Elevated Lp(a): What It Means and What to Do

Lp(a) is a particle built like LDL, that is, with a cholesterol-rich core and one molecule of apolipoprotein B, to which a large protein, apolipoprotein(a), is additionally attached. This 'extra' part gives it special properties: it favors deposition in the vessel wall, carries oxidized phospholipids with a pro-inflammatory effect and can take part in calcification of the aortic valve. Because it carries ApoB, it is counted in the ApoB test, but its significance goes beyond ordinary LDL.

The Lp(a) level is determined overwhelmingly by the LPA gene and is set in early childhood. It changes little over a lifetime, depends little on diet and physical activity, and two people with a similar lifestyle can have values that differ many-fold. According to reviews, an elevated Lp(a) occurs in about 20-30% of people, which makes it one of the most common inherited lipid disorders, one that most patients have never heard of.

Why is it rarely measured? Until recently there were no drugs meant to lower it, and a standard lipid panel does not include it. The 2022 consensus of the European Atherosclerosis Society (EAS), however, recommends measuring Lp(a) at least once in every adult, because the information about risk alone changes management.

The biggest trap is units. Laboratories report Lp(a) in mg/dL (mass) or in nmol/L (particle number), and no simple conversion exists, because results depend on the method used. The EAS consensus does not recommend a fixed factor; as a 'best guess' it points to a multiplier of about 2-2.5 (for general comparisons it adopts about 2.5), and for tests based on polyclonal antibodies it recommends reporting in mg/dL if no molar option is available. So when comparing results over time, have them done in the same laboratory with the same method, and when assessing thresholds always check which unit the result is in.

It is also worth remembering that Lp(a) contains cholesterol, which in an ordinary lipid panel can be counted within LDL-C. The consensus does not recommend routinely correcting LDL-C for the cholesterol contained in Lp(a), with the exception of suspected familial hypercholesterolemia coexisting with a high Lp(a).

These two analyses say the same thing in two ways. First, risk rises gradually with Lp(a) rather than abruptly at a single boundary, so a value just below 50 mg/dL is not safe and one just above is not catastrophic. Second, Lp(a) and LDL-C act independently and add together, and a low LDL-C on treatment does not 'zero out' the risk from Lp(a). It is worth remembering, however, that these are data from studies involving mainly middle-aged and older people with established disease or high risk, and a relative risk (for example HR 1.3-1.9) does not by itself say what the absolute risk is for a specific person.

The 2022 EAS consensus stresses that a high Lp(a) is linked not only to changes in the arteries but also to micro- and macrocalcification of the aortic valve, which leads to aortic stenosis. With valve narrowing, symptoms often appear late: exertional shortness of breath, chest pain or fainting. A high Lp(a) is therefore one of the reasons why, in people with a family burden, it is worth paying attention to listening to the heart and to echocardiography, if the doctor considers it justified.

The good news concerns the veins: according to the consensus authors, existing data do not support Lp(a) as a risk factor for venous thromboembolism or for impaired fibrinolysis, even though apolipoprotein(a) resembles plasminogen. A high Lp(a) alone should therefore not prompt worry about deep vein thrombosis. The consensus also mentions that a very low Lp(a) may be linked to a higher risk of diabetes, which needs further study, and is not a reason to try to 'raise' low results.

Several practical conclusions follow. PCSK9 inhibitors lower Lp(a) by about a quarter to a third, but in practice they are used for LDL-C, not Lp(a). Niacin lowers Lp(a), but in large trials it did not reduce cardiovascular events and had adverse effects, as we write in our article on low HDL, so it is not a sensible way to lower Lp(a). Diet, exercise, supplements and most popular 'Lp(a) remedies' have practically no documented, significant effect on its level.

Statins are a separate matter. The meta-analysis by Tsimikas et al. (2020), covering 5256 patients, showed that statins raise Lp(a) concentration by about 9-20% (ratio of geometric means 1.11), suggesting a possible modest rise during treatment, whereas the analysis by Willeit et al. showed no significant change after starting a statin. The EAS consensus explicitly advises not stopping a statin for this reason, because the benefit of lowering LDL-C in people with a high Lp(a) outweighs the small rise in Lp(a). This matters, because some patients, having read about this rise, stop their statin. Do not do that without talking to your doctor.

This explains why the 'moderate' Lp(a) reductions achieved so far, of 20-30% (PCSK9, niacin), did not give a visible benefit: with an Lp(a) of around 100 mg/dL, a 25% fall is only 25 mg/dL, which according to the model is roughly a quarter of the effect given by lowering LDL-C by 1 mmol/L. That is why drugs were developed that are meant to lower Lp(a) far more strongly, based on antisense oligonucleotides or small interfering RNA (siRNA).

What does this mean for the patient? First, it does not mean that Lp(a) has stopped being a risk factor: the genetic and observational data on its causal role remain in force. Second, at this point (October 2026) there is no approved drug that specifically lowers Lp(a) and has demonstrated clinical benefit, so the strategy must rest on controlling the other factors. Third, it is worth following the publication of the full results, because they may bring important distinctions (for example for younger people or those with a higher baseline Lp(a)).

We write more about how ApoB counts all atherogenic particles (including Lp(a)) and the goals for different risk categories in our article on LDL and ApoB. A knowledge-base entry covers statins, including their side effects, at length. For those interested in imaging, we describe coronary artery calcium in a separate entry.

Because Lp(a) is inherited, in close relatives of a person with a high result the probability of a similarly high level is considerably raised. The EAS consensus recommends cascade testing of families in three situations: with familial hypercholesterolemia, with a family or personal history of (very) high Lp(a), and with a personal or family history of premature atherosclerotic disease. It advises informing close relatives about the risk and offering them testing, with counselling before and after the test.

For children and adolescents the matter is more delicate. The Lp(a) level is set early, so a one-off measurement can be informative, but decisions about testing children should be taken by a doctor together with the parents, especially when premature heart disease or familial hypercholesterolemia runs in the family. For young people the most important thing is healthy habits from an early age and avoiding smoking; drug therapy in children requires an individual decision by a specialist.

Equally mistaken is the opposite belief that, since nothing lowers Lp(a), there is no point measuring it. Knowing about a high Lp(a) changes risk assessment, the intensity of LDL and ApoB control and decisions about further tests, and makes it possible to test relatives, so it makes sense even without a specific drug.

An elevated Lp(a) is an inherited risk factor that you did not choose and over which you have little influence, but which you can compensate for with what is up to you. The best strategy at present is a dull and therefore effective one: know your result, match the intensity of LDL and ApoB control to it, look after blood pressure, glycemia and not smoking, test your family and follow news of new drugs. There is as yet no drug that demonstrably reduces heart attacks by lowering Lp(a), but that is not a reason to give up acting.