Non-HDL Cholesterol: Your Risk as You Get Older

Non-HDL cholesterol is not measured separately. The laboratory subtracts your HDL cholesterol from your total cholesterol, and what remains is non-HDL. That simple subtraction has a precise biological meaning.

Cholesterol does not dissolve in blood and therefore travels packaged inside lipoproteins. Some of those particles carry cholesterol away to the liver: those are the HDL particles. Every other lipoprotein can leave cholesterol behind in the artery wall, and each of them carries exactly one molecule of apolipoprotein B. Subtracting HDL from the total therefore leaves you with exactly the cholesterol sitting inside those ApoB-bearing particles: LDL, IDL, VLDL, the remnant particles left after fat is digested, and Lp(a).

An LDL value tells you only what is inside the LDL particles. Non-HDL counts the whole family. That makes it the cheapest measure of total atherogenic burden in existence: it costs no extra tube and no extra assay, because total cholesterol and HDL already appear on every lipid panel.

Non-HDL is a derived value, but not every derived number is equally fragile. The difference from a calculated LDL is fundamental, and that comparison explains almost everything worth saying about non-HDL.

The table shows why non-HDL is so usable. It is a subtraction of two directly measured values and it makes no assumption at all about your triglycerides. A calculated LDL does make one, because the Friedewald formula subtracts your triglycerides divided by 2.2 from the non-HDL. LDL therefore inherits a weakness that non-HDL simply does not have.

Atherosclerosis develops because ApoB-bearing particles enter the artery wall and deposit cholesterol there over decades. The more of those particles pass by, the greater the chance that something stays behind. A measure that includes every atherogenic particle therefore matches the underlying process better than one that looks only at LDL.

This is not merely theory. SCORE2, the risk model the European cardiology guideline has used since 2021 to estimate 10-year cardiovascular risk, works from exactly five inputs: age, sex, smoking, systolic blood pressure and non-HDL cholesterol. LDL is not among them, and the total-to-HDL ratio that preceded it has been dropped as well. Non-HDL is therefore not some alternative value at the margins; it is the lipid number Europe estimates risk with.

Non-HDL makes the clearest difference in a triglyceride-rich profile. With insulin resistance, central adiposity or type 2 diabetes, the LDL often looks perfectly tidy while a great deal of VLDL and remnant particles circulate alongside it. Those particles are counted in non-HDL and not in LDL. An LDL of 2.8 mmol/l next to triglycerides of 3.0 mmol/l can therefore coincide with a clearly raised non-HDL. That is not a measurement error or a quirk of the arithmetic, but a genuine signal that LDL alone does not pick up.

Then the target, and this is where most results are misread. The 3.3 mmol/l on your report is a population reference interval: it describes what is usual, not what is desirable. The European dyslipidaemia guideline ties the non-HDL target to your risk category, and sets it 0.8 mmol/l above the corresponding LDL target every time.

The fact that the 3.3 mmol/l reference value happens to sit close to the moderate-risk goal makes it all the more treacherous. At high or very high risk, 3.3 mmol/l is not reassurance but comfortably above target, while the result itself flags nothing at all. Which category applies to you is judged by your doctor on the basis of age, blood pressure, smoking, diabetes, family history and any previous cardiovascular disease.

Finally, the relationship with ApoB. The two numbers usually say the same thing. Where they diverge, ApoB is the purer measure, because ApoB counts the particles themselves while non-HDL still measures the cholesterol inside them. Non-HDL is thus the free, always-available approximation of ApoB.

Non-HDL cholesterol falls out of any lipid panel automatically, because total cholesterol and HDL are on it anyway. Nothing extra has to be drawn for it, and you do not need to fast: across the whole lipid panel only triglycerides are genuinely meal-sensitive, and that value does not enter the subtraction.

The value is especially useful when your triglycerides are raised, when you have diabetes or metabolic syndrome, and when the sample was taken non-fasting. In all of those situations the calculated LDL on the very same report is less trustworthy than the non-HDL sitting next to it. If a non-fasting triglyceride result comes back above roughly 4.5 mmol/l, have the panel repeated fasting before anything is read into it.

Do not test during or shortly after an acute illness. After a heart attack, major surgery or a significant infection, cholesterol values fall for weeks, so a panel drawn in that window underestimates your habitual value. Wait a few weeks after recovery.

A single result remains a snapshot. Total cholesterol and LDL vary by some 5 to 10 percent from day to day, triglycerides by as much as 20 to 25 percent. A difference of 0.2 mmol/l between two measurements is therefore noise, not a trend. If you follow the value over time, have it measured at the same laboratory each time.

If your non-HDL is unexpectedly high, the first question should be whether an underlying cause is at play. An underactive thyroid is the classic missed explanation, but poorly controlled diabetes, kidney disease, liver disease and cholestasis, heavy alcohol use, pregnancy and a range of medicines can all disturb a lipid profile. That is a conversation with your doctor, not a puzzle to solve on your own.