THE PARTICLE COUNT // ApoB & LIPOPROTEIN(a)
For decades the world measured the wrong thing. A standard cholesterol panel reports LDL-C — the cholesterol carried inside your LDL particles — but the artery wall does not count cholesterol, it counts particles. Every atherogenic particle carries exactly one apolipoprotein B, so apoB is a direct headcount of the particles that invade the wall. When apoB and LDL-C disagree — which happens in a large minority of people, especially the metabolically stressed and the statin-treated — apoB wins every time: it predicts events when LDL-C says you are fine (Richardson et al., 2020; Johannesen et al., 2021). Multivariable Mendelian randomization is blunt about it: adjust for apoB and LDL-C's causal signal collapses to nothing — apoB is the trait that actually causes coronary disease (Richardson 2020; Zuber 2020). Sitting on top of this is lipoprotein(a) — a mostly-genetic, largely-fixed apoB particle that about one in five people carry at high levels, is roughly six-fold more atherogenic per particle than ordinary LDL, and is invisible on a standard panel (Björnson et al., 2024; Reyes-Soffer et al., 2021). The catch that keeps the frontier honest: apoB is proven and cheap to lower today, but no drug has yet been shown to cut events by lowering Lp(a) specifically — those outcome trials are running now (Malick et al., 2023). Measure the particle count. It is the most important number your annual physical probably never showed you.
Atherosclerotic cardiovascular disease is the single largest killer of the population this file is written for, and the standard lipid panel measures the wrong thing about it. It reports LDL-C — the mass of cholesterol carried inside your LDL particles. But an artery wall is not infiltrated by cholesterol in the abstract; it is infiltrated by particles. Each atherogenic particle carries exactly one apolipoprotein B, so apoB is not another line on the panel — it is a literal headcount of the particles doing the damage.
Most of the time apoB and LDL-C agree, and it does not matter which you read. The problem is the large minority in whom they disagree — the metabolically stressed, the person on a statin, the one with high triglycerides — because in exactly those people LDL-C reads reassuringly normal while apoB, and the risk, is high. When the two disagree, the particle count is right and the cholesterol mass is wrong.
This file grades the science, not the hype. apoB's causal role and its superiority as a marker are as settled as cardiovascular medicine gets. Lipoprotein(a) is causal too — and mostly written into your genes. What is not yet settled is whether lowering Lp(a) with a drug cuts events; that answer is being written in trials right now. We publish the proven part as proven and the pending part as pending.
The Particle, Not The CholesterolTIER A
Multivariable Mendelian randomization — using inherited gene variants as a lifelong natural experiment — is unusually blunt here. Assessed alone, LDL-C, triglycerides and apoB all track with coronary disease. But put them in the model together and only apoB keeps a robust causal effect; the LDL-C signal actually reverses toward null once apoB is accounted for (Richardson et al., PLoS Medicine 2020). An independent Bayesian analysis across 30 lipid measures reached the same verdict from a different direction: apoB is the single trait the data keep selecting as the primary lipid cause of coronary artery disease (Zuber et al., 2020). The number of particles is the cause; the cholesterol inside them is a passenger.
When apoB and LDL-C disagree, apoB wins at the bedside. In 13,015 statin-treated adults followed eight years, high apoB — but not high LDL-C — flagged residual risk of both myocardial infarction and all-cause mortality (Johannesen et al., JACC 2021). The physiology behind that superiority is now well characterized (Glavinovic et al., JAHA 2022). Practical translation: a normal LDL-C is not an all-clear. Only apoB tells you whether your particle burden is actually low — and it is a cheap, standardized test that needs no fasting.
Lipoprotein(a) — The Silent MultiplierTIER A (RISK) / TIER C (DRUG-LOWERING OUTCOMES)
Lipoprotein(a) is an LDL-like particle with an extra protein bolted on, and it is one of the most common inherited cardiovascular risk factors in existence: 70–90% of your level is set by genetics, it is largely fixed for life, and roughly one in five people carry it high enough to meaningfully raise risk (Reyes-Soffer et al., AHA Scientific Statement, ATVB 2021). It is causal — for coronary disease and, distinctively, for calcific aortic-valve stenosis (Tsimikas, JACC 2017). And per particle it is not slightly worse than LDL, it is dramatically worse: a 2024 apoB-anchored genetic analysis put Lp(a) at roughly six-fold more atherogenic than ordinary LDL, particle for particle (Björnson et al., JACC 2024).
Two things make Lp(a) uniquely under-managed. First, it is invisible on a standard panel — it must be ordered specifically, and most people never are. Second, the tools that crush LDL barely touch it: statins nudge Lp(a) slightly up, PCSK9 inhibitors lower it ~20–30%, and only the new RNA therapies (pelacarsen, olpasiran, SLN360) drop it 80%+ (Tsimikas 2017; Koren et al., Nature Medicine 2022). Here is the border we will not cross: as of 2026 no trial has yet shown that lowering Lp(a) with a drug reduces cardiovascular events. The phase-3 outcome trials are running now (Malick et al., JACC 2023). Until they report, a high Lp(a) is a reason to measure — and to lower every other apoB particle harder — not a reason to buy an unproven injection.
Is a standard cholesterol panel enough to know your cardiovascular risk?
LDL-C is cheap, universal, and adequate when it agrees with the particle count. But it measures cholesterol mass, not particle number, so in the metabolically stressed and the statin-treated it can read normal while risk is high. A proxy that is wrong exactly when being right matters most.
Johannesen et al. — apoB & non-HDL reflect residual risk better than LDL-C (JACC, 2021) ↗apoB counts every atherogenic particle directly. Genetics say it is the causal trait; outcome data say it wins whenever it disagrees with LDL-C. A cheap, standardized test that upgrades a guess into a measurement — and one you can lower with proven drugs.
Richardson et al. — multivariable MR: apoB is the predominant causal lipid trait (PLoS Medicine, 2020) ↗Neither LDL-C nor a routine apoB reports your Lp(a) — a separate, mostly-genetic, ~6× more atherogenic particle carried by ~1 in 5 people. It must be ordered on purpose, once in a lifetime. Its risk is proven; a drug that lowers its events is not yet.
Björnson et al. — Lp(a) ~6-fold more atherogenic than LDL per particle (JACC, 2024) ↗The panel is enough only when apoB and LDL-C concord and Lp(a) has been checked once and is low. Absent those two facts, 'my cholesterol is fine' is an assumption, not a measurement. Get apoB; get Lp(a) once; then optimize.
Before any drug, the particle count answers to the same Tier-0 substrate the rest of this compendium is built on. Visceral fat, refined-carbohydrate load, and inactivity all raise apoB-particle number; losing visceral fat, training, and shifting the diet toward fiber, unsaturated fat and protein lower it — as do soluble fiber and plant sterols. None of this touches Lp(a) — that is the one number your habits cannot move, which is exactly why it is worth knowing. The sequence is not glamorous and it is not optional: fix the substrate, measure the particle count, and only then reach for pharmacology, with a physician, against a real number.
We sell nothing here and grade against ourselves. apoB and Lp(a) are ordinary, cheap, standardized blood tests — there is no premium product to push, and that is the point: the highest-leverage move in this entire file costs less than a month of most supplement subscriptions. If we ever link a testing service through a vetted partner we will disclose it plainly, and it will change nothing about the grade. apoB is graded CLINICAL because the causal science and the lowering evidence are closed; Lp(a) risk is CLINICAL and Lp(a) drug-lowering is held at Tier C because the outcome trials have not reported. The data drive the grade. Nothing else does.
Peer-reviewed trials — sample size, effect size, stage.
Applies to healthy apex, not only to clinical deficit.
Survives a demanding calendar. Zero executive friction.
Cleared all three layers. Protocol-grade.
The score is a geometric mean — a single failed layer collapses it. Excellence in two cannot rescue a gap in the third. That is why hype scores low and proven, feasible, broadly-applicable work scores high. The restraint is the product.