TL;DR: Apolipoprotein B, usually shortened to apoB, is a protein that sits on the surface of every lipoprotein particle capable of depositing cholesterol in an artery wall. Exactly one apoB molecule per particle, which means an apoB measurement is a direct count of atherogenic particles in your blood. That is a meaningfully different thing from LDL cholesterol, which measures how much cholesterol those particles are carrying. Most of the time the two track each other closely and either works fine. The problem is the minority of cases where they diverge, and those cases are disproportionately common in people with insulin resistance, high triglycerides, obesity, or metabolic syndrome. In those people, LDL cholesterol can look acceptable while particle count is high, and the particle count is what predicts events. Mendelian randomisation work and large cohort analyses have converged on the view that apoB is the better single measure of atherogenic lipoprotein burden. Several major guideline bodies now recognise apoB as a risk-enhancing or refinement measurement, though most still anchor primary decisions to LDL cholesterol. ApoB is cheap, standardised, and does not require fasting. If you are getting lipids checked and can add it, it is worth having.
What is apoB and why does it matter more than LDL cholesterol?
ApoB is a structural protein. Every lipoprotein particle that can cross into an artery wall and lodge there carries exactly one apoB molecule: LDL, VLDL, IDL, chylomicron remnants, and lipoprotein(a). HDL particles carry apoA-I instead, and HDL is not part of this count. So when a laboratory reports your apoB in milligrams per decilitre, it is functionally a headcount of the particles that can cause atherosclerosis.
Why does the count matter more than the cargo? The current understanding of atherosclerosis is that the disease starts when an apoB-containing particle crosses the endothelium and gets retained in the artery wall. It is the particle that gets stuck. How much cholesterol that particle happened to be carrying at the time is secondary. Two people can have identical LDL cholesterol values with quite different particle numbers, because particles vary in how much cholesterol they carry. The person with more particles, each carrying less cholesterol, has more chances for retention and, on the evidence, more risk.
This is not a fringe position. Mendelian randomisation studies, which use genetic variants as natural experiments to get around confounding, have supported the interpretation that the causal relationship with coronary disease tracks apoB particle number rather than the cholesterol content per particle. That kind of genetic evidence is one of the more persuasive tools available in cardiovascular epidemiology, because your genotype is fixed at conception and is not confounded by lifestyle in the way an observational cohort is.
When do apoB and LDL cholesterol disagree, and why does it matter?
They disagree most in people with metabolic dysfunction, and that is exactly the group where getting it right matters most.
The pattern is called discordance. In a person with insulin resistance, raised triglycerides, and central adiposity, LDL particles tend to become smaller and depleted of cholesterol. You end up with many particles each carrying a modest cholesterol load. Total LDL cholesterol can therefore land in a range that looks reassuring while particle count, and thus apoB, is high. That person's LDL cholesterol result underestimates their risk. Studies examining discordant cases have generally found that when the two markers disagree, cardiovascular events track the apoB value, not the LDL cholesterol value.
The reverse discordance also happens: some people have relatively few, large, cholesterol-rich particles, giving a high LDL cholesterol alongside a lower apoB. Their risk tends to track the lower apoB.
Roughly speaking, discordance affects a minority of people, but it is not rare, and it clusters in exactly the population that has expanded most in recent decades. If you have a raised waist circumference, high triglycerides, low HDL, or a diagnosis of prediabetes or type 2 diabetes, your LDL cholesterol is more likely to mislead you. The dietary pattern evidence in our DASH diet and lifespan piece is relevant here, because the metabolic picture and the lipid picture tend to move together.
There is a related measurement, non-HDL cholesterol, which you get by subtracting HDL cholesterol from total cholesterol. It is a good, free approximation of atherogenic burden and it captures much of what apoB captures without an extra test. If apoB is unavailable or unaffordable, non-HDL cholesterol is a reasonable substitute and is already on your standard lipid panel. It is not quite as good as apoB, but the gap is smaller than enthusiasts sometimes imply.
What does the evidence actually show about apoB and mortality?
Let us be precise about what is established and what is not.
Established: apoB is causally related to atherosclerotic cardiovascular disease. This comes from multiple lines of evidence that converge. Genetic conditions that raise apoB-containing particles from birth, such as familial hypercholesterolaemia, cause premature coronary disease. Mendelian randomisation across many variants shows a dose-dependent relationship with coronary risk. And randomised trials of therapies that lower apoB-containing particles, most prominently the statin trials conducted through the 1990s and 2000s, reduce cardiovascular events. The Cholesterol Treatment Trialists' Collaboration, which pooled individual participant data from the major statin trials, established this repeatedly across different populations and baseline risk levels.
Also established: in head-to-head comparisons within large cohorts, apoB has generally been at least as good as, and often modestly better than, LDL cholesterol at predicting cardiovascular events. Analyses in UK Biobank and in earlier cohort work have supported this.
Less established, and worth saying plainly: the incremental predictive gain from adding apoB to a well-constructed risk model containing standard lipids is real but modest at the population level. This is the reason guideline committees have been slower to adopt apoB as a primary target than online commentary would suggest they should be. From a public health perspective, replacing a cheap universal test with a slightly better one across an entire population is a different calculation from an individual wanting the most informative number about themselves. Both perspectives are defensible.
Not established: that lowering apoB specifically, as opposed to lowering atherogenic particles generally, produces benefits beyond what lipid-lowering trials already show. The trials lowered particles. ApoB is how we count them. There is no separate apoB intervention.
And a caveat on the phrase "predicts mortality". ApoB predicts cardiovascular events well. All-cause mortality is a broader target that includes cancer, dementia, accidents, and much else that lipids do not touch. In older people especially, the association between cholesterol markers and all-cause mortality weakens and in some cohorts inverts, partly because illness lowers cholesterol. Do not expect one lipid marker to summarise your total mortality risk. Our guide to the science of biological ageing discusses why composite measures beat single markers for that purpose.
What is a good apoB level?
There is no single answer, because the right target depends on your overall risk, and that is a clinical judgement rather than a number you can look up.
Some orientation is still useful. Laboratory reference ranges for apoB typically span roughly 40 to 125 mg/dL, but a reference range describes what is common in the tested population, not what is optimal. In a population where atherosclerosis is near-universal by later life, "normal" is not a target worth aspiring to. Population averages in Western countries sit around the 90 to 110 mg/dL region, and that population has a great deal of coronary disease.
European lipid guidance from the ESC and EAS has published apoB goals that mirror their LDL cholesterol goals across risk categories, with progressively lower targets as risk rises: broadly, under about 100 mg/dL for low risk, under about 80 for high risk, and under about 65 for very high risk, with even lower thresholds discussed for those with recurrent events. American guidance from the AHA and ACC has treated apoB above roughly 130 mg/dL as a risk-enhancing factor that shifts a borderline treatment decision, rather than as a primary target.
The unhelpful truth is that a "good" apoB for a 30-year-old with no other risk factors is different from a "good" apoB for a 62-year-old with type 2 diabetes and a coronary calcium score in the hundreds. Lower is better in terms of atherosclerosis risk, and there is no known threshold below which apoB stops mattering, but how aggressively to pursue lower is a conversation to have with a clinician who knows your history, not a number to chase from a blog post.
One thing worth flagging: if your apoB is very high, particularly if it has been high since youth or if there is a family history of early heart attacks, familial hypercholesterolaemia is worth ruling out. It is more common than most people assume, affecting roughly one in 250 to 300 people, and it is substantially underdiagnosed. It is also very treatable once identified.
How do you get apoB tested?
Ask for it. In most healthcare systems apoB is a standard, inexpensive, well-standardised assay, and the main obstacle is that nobody orders it by default.
Practical points. Fasting is not required. This is a genuine advantage over calculated LDL cholesterol, which uses a formula that becomes unreliable when triglycerides are high or when you have eaten recently. ApoB is measured directly by immunoassay and is not affected by a recent meal in the way calculated LDL is.
Timing and consistency matter. Lipid values fluctuate with acute illness, recent inflammation, and substantial weight change. A single result during or shortly after an infection can mislead. If you are tracking a change after a dietary or medication adjustment, allow enough time, generally six to twelve weeks, and try to test under comparable conditions.
Lipoprotein(a) is a separate matter but deserves a mention alongside apoB. Lp(a) is an apoB-containing particle whose concentration is largely genetically determined and does not respond meaningfully to diet or lifestyle. It is included in your apoB count but a very high Lp(a) can add risk that the apoB number alone does not fully convey. Most guidance now suggests measuring Lp(a) at least once in adulthood, since the result is essentially lifelong and changes how the rest of the risk picture is interpreted.
What actually lowers apoB?
Two categories: what you do, and what you take.
On the lifestyle side, the levers are known and their effect sizes are honest but limited. Reducing saturated fat intake and replacing it with unsaturated fat reliably lowers apoB-containing particles, and this is one of the more replicated findings in nutrition science, established through decades of controlled feeding studies. Our pieces on olive oil and lifespan and red meat and mortality cover the relevant food-level evidence. Soluble fibre, from oats, legumes, and psyllium, produces modest reductions. Losing excess weight, particularly visceral fat, improves the whole discordance picture by shifting particles back towards fewer and larger. Regular aerobic exercise has a small direct effect on apoB and a larger effect on the metabolic context, as covered in zone 2 cardio and longevity. Dietary cholesterol itself, the thing everyone worried about for forty years, has a smaller effect on blood lipids than saturated fat does in most people.
Realistically, a determined and well-executed diet and exercise change might lower apoB by something in the region of 10 to 20 percent. For many people that is enough. For someone with familial hypercholesterolaemia or established coronary disease, it is not, and that is not a moral failing. Genetics set a large part of the starting point.
On the medication side, statins remain the best-evidenced drug class in cardiovascular medicine, with mortality and event reductions demonstrated across an enormous trial literature. Ezetimibe, PCSK9 inhibitors, and bempedoic acid all lower apoB-containing particles and all have outcome trial evidence behind them. Whether any of these is appropriate for you is a clinical decision that depends on your absolute risk, not on a number in isolation.
Frequently Asked Questions
Should I ask my doctor for an apoB test? It is a reasonable request, particularly if you have raised triglycerides, a large waist, prediabetes or diabetes, a family history of early heart disease, or a borderline risk estimate where the decision about treatment could go either way. In those situations apoB is most likely to change the picture. If your standard lipid panel is unremarkable and you have no metabolic risk factors, apoB will most likely tell you roughly what your LDL cholesterol already told you.
Is apoB better than a coronary calcium score? They answer different questions and are complementary rather than competing. ApoB measures the causal exposure, how many atherogenic particles are circulating right now. A coronary artery calcium score measures the accumulated result, how much calcified plaque has already formed over decades. A young person can have high apoB and a calcium score of zero simply because the damage has not accumulated yet, and that zero should not be reassuring. Clinicians often use both.
Does a high apoB with normal LDL cholesterol mean something is wrong? It means your particles are on average carrying less cholesterol each, which is characteristic of insulin resistance and metabolic syndrome. It is a real signal worth taking to a clinician. It usually points towards addressing the metabolic picture, weight, glycaemic control, triglycerides, alongside considering the lipid picture directly. It is precisely the scenario where relying on LDL cholesterol alone would have underestimated your risk.
Do I need to fast before an apoB test? No. ApoB is measured directly and is not meaningfully affected by a recent meal in the way that calculated LDL cholesterol is. This is one of its practical advantages. If your blood draw includes other tests such as glucose, those may have their own fasting requirements, so follow the instructions you are given for the panel as a whole.
Can diet alone get my apoB into a good range? Sometimes, depending on where you start and what is driving the number. Replacing saturated fat with unsaturated fat, increasing soluble fibre, and losing visceral fat can produce reductions in the region of 10 to 20 percent, which is enough for many people with modest elevations. If your apoB is very high, or driven by a genetic condition such as familial hypercholesterolaemia, diet will not be sufficient on its own and pretending otherwise delays effective treatment.
Medical Disclaimer
This article is for general information only and does not constitute medical advice, diagnosis, or treatment. It is not a substitute for consultation with a qualified healthcare professional who knows your medical history. Biomarker results, including apoB and other lipid measurements, cannot be interpreted meaningfully in isolation and require clinical context to be useful. Do not start, stop, or change any medication or supplement on the basis of this content, and do not act on a laboratory result without discussing it with a clinician who can interpret it alongside your full risk picture. This is particularly important if you have existing cardiovascular disease, diabetes, kidney disease, a family history of early heart disease, or if you are pregnant or breastfeeding. If you have symptoms that concern you, especially chest pain or breathlessness, seek medical attention promptly rather than relying on information found online.