ApoB: The Real Cardiovascular Number

Two people can walk out of the same clinic with the same LDL cholesterol and very different cardiovascular risk. That is not a measurement error, and it is not an exotic edge case — it is a routine consequence of measuring the cargo instead of counting the vehicles. Atherosclerosis is caused by lipoprotein particles getting into the artery wall and staying there. The standard lipid panel does not count particles. ApoB does, and it costs about as much as the test you are already having.

What ApoB actually is

Apolipoprotein B is a large structural protein wrapped around every atherogenic lipoprotein particle in circulation. There is exactly one ApoB molecule per particle, it is added when the particle is assembled, and it stays with that particle for its entire life. It is not an estimate or a ratio. Measuring ApoB concentration is counting particles.

The particles carrying ApoB are:

• LDL, low-density lipoprotein
• VLDL, very-low-density lipoprotein
• IDL, intermediate-density lipoprotein
• Lp(a), lipoprotein little a
• Chylomicron remnants

HDL carries a different apolipoprotein, ApoA1, and does not appear in an ApoB measurement. So one number captures the entire atherogenic population, including the remnant particles that a standard panel handles poorly and the Lp(a) it ignores entirely.

Why particle count is the causal variable

The standard lipid panel measures the cholesterol carried by these particles. ApoB measures the particles themselves (Sniderman et al., JAMA Cardiology 2019). The distinction matters because of how atherosclerosis physically begins.

An ApoB-containing particle crosses the endothelium into the arterial intima. Once inside, it can bind to proteoglycans in the wall and be retained. Retained particles are oxidised, taken up by macrophages, and the resulting foam cells are the earliest lesion. Everything downstream — plaque growth, inflammation, eventual rupture — follows from particles entering and being retained.

Entry is driven largely by the concentration gradient across the endothelium, and that gradient is set by how many particles are on the arterial side, not by how much cholesterol each one happens to contain. A large cholesterol-rich particle and a small cholesterol-poor one each count once at the wall. Genetic evidence supports the causal reading rather than merely a correlational one: variants that lower ApoB lower cardiovascular risk in proportion, and the effect tracks particle number rather than cholesterol content (Richardson et al., PLoS Medicine 2020; Marston et al.).

Discordance, and who it happens to

Discordance is the situation where LDL-C and ApoB disagree about risk. It is not rare, and it is not random — it clusters in exactly the population most likely to be reassured by a normal-looking panel.

People with insulin resistance, type 2 diabetes or metabolic syndrome tend to produce small, dense LDL particles. Each carries less cholesterol than a large buoyant particle. To carry the same total cholesterol you need more particles, so LDL-C can read normal while ApoB is elevated and true risk is higher than the panel suggests (Witt et al., Diabetes Obes Metab 2025).

The reverse also occurs. Someone with large, cholesterol-rich particles can have a mildly elevated LDL-C with a normal ApoB, and their risk is lower than the number implies. Both directions matter, but the first is the clinically dangerous one because it produces false reassurance in the group that needs attention most.

A useful shortcut: high triglycerides with low HDL is the signature of the small-dense pattern. If that combination is on your panel, LDL-C is probably underselling your particle count. The triglyceride-to-HDL ratio and particle number versus concentration both cover this from different angles, and HOMA-IR covers measuring the insulin resistance underneath it.

Reading your own number

ApoB is reported in mg/dL and the interpretation is straightforward, because unlike LDL-C there is no derived calculation to distrust.

• Population average sits around 95 mg/dL — which is a description of a population with a great deal of cardiovascular disease in it, not a target
• Under 80 mg/dL for primary prevention
• Under 70 mg/dL at moderate risk
• Under 60 mg/dL in high-risk patients, meaning established cardiovascular disease or diabetes
• Under 50 mg/dL where prevention is being pursued aggressively

The relationship between particle count and risk is continuous and has no threshold below which it flattens out. Which of these targets applies to an individual is a clinical judgement about overall risk, not something to be selected from a list — and lower is not automatically better for every person once other considerations enter.

What ApoB does not tell you

ApoB is the best single lipid number available. It is not the only number that matters, and treating it as sufficient produces its own blind spots.

It does not distinguish Lp(a) from ordinary LDL. Lp(a) is genetically determined, largely unresponsive to lifestyle, and carries risk beyond its particle count — so someone with high Lp(a) hidden inside an acceptable ApoB is being under-estimated. It should be measured once in a lifetime; Lp(a), the genetic risk covers why.

It also says nothing about the vessel wall's willingness to retain particles, which is where inflammation and endothelial function come in — hs-CRP is the usual first look at that. And it is a snapshot of exposure now, whereas plaque reflects cumulative exposure over decades. A good ApoB at 50 does not undo thirty years of a bad one, which is the argument for measuring it early rather than at the age when the conversation usually starts.

What moves it

Roughly in order of the size of effect available:

• PCSK9 inhibitors — the largest reductions of any available agent
• Statins — the best-evidenced, with outcome data across decades
• Ezetimibe — modest alone, useful added to a statin
• Reducing visceral fat — acts on the small-dense pattern at its source rather than on the number downstream
• Reducing refined carbohydrate and alcohol — primarily through triglycerides and remnant particles
• Regular exercise — modest direct effect, larger indirect effect via insulin sensitivity
• GLP-1 therapy — a modest reduction, largely secondary to weight and metabolic change

The medication rows are physician decisions and belong in a physician conversation. The lifestyle rows are worth understanding as acting on a different part of the problem: they change the kind of particles being produced, whereas the drugs mostly change how many are cleared. In someone with insulin resistance, improving insulin sensitivity often shifts the whole lipid pattern in a way that no single lipid number fully captures.

Practical points on the test itself

ApoB does not require fasting, which removes the main reason lipid panels get postponed. It is measured directly by immunoassay rather than calculated, so it does not inherit the estimation error that affects calculated LDL-C at high triglycerides. It is well standardised across laboratories, which makes serial values genuinely comparable over years. And it is inexpensive. The main obstacle is that it is often not on the default panel, which means asking for it specifically.

Bottom line

ApoB counts the particles that cause atherosclerosis, one molecule per particle, across every atherogenic class including the remnants and Lp(a) that a standard panel handles badly. It predicts events better than LDL-C, and the gap between them is widest in people with insulin resistance — the group most likely to be falsely reassured. Under 80 mg/dL is the usual primary-prevention target, lower with higher risk, and the relationship is continuous rather than threshold-based. It is cheap, non-fasting, directly measured and comparable over time. It should be on the panel of any adult who wants a real answer about cardiovascular risk, and what any given result means for you is a conversation with a physician who has your full picture in front of them.

Educational content, not medical advice. Laboratory interpretation and any treatment decision are made by a licensed physician after individual evaluation. Individual results vary.