LDL vs Non-HDL vs ApoB: Which Cholesterol Number Actually Predicts Risk?
Bottom line: LDL cholesterol measures the cholesterol inside your particles. ApoB counts the particles themselves. When those two disagree, the particle count is the better predictor of risk, and you can be at target on LDL while ApoB says otherwise. Non-HDL sits between them, costs nothing extra, and you can work it out from a standard panel right now: total cholesterol minus HDL.
What each number actually measures
All three describe the same underlying problem, cholesterol-carrying particles lodging in artery walls. They just count it differently.
LDL cholesterol measures the mass of cholesterol carried inside your low-density lipoprotein particles. On most lab reports it is not measured directly at all, it is estimated using the Friedewald equation from total cholesterol, HDL and triglycerides.
Non-HDL cholesterol is every atherogenic particle's cholesterol added together, LDL plus VLDL plus remnants plus Lp(a). It is total cholesterol minus HDL. Nothing extra needs measuring, which is why it is on this page.
ApoB (apolipoprotein B) counts particles. Every atherogenic particle carries exactly one ApoB molecule, so ApoB is a direct particle count rather than a cholesterol weight.
Side by side
| LDL-C | Non-HDL-C | ApoB | |
|---|---|---|---|
| What it counts | Cholesterol inside LDL particles | Cholesterol in all atherogenic particles | The particles themselves |
| On a standard panel? | Yes, usually estimated | Yes, you calculate it | No, must be ordered |
| Affected by high triglycerides? | Yes, becomes unreliable | No | No |
| Needs fasting? | Ideally | No | No |
| Desirable, general population | <100 mg/dL | <130 mg/dL | Discuss with your clinician |
| Very high risk target | <55 mg/dL | <85 mg/dL | Discuss with your clinician |
Non-HDL bands, 2026 ACC/AHA dyslipidemia guideline: desirable below 130 mg/dL, borderline high 130 to 159, high 160 to 189, very high 190 or above. LDL bands are the long-standing NCEP ATP III classification.
How to work out your own non-HDL
You almost certainly already have the two numbers you need on a report you have been sent.
Non-HDL cholesterol = total cholesterol − HDL cholesterol
A useful rule of thumb from the guidelines: your non-HDL target is your LDL target plus 30 mg/dL. If your LDL goal is under 100, your non-HDL goal is under 130. If you are very high risk and your LDL goal is under 55, your non-HDL goal is under 85.
This matters most when triglycerides are raised. The Friedewald equation that estimates your LDL becomes unreliable as triglycerides climb, so a reassuring LDL can be an artefact of the maths. Non-HDL does not use that equation and is unaffected.
Look up your own values: total cholesterol, HDL, LDL and triglycerides.
Why ApoB wins when the numbers disagree
Most of the time these three markers agree, and when they agree the distinction is academic. The interesting case is discordance, where LDL says one thing and the particle count says another.
In a systematic review of discordance analyses, ApoB was superior to LDL-C in 9 out of 9 studies, and superior to non-HDL-C in 7 out of 9. That is about as consistent as this kind of evidence gets.
The mechanism is simple. Particles vary in how much cholesterol they carry. Someone with many small, cholesterol-depleted particles can post a perfectly respectable LDL number while carrying a high particle count, and it is the particle count that drives risk. LDL cannot see that. ApoB can.
The 2026 ACC/AHA dyslipidemia guideline reflects this: ApoB testing is now an official recommendation, particularly for people already on cholesterol-lowering medication, and for those with existing cardiovascular disease, type 2 diabetes, raised triglycerides, or an already-low LDL. Where LDL and non-HDL targets are met but ApoB stays elevated, the guideline recommends intensifying treatment.
When a normal LDL still hides risk
Four situations where a reassuring LDL deserves a second look. None of these is a diagnosis, and none of them means anything is wrong. They are reasons to ask a question.
- Raised triglycerides. The estimate behind your LDL degrades. Check non-HDL.
- Type 2 diabetes or metabolic syndrome. Particle counts often run high relative to the cholesterol they carry.
- Already on a statin with a low LDL. This is precisely the group the 2026 guideline names for ApoB testing.
- Family history that the numbers do not explain. Lp(a) is largely inherited, is included in non-HDL, and is not captured by LDL. It is also in the 2026 guideline.
What to ask your doctor to test
Non-HDL is free, you can calculate it today from a report you already have. ApoB and Lp(a) have to be ordered, are widely available, and are inexpensive relative to the rest of a workup. A reasonable question at your next appointment: "my LDL is X and my non-HDL is Y, would ApoB or Lp(a) add anything for someone like me?"
What none of this replaces is the conversation. Treatment decisions rest on your whole risk picture, not one marker, and that is a judgement for your clinician.
Frequently Asked Questions
Is non-HDL better than LDL?
For predicting risk it is at least as good, and it has two practical advantages: it captures every atherogenic particle rather than LDL alone, and it does not rely on the Friedewald estimate, so it stays reliable when triglycerides are raised. It also costs nothing extra, because it is total cholesterol minus HDL.
How do I calculate non-HDL cholesterol?
Subtract your HDL from your total cholesterol. If total is 210 mg/dL and HDL is 50 mg/dL, your non-HDL is 160 mg/dL. Desirable is under 130 mg/dL in the general population.
Is ApoB worth paying for?
It is the most consistent predictor of the three: superior to LDL-C in 9 of 9 discordance studies and to non-HDL-C in 7 of 9. The 2026 ACC/AHA guideline recommends it particularly for people on cholesterol-lowering medication, or with cardiovascular disease, type 2 diabetes, raised triglycerides, or a low achieved LDL. Whether it changes anything for you is a question for your clinician.
Can I have a normal LDL and a high ApoB?
Yes, and that is the whole reason the distinction matters. Particles differ in how much cholesterol each one carries, so a person with many cholesterol-poor particles can show an acceptable LDL alongside a high particle count. The 2026 guideline addresses this case directly, recommending intensified treatment where LDL and non-HDL targets are met but ApoB remains elevated.
Do I need to fast for these tests?
Non-HDL and ApoB do not require fasting. LDL is ideally measured fasting because the estimate depends on triglycerides, which rise after eating.
BloodMarker does not provide medical advice, diagnose conditions, or recommend treatments. Reference ranges used across this site are published in full at doi.org/10.5281/zenodo.21694169. Talk to a clinician about your own results.