Am I statin eligible under the 2026 guideline? How to run your own numbers

By Marcus Renshaw · · updated

My lipid panels came back unremarkable for years and I filed them as done. What I never did was run the numbers through a risk calculator, because the output felt like something that happened in a clinic rather than something I could compute. That was a mistake, and the 2026 dyslipidemia guideline is a good reason to fix it, because the calculation changed underneath everyone and the inputs are mostly things a self-tracker already has on a spreadsheet.

The 2026 ACC/AHA guideline replaced the old Pooled Cohort Equations with the PREVENT-ASCVD equations and made an estimated 21.5 million more US adults statin eligible. PREVENT actually rates most people as lower risk than the old calculator did. Eligibility grew because the guideline added a 30 year risk pathway for ages 30 to 59.

What actually changed in the 2026 guideline?

Two things, and they pull in opposite directions.

The first is the calculator. The guideline now says the PREVENT-ASCVD equations should be used for risk assessment in adults aged 30 to 79 without existing cardiovascular disease whose LDL-C sits between 70 and 189 mg/dL. That is a Class 1 recommendation, the strongest the guideline issues. PREVENT stands for Predicting Risk of cardiovascular disease EVENTs, and the American Heart Association published the equations in Circulation in 2023 after developing them on 3.3 million adults and validating them on another 3.3 million.

The second is the thresholds. Because PREVENT produces different numbers than the old Pooled Cohort Equations, the guideline remapped the risk bands to match. Its own slide set labels the pairing as approximate equivalent ranges:

Risk categoryPooled Cohort EquationsPREVENT-ASCVD
Lowunder 5%under 3%
Borderline5 to under 7.5%3 to under 5%
Intermediate7.5 to under 20%5 to under 10%
High20% or more10% or more

Read that table carefully, because it is the whole story in one image. The thresholds were cut roughly in half. That is not the guideline getting more aggressive. It is the guideline compensating for a calculator that returns smaller numbers, so that the same person lands in roughly the same band.

Why did eligibility grow if the new calculator scores you lower?

This is the part that gets reported backwards, so it is worth being precise.

PREVENT scores most people as lower risk. When researchers applied it at the old thresholds, before the guideline rebalanced them, they found it would reclassify about half of US adults to lower risk categories and cut statin eligibility by 14.3 million, not raise it.

And yet the JAMA analysis published on 20 July 2026 found the 2026 guideline makes 87.5 million US adults aged 30 to 79 eligible for primary prevention statins, which is 56.6% of that population, including 21.5 million who were not eligible before. Its own conclusion is that the guideline “substantially expands the US population recommended for primary prevention statin therapy, predominantly in lower-risk individuals.” The newly eligible group has a mean 10 year risk of 3.1%, against 6.1% for people who were already eligible.

So where does the expansion come from, if not from the calculator being more alarmist? From a second clock. For adults aged 30 to 59 the guideline now asks you to calculate both 10 year and 30 year risk, and a high 30 year risk becomes its own route to a statin even when your 10 year number is low. An analysis of the guideline text puts that bar at a 30 year risk of 10% or more, and finds the 30 year criterion is the dominant driver of the expansion, concentrated in people in their forties and fifties. That analysis is a preprint and has not been peer reviewed, so treat the mechanism as well supported and the exact split as provisional.

The logic is defensible. A 45 year old with a mildly elevated LDL has a small ten year risk and a large lifetime one, and judging that person only on the next decade has always understated their exposure. It is the same argument for measuring Lp(a) once: what matters is cumulative exposure, not this year’s snapshot.

What does “eligible” actually mean here?

Not what the headline number implies, and this is the single most important thing to get right.

The guideline sorts primary prevention into four bands and attaches different recommendation strengths to each. Class 1 means the guideline recommends it. Class 2a means it is reasonable, a weaker signal that leans on a discussion rather than a directive.

10 year PREVENT riskWhat the guideline saysClass
Low, under 3%, LDL-C under 160Health behavior counseling1
Low, but LDL-C 160 to 189 or high 30 year riskStart a moderate statin2a
Borderline, 3 to under 5%A moderate statin is reasonable2a
Intermediate, 5 to under 10%Start a moderate to high intensity statin1
High, 10% or moreStart a high intensity statin1

Now put the JAMA finding next to that table. The newly eligible group averages 3.1% ten year risk, which lands in the borderline band or in the low band reached through the 30 year pathway. Both of those are Class 2a. So the accurate sentence is that 21.5 million more people now qualify for a conversation in which a statin is a reasonable option, not that 21.5 million more people should be taking one. The same preprint found that if you apply only the Class 1 criteria, the guideline actually reduces the number of people recommended a statin.

The guideline itself is built as a sequence rather than a verdict. It calls the structure CPR: calculate and classify, personalize with risk enhancers, then reclassify and reassess before deciding. The number is the first step, not the answer.

What do you need to run your own numbers?

Here is the part that rewards anyone already keeping records, because the required inputs are unglamorous and you probably hold most of them.

PREVENT needs age, sex, systolic blood pressure, total cholesterol, HDL-C, eGFR, smoking status, diabetes status, BMI, and whether you take blood pressure medication or a statin. It will also take three optional inputs that sharpen it: urine albumin to creatinine ratio, HbA1c, and a ZIP code based social deprivation index.

Two features of that list are worth naming. It includes eGFR, so kidney function now feeds a cardiovascular risk score directly, which the old equations never did. And it dropped race entirely. The authors were explicit about why: race and ethnicity “are social constructs and, thus, were not considered as predictors in risk modeling to eliminate propagation of race-based risk algorithms.”

Notice how much of that is blood pressure and bloodwork. Systolic pressure is an input, and so is whether you are on a BP medication. I am on one, and my morning readings are the number I have the most of, which means two of the inputs to my own cardiovascular risk estimate come straight off the log I keep anyway. If your systolic comes from one hurried reading in a waiting room, your risk estimate inherits that error, which is one more reason to measure at home properly rather than accept a single clinic number.

The rest comes from a standard lipid panel plus a metabolic panel for eGFR and glucose. That is an ordinary annual draw, not an exotic one. You can run the equations yourself using the AHA’s PREVENT calculator or an independent implementation of the published equations. The AHA tool sits on their clinician-facing site and is written for use in a consultation, but it is open to read and run.

Two things to get right before you trust the output. The calculator returns separate estimates for total cardiovascular disease, for ASCVD, and for heart failure, and it displays total CVD by default. Every threshold in this article is an ASCVD threshold, so switch the output to PREVENT-ASCVD before comparing your number to anything here. The AHA also notes that because ASCVD and heart failure are modeled independently, adding them together can exceed the total CVD figure, so do not try to reconcile the three.

The second is who the equations were not built for. The AHA says not to use them in anyone with known cardiovascular disease, with severe subclinical disease including a coronary calcium score at or above 300, with a positive genetic test for an inherited cardiac condition, with end stage kidney disease, or with a life expectancy under a year. Several of those describe people who are already past the question this calculator answers. If you are in one of those groups the number it hands you will not mean what you think it means.

Here is where the guideline is the floor rather than the answer. Clinical risk assessment runs on an annual or less frequent cadence because it is built around a scheduled appointment. If you are optimizing across decades, the useful move is to recompute when an input actually moves, and to know which input moved. My own honest position is that I have been putting off the full panel that would let me do this properly, which is a common enough failure that it is worth naming: the calculator is free and takes two minutes, and the reason most people never run it is a missing lipid panel, not a missing calculator.

What breaks the tie when your number lands in the middle?

The borderline band is deliberately unresolved, and the guideline gives two tools for resolving it.

The first is risk enhancers, which the guideline says it is reasonable to consider when you land at borderline risk. The list includes several markers this audience often already has: Lp(a) at 125 nmol/L or above, ApoB at 120 mg/dL or above, hsCRP at 2 mg/L or above on more than one occasion, triglycerides persistently at or above 150 mg/dL fasting, a parent or sibling with premature cardiovascular disease, and chronic inflammatory disease. This is the concrete payoff for putting ApoB and Lp(a) on the panel rather than settling for a standard lipid profile. Those two markers are not just extra numbers, they are formal tie breakers inside the decision the guideline is asking you to make.

The second is a coronary artery calcium scan, which the guideline says to consider when there is clinical or patient uncertainty after the risk discussion. It measures calcified plaque directly rather than predicting it, which makes it the closest thing available to checking the answer. The published tiers start at 1 to 99 Agatston units, where a moderate statin is worth considering, and escalate from there, each step tightening the LDL-C target until a score above 1000 is handled like established disease.

That ordering matters for a self experimenter. A risk score is a population estimate applied to you. A CAC score is a measurement of you. When the two disagree, the measurement is the better information, which is the same reason a tracked trend beats a single reference range.

How much does a statin actually buy someone at 3% risk?

Being honest about this is the difference between a useful article and a press release.

Statins cut risk by roughly a constant proportion, which is the assumption the whole risk-based framework rests on, and it means the absolute benefit shrinks as your baseline risk falls. The Cochrane review of statins for primary prevention puts the size of it plainly: of 1000 people treated for five years, 18 would avoid a major cardiovascular event. That is a number needed to treat of about 56, and it is a real effect, with all cause mortality reduced at an odds ratio of 0.86 (95% CI 0.79 to 0.94).

Push into genuinely low risk populations and the numbers thin out. A 2017 AAFP review of statins in people at low cardiovascular risk reports a number needed to treat of 217 to prevent one nonfatal heart attack and 313 for one nonfatal stroke, against a number needed to harm of 21 for muscle pain and 204 for new onset diabetes over five years. That same review found no statistically significant mortality benefit in the low risk group, which sits directly against the Cochrane figure above. Both are defensible readings of overlapping trial sets, and which one you get depends on who gets pooled in, so treat a confident claim in either direction with suspicion. A meta-analysis of survival postponement found a median postponement of death of 10.2 days in primary prevention, though that rests on only two trials and measures the trial’s duration rather than a lifetime, so it understates the case as much as a relative risk headline overstates it.

None of that makes statins a bad drug. It makes them a drug whose value depends heavily on where you actually sit, which is the entire reason the guideline asks you to calculate first. At 10% risk the arithmetic is comfortable. At 3% it is a genuine judgment call, and a judgment call is exactly what a Class 2a recommendation is supposed to signal.

If you are in the 30 to 59 band, the specific thing worth doing is running both numbers rather than just the ten year one, because the 30 year figure is the one that changed and the one most likely to reclassify you. Book the panel first if you do not have a current lipid result and an eGFR, since without those the calculator has nothing to work with.

FAQ

What is the PREVENT calculator?

PREVENT stands for Predicting Risk of cardiovascular disease EVENTs. It is a risk model published by the American Heart Association in 2023 and adopted by the 2026 dyslipidemia guideline to replace the Pooled Cohort Equations. It estimates 10 year and 30 year risk of cardiovascular disease, atherosclerotic cardiovascular disease, and heart failure, for adults aged 30 to 79.

How is PREVENT different from the Pooled Cohort Equations?

It covers ages 30 to 79 rather than 40 to 79, adds a 30 year horizon, includes kidney function as an input, adjusts for the competing risk of dying from something else, and removes race as a predictor. It also produces lower risk estimates for most people, which is why the 2026 guideline paired it with lower treatment thresholds.

Am I statin eligible under the 2026 guideline?

If your 10 year PREVENT-ASCVD risk is 5% or above, the guideline recommends starting a statin. Between 3% and 5% it says a moderate statin is reasonable after personalizing with risk enhancers. Below 3% the recommendation is health behavior counseling, unless your LDL-C is 160 to 189 mg/dL or your 30 year risk is high, in which case a moderate statin is reasonable.

Does eligible mean I should take a statin?

No. Eligible means you meet criteria for a decision, and for most of the newly eligible group that decision carries Class 2a language, meaning a statin is reasonable rather than recommended. The newly eligible average a 3.1% ten year risk. Class 1 recommendations begin at 5%.

What numbers do I need to run it?

Age, sex, systolic blood pressure, total cholesterol, HDL-C, eGFR, BMI, smoking status, diabetes status, and whether you take blood pressure or statin medication. HbA1c, urine albumin to creatinine ratio, and a ZIP code based deprivation index are optional and improve the estimate.

Who should not use the PREVENT calculator?

The AHA says not to use the equations in anyone with known cardiovascular disease, severe subclinical disease including a coronary calcium score at or above 300, a positive genetic test for an inherited cardiac condition, end stage kidney disease, or a life expectancy under a year. The equations were derived and validated in adults aged 30 to 79 without known cardiovascular disease, so outside that group the estimate is not doing the job it was built for.

What if my result lands in the borderline band?

The guideline points to two tie breakers. Risk enhancers including Lp(a) at 125 nmol/L or above, ApoB at 120 mg/dL or above, hsCRP at 2 mg/L or above, and premature family history. And a coronary artery calcium scan when uncertainty remains, which measures plaque directly and sorts you into tiers running from 1 to 99 Agatston units upward, each with a tighter LDL-C target.

Why did eligibility increase if PREVENT scores people lower?

Because of the 30 year pathway. For ages 30 to 59 the guideline calculates both horizons, and a high 30 year risk qualifies someone whose 10 year risk is low. Applying only the stronger Class 1 criteria would have reduced the number of people recommended a statin.

How much benefit does a statin give at low risk?

Cochrane reports 18 major cardiovascular events avoided per 1000 people treated for five years in primary prevention, with all cause mortality reduced at an odds ratio of 0.86. In lower risk groups specifically, a 2017 review puts the number needed to treat at roughly 217 for one nonfatal heart attack against a number needed to harm of 21 for muscle pain, and found no significant mortality benefit there. Absolute benefit falls as baseline risk falls, even though the relative reduction holds.

Sources