Lp(a) mg/dL to nmol/L: the conversion, and why it is not exact

If you need a rough figure, the factor the European Atherosclerosis Society (EAS) uses for its own general statements is 2.5: multiply mg/dL by 2.5 for an approximate nmol/L, or divide nmol/L by 2.5 for an approximate mg/dL. The same 2022 consensus statement is explicit that this is a stopgap. Its words: "This panel does not recommend using a standard factor to convert between mg/dL and nmol/L, as assays vary extensively." Where both units are in use, it describes a factor of 2 to 2.5 as only a "best guess", and calls the 2.5 it uses an approximated "interim" factor.

The American Heart Association's 2022 scientific statement on Lp(a) goes further: "there is no unbiased conversion factor from milligrams per deciliter to nanomoles per liter or vice versa." So any calculator that turns your 48 mg/dL into exactly 120 nmol/L is giving you more precision than the measurement contains.

The two units answer different questions. A result in nmol/L counts particles: it reports the molar concentration of apolipoprotein(a), which the AHA statement describes as reflecting the number of Lp(a) particles. A result in mg/dL estimates weight: how much Lp(a) mass is in a deciliter of blood. Converting between a count and a weight works only when every item weighs the same, and Lp(a) particles do not.

The variable part is apo(a) itself. It contains a repeating segment called kringle IV type 2, and the number of repeats is inherited. The EAS statement explains that a low number of repeats (fewer than 23) is characterized by small apo(a) isoforms and a markedly higher Lp(a) concentration, and that this repeat polymorphism explains about 30 to 70 percent of the variation in Lp(a) between people. A review by Marcovina and Albers in the Journal of Lipid Research adds that apo(a) varies in size not only between people but within one person, because a person can carry two different, genetically determined isoform sizes (we read the abstract). A particle with a large apo(a) weighs more than one with a small apo(a), so the same particle count corresponds to different masses in different people.

History made it worse. The AHA statement recounts that the first sensitive Lp(a) assay, reported in the 1970s, was calibrated against Lp(a) purified from a single donor, and later immunoassays were calibrated in mg/dL of total Lp(a) mass even though they measure only the apo(a) component. The statement calls that approach, which assumed every person's particle had the same mass, "unacceptable". The reference ELISA it describes as the gold standard was calibrated in nmol/L.

The clearest test of the 2 to 2.5 rule is a 2018 study by Tsimikas, Marcovina and colleagues in the Journal of Clinical Lipidology, which measured Lp(a) in nmol/L at a reference laboratory and in mg/dL with six other assays (80 to 2,673 samples per assay), then calculated the ratio between the two. We read the abstract, which reports:

Three things follow from that table. The factor depends on the assay: against the same reference laboratory, the three comparisons with reported averages gave mean ratios from 1.64 to 2.42. It depends on the level: in each of those three comparisons the ratio was higher above 125 nmol/L than below 75 nmol/L, which on our reading fits the biology, since the EAS links small apo(a) isoforms to higher concentrations. And for some commercial methods no single number describes it at all. The authors' conclusion: "A single conversion factor between assays is not appropriate."

Even the EAS's own paired cutoffs do not share one factor. Its rule-out and rule-in thresholds use 2.5 (30 mg/dL with 75 nmol/L, 50 mg/dL with 125 nmol/L), while its very-high line pairs 180 mg/dL with 430 nmol/L, which on our arithmetic is a factor of about 2.39.

The first table below shows the pairs as the EAS publishes them. The second is our arithmetic: what the 2 and 2.5 ends of the EAS "best guess" produce for the same mg/dL values. Treat the second as a range of plausible equivalents, not a conversion of your result.

The 2026 ACC/AHA multisociety dyslipidemia guideline, which replaced the 2018 US cholesterol guideline, covers elevated Lp(a), but its full text was closed to us when we checked on 28 October 2026, so we do not quote its wording. The American College of Cardiology's own summary of the guideline, a podcast page featuring the guideline's first author, Roger S. Blumenthal, describes risk at levels around 125 nmol/L (50 mg/dL) and a substantially higher-risk group at about 250 nmol/L (100 mg/dL). Those are the same 2.5 pairings, and they come with the same caveat: a pair of thresholds is not a conversion factor for your sample.

Far from a threshold the factor matters less. On our arithmetic with the EAS range of 2 to 2.5, 12 mg/dL (24 to 30 nmol/L) sits under the rule-out line of 75 nmol/L and 220 mg/dL (440 to 550 nmol/L) sits over the very-high line of 430 nmol/L either way. The trouble is near a threshold. Take a reading of 45 mg/dL. On our arithmetic, times 2.0 gives 90 nmol/L and times 2.5 gives about 113 nmol/L: both sit inside the EAS gray zone of 75 to 125. Apply the 2.80 ratio the 2018 study found above 125 nmol/L for one assay pair and, again on our arithmetic, the same 45 mg/dL becomes 126 nmol/L, just over the line. A converted number near a cutoff cannot tell you which side of it you are on.

It matters less than it seems for one reason the EAS spells out: risk rises continuously with Lp(a) concentration, already at levels below the reported thresholds, so the cutoffs are described as pragmatic tools rather than a line where risk begins. On our reading, a result just under 125 nmol/L does not carry meaningfully less risk than one just over it. What changes is how the number is used in a risk discussion, which is why the unit should match the threshold being applied.

Read the unit before the number. Lab reports print it next to the value. Compare an nmol/L result only with nmol/L thresholds and an mg/dL result only with mg/dL thresholds.

Look for the assay name. The EAS statement says laboratories should include the assay name in reports, so discrepant results can be tracked. If yours does not, the lab can tell you which method it used.

Do not compare two labs as if they were one. The AHA statement's clinical-use table says Lp(a) should be measured with an isoform-insensitive assay, traceable to the WHO/IFCC reference material SRM-2B, and reported in nmol/L, and adds: "If measurements are not uniformly calibrated, one cannot compare measurements generated by different assays." A result that moves between two labs may reflect the assays, not you.

Ask for nmol/L if you have the choice. Both the EAS and the AHA statements prefer molar units, and the EAS says assays based on polyclonal antibodies, which strictly cannot report in molar units, should report in mg/dL when they cannot be compared with a molar reference method.

Our Lp(a) reference page lists the thresholds with their source and date, and our guide to what counts as a high Lp(a) explains why one test is usually enough and where the drug trials stand. Lp(a) also adds to the particle count that ApoB measures, which our guide to ApoB versus LDL cholesterol covers, and which blood tests are worth repeating explains why Lp(a) is a once-in-adulthood test. If you want the checks that apply at your age and sex in one place, your personal screening plan pulls them together.