Lp(a): the test your father's cardiologist didn't order

Eduardo is 39, a lawyer, lean, runs triathlons on the weekend. He had a complete corporate checkup two months ago. Total cholesterol 192 mg/dL, LDL 118 mg/dL, HDL 54 mg/dL, triglycerides 88 mg/dL. All normal, doctor. Why did I come back here? He came back because his father had a heart attack at 55, his paternal uncle at 49, his paternal grandfather at 53. They told me it was hereditary, but that for me it was different because I train.

I ordered an Lp(a) test. Result: 187 nmol/L (reference: <75 nmol/L). To put it in scale: he sits above the 90th percentile of the population. I ordered a coronary calcium score. Score 38 — subclinical plaque already present, in someone 39 years old, lean, and an athlete.

Eduardo is the common case. It is not lack of information. It is a test that drops out of the checkup for the wrong reason: no one ordered it.

The story in one sentence

Lipoprotein(a), abbreviated Lp(a), is a particle similar to LDL — “bad” cholesterol — but with an extra protein attached, called apolipoprotein(a). It was described in 1963. It took another fifty years to enter the routine cardiology conversation in earnest, even as the genetic and epidemiologic literature kept stacking up: those with elevated Lp(a) carry significantly higher atherosclerotic risk, independent of LDL, blood pressure, glucose, or lifestyle.

The European Atherosclerosis Society consensus, published by Florian Kronenberg and colleagues in the European Heart Journal in 2022, is the current reference document. Central recommendation: measure Lp(a) at least once in a lifetime in every adult.

Why once in a lifetime? Because the level is genetically determined, encoded by the LPA gene, and remains practically constant throughout adult life. Lifestyle, diet, statin — they do not significantly change the value. It is a biological calling card you inherited.

The size of the problem

Sotirios Tsimikas's review in the Journal of the American College of Cardiology in 2017 captured the magnitude well. Approximately 20 to 25% of the general population has Lp(a) above the threshold considered high-risk (>50 mg/dL or >125 nmol/L). In people of African descent, that proportion rises to 30–40%.

And the impact is real. The Robert Clarke study in NEJM in 2009 showed, in a genetic analysis of more than 26,000 people, that carriers of LPA variants associated with elevated Lp(a) had coronary disease risk increased by approximately 1.5-fold. People in the top quintile of Lp(a) carry 2 to 3 times the risk of heart attack, ischemic stroke, and calcified aortic stenosis — in some cuts, equivalent to smoking.

The Mendelian randomization analysis by Stephen Burgess and colleagues, published in JAMA Cardiology in 2018, was one of the works that established the relationship as causal, not merely associative. Genetic variants that naturally lower Lp(a) reduce cardiovascular risk in the expected proportion — the classical signature of causality.

Why it does not appear on the checkup

Three practical reasons.

First: the traditional Brazilian checkup protocol was designed around total cholesterol, LDL, HDL, and triglycerides — the classical “lipid panel.” Lp(a) is a separate test and was never strongly standardized in local guidelines.

Second: for decades, elevated Lp(a) was viewed as “incurable” — there was no specific drug to lower it. Statins lower LDL by 30–50%, but typically raise Lp(a) slightly. With no direct therapeutic tool, the test seemed like diagnosis without treatment. The objection made some sense in 1990. It makes less sense today, with specific therapies (antisense oligonucleotides and anti-Lp(a) siRNAs) already in phase 3 and the clinical reading clear: knowing Lp(a) changes the intensity of the overall preventive strategy.

Third: family-history screening is poorly done. Most patients do not bring the paternal history spontaneously; most physicians do not have time to map the family tree in a fifteen-minute visit.

What to do with the result

The 2019 ESC/EAS guideline (Mach and colleagues, European Heart Journal 2020) has already incorporated Lp(a) into the cardiovascular risk stratification algorithm. The practical reading:

• Lp(a) <75 nmol/L (~30 mg/dL): desirable value, no specific action.
• Lp(a) 75–125 nmol/L (~30–50 mg/dL): intermediate zone, consider as a risk modifier.
• Lp(a) >125 nmol/L (~50 mg/dL): elevated risk, intensify prevention.
• Lp(a) >250 nmol/L (~100 mg/dL): very high risk, equivalent to familial hypercholesterolemia in some cases.

When the value is elevated, the action is not “wait for specific therapy.” The action is knock down every other modifiable risk factor more aggressively:

• LDL/ApoB: more ambitious targets. For a person with elevated Lp(a) and an additional risk factor, consider ApoB <80 mg/dL as the goal (some experts suggest <60 mg/dL in very high risk). The statin discussion now happens earlier.
• Blood pressure: consistent <130/80 mmHg target.
• Glucose: strict control of pre-diabetes or diabetes.
• Smoking: zero.
• Coronary calcium score: consider it to refine risk — plaque already present changes the tone of the conversation.
• Cascade family screening: children, siblings, nieces and nephews. If you have high Lp(a), there is a 50% chance each child carries it.

Who should measure, and when

The current European recommendation is to measure once in every adult. In Brazilian practice, the pragmatic consensus we adopt:

• In every adult with a family history of premature cardiovascular disease (man <55 years, woman <65 years with heart attack, stroke, or sudden death).
• In every adult approaching 40 doing their first complete cardiovascular panel.
• In a person with documented atherosclerotic disease earlier than expected — coronary plaque on a calcium score above the 75th percentile for age, early clinical event.
• In a person with known familial hypercholesterolemia — frequently coexists.
• In idiopathic calcified aortic stenosis.
• In first-degree relatives of someone with an already-measured elevated Lp(a).

Collection is simple: fasting blood is not strictly necessary. The ideal assay is isoform-independent, reported in nmol/L (not in mg/dL — the conversion between units is not linear, because it depends on particle size). Ask the lab to report in nmol/L if there is an option.

The reading the Continuum makes

In the initial Continuum Plenya panel, Lp(a) enters alongside ApoB, hs-CRP, fasting insulin, HbA1c, vitamin D, ferritin, B12, and homocysteine. Not as an “extra test” — as structural information about the patient's cardiovascular risk for the next forty years. Eduardo, from the beginning, today is on a high-potency statin with an ApoB target <70, with the calcium score being monitored every three years, with his siblings screened (one of them also with elevated Lp(a), brought into preventive workup), and with him in a structured training plan and a sustained Mediterranean diet. Doctor, no one had ever told me this. I believed I was bulletproofed by training.

Training does not bulletproof. Genetics deals the card. Knowing which card came is the difference between training in the dark and training with a map.