Lipoprotein(a) — Lp(a)

Why it matters (the biology, briefly)

Lp(a) is an LDL particle wearing an extra protein — apolipoprotein(a) — bolted onto it. That single addition makes it nastier than ordinary LDL in three ways at once:

• Pro-atherogenic: the LDL-like core deposits cholesterol into artery walls, same as regular LDL.
• Pro-thrombotic: the apo(a) tail mimics plasminogen, a clot-dissolving protein — so it gums up your ability to break down clots.³
• Pro-inflammatory: it ferries oxidized phospholipids, which inflame the artery wall.⁴

The risk it carries is independent of your LDL, your ApoB, and the usual lifestyle factors — it stacks on top of them. The cleanest evidence comes from genetic studies (Mendelian randomization), which use people’s inherited Lp(a) genes as a natural lifelong experiment: more Lp(a) means a dose-dependent rise in coronary disease and large-artery ischemic stroke. It’s about as close to “causal” as cardiology gets without a trial.⁵

One number to anchor on: above 125 nmol/L (~50 mg/dL), risk of heart attack or stroke runs roughly 1.4× higher; around 250 nmol/L it roughly doubles.¹

Lp(a) and your aortic valve. Worth its own line: elevated Lp(a) is the strongest known genetic risk factor for calcific aortic stenosis — a slow stiffening of the heart’s main outflow valve.⁶ If your Lp(a) is high and you’ve been told you have a heart murmur or valve sclerosis, that’s a reason for closer cardiology follow-up.

Should you test it? (yes — once)

Short answer: yes, exactly once, and the result is good for life. As of 2026, major guidelines recommend that every adult measure Lp(a) at least once — in the new ACC/AHA guideline it’s a Class I (highest) recommendation.¹ It’s a one-time blood draw that can reclassify your entire lifetime cardiovascular risk.

• It’s a separate order. Lp(a) isn’t on a standard lipid panel — you have to ask for it specifically. Some doctors are reluctant to order it precisely because “there’s nothing to do about the number” — if you hit that wall, direct-to-consumer labs run it for about $25–75.
• No fasting, no timing tricks. There’s no diurnal variation and the reading is genetically stable, so there’s nothing to prepare.
• Ask for nmol/L. It counts particles and is the modern, isoform-independent standard.⁸
• It’s a “test once” test, not a tracking metric. One accurate adult result is enough unless your clinical situation changes or you go on a drug meant to lower it.

Two footnotes: a recent illness or surgery can transiently bump Lp(a), so repeat a borderline result in 4–8 weeks; and the genetic stability is strongest at the tails (clearly low or high), while the grey zone (~75–125 nmol/L) is where a single reading wobbles most.

Reading your result (and the unit trap)

“Mine came back high. Am I doomed?”

No — and this is the most important paragraph on the page. A high Lp(a) is a risk multiplier, not a verdict. It multiplies the risk from your other factors — LDL/ApoB, blood pressure, smoking, diabetes. That’s actually good news, because it means you have a lever: shrink the things it multiplies, and you shrink the total. Two people with identical high Lp(a) can have wildly different actual risk depending on how well everything else is controlled. You get to decide which of those people you are.

What you should not do:

• Don’t hunt for a drug to lower the number. As of 2026, none has been proven to cut heart attacks by lowering Lp(a). Lowering the number is not yet the same as lowering the risk.
• Don’t stop your statin because you read it “raises Lp(a).” The effect is small or null in the best data, and the statin’s LDL benefit dwarfs it.¹²
• Don’t lose sleep over diet and exercise “not working” on Lp(a). They genuinely don’t move this number — but they move everything else it multiplies.

The drugs everyone’s reading about (pipeline — outcome data pending)

If you’ve gone down the Lp(a) rabbit hole, you’ve seen names like pelacarsen, olpasiran, zerlasiran, lepodisiran, and muvalaplin. The honest picture, hype removed: these are purpose-built drugs that lower Lp(a) by 70–97% — dramatically more than anything else.²³ But as of mid-2026, not one has been shown to actually reduce heart attacks or strokes. They move the number; whether moving the number moves outcomes is the open question — and it’s a real one, because cardiology is littered with drugs that improved a biomarker and did nothing (or harm) for patients.