The short answer: LDL cholesterol (LDL-C) measures how much cholesterol is riding inside your LDL particles. ApoB counts the particles themselves, because every particle that can get into an artery wall carries exactly one ApoB protein.
Most of the time the two numbers tell the same story. When they don't, the 2026 US cholesterol guideline says ApoB "predicts ASCVD risk more accurately than LDL-C." That gap, and who tends to fall into it, is the whole reason this test exists.
Status as of October 30, 2026: this post reports what published guidelines, drug labels and studies say. It isn't medical advice, and the numbers below are the targets each guideline sets for its own risk groups, not personal goals. What to test and what to do about a result is a conversation with your clinician.
Want the short version of what's actually approved, legal and tested? Get the free BHN cheat sheet.
Is ApoB the same as LDL?
No. They measure two different things about the same particles: one counts cholesterol, the other counts carriers.
The National Lipid Association (NLA) puts it plainly in its 2024 consensus statement: "There is a single apoB molecule found on each atherogenic lipoprotein," which covers LDL, VLDL, IDL and Lp(a) particles, so "the apoB concentration is a direct measure of the circulating burden of atherogenic lipoprotein particles" (Soffer et al., J Clin Lipidol, Sept 2024).
LDL-C is different in two ways:
It's a cholesterol weight, not a count. Two people with the same LDL-C can carry very different numbers of particles if one has smaller, cholesterol-poor particles.
It's usually calculated, not measured. The NLA notes that routine labs estimate LDL-C from an equation. For decades that was the Friedewald formula, which works backward from total cholesterol, HDL and triglycerides.
LDL-C | ApoB | |
|---|---|---|
What it measures | Cholesterol carried inside LDL particles (mg/dL) | Number of potentially artery-clogging particles: LDL, VLDL, IDL and Lp(a) |
How labs get it | Usually calculated from a standard lipid panel | Measured directly as its own test |
In a standard lipid panel? | Yes | No, ordered separately |
Fasting needed? | Depends on the panel | "Unaffected by fasting status," per the 2026 guideline |
Main guideline role | Primary treatment goal in US and European guidelines | Add-on to refine risk, mostly once LDL-C goals are met |
Why do some guidelines and cardiologists prefer ApoB?
Because when the two numbers disagree, risk tends to follow ApoB. Every major guideline now gives ApoB a role, but none has made it the main target over LDL-C.
Here's what each one says, in its own words:
2018 AHA/ACC cholesterol guideline (now replaced). ApoB at or above 130 mg/dL counted as a "risk-enhancing factor," with "a relative indication for its measurement" being triglycerides of 200 mg/dL or more (Grundy et al., Circulation, 2019).
2019 ESC/EAS guideline (Europe). ApoB "is the preferred measurement to further refine the estimate of ASCVD risk," and measuring it with non-HDL cholesterol "is recommended as part of routine lipid analysis" in people with high triglycerides (Mach et al., Eur Heart J, 2019).
2024 NLA consensus. ApoB "has been shown to be superior to LDL-C in risk assessment both before and during treatment" (NLA, 2024).
2026 ACC/AHA dyslipidemia guideline (current US guideline). For adults already on cholesterol-lowering treatment, "particularly those with ASCVD, CKM syndrome, type 2 diabetes, and/or elevated TG," measuring ApoB "is reasonable to guide decisions regarding further therapeutic intensification once LDL-C and/or non-HDL-C goals are achieved." That's a Class 2a recommendation, one step below the strongest (Circulation, Mar 2026).
Read that last one carefully. LDL-C is still the first number, and the American College of Cardiology's own summary frames ApoB as a check on "residual ASCVD risk" in people "who have reached their LDL-C and non-HDL-C goals" (ACC, Mar 13, 2026). A second opinion, not a replacement.
When does LDL look fine but ApoB doesn't?
Mostly in people with metabolic problems. The 2026 guideline says discordance "is most common in individuals with cardiometabolic disease, including ASCVD, CKM syndrome, diabetes, and/or TG ≥150 mg/dL," and that "LDL-C may appear at goal while apoB remains elevated, masking residual risk" (Circulation, Mar 2026).
How often? The 2019 European guideline says "in around 20% of patients there may be discordance between measured LDL-C and ApoB levels" (Eur Heart J, 2019).
The clearest data comes from Denmark. Researchers followed 13,015 adults on statins in the Copenhagen General Population Study for a median of 8 years and split each marker at its median (Johannesen et al., JACC, 2021):
High ApoB with low LDL-C: 21% higher risk of death from any cause (hazard ratio 1.21) and 49% higher risk of heart attack (1.49), compared with people low on both.
High LDL-C with low ApoB: no increased risk of death or heart attack.
The same group later looked at 95,108 adults not on statins and found that "excess" ApoB, the amount above what their LDL-C predicted, raised heart attack and ASCVD risk in a dose-dependent way in both women and men, "across the entire LDL-C spectrum" (Johannesen et al., JACC, 2024). These are observational studies. They show association, not that lowering ApoB specifically fixes it.
One wrinkle most ApoB content skips: the LDL formula matters. The 2026 guideline says the newer Martin/Hopkins equation "markedly reduces discordance with apoB compared with the Friedewald equation." Among people with an LDL-C under 70 or under 100 mg/dL by Martin/Hopkins, only about 2% and 1% had non-HDL-C and ApoB above guideline targets. The same guideline now prefers Martin/Hopkins or the Sampson/NIH equation over Friedewald. If your lab report says how LDL-C was calculated, that line is worth reading.
What is a good ApoB level?
There isn't one universal number. Each guideline sets ApoB numbers by risk group, and they don't all match.
Guideline | ApoB numbers it names | What the number is for |
|---|---|---|
2019 ESC/EAS (Europe) | Under 65 mg/dL very high risk; under 80 high risk; under 100 moderate risk | "Secondary goals," which the guideline says "have not been extensively studied in RCTs" |
2024 NLA consensus | 60 very high risk; 70 high risk; 90 borderline to intermediate risk | "Suggested" thresholds at or above which to consider stepping up treatment |
2026 ACC/AHA (US) | 120 or higher is a risk-enhancing factor; 140 or higher counts as severe hypercholesterolemia | Flags higher lifetime risk; not a goal |
2026 ACC/AHA (US) | Under 100 alongside an LDL-C goal under 100, for people with mild coronary calcium; under 55 in familial hypercholesterolemia with ASCVD | Goals in those specific groups |
Sources: ESC/EAS 2019; NLA 2024, Table 4; ACC/AHA 2026.
The 2025 ESC/EAS focused update kept the 2019 LDL-C goals and didn't publish new ApoB goals (Mach et al., Eur Heart J, 2025).
For context on where people sit, the NLA reports that in untreated US adults (NHANES 2005 to 2016, n = 12,696), the median ApoB was 90 mg/dL and the 20th percentile was 70 mg/dL (NLA, Table 1). So the European "high risk" goal of under 80 is below what most American adults run. That's a population snapshot, not a target.
ApoB vs Lp(a): what's the difference?
Lp(a) is one type of particle. ApoB is the count of all of them, Lp(a) included.
The 2026 guideline describes Lp(a) as "an LDL-like particle" that "carries a single apoprotein(a) strand bound to its apoB-100 component," with levels "mostly genetically determined by the LPA gene." The 2019 European guideline puts it at "about 90%" inherited. That's why the two tests answer different questions:
ApoB tells you your total particle load, which moves with diet, weight, medications and metabolic health.
Lp(a) tells you about one inherited particle type. The NLA notes elevated Lp(a) "is not always accompanied by apoB elevation," so a normal ApoB doesn't rule it out.
Both current guidelines say to test Lp(a) at least once. The 2026 ACC/AHA guideline makes it a Class 1 recommendation for all adults and calls 125 nmol/L (50 mg/dL) or higher a risk-enhancing factor, tied to about 1.4 times the ASCVD risk, with 250 nmol/L (100 mg/dL) or higher tied to 2 times or more (Circulation, Mar 2026). The 2025 European update says Lp(a) measurement "should be considered at least once in every adult's lifetime" and uses above 50 mg/dL (above 105 nmol/L) as its cut point (Eur Heart J, 2025).
Notice the two guidelines convert 50 mg/dL to different nmol/L numbers. Lp(a) is reported in either unit depending on the lab, so check which one is on your report before comparing it to anything.
How do you get an ApoB test, and what does it cost?
Ask your clinician to add it to a lipid panel, or buy it direct where your state allows. The test itself is cheap to run. Getting it covered is the hard part.
The NLA says the assay's "cost of performing the test is low," but that "health insurance payment denials or prior authorization requirements often occur for apoB testing in routine practice" (NLA, 2024). Its authors call for better "access and reimbursement."
Here's what the two big US labs list on their own consumer sites, checked October 30, 2026:
Lab (consumer site) | ApoB | Lp(a) | Notes on the page |
|---|---|---|---|
Labcorp OnDemand | $69 standalone (shown at $51.75 in an October sale) | $49 | A healthcare provider reviews the order, no visit needed; drawn at a Labcorp location; "No special preparation needed" |
Quest (questhealth.com) | No standalone ApoB test listed in its heart health category. ApoB comes in a panel at $215 plus a $6 physician service fee | $45 plus a $6 physician service fee | The panel also includes cholesterol, triglycerides, hs-CRP, Lp(a) and Lp-PLA2 |
Prices change and sales end. Consumer ordering also isn't available everywhere; the state rules are covered in blood work before peptides.
Does ApoB matter more if you're on a GLP-1 or peptides?
For GLP-1s, there's a reasoned case, not a proven one. For research peptides, there's no data either way.
The GLP-1 labels show the pattern. Triglycerides fall a lot more than LDL-C:
Zepbound (tirzepatide): in its 72-week trial in adults with obesity and no diabetes, LDL-C fell 2.9% to 5.5% more than placebo across doses, while triglycerides fell 16.5% to 24.9% more. The label marks these as not controlled for statistical error, meaning they weren't formal trial endpoints (Zepbound label, rev. 08/2026).
Wegovy (semaglutide): at 68 weeks in adults with obesity or overweight and another condition, LDL-C fell 3.8% more than placebo and triglycerides 15.8% more. Also not part of the pre-specified testing (Wegovy label, rev. 06/2026).
Neither label reports ApoB. So what happens to particle count on these drugs isn't something the labels can tell you.
Here's the reasoning, and it's mine, not a guideline's: the people GLP-1s are prescribed for are often the same people the 2026 guideline names as most likely to have discordance (diabetes, high triglycerides, cardiometabolic disease). If your LDL-C looked "fine" before and you want to know whether your particle count agrees, ApoB is the test that answers that.
For research peptides, I found no published study of ApoB on any of them. Retatrutide has no approved label at all. If you're running anything, the case for a baseline is the same one in the bloodwork post: one result tells you where you are, two tell you what changed.
That baseline is less common than you'd think. In BHN's October reader survey (38 readers, self-selected, not a study), 8 said they get no bloodwork at all, and 5 more test only before starting.
The short version
LDL-C weighs the cholesterol. ApoB counts the particles. When they disagree, studies find risk tracks ApoB, most often in people with diabetes, high triglycerides or metabolic disease.
The 2026 US guideline keeps LDL-C first and calls ApoB "reasonable" once LDL-C goals are met. No guideline gives one "good" ApoB number for everyone.
Lp(a) is a separate, mostly inherited particle. Both guidelines say test it at least once.
For the short take on why a normal cholesterol number can miss this, see why "normal cholesterol" can still miss the problem. For the core panel, see the minimum bloodwork stack.
If you've had ApoB and LDL-C tested on the same draw, reply and tell me whether they agreed. And if you're on a GLP-1, whether either one moved. Nobody publishes that for real-world users.
For research-grade context only, here's how I pick a peptide vendor.
The vendor links on that page are affiliate links. If you buy through them, BHN earns a commission.
Want the one-page version of what's approved, what's legal and what's actually been tested? Get the free BHN cheat sheet.