There’s a type of patient I keep seeing, and I think about them more than any other.

Late 30s or early 40s. Active. No smoking, no diabetes, blood pressure fine. Father had bypass surgery at 55, maybe an uncle who had a heart attack “young” — but the patient in front of me looks nothing like that story. Their lipid panel is unremarkable. Low-density lipoprotein (LDL) under 100, sometimes under 90. High-density lipoprotein (HDL) decent. Triglycerides fine. Every calculator I run puts them at low to intermediate risk, and the standard of care says reassure them, talk about lifestyle, see them back in a year.

I started feeling uneasy about that pattern a few years ago. Not because the numbers were wrong, but because I kept seeing that same patient show up in cardiology referrals — or worse, coming back from a calcium score ordered for something else with findings that didn’t match anything their lipid panel had suggested. At some point, I decided to start looking harder. I added apolipoprotein B (ApoB) and lipoprotein a (Lp(a)) to the workup on anyone who fit that profile: young, fit, family history, clean standard panel.

What I found genuinely surprised me. Elevated ApoB in patients whose LDL I would have called controlled. Lp(a) levels well above 125 nmol/L in patients who had never heard the term. Risk that the standard lipid panel had no mechanism to capture, sitting right there underneath numbers that looked fine.

This spring, the framework I’d been working around finally caught up. The 2026 ACC/AHA Dyslipidemia Guideline — the first major overhaul since 2018 — formalized what a lot of us had been quietly doing: looking beyond LDL.

The return of the number

The first thing the guideline restores is something deceptively simple: a target. After years of percentage-based LDL reduction, the 2026 update brings back explicit LDL-C goals. Under 100 mg/dL for primary prevention in intermediate-risk patients. Under 70 for high risk. Under 55 for those with established atherosclerotic cardiovascular disease (ASCVD) at very high risk. That last threshold is aggressive and intentionally so — the underlying argument is that lower LDL for longer, starting earlier in life, compounds protection in ways that treating a number at age 55 simply can’t replicate.

Having a target also changes how the conversation goes in the room. When I can tell a patient we’re aiming for a specific number and we’re not there yet, it’s harder for either of us to settle. That matters more than it sounds.

The particle problem

Here is the thing about LDL-C that the guideline now formally acknowledges: It measures the cholesterol mass being carried by atherogenic lipoproteins, not the number of particles doing the carrying. Two patients can have an identical LDL-C and completely different particle counts. The one with more particles — smaller, denser, more numerous — has more opportunities for arterial wall penetration regardless of what the lab report says. That is exactly the patient I kept seeing. Reassuring on paper. Not reassuring underneath.

ApoB solves that problem directly. Every atherogenic lipoprotein — LDL, VLDL, IDL, Lp(a) — carries exactly one ApoB molecule, making it a direct count of circulating atherogenic particles. The 2026 guideline recommends measuring ApoB once LDL-C and non-HDL-C goals are met, particularly in patients with elevated triglycerides, diabetes, CKM syndrome, or an achieved LDL-C below 70 mg/dL. It doesn’t require fasting. It doesn’t rely on the Friedewald equation, which becomes unreliable at low LDL levels. In the patients where standard lipid panels mislead most — metabolic disease, insulin resistance, the young, fit patient with a family history — ApoB often tells a different story.

We’re not ordering different drugs because of it. Statins, ezetimibe, bempedoic acid, and PCSK9 inhibitors all reduce ApoB alongside LDL-C. The question ApoB answers isn’t how to treat. It’s who still has residual risk after we think we’ve already addressed it.

The test nobody was ordering

When I started checking Lp(a) routinely, I expected to find it elevated occasionally. I didn’t expect to find it as often as I did. Roughly 1 in 5 people carry elevated levels, it’s set almost entirely by genetics, and it doesn’t respond meaningfully to lifestyle or most existing lipid-lowering therapies. For years, the clinical response when it came back high was essentially: There isn’t much we can specifically do. That was always an incomplete answer. The 2026 guideline makes it a nonanswer.

Lp(a) measurement is now a Class I recommendation for all adults — not selected high-risk patients, everyone, at least once. The thresholds matter: above 125 nmol/L, ASCVD risk increases roughly 1.4-fold. Above 250 nmol/L, risk more than doubles, approaching the cardiovascular hazard of heterozygous familial hypercholesterolemia.¹ When it’s elevated, the response is intensified LDL-C lowering, aggressive management of every other modifiable risk factor, and cascade screening of first-degree relatives — because this is an inherited condition and the patient’s siblings and children may be carrying the same number.

The specific therapies are coming. Pelacarsen’s Lp(a) HORIZON trial — the first cardiovascular outcomes trial ever run for an Lp(a)-lowering drug, over 8,000 patients — is reporting results this year. Olpasiran, an siRNA therapy that reduced Lp(a) by over 95% in phase 2, has outcomes data expected in 2027. Muvalaplin, the first oral agent in this class, reduced Lp(a) by up to 86% in its phase 2 trial. None are approved yet. But the patients who will eventually benefit from them are already in practice. Most have never had the test that would identify them. That’s the gap the guideline is trying to close.

When imaging settles it

The last clarification in the guideline is when coronary artery calcium (CAC) scoring should enter the conversation. For men 40 and older and women 45 and older, CAC is now explicitly recommended when the treatment decision is genuinely uncertain — not as routine screening, but as a tiebreaker when a patient’s risk is borderline and you’re not sure whether to treat.

For the young, fit patient with a family history and a clean lipid panel, this is directly relevant. A CAC of zero carries real prognostic weight and supports deferring pharmacotherapy in appropriately selected patients. A score of 100 or above — or placement at the 75th percentile or higher for age and sex — supports high-intensity treatment even when everything else looks mild. The guideline also now addresses incidental coronary calcium found on noncardiac imaging: If it shows up on a CT ordered for something else entirely, it should start a lipid management conversation.

That last point alone would have changed a few conversations I’ve had over the years.

What the guideline is actually saying

I don’t think the 2026 guideline is asking us to practice fundamentally different medicine. Statins still work. LDL-C reduction still saves lives across nearly every population studied. That foundation isn’t moving.

What it’s asking is that we stop letting a reassuring LDL-C end the conversation in patients where the conversation isn’t actually over. The young, fit patient with a family history and a Lp(a) of 180 nmol/L and an ApoB that doesn’t match their LDL — that patient has always existed. The guideline just gives us the framework, and the clinical mandate, to find them.

I’m glad it finally does.

How will the 2026 ACC/AHA cholesterol guidelines change your approach to preventive cardiology? Share in the comments.