The Number You Were Born With: Lp(a) and the Case for One Test in a Lifetime

There is a particular unfairness buried in cardiovascular risk: a person can eat carefully, exercise, watch their LDL come back fine — and still carry a level of inherited danger that no standard panel reports. Lipoprotein(a) is that danger. Its blood concentration is roughly 80 to 90 percent determined by variation in the LPA gene, stays relatively stable across a lifetime, and differs widely between individuals and populations. About one in five people carry high levels. Most have never been told, because a standard cholesterol panel does not capture it.

The emotional stakes run in both directions. Never testing means false reassurance — risk written before birth, discovered at the first event. Testing means possibly learning something you cannot yet treat directly. Whether the knowledge is worth having is the question this piece answers.

The causal evidence is unusually strong

Lp(a) is an LDL-like particle — an apolipoprotein(a) molecule covalently bound to apolipoprotein B-100 — that promotes atherosclerosis, inflammation, and valve calcification. Large meta-analyses, prospective cohorts, and Mendelian randomization studies establish elevated Lp(a) as a causal, independent risk factor for coronary heart disease, myocardial infarction, ischemic stroke, peripheral artery disease, and calcific aortic valve stenosis. A 2009 JAMA study and a 2011 meta-analysis of prospective studies in Clinical Laboratory anchored the epidemiology; a 2021 Clinical Chemistry review frames the current consensus; a 2022 Journal of the American College of Cardiology paper extended the reach to atrial fibrillation; and a 2024 participant-level meta-analysis in Circulation showed that Lp(a)-mediated risk is independent of LDL-cholesterol-mediated risk. Very high levels carry risk comparable to heterozygous familial hypercholesterolemia. The Mendelian randomization evidence matters: because the exposure is genetic, the causation case escapes the association-versus-causation caveat that weakens most biomarker stories.

What a test can and cannot change

Here is the honest constraint. RNA-based therapies — and now a first oral inhibitor — lower Lp(a) by roughly 70 to 100 percent in trials. But cardiovascular outcome trials are still ongoing, and no Lp(a)-specific drug is yet approved. Measuring Lp(a) today therefore guides risk stratification more than treatment: a high result argues for tightening everything that can move — LDL cholesterol, blood pressure, the rest of the modifiable ledger — rather than for a drug that does not yet exist. The association with disease is continuous and largely independent of LDL, though effect sizes for stroke and all-cause mortality are more modest than for coronary disease. A high result is not a diagnosis; it is a recalibration.

That is exactly why major guidelines now recommend measuring Lp(a) at least once in every adult's lifetime: the number barely changes, so one measurement settles it — a cheap answer to an expensive question. For the roughly four in five adults whose level is not elevated, that single measurement closes the question for life; for the one in five, it opens a more useful one. Few risk markers in medicine arrive with the pedigree this one carries — genetic in origin, stable across decades, causal in the Mendelian evidence. What it still lacks is a therapy of its own, and that gap is the honest summary of the entire field. The outcome trials now running will decide whether lowering the number lowers the events. Knowledge first; drugs later, if the trials cooperate. And the evidence concerns the biomarker and its measurement in general, not any specific commercial mail-in test.

The unfairness does not disappear with knowledge. But a risk that is known can be surrounded — every modifiable factor tightened around the one that is fixed. The gene was dealt at birth; the response to it is not. Educational, not medical advice.