Reading Your Cholesterol Panel: What's Worth Knowing

A lipid panel is one of the most commonly ordered blood tests in the country, and one of the most commonly contested. Parts of it anyway. Most people get theirs back, scan for the red text, and land in one of three camps. Panic, or dismissal, or wait for their doctor’s office to call.

This comes up a lot in two groups in particular: women moving through perimenopause, and people eating lower carb. Both tend to see their numbers shift, and both deserve better than "your cholesterol is high, here is a statin" on one side, or "cholesterol does not matter" on the other.

Here is what is worth knowing before your next blood test - save this for later just in case!

Start with non-HDL, not LDL

On most panels, LDL is calculated rather than measured. The standard Friedewald equation estimates it as total cholesterol minus HDL minus triglycerides divided by 2.2.

So, despite confidence in that number (and the focus on it), your lab doesn't actually measure your LDL. It measures three things: total cholesterol, HDL, and triglycerides. Then it works out LDL by subtraction. Take the total, subtract the HDL, subtract an estimate of the cholesterol sitting in a third particle called VLDL, and whatever is left over gets reported as your LDL.

That estimate of VLDL is the weak link. There's no test for it on a standard panel, so the lab guesses it from your triglycerides using a fixed rule of thumb. The rule works fine when triglycerides are in a normal range. When triglycerides are high, the guess comes out too big, and because that guess is being subtracted, too much gets taken off the total. Your LDL is reported lower than it really is and is a known limitation of the Friedewald equation.

It's a bit like working out what's inside a parcel by weighing the whole thing and subtracting your estimate of the packaging. Guess the packaging too heavy and the contents look lighter than they are.

It’s good to know that the error only runs one way. High triglycerides make LDL look better than it is, never worse. So if your triglycerides are up and your LDL came back reassuring, that's the number to be careful with. A high LDL alongside normal triglycerides is a number you can trust.

Non-HDL cholesterol is often already on your panel, and if it is not, you can work it out in about two seconds: total cholesterol minus HDL. It captures every potential atherogenic (artery-damaging) particle in the sample rather than just LDL, it is not affected by whether you fasted, and it costs nothing extra.

ApoB

Every atherogenic particle carries exactly one apolipoprotein B molecule. So apoB is a direct count of the particles that can lodge in an artery wall, rather than an estimate of what is inside them.

When apoB and LDL cholesterol disagree, and they disagree more often than you would expect, apoB is the one that tracks risk. A 2022 analysis in JAMA Cardiology pooled UK Biobank data with two large lipid-lowering trials and found that once apoB was accounted for, cholesterol concentration and triglyceride concentration added essentially nothing to risk prediction.

ApoB is not standard on a New Zealand lipid panel, but you can ask for it and it is inexpensive. If your doctor is concerned about your cholesterol levels, it makes sense to also get this tested.

The triglyceride to HDL ratio

Plenty of people use the TG to HDL ratio as a useful check for insulin sensitivity.

What it is not is a measure of how many atherogenic particles you are carrying. A favourable ratio does not cancel out a high non-HDL. You can be insulin sensitive and still be carrying a lot of apoB. The ratios are different depending on where you live, because they have different conversion factors.

LDL particle size

The TG:HDL ratio also tells you about LDL particle size. I used to understand that if your LDL particles are large and buoyant (as a TG:HDL at optimal would indicate) rather than small and dense, a high LDL is nothing to worry about.

The EPIC-Norfolk study though found LDL particle size association disappeared entirely once particle number was accounted for. Every particle carries one apoB whether it is large or small, therefore it isn’t as protective a factor as I once thought (and what you might hear).

If you are in perimenopause, expect a shift

The SWAN study followed just over a thousand women through their final menstrual period specifically to separate what changes because of ageing from what changes because of menopause. Total cholesterol, LDL and apoB all rose sharply in the year either side of the final period. After that first year they stayed high but stopped climbing. This was in contrast to glucose, clotting factor and other risk factors which continued to climb, suggesting the transition accounted for the change in cholesterol numbers, rather than simply getting older.

Therefore, if your numbers move in your late forties or early fifties, as often seen, this increased exposure to cholesterol does increase your risk, and would be a reason to look at the panel properly rather than dismiss it and think that worrying about your cholesterol is some kind of big pharma conspiracy.

If you eat low carb

Cholesterol can shift when carbohydrate restricted, but this doesn’t immediately mean anyone with cholesterol changes is high risk. Clinically speaking (and from the literature) we see many people with poor metabolic health and carry excess body fat have an improvement in their lipid profile with weight loss. Conversely, LDL can rise on carbohydrate restriction, sometimes substantially, and it tends to happen most in lean, metabolically healthy people.

Currently, it is uncertain as to the long term implications of this. A study on lean mass hyper responders was retracted from JACC: Advances earlier this year, at the request of its own authors, after problems emerged with the plaque analysis. The authors raised the concerns themselves and pushed for the retraction, which is to their credit. In the studied cohort, it is unclear whether the high cholesterol numbers are problematic or not. We do not have long-term outcome data in either direction.

One panel is a snapshot, not a story

Possibly the most useful thing you can do is dig out your old results. A non-HDL of 5.5 that was 3.5 five years ago is a completely different situation from one that has sat at 5.5 your entire adult life. The first points to something acquired: a diet change, thyroid function, menopause, a medication. The second raises the question of familial hypercholesterolaemia, an inherited condition that is more common than most people assume and potentially significantly under-diagnosed.

Family history matters here too. Heart disease in a parent or sibling before about 55 for men or 60 for women changes the picture considerably.

What to ask for

• Your previous lipid results over several years, so you can see the direction of travel rather than a single point
• ApoB, once, to see whether it agrees with your LDL
• Lp(a), once in your life. It is genetic and it does not respond to diet. This can be done through Nutrisearch here in NZ (likely other places too)
• A proper global risk assessment rather than a decision made off one red flag on one line

None of this is an argument for ignoring a high LDL, and none of it is an argument for treating every high reading as an emergency. We just need a full picture. It’s worth noting, also, that cholesterol is labile and, as part of our innate immune system, reactive to situations of inflammation, illness, infection, injury and also weight loss (as free fatty acids are mobilised this can transiently push LDL up), therefore a history of trends is worthwhile.