What a Lipoprotein Panel Measures and How to Read One

A lipoprotein panel measures the particles that carry cholesterol through your blood, rather than the cholesterol itself. A standard lipid panel gives you four numbers (total cholesterol, HDL-C, triglycerides, and an LDL-C value usually calculated rather than measured). A lipoprotein panel adds particle-level measurements: apolipoprotein B (apoB, one molecule per atherogenic particle), lipoprotein(a), and in the expanded versions, particle counts and sizes by NMR or ion mobility, plus apoA-I and sometimes oxidized LDL. If you order one test beyond the basic lipid panel, order apoB. If you order two, add Lp(a) once in your life, because it is set almost entirely by genotype.

Interpretation of these numbers, and any decision about what to do with them, belongs with a clinician, ideally a lipidologist or preventive cardiologist. What follows is about measurement: what the assays do, where they fail, and how to make the numbers comparable over time.

Why particle count beats cholesterol mass

LDL-C tells you how much cholesterol is being carried by LDL particles. It does not tell you how many particles are doing the carrying. The two numbers track each other loosely in most people and come apart badly in the people you most want to identify: insulin-resistant patients with high triglycerides, who tend to have small, cholesterol-poor LDL particles. A person with LDL-C of 95 mg/dL and 1,400 nmol/L of LDL particles is carrying a very different arterial burden than one with the same LDL-C and 900 nmol/L. Discordance between LDL-C and particle measures is where the risk information lives, and apoB tracks the outcome better than LDL-C when the two disagree ¹.

ApoB is the cleanest way to get at this. Every LDL, IDL, VLDL, and Lp(a) particle carries exactly one apoB-100 molecule, so serum apoB in mg/dL is a direct proxy for the count of atherogenic particles. It is measured by immunoassay (immunoturbidimetric or nephelometric), it is standardized against WHO/IFCC reference material, the CV is typically 3–5%, and it does not require fasting. Compare that to calculated LDL-C, which inherits error from three other measurements and from whichever equation the lab used.

Check which equation your lab uses. Friedewald (LDL-C = TC − HDL-C − TG/5) is still common and systematically underestimates LDL-C at low LDL-C and high triglycerides. Martin-Hopkins uses an adjustable divisor and is meaningfully better in that region. Sampson/NIH equation 2 was designed for hypertriglyceridemia. If your LDL-C value drops 12 mg/dL between two draws and nothing else changed, check whether the lab switched equations. This is the single most common artifact in longitudinal lipid tracking.

Lp(a): the one test you get once

Lipoprotein(a) is an LDL-like particle with apolipoprotein(a) covalently bound to apoB-100. Plasma concentration is roughly 70–90% heritable, driven mostly by the number of kringle IV type 2 repeats in the LPA gene. Fewer repeats give a smaller isoform and higher plasma Lp(a). It is stable across adult life apart from modest increases with kidney disease and menopause, and it barely moves with diet, exercise, or statins. That is why the advice is to measure once and record it.

The units problem is real and trips up almost everyone. Lp(a) is reported either in mg/dL (mass, including the cholesterol, phospholipid, and protein of the whole particle) or in nmol/L (molar particle concentration). There is no valid universal conversion factor, because the mass of apo(a) varies with isoform size. The commonly cited ×2.15 or ×2.5 multipliers are approximations that fail worst in the people with the most extreme isoforms. Ask for nmol/L, from an isoform-insensitive assay calibrated to the WHO/IFCC SRM 2B reference material. If you have a historical mg/dL value, keep it as mg/dL in your records and do not convert.

For the common question about a value of 42: that number is uninterpretable without units. 42 mg/dL sits around the typical threshold where risk begins to rise steeply. 42 nmol/L is well within the range most labs flag as normal. Thresholds in wide use are roughly 30 mg/dL or 75 nmol/L for “elevated” and 50 mg/dL or 125 nmol/L for “high,” and they are population cutpoints rather than biological switches. Distribution is strongly skewed and varies by ancestry, with substantially higher median levels in people of African descent. Take the interpretation to a clinician who works with Lp(a) regularly.

Doctors do not order it routinely for a few reasons that made sense historically and make less sense now: assay standardization was poor until recently, there was no therapy that lowered it specifically, and reimbursement was inconsistent. Assay quality has improved and apo(a)-targeted agents are in late-phase trials, so ordering patterns are shifting.

The expanded panel: NMR, ion mobility, oxLDL and inflammatory markers

NMR (LabCorp NMR LipoProfile) reads lipoprotein subclass concentrations from the methyl signal lineshape and reports LDL-P in nmol/L, small LDL-P, HDL-P, VLDL size, and a derived insulin resistance score. Ion mobility (Quest Cardio IQ) separates particles by size in the gas phase and counts them directly. Both give you more resolution than apoB. Neither is as well standardized between platforms, so comparing an NMR LDL-P from 2021 against an ion mobility result from 2026 is not a comparison. Pick a platform and stay on it. The classical methods behind all of this, ultracentrifugation and electrophoresis, are still the reference frame the immunoassays are calibrated against ².

Oxidized LDL is measured by ELISA against a monoclonal antibody (usually 4E6 or DLH3) recognizing oxidized apoB epitopes. It is biologically interesting: oxLDL is what macrophage scavenger receptors take up to become foam cells, and circulating levels associate with plaque burden and event risk ³. The problem is preanalytical. Different antibodies measure different things, there is no reference standard, and values are sensitive to sample handling and freeze-thaw. Treat oxLDL as a research-grade number within a single lab. The value is not a clinical threshold.

hs-CRP is worth adding because it captures a risk axis that lipoproteins miss entirely. It is inexpensive, standardized, and the residual inflammatory risk story is well characterized ⁴. Get it when you are not sick: any intercurrent infection can push CRP up tenfold and wash out the signal you wanted. Repeat and take the lower of two values at least two weeks apart.

Making the numbers comparable over time

• Same lab, same assay, same platform. Method differences between labs exceed most of the biological changes you are trying to detect.
• Fasting status matters for triglycerides and therefore for calculated LDL-C. ApoB, Lp(a), and hs-CRP do not require fasting. If you want a consistent series, standardize on 12-hour fasted morning draws anyway.
• Biological variability for triglycerides is large (within-person CV around 20%). Two draws a week apart, averaged, will save you from chasing noise.
• Record the units, the assay name, the equation, and the reference interval alongside every value. A CSV with columns date, analyte, value, unit, assay, lab, fasting_hours, ref_low, ref_high costs nothing to maintain and makes every later analysis possible.
• Avoid drawing within a few weeks of an acute illness, surgery, or a major weight change. Lipoproteins are acute-phase reactive in both directions.

Diet and lifestyle move apoB, triglycerides, and HDL function meaningfully, and the response varies between people in ways partly predictable from genotype and baseline metabolic state ¹. Controlled feeding trials show measurable shifts in lipoprotein and vascular biomarkers over months rather than weeks ⁵. Lp(a) is the exception: it mostly does not move. If you want to know what an intervention did to your lipoproteins, measure before, hold everything else steady, and measure again at 12 weeks.

Questions people also ask

Can I live a long life with high Lp(a)? Yes. Elevated Lp(a) raises lifetime risk of atherosclerotic events and aortic valve calcification, and plenty of people with high levels never have an event. It is one input to a risk estimate, alongside apoB, blood pressure and smoking. Glucose and family history belong in the same estimate. Discuss the aggregate with a clinician rather than reacting to the single number.

What not to eat with high Lp(a)? No dietary pattern reliably lowers Lp(a), and some low-saturated-fat patterns nudge it slightly upward even as they lower apoB. Diet is worth attending to for the other lipoprotein measures. Eating a certain way is not an Lp(a) strategy.

Should I take a statin if I have high Lp(a)? That is a clinical decision, and not one to make from a web page. Statins lower apoB substantially and do not lower Lp(a). Bring the full panel to a physician.

Are there symptoms of high lipoproteins? No. Atherosclerosis is silent until it is not. Extreme familial hypercholesterolemia sometimes produces visible tendon xanthomas or corneal arcus before age 45, and those warrant prompt medical evaluation, but the ordinary case has no symptoms at all.

Why is Lp(a) not routinely tested? Historically poor assay standardization, absence of a specific therapy, and inconsistent coverage. All three are changing, and several guidelines now recommend measuring once in every adult.

Footnotes

1. Catherine J. Andersen, María Luz Fernández. Emerging Biomarkers and Determinants of Lipoprotein Profiles to Predict CVD Risk: Implications for Precision Nutrition. Nutrients, 2024. https://doi.org/10.3390/nu17010042 ↩ ↩²

2. Nader Rifai, G. Russell Warnick, Marek H Dominiczak. Handbook of Lipoprotein Testing. Medical Entomology and Zoology, 2001. http://ci.nii.ac.jp/ncid/BA36965851 ↩

3. Andreja Trpković, Ivana Resanović, Julijana Stanimirović, et al. Oxidized low-density lipoprotein as a biomarker of cardiovascular diseases. Critical Reviews in Clinical Laboratory Sciences, 2014. https://doi.org/10.3109/10408363.2014.992063 ↩

4. Sotirios Tsimikas, James T. Willerson, Paul M. Ridker. C-Reactive Protein and Other Emerging Blood Biomarkers to Optimize Risk Stratification of Vulnerable Patients. Journal of the American College of Cardiology, 2006. https://doi.org/10.1016/j.jacc.2005.10.066 ↩

5. Peter J. Curtis, Vera van der Velpen, Lindsey Berends, et al. Blueberries improve biomarkers of cardiometabolic function in participants with metabolic syndrome—results from a 6-month, double-blind, randomized controlled trial. American Journal of Clinical Nutrition, 2018. https://doi.org/10.1093/ajcn/nqy380 ↩