Is ApoB better than LDL cholesterol for predicting heart disease?

Is ApoB better than LDL cholesterol? In studies that put the two numbers against each other, ApoB is the more accurate predictor of heart attacks and strokes, and the gap opens widest in people whose two results disagree.

Apolipoprotein B, or ApoB, is a protein carried by every cholesterol particle that can lodge in an artery wall, one copy per particle, so measuring it counts the particles instead of weighing what is inside them. A systematic review of 15 studies covering 593,354 adults found ApoB more accurate than LDL cholesterol in all nine studies that compared the two directly [4]. Against non-HDL cholesterol, which adds up the cholesterol in those same particles rather than counting them, ApoB was the better marker in seven studies, equivalent in one and worse in one [4].

The two numbers usually move together, so what a large cohort can add is what happens to the people in whom they come apart. Denmark's Copenhagen General Population Study followed 53,484 women and 41,624 men who were not taking statins, the prescription drugs that lower cholesterol, for a median of 9.6 years, meaning half were tracked for longer than that [5]. It recorded 2,048 heart attacks and 4,282 cases of atherosclerotic cardiovascular disease, the umbrella term for heart attacks, strokes and the clogged arteries behind them [5]. The researchers measured excess ApoB, the amount by which someone's particle count runs above what their LDL cholesterol alone would predict, and risk climbed with it step by step.

In men, set against those whose excess ApoB was under 11 mg/dL (milligrams per decilitre, the unit printed on a cholesterol report), the risk of that artery disease was:

• about 14% higher with 11 to 25 mg/dL of excess ApoB
• about 41% higher with 26 to 45
• about 41% higher with 46 to 100
• about 52% higher above 100 [5]

Women climbed the same way from the second band upwards, about 30%, then 34%, then 75% higher, though the smallest excess did not clearly raise their risk [5]. The pattern held across the whole range of LDL cholesterol, including in people whose LDL number read as reassuring [5].

The National Lipid Association reached the same conclusion in its 2024 expert consensus. ApoB reflects the total number of artery-clogging particles in the blood more accurately than LDL cholesterol does, before lipid-lowering treatment and during it, and where the two disagree in one person, risk generally follows ApoB [10]. That document also names the practical obstacle: access to the test and reimbursement for it both need to improve [10].

Counting particles treats every particle as equally dangerous, and a couple of them are not. Using inherited gene variants as a natural experiment in UK Biobank, a large British health research dataset, researchers weighed lipoprotein(a), or Lp(a), an inherited particle type that also carries one ApoB each, against ordinary LDL. Per particle, Lp(a) was roughly six times more likely to cause coronary heart disease, the narrowing of the arteries that feed the heart [3].

Remnant cholesterol slips through from the other direction. It is the cholesterol carried in particles rich in triglycerides, the main type of fat in the blood, and in 17,532 adults who started with no artery disease and were followed a median of 18.7 years, it predicted disease even after LDL cholesterol and ApoB had both been accounted for [1]. People with high remnant cholesterol but low LDL cholesterol were about 21% more likely to have an event than people whose two numbers agreed [1].

A 2026 analysis tried to fold all three particle types into one number, weighting ApoB by how dangerous each type is per particle [13]. Ranking people by that weighted figure rather than by plain ApoB shifted 52% of them by ten percentiles or more, that is, at least ten places in every hundred when everyone is lined up from lowest risk to highest. About a third of those in the top fifth on the weighted measure, a group in which 5.4% went on to have a coronary event, had been rated lower risk by ApoB alone; people in the top fifth by plain ApoB who carried little Lp(a) and few triglyceride-rich particles had an event rate of 3.9% [13].

Cholesterol does not travel through the blood on its own. It rides inside particles, and each of those particles, whether LDL, triglyceride-rich or Lp(a), carries exactly one ApoB molecule [6] [7]. One particle, one ApoB. That is what turns the measurement into a count rather than an estimate.

How much cholesterol is packed into each particle varies from person to person [6]. Two people can therefore report an identical LDL cholesterol while one of them carries many small, cholesterol-poor particles and the other fewer large, cholesterol-rich ones. What ends up deposited in the artery wall depends on how many particles get in and stay there [6].

Getting stuck is a physical process. LDL crosses the vessel lining and is held in the layer underneath by binding to proteoglycans, sugar-and-protein molecules in the wall, at particular sites on the ApoB protein itself [7]. Trapped particles are then chemically modified, which releases lipid products that set off inflammation. Immune cells swallow the modified particles, become fat-stuffed foam cells and eventually die of the load, and the dead cells and cholesterol crystals pile up into the core of a plaque, the fatty deposit that narrows an artery [7].

Two effects pull against each other here. Small, cholesterol-poor particles bind those wall molecules more readily and so are trapped more often, while a cholesterol-rich particle deposits more cholesterol once it is trapped [6]. The net result, that review argues, is that ApoB particles are about equally dangerous, with Lp(a) and the abnormal remnants seen in one inherited disorder as the exceptions [6].

Almost all of this is observational. People had their blood measured and were then followed for years; nobody was assigned at random to anything [1] [5]. No trial has managed one group of people by their ApoB and another by their LDL cholesterol and then counted the heart attacks in each, so ApoB is far better established as a predictor of risk than as a target to steer treatment by.

Where the evidence does not reach:

• the Copenhagen participants were not taking statins, so those results say nothing about how the two numbers compare in someone already on lipid-lowering treatment [5]
• three of the largest analyses here rest on two national datasets, UK Biobank and the Copenhagen General Population Study [3] [5] [13]
• in women, the smallest excess of ApoB was not clearly linked to higher risk, and the association only became solid in the bands above it [5]

The Lp(a) comparison is a genetic model of risk per particle rather than a measurement of what happens when Lp(a) is lowered [3]. The National Lipid Association statement is an expert consensus built on the evidence, not a trial of its own [10]. The weighted ApoB figure has been tested for how well it sorts people by risk, not for whether acting on it changes what happens to them [13]. Which lipid tests to run, and what any result means for one person, is a decision made with a doctor.

ApoB is a blood test, run on the same sample a standard cholesterol panel comes from. Longevity clinics in this directory list comprehensive biomarker panels that add cholesterol particle counts to what a routine check-up measures, and a first visit is normally a blood draw followed by an appointment where a clinician goes through the numbers with you. Discordance is the reason to look: the person whose LDL cholesterol reads reassuringly while their particle count runs high is the one a standard panel misses [5] [10]. Whether a high result should change anything is a separate question, and that answer comes from a doctor rather than from the test.