LDL or ApoB: Which Result Better Predicts Cardiovascular Risk?

Cholesterol does not circulate in the blood on its own but in lipoproteins. LDL-C is the amount of cholesterol contained in LDL particles, calculated by a formula or measured directly. Every VLDL, IDL, LDL and lipoprotein(a) particle, however, carries exactly one ApoB molecule, so the ApoB concentration in practice corresponds to the number of all atherogenic particles, not only LDL. Two people with an identical LDL-C can have a different number of particles: one a few large ones rich in cholesterol, the other many small ones poor in cholesterol.

The intuition is this: arteries do not 'see' the mass of cholesterol, only the particles that get trapped in them. If there are many particles and each carries little cholesterol, LDL-C can look innocent while the particle count is high. That is why ApoB is sometimes called a 'particle counter'. It is worth remembering, however, that this is a theoretical model supported by data, not proof that LDL-C is useless.

Discordance means that LDL-C and ApoB 'point to different places on the risk scale'. In Mora's study about one in five people had an ApoB discordant with LDL-C relative to the median. So it is not a margin, but it is not the majority either: in about 80% of people the two measures coincide, which explains why LDL-C served well in studies and practice for decades.

This analysis is one of the most frequently cited arguments for ApoB: when ApoB, non-HDL-C and triglycerides were put into one model, only ApoB retained an association with heart attack. The observational nature, however, means the result describes a statistical association and does not by itself prove that changing ApoB will change risk. The result does match the direction of the genetic data, which we discuss below.

This is one of the strongest studies supporting ApoB, because it uses genetics as a natural randomization and is not susceptible to the usual observational confounding. It shows that the reduced risk is explained more by the fall in the number of ApoB particles than by whether cholesterol or triglycerides fell. In practice this translates into a rule: the lower the ApoB (sustained for longer), the lower the risk, regardless of how the fall was achieved.

An honest presentation of the topic requires showing data that do not fit the simple narrative that 'ApoB wins'. First, a 2022 Mendelian randomization analysis (Helgadottir et al.) used 235 genetic variants and found that 35% of them had discordant effects on non-HDL-C and ApoB. Adding non-HDL-C to a model that already contained ApoB significantly improved the prediction of coronary disease (p = 3.9 × 10^-5), whereas adding ApoB to a model with non-HDL-C did not (p = 0.69). The authors conclude that the atherogenic mechanism of ApoB particles is more strongly tied to non-HDL-C than to ApoB particle concentration. In other words, the amount of cholesterol in the particles may also matter.

Second, in an analysis of the MESA cohort (Cao et al. 2020) in adults over 45 not taking statins, higher ApoB was associated with the presence, new appearance and progression of coronary artery calcium, but ApoB gave only modest additional predictive value compared with LDL-C and non-HDL-C. Discordantly high ApoB was associated with CAC progression, but the associations with discordance were inconsistent, and the authors stressed the need for further research.

The 2019 European dyslipidemia guidelines (Mach et al., ESC/EAS) take ApoB into account. They recommend measuring it as part of routine lipid analysis, together with non-HDL-C, in people with raised triglycerides, diabetes and a very low achieved LDL-C. They also point out that in people with obesity LDL-C may underestimate the number of atherogenic particles. The main treatment goal is still expressed in LDL-C, and ApoB serves as a secondary goal set by inference rather than direct randomized trials.

Pay attention to units: in Poland laboratories report ApoB in mg/dL or g/L (1 g/L is 100 mg/dL). When comparing a result with the goals, always check the unit and the laboratory's reference range. The goals depend on an individual risk category assessed by a doctor, not on the test result alone.

The ApoB test is done on serum or plasma and does not require fasting, which is also supported by the pooled ERFC analysis, in which non-fasting lipids showed at least as strong an association with risk as fasting ones. It is available in most large commercial laboratories in Poland, while cost and availability under the public health fund depend on the facility and indication, so check the laboratory's price list or ask your doctor. The result is best interpreted together with the full lipid panel, Lp(a) (once in a lifetime) and a risk assessment, not as an isolated number.

What to do about a high ApoB? The conclusions are the same as for a high LDL-C: address the causes (insulin resistance, weight, diet, alcohol, thyroid) and, where needed, consider drug treatment with your doctor. We write about how to lower LDL naturally in our article on ten ways, and about statins in a separate knowledge-base entry.

The scenarios below are simplified, hypothetical illustrations, not descriptions of specific patients or treatment recommendations. They show how ApoB can change the picture of risk when LDL-C alone is misleading.

Scenario one: a 45-year-old man with a waist above 100 cm, triglycerides of about 250 mg/dL and an LDL-C 'almost normal'. Insulin resistance favors the formation of numerous small particles rich in triglycerides, so ApoB may be clearly higher than LDL-C would suggest. Mora et al. showed that with such discordance, risk based on LDL-C alone can be underestimated. The ApoB result then shifts attention from 'cholesterol' to the cause, which is metabolism: weight, exercise, alcohol, simple sugars and glycemia.

Scenario two: a person on intensive statin and ezetimibe therapy with an LDL-C below 55 mg/dL. At such low values the calculated LDL-C is imprecise, and ApoB makes it possible to check whether the number of atherogenic particles has also fallen to the intended goal (below 65 mg/dL for very high risk in the ESC/EAS guidelines). If ApoB is still above the goal, the conversation with the doctor may concern intensifying treatment or looking for causes such as a high Lp(a).

Scenario three: a lean person with no risk factors, with normal triglycerides, and LDL-C and ApoB falling in the same percentiles. Here ApoB adds nothing new, and the extra test is merely confirmation. This is exactly the situation in which the 'ApoB or LDL' debate has the least practical significance, and risk is determined rather by whether someone smokes, what the blood pressure is, the family history and whether Lp(a) is elevated.

The common conclusion: ApoB is a tool for settling ambiguous situations and monitoring the effectiveness of treatment, not a stand-alone 'verdict'. The best decisions are made when the doctor combines ApoB with the risk category, Lp(a), blood pressure, glycemia and, where needed, an assessment of coronary artery calcium.

The second myth is the belief that a 'normal ApoB' means no risk. ApoB is one factor. Blood pressure, diabetes, smoking, age, family history and the state of the arteries (for example CAC) also determine total risk. That is why how low ApoB should be in a given person is decided by the risk category set by the doctor.

If you have normal triglycerides, no metabolic disorders and a low LDL-C, a standard lipid panel is usually enough. If, however, you have insulin resistance, excess weight, high triglycerides, diabetes, a very low LDL on treatment or simply an inconsistent picture, a one-off ApoB measurement is a cheap way to reduce uncertainty. The safest stance is to treat LDL-C, non-HDL-C and ApoB as complementary measures of the same phenomenon and to remember that none of them replaces an assessment of total risk.