylomicrons, but there are roughly nine LDL particles for every VLDL particle and roughly nine VLDL particles for every chylomicron, so the postprandial contribution is lost in the noise.[4]

    What changed in the United States this year

    In March 2026 the American College of Cardiology, the American Heart Association and nine other societies published a new dyslipidemia guideline that formally retires and replaces the 2018 blood cholesterol guideline.[8] The 2018 document had given apoB only limited endorsement, treating an elevated apoB as a risk-enhancing factor and little more.

    The new one goes further without going all the way. According to the College's own summary, apoB “may be used to assess any residual ASCVD risk and guide treatment” in people with cardiovascular-kidney-metabolic syndrome, type 2 diabetes, elevated triglycerides or known cardiovascular disease who have already reached their LDL-C and non-HDL-C goals — and in those groups, “apoB may be a more accurate risk marker than LDL-C.”[11] A review of the guideline's primary-prevention takeaways describes it as recommending selective rather than universal apoB testing, and as reintroducing treatment goals for LDL-C, non-HDL-C and apoB in selected cases, where the 2018 guideline had relied on percentage reduction alone.[12]

    European guidance went furthest and went there earliest. The 2019 European Society of Cardiology and European Atherosclerosis Society guidelines concluded that apoB is a more accurate measure of cardiovascular risk and a better guide to the adequacy of lipid lowering than either LDL-C or non-HDL-C.[7] That summary comes from Allan Sniderman of McGill, who has argued the apoB case for four decades — a position worth reading as a position, not as a neutral précis.

    Three guideline bodies, three different degrees of enthusiasm, all moving the same direction. And all three still keep LDL-C as the primary metric.

    The honest case against switching

    Here is the fact that cuts against everything above, and it belongs in the open.

    The EAS/EFLM consensus statement puts it bluntly: “There is no consistent evidence yet that non-HDL cholesterol-, apoB-, or LDL particle-targeted treatment reduces the number of cardiovascular events and healthcare-related costs than treatment targeted to LDL cholesterol.”[5]

    The trial that would settle this has never been run. Nobody has randomised a large group of people to apoB-guided management versus LDL-C-guided management and counted the events. Everything in this article concerns which number measures atherosclerotic risk more faithfully. Whether managing to it produces better outcomes is a separate claim, and it is unproven.

    What follows practically is less dramatic than the enthusiasm sometimes suggests. ApoB is a better instrument. The drugs are the same drugs. Nobody will treat you differently on the strength of an apoB result unless that result disagrees with your LDL-C — which is precisely why the test earns its place in the people where disagreement is likely, and earns very little in everyone else.

    So should you ask for it?

    There is a reasonably well-defined list. ApoB is superior to non-HDL cholesterol specifically in people with mild-to-moderate hypertriglyceridemia, diabetes, obesity or metabolic syndrome, or a very low LDL-C on treatment.[6] Canada's own guideline adds the simplest trigger of all: triglycerides above 1.5 mmol/L.[3]

    There are two diagnostic uses as well. ApoB is needed to tell hypertriglyceridemia driven by excess VLDL particles apart from hyperchylomicronemia, and it is necessary for identifying familial dysbetalipoproteinemia — type III hyperlipoproteinemia — a highly atherogenic and very treatable condition that a standard lipid panel alone will not surface.[4]

    And the other half of the honest answer: if your triglycerides are normal, you do not have diabetes or metabolic syndrome, and you are not on intensive lipid-lowering therapy, an apoB will almost certainly tell you what your LDL-C already told you. It is a sensible thing to measure once, alongside the once-in-a-lifetime lipoprotein(a) that Canadian expert guidance now recommends for every adult.[9] It is not something most people need repeated. You can read more about the markers themselves in our cardiovascular biomarker database and the metabolic biomarkers that sit alongside them.

    What it costs, and whether Ontario pays for it

    The 2021 Canadian guideline states, verbatim, that “ApoB is also available as an insured laboratory test in all provinces except Ontario.”[3]

    That sentence is now out of date. Ontario's Schedule of Benefits for Laboratory Services, in force since 1 April 2026, lists Apolipoprotein B as code L042 at $18.00.[10] For scale, the three measured components of a standard lipid panel are listed at $1.28 for total cholesterol, $2.00 for HDL cholesterol and $1.28 for triglycerides — so apoB costs the public system roughly four times the panel it supplements, and about two-thirds of what a lipoprotein(a) costs.

    Being listed in a schedule and being paid for in your particular case are not the same thing. Whether your apoB is covered still depends on your physician ordering it and your laboratory performing it, and practice varies between labs and between provinces. Ask before you assume. Where it is not covered, Canadian community laboratories generally quote the test in the tens of dollars.

    Figure 3

    The three numbers, side by side

    LDL-C, non-HDL cholesterol and apoB compared
    FeatureLDL-CNon-HDL-CApoB
    What it measuresCholesterol carried inside LDL particlesCholesterol in every atherogenic particle (total cholesterol minus HDL-C)The number of atherogenic particles — one apoB per particle
    How you get itUsually calculated (Friedewald, Martin/Hopkins or Sampson)Calculated: total cholesterol minus HDL-CMeasured directly by immunoassay
    Fasting required?Non-fasting is recommended for most adults; can read up to 10% low after a mealNoNo
    StandardisationDepends which formula the lab usesDepends on its inputsStandardised; IFCC/WHO reference material since 1994
    CCS threshold to intensify, established CVD≥ 1.8 mmol/L≥ 2.4 mmol/L≥ 0.70 g/L
    CCS threshold, intermediate-risk primary prevention≥ 3.5 mmol/L≥ 4.2 mmol/L≥ 1.05 g/L
    Ontario public fee (April 2026 schedule)No separate code — calculatedNo separate code — calculatedL042, $18.00
    Preferred whenMost people, most of the timeA non-fasting alternative when apoB is unavailableTriglycerides > 1.5 mmol/L, diabetes, obesity or metabolic syndrome, or a very low LDL-C on treatment

    Thresholds from the 2021 Canadian Cardiovascular Society dyslipidemia guideline; fees from Ontario's Schedule of Benefits for Laboratory Services. Accurate as of November 2026. Sources: references 3, 4, 5 and 10.

    The practical bottom line

    • ApoB counts the particles that cause atherosclerosis. LDL-C estimates the cholesterol inside them, and in most labs it is calculated rather than measured.
    • Where the two disagree — in roughly 5% to 25% of adults, depending on the definition — the particle count is the number that tracks events.
    • Disagreement concentrates in men, and in people with high triglycerides, diabetes, obesity or metabolic syndrome. It is not confined to them.
    • Canadian guidelines have recommended apoB or non-HDL-C instead of LDL-C for anyone with triglycerides above 1.5 mmol/L since 2012.
    • No trial has randomised apoB-guided treatment against LDL-C-guided treatment. ApoB is a better measurement, not a different medicine.
    • Ontario's current laboratory schedule lists apoB at $18.00, contrary to the 2021 guideline's statement that Ontario was the exception. Coverage in practice still varies.
    • None of this is medical advice. Interpret any lipid result with the clinician who ordered it.

    Frequently asked questions

    Is apoB the same thing as LDL cholesterol?

    No. They measure different quantities on the same particles. LDL-C is the total mass of cholesterol carried by your LDL particles; apoB is a count of all the particles capable of causing atherosclerosis, including VLDL, IDL and their remnants as well as LDL. In most people the two track each other closely. In a meaningful minority they do not.

    Do I need to fast before an apoB test?

    No. ApoB is essentially unchanged by a recent meal, which is one of its practical advantages. Canadian guidelines recommend non-fasting lipid testing for most adults in any case, with the exception of people already known to have triglycerides above 4.5 mmol/L.

    What is a normal apoB level?

    There is no single agreed answer, and that is part of why the test has been slow to spread. Different guideline bodies have proposed different apoB numbers for the same LDL-C target, because there is no universally accepted conversion between the two. Canada's guideline sets thresholds for acting rather than targets to reach: 0.70 g/L in established cardiovascular disease on maximal statin therapy, and 1.05 g/L for starting a statin in intermediate-risk primary prevention. Interpret your own number with the clinician who ordered it.

    If I get an apoB, should I stop tracking LDL-C?

    No. Every major guideline — Canadian, American and European — still keeps LDL-C as the primary metric, and almost all of the treatment evidence is expressed in LDL-C terms. The argument for apoB is that it measures the causal quantity more faithfully, not that LDL-C is uninformative.

    Is apoB covered by provincial health insurance in Canada?

    Mostly, and increasingly. The 2021 Canadian guideline named Ontario as the one exception, but Ontario's Schedule of Benefits for Laboratory Services has since listed apolipoprotein B as an insured test (code L042, $18.00), so that statement no longer holds. Whether a given test is paid for still depends on the ordering physician and the laboratory, so confirm rather than assume.

    My LDL-C is normal but my apoB is high. Does that mean I need medication?

    Not by itself. A discordantly high apoB means your risk is probably higher than your LDL-C suggests, and that is useful information — but a treatment decision rests on your overall cardiovascular risk, not on a single number. It is a reason for a careful conversation, not an automatic prescription.

    References

    1. [1]Marston NA, Giugliano RP, Melloni GEM, et al. Association of Apolipoprotein B-Containing Lipoproteins and Risk of Myocardial Infarction in Individuals With and Without Atherosclerosis. JAMA Cardiol. 2022;7(3):250–256.
    2. [2]Ference BA, Kastelein JJP, Ray KK, et al. Association of Triglyceride-Lowering LPL Variants and LDL-C-Lowering LDLR Variants With Risk of Coronary Heart Disease. JAMA. 2019;321(4):364–373.
    3. [3]Pearson GJ, Thanassoulis G, Anderson TJ, et al. 2021 Canadian Cardiovascular Society Guidelines for the Management of Dyslipidemia for the Prevention of Cardiovascular Disease in Adults. Can J Cardiol. 2021;37(8):1129–1150.
    4. [4]De Oliveira-Gomes D, Joshi PH, Peterson ED, Rohatgi A, Khera A, Navar AM. Apolipoprotein B: Bridging the Gap Between Evidence and Clinical Practice. Circulation. 2024;150(1):62–79.
    5. [5]Langlois MR, Chapman MJ, Cobbaert C, et al. Quantifying Atherogenic Lipoproteins: Current and Future Challenges in the Era of Personalized Medicine and Very Low Concentrations of LDL Cholesterol. A Consensus Statement from EAS and EFLM. Clin Chem. 2018;64(7):1006–1033.
    6. [6]Langlois MR, Sniderman AD. Non-HDL Cholesterol or apoB: Which to Prefer as a Target for the Prevention of Atherosclerotic Cardiovascular Disease? Curr Cardiol Rep. 2020;22(8):67.
    7. [7]Sniderman A, Langlois M, Cobbaert C. Update on apolipoprotein B. Curr Opin Lipidol. 2021;32(4):226–230.
    8. [8]Blumenthal RS, Morris PB, Gaudino M, et al. 2026 ACC/AHA/AACVPR/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline on the Management of Dyslipidemia. Circulation. 2026;153(17):e1154–e1276.
    9. [9]Thanassoulis G, Anand S, Arsenault BJ, et al. Clinical Care for Patients With Elevated Lipoprotein(a) Level in Canada: A Practical Framework for Screening, Risk Assessment, and Management. Can J Cardiol. 2026;42(4):867–884.
    10. [10]Ontario Ministry of Health. Schedule of Benefits for Laboratory Services, effective 1 April 2026.
    11. [11]American College of Cardiology. ACC, AHA Release New Clinical Guideline For Managing Dyslipidemia. 13 March 2026.
    12. [12]Abramov D, Minhas AMK, Shapiro MD, et al. Primary Prevention of Dyslipidemia: 10 Practice-Changing Takeaways from the 2026 ACC/AHA Multisociety Guideline. Curr Atheroscler Rep. 2026;28(1).
    Disclaimer: This article is for educational purposes only and is not medical advice. It does not establish a physician–patient relationship. Decisions about lipid testing, interpreting your results and any treatment should be made with your own physician.