Key takeaways
- Lp(a) is an LDL-like particle carrying apolipoprotein(a), making it atherogenic, inflammatory and possibly pro-thrombotic at once.
- The level is roughly 90% genetic, set by inherited isoform size, and does not move with diet, exercise or statins.
- Because it is stable for life, one measurement answers the question permanently — but it must be requested by name.
- Elevated Lp(a) raises cardiovascular risk 1.5 to 2-fold and is specifically associated with calcific aortic stenosis.
- With no established treatment yet, the value of knowing is that it justifies pushing every other risk factor harder — and testing relatives.
There is a cardiovascular risk factor carried by roughly one adult in five, largely fixed at birth, unmoved by diet, exercise or statins, and absent from the lipid panel almost everyone has had. Lipoprotein(a) is not obscure — it appears in every major lipid guideline — but it is not measured unless somebody asks for it by name. The result is a lot of people reassured by cholesterol results that were never looking at the thing most likely to explain their family history.
What the particle actually is
Start with an LDL particle: a core of cholesterol and triglyceride inside a shell, with one large protein, apolipoprotein B, wrapped around it. Lp(a) is that same particle plus a second protein, apolipoprotein(a), bonded to the apoB by a disulphide bridge.
Apo(a) is a strange molecule. It is built from repeating loop-shaped domains called kringles and is structurally homologous to plasminogen, the precursor of the enzyme that dissolves fibrin clots. The resemblance is not decorative: apo(a) can occupy sites on fibrin and cell surfaces where plasminogen would otherwise bind, potentially slowing clot breakdown. So the particle carries two liabilities at once — it delivers cholesterol into the artery wall like LDL, and it interferes with fibrinolysis like nothing else on a lipid panel (Schmidt et al., J Lipid Res 2016). It is also the main plasma carrier of oxidised phospholipids, which are directly inflammatory to the vessel wall and the aortic valve — the likeliest reason Lp(a) is associated with valve disease when other lipoproteins are not.
Why the level is set genetically
The number of kringle repeats in apo(a) varies between people and is inherited. Shorter isoforms are assembled and secreted by the liver far more efficiently than long ones, so isoform size is inversely related to plasma concentration: two short alleles means a high Lp(a) for life.
Roughly 90% of the variance is explained by variation at the LPA locus, and almost nothing else moves it. Weight loss, exercise, dietary fat quality and fibre all move LDL and none of them move this. Statins do not lower Lp(a) and pooled analysis suggests they raise it slightly (Tsimikas et al., Eur Heart J 2020), though that has not translated into worse outcomes on statins, which reduce risk overall.
That stability is what makes this test different from the rest of a lipid panel. The level at 25 is essentially the level at 65 — a once-in-a-lifetime measurement rather than something to track, and a normal result can genuinely be filed away.
What elevation does to risk
Elevated Lp(a), usually above 50 mg/dL or 125 nmol/L, raises cardiovascular risk in the region of 1.5 to 2-fold, and the evidence is unusually strong on causation rather than association. Mendelian randomisation — which exploits the fact that LPA variants are allocated randomly at conception, before any lifestyle confounding — shows that genetically determined high Lp(a) predicts myocardial infarction (Kamstrup et al., JAMA 2009). That design is as close to a randomised trial of lifelong exposure as observational epidemiology gets.
The consequences that follow:
- Coronary artery disease, with events occurring at younger ages than the rest of the risk profile would predict
- Calcific aortic valve stenosis — the association Lp(a) has and other lipoproteins largely do not
- Atherothrombotic stroke, particularly the subtypes involving large-artery disease
- Amplified risk in familial hypercholesterolaemia, where the two inherited problems compound each other
Prevalence is around 20% globally and varies by ancestry, being higher in people of African descent and lower in East Asian populations (Nordestgaard et al., Eur Heart J 2010). Most of those people have never been tested.
Who should be tested, and when
European consensus now argues for measuring Lp(a) once in every adult: it is stable, causal, common and cheap, and knowing it changes how aggressively everything else is managed (Kronenberg et al., Eur Heart J 2022). Where universal testing is not routine, the priority cases are unambiguous:
- A family history of premature cardiovascular disease — a parent or sibling with an event before about 55 in men or 65 in women
- Personal cardiovascular disease that the standard risk factors do not adequately explain
- Known or suspected familial hypercholesterolaemia
- Aortic stenosis, or a calcified valve found incidentally
- Any situation where the decision to treat other risk factors is borderline and could be tipped either way
It has to be requested explicitly, and is not part of a standard lipid panel — the single commonest reason people with high Lp(a) do not know, and a version of the point made in why bloodwork comes first.
The units problem, and how to avoid being misled
Lp(a) is reported in mass units (mg/dL) or molar units (nmol/L), and the two do not convert reliably. The reason is the isoform variation itself: mass assays measure total protein, so someone with long apo(a) isoforms carries more protein per particle and a mass result over-states particle count. Molar assays count particles and are preferred.
Practically: read the result against the reference range for the units it was reported in, never apply a conversion factor found online, and check that any two results being compared used the same assay. The commonly quoted bands are below 30 mg/dL or 75 nmol/L as low, 30-50 mg/dL or 75-125 nmol/L as intermediate, above 50 mg/dL or 125 nmol/L as elevated, and above 125 mg/dL or 300 nmol/L as very high.
What to do with a high result
The instinctive response — how do I lower it — is currently the wrong question, because no widely available therapy lowers it much and none has been shown to reduce events by doing so. The right question is what a high result changes about everything else.
Risk is multiplicative, so an unmodifiable factor makes the modifiable ones matter more, not less. A high Lp(a) is an argument for driving ApoB lower than you otherwise would, treating blood pressure earlier, taking smoking cessation as non-negotiable, and taking glucose and inflammation seriously. ApoB is the right lipid target because it counts every atherogenic particle rather than the cholesterol inside them — set out in ApoB, the real cardiovascular number and compared with conventional LDL in ApoB versus LDL and particle number versus cholesterol. hs-CRP is the accessible inflammation measure to run alongside.
A high result also carries a family obligation. Lp(a) is inherited in a simple, dominant-behaving way, so first-degree relatives have a substantially raised probability of the same. Cascade testing of siblings and children is the highest-yield thing a positive result generates.
On existing drugs: PCSK9 inhibitors lower Lp(a) by around 25% as a side effect of their main action, niacin by around 30% with a poor risk-benefit record, and apheresis is reserved for very high levels with established disease. None is prescribed for Lp(a) alone — these are whole-risk-picture decisions made by a physician, and the 60-second assessment is where that starts.
The clinical pearl: test Lp(a) once, deliberately, by name. A normal result can be filed permanently. A high one does not currently come with a drug, but it changes how hard every other risk factor should be pushed — and it tells your siblings and children to get tested too.
The emerging drugs, described honestly
Therapies designed to reduce Lp(a) specifically — antisense oligonucleotides and small interfering RNA agents that suppress hepatic apo(a) production — have produced large reductions in the target, in the region of 80-90%.
What has not been shown is that lowering Lp(a) lowers events. The outcome trials are running; until they report, the causal case rests on genetics rather than intervention, and the history of lipid medicine contains at least one class that lowered its target convincingly and delivered nothing. Following the field is reasonable. Assuming the outcome is not.
Bottom line
Lp(a) is an LDL-like particle carrying an extra protein that makes it atherogenic, pro-inflammatory and possibly pro-thrombotic at once. The level is set almost entirely by inherited isoform size, does not respond to lifestyle, and is stable across a lifetime — one measurement answers the question permanently. About one adult in five is elevated and most do not know, because the test is not on a standard panel and must be asked for. A high result does not yet come with a specific treatment, but it justifies pushing ApoB, blood pressure, smoking and glucose harder than the rest of the profile alone would suggest, and it is a strong reason to have first-degree relatives tested.
Educational content, not medical advice. Laboratory interpretation and any treatment decision are made by a licensed physician after individual evaluation. Individual results vary.
