Lp(a): The Inherited Risk Factor Almost Nobody Measures

A man of forty-seven arrives at the emergency department with an anterior myocardial infarction. He has never smoked. His blood pressure is normal, his weight is normal, his cholesterol has been reported as acceptable at every check-up since his thirties, and he runs three mornings a week. His father had a bypass at fifty-two. Nobody in the family has ever been told why.

In a meaningful share of such cases the explanation is a particle that almost nobody measures: lipoprotein(a), written Lp(a). It is inherited, it is essentially fixed for life, it does not appear on a standard lipid panel, and it is among the most common genetic risk factors for cardiovascular disease anywhere in the world. A single blood test identifies it. Most people never have that test done.

What lipoprotein(a) actually is

Start with an ordinary LDL particle: a sphere of cholesterol and phospholipid wrapped around one molecule of apolipoprotein B. Now attach a second, much stranger protein to it — apolipoprotein(a), or apo(a) — bound covalently to the ApoB. That hybrid is Lp(a). It carries everything an LDL particle carries, and something extra besides.

The extra part is peculiar. Apo(a) is structurally close to plasminogen, the precursor of the enzyme that dissolves clots. It is built from repeated looped domains called kringles, and the number of one repeat varies considerably between people. That variation is most of the story: those who inherit fewer repeats make a smaller apo(a), secrete it more efficiently, and carry higher Lp(a) for life.

So Lp(a) is best pictured as an atherogenic particle wearing a clotting protein's coat — which is roughly how it behaves.

Why the number is fixed for life

Plasma Lp(a) is on the order of 80 to 90 per cent determined by a single gene, LPA. Adult concentrations are reached in early childhood and remain remarkably stable thereafter. Diet barely moves it. Nor does exercise, nor weight loss. Statins do not lower it and may nudge it slightly upward.

The practical consequence is unusual in laboratory medicine and worth stating plainly: Lp(a) needs to be measured once. It is not a tracking marker like LDL-C or HbA1c, where the point is to watch a number respond to treatment. It is closer to a genetic result that happens to be reported in units of concentration.

A few caveats. Levels rise modestly after menopause, in advanced kidney disease and the nephrotic syndrome, and in untreated hypothyroidism. Lp(a) is also a mild acute-phase reactant, so a level drawn in the weeks after a heart attack, an operation or a serious infection can read above true baseline. Otherwise one properly timed measurement settles the question permanently.

Three ways it does harm

Lp(a) is unusual in carrying three distinct mechanisms of vascular injury in a single particle.

• Atherogenic. Like LDL, it penetrates the arterial wall, becomes trapped in the subendothelial space, and contributes directly to plaque. Each Lp(a) particle carries its own ApoB, so a high Lp(a) also inflates total atherogenic particle count.
• Pro-inflammatory. Lp(a) is the principal carrier of oxidised phospholipids in plasma — potent inflammatory signals within the arterial wall, thought to account for a substantial part of the risk. The same mechanism appears to drive calcification of the aortic valve: Lp(a) is one of very few clearly causal risk factors for calcific aortic stenosis, a condition otherwise dismissed as wear and tear.
• Prothrombotic. Because apo(a) so closely resembles plasminogen, it is expected to interfere with the breakdown of fibrin. Mechanistically compelling; how much it contributes in living patients is still genuinely debated.

Human genetic evidence — in particular Mendelian randomisation work from the large Danish population cohorts — indicates that this relationship is causal, not merely associative. People who inherit variants that raise Lp(a) have more disease, in proportion to how much it is raised.

A common problem that hides in plain sight

Roughly one adult in five carries an elevated Lp(a). The distribution is heavily skewed: most people sit at low levels, and a substantial minority sit a long way out on the right-hand tail. It also differs by ancestry — average levels are highest in people of West African descent, intermediate in South Asian populations, lower in European and East Asian ones — while most published thresholds derive from European cohorts. Worth remembering when a result comes back borderline.

Because Lp(a) is absent from routine panels and produces no symptoms until it produces an event, the overwhelming majority of those people do not know. This is where a great deal of so-called unexplained premature coronary disease comes from: families with early heart attacks, unremarkable cholesterol, and no diagnosis beyond bad luck.

The measurement problem

There are two kinds of Lp(a) assay, and the difference is not cosmetic.

Mass assays report mg/dL — the total mass of Lp(a) material in the sample. Because apo(a) size varies between people, that mass depends partly on how large each particle is rather than on how many there are. Molar assays report nmol/L and count particles independent of isoform size. Nmol/L is the preferred unit and the one to request.

There is no reliable conversion between the two. Approximate factors circulate — roughly two to two and a half nmol/L per mg/dL — but they are population averages, not valid for an individual. Comparing a new result with an old one means checking the unit, and ideally the assay, first.

What to do when Lp(a) is high

Here is the honest position. No widely approved medication lowers Lp(a) and has been shown to reduce cardiovascular events. Niacin lowers it modestly, but large outcome trials of niacin showed no cardiovascular benefit and meaningful harm. PCSK9 inhibitors lower it by around a fifth, incidentally to their main action. Lipoprotein apheresis — physically filtering the particle out of plasma — exists at a few specialist centres for extreme cases with progressive disease.

So management is not about the number. It is about everything else, pursued harder than it otherwise would be:

• Lower ApoB aggressively. You cannot remove the inherited particle excess, which makes the acquired excess more worth removing. In practice that means adopting the stricter ApoB targets reserved for high-risk patients rather than the population-average ones.
• Treat blood pressure to a tight target. Hypertension and elevated Lp(a) together are considerably worse than either alone.
• Do not smoke. Smoking amplifies inherited vascular risk more than almost anything else within your control.
• Look, rather than only predict. Coronary calcium scoring, carotid ultrasound or CT angiography answer a question no risk calculator can: is there disease present now? A high Lp(a) with clean arteries at fifty is a different clinical situation from a high Lp(a) with established plaque.
• Address insulin resistance and inflammation. Neither changes Lp(a), but both compound what it does.
• Test first-degree relatives. Parents, siblings and children each have roughly a one-in-two chance of sharing the trait. This is among the highest-yield acts in preventive cardiology: an inexpensive test that can identify a relative decades before an event.
• Keep the aortic valve in view. Given the causal link with calcific aortic stenosis, periodic clinical attention to the valve in later life is reasonable.

Lp(a) is the rare test whose result you cannot change and should want anyway. It does not alter your biology. It alters how much every modifiable risk factor is worth attacking — and how early.

What is coming, and what remains uncertain

This is an area where the ground is genuinely moving. RNA-targeted agents — antisense oligonucleotides and small interfering RNAs that suppress hepatic apo(a) production — can lower Lp(a) by eighty per cent or more, an effect size unlike anything previously available. Late-stage cardiovascular outcome trials are under way.

What is not yet established is whether lowering Lp(a) pharmacologically reduces events, and by how much it must fall to matter. Genetic evidence suggests a large absolute reduction may be required. Until those outcome data are in hand, treating Lp(a) directly is not standard care, and anyone telling you otherwise is ahead of the evidence.

Two further caveats. The thresholds above are conventions imposed on a continuous variable — there is no biological cliff at 125 nmol/L. And how much weight an elevated Lp(a) should carry in an otherwise low-risk fifty-year-old remains clinical judgement rather than settled guideline.

What to ask your physician

• Ask for Lp(a) measured once, reported in nmol/L. If a previous result exists in mg/dL, do not attempt to convert it — note which assay was used.
• Ask what your ApoB target should be given the Lp(a) result, not independently of it.
• Ask whether imaging is warranted to establish whether atherosclerosis is already present, and at what age to start.
• Ask what to tell your parents, siblings and children — and make sure the conversation actually happens.

There is a natural objection to all of this: why measure something that cannot be changed? Because the number was never the target. A high Lp(a) does not tell you to do something new. It tells you that the ordinary things — particle burden, blood pressure, tobacco, imaging, family screening — matter more in your case than the population average implies, and that the margin for complacency is smaller. That is decision-changing information, and it costs one blood draw, once, in a lifetime.