ACC/AHA 2026: dyslipidaemia
International ACC/AHA Current
2026 ACC/AHA/Multisociety Guideline on the Management of Dyslipidemia
Circulation. 2026;153:e1154–e1276
DOI 10.1161/CIR.0000000000001423 · Summary checked against the source: 2026-11
Replaces the 2018 AHA/ACC cholesterol guideline. The big changes: LDL-C goals are back, risk is estimated with the PREVENT equations, Lp(a) is measured once in everyone, and a very-high-risk group gets an LDL-C goal <55 mg/dL.
At a glance
- Screen with a lipid profile from age 19, then at least every 5 years; children once at 9–11 years. [1, B-NR · e1164]
- Measure Lp(a) at least once in every adult. High = ≥125 nmol/L (≥50 mg/dL). [1, B-NR · e1168; e1234]
- Estimate 10-year risk with PREVENT-ASCVD (age 30–79, LDL-C 70–189 mg/dL): low <3%, borderline 3–<5%, intermediate 5–<10%, high ≥10%. [1, B-NR · e1183]
- Goals: primary prevention <100 (borderline or intermediate) or <70 mg/dL (high risk, diabetes with risk factors); ASCVD <70; very-high-risk ASCVD <55 mg/dL. [e1192–e1208]
- CAC score settles uncertain decisions: CAC 0 may defer therapy; CAC >0 means start; ≥1000 treat like very high risk. [e1189–e1213]
- Lipid profile 4–12 weeks after starting or changing therapy, then every 6–12 months. [1, A · e1169]
Measurement
- Fasting or non-fasting profile for baseline; fasting if TG ≥400 mg/dL or a TG disorder is suspected. [1, B-NR · e1166]
- Calculate LDL-C with the Martin/Hopkins or Sampson/NIH equation, not Friedewald. Report non-HDL-C. [1, B-NR · e1166]
- ApoB is reasonable to guide intensification on therapy, especially with diabetes, CKM syndrome or high TG. [2a, B-NR · e1167]
- With TG ≥150, prefer non-HDL-C or apoB to LDL-C for decisions. [2a, B-NR · e1228]
Primary prevention
| 10-year PREVENT risk | Treatment | LDL-C goal | Grade |
|---|---|---|---|
| Low <3% (age 30–59), LDL <160, 30-year risk <10% | Healthy lifestyle | 1, A | |
| Low, but LDL 160–189 or 30-year risk ≥10% | Moderate-intensity statin reasonable | 2a, C-LD | |
| Borderline 3–<5% | Consider risk enhancers; moderate statin if treating (≥30–49% LDL fall) | <100 (non-HDL <130) | 2a |
| Intermediate 5–<10% | At least moderate-intensity statin | <100 (non-HDL <130) | 1, A |
| High ≥10% | High-intensity statin (≥50% LDL fall); add ezetimibe, then a PCSK9 mAb or bempedoic acid if not at goal | <70 (non-HDL <100) | 1, A; 2a; 2b |
Adapted from ACC/AHA 2026 recommendations 5.1.1–5.1.11 [e1192]
- Borderline risk: risk enhancers, and hsCRP ≥2 mg/L twice, favour treatment; consider reproductive history (early menopause, pre-eclampsia, gestational diabetes, preterm delivery). [2a · e1186, e1189]
- CAC for uncertain decisions (intermediate, or selected borderline): CAC 0 with a preference to avoid drugs may defer therapy; CAC >0 start, especially if ≥100 or ≥75th percentile. Incidental CAC on any CT means atherosclerosis is present. [1, B-R; 2a; 1, B-NR · e1189–e1190]
| CAC score (AU) | Treatment | LDL-C goal |
|---|---|---|
| ≥1000 | Statin-based, ≥50% fall | <55 (1, B-NR) |
| 300–999 | Statin-based, ≥50% fall | <70 (1, B-R); <55 reasonable |
| 100–299 or ≥75th percentile | Statin-based, ≥50% fall | <70 (1, B-R) |
| 1–99 and <75th percentile | Moderate-intensity statin reasonable | <100 (2a, B-R) |
Adapted from ACC/AHA 2026 section on CAC [e1213]
Special groups
| Group | Recommendation | Grade |
|---|---|---|
| LDL-C ≥190 mg/dL | Exclude secondary causes; maximally tolerated statin; add ezetimibe, PCSK9 mAb or bempedoic acid as needed; inclisiran reasonable if LDL ≥100 despite statin ± ezetimibe | 1; 2a |
| Familial hypercholesterolaemia | Panel-based genetic testing; don't use general-population risk scores; cascade screening of relatives from age 2 | 1, B-NR; 3: Harm |
| Homozygous FH | Lipid specialist; statin + ezetimibe ± PCSK9 mAb; evinacumab or lomitapide; apheresis | 1; 2b |
| Diabetes, age 40–75 | Moderate-intensity statin, goal <100; high-intensity, goal <70, with multiple risk factors | 1, A; 2a, B-R |
| CKD stage 3+, age 40–75 | Moderate statin ± ezetimibe; with ASCVD, high-intensity, goal <55 | 1, B-R |
| Haemodialysis | Continuing a statin may be reasonable | 2b |
| HIV on ART, age 40–75 | Statin | 1, B-R |
| Older than 75 | Moderate statin may be reasonable with life expectancy ≥2.5 years; CAC can help | 2b |
| Life expectancy <1 year | Stopping may be reasonable | 2b |
Adapted from ACC/AHA 2026 [e1197–e1225]
Clinical ASCVD
Very high risk = ≥2 major ASCVD events (ACS in the past 12 months, previous MI, ischaemic stroke, symptomatic PAD), or 1 major event + ≥2 high-risk features (age ≥65, CABG or PCI, smoking, diabetes, heart failure, hypertension, LDL ≥100 despite statin + ezetimibe). [e1163]
| Not very high risk | Very high risk | |
|---|---|---|
| Statin | High-intensity, ≥50% fall (1, A) | High-intensity, ≥50% fall (1, A) |
| LDL-C goal | <70 mg/dL (non-HDL <100); <55 reasonable | <55 mg/dL (non-HDL <85) |
| Add-on | Ezetimibe, PCSK9 mAb or bempedoic acid reasonable (2a) | Ezetimibe and/or PCSK9 mAb should be added (1, A); bempedoic acid or inclisiran reasonable (2a) |
Adapted from ACC/AHA 2026 recommendations for clinical ASCVD [e1208]
Triglycerides and Lp(a)
| TG (fasting) | Management |
|---|---|
| 150–499 mg/dL | Lifestyle (low added sugar and refined carbohydrate, less alcohol, 5–10% weight loss, exercise); treat secondary causes; use PREVENT for statin decisions |
| ASCVD or diabetes + risk factor, LDL <100 on a statin: icosapent ethyl may be considered (2b) | |
| 500–999 mg/dL | No alcohol, limit fat; fibrate or omega-3 reasonable |
| ≥1000 mg/dL | Very low-fat diet, no alcohol or added sugar, dietitian; risk of pancreatitis; olezarsen for familial chylomicronaemia (1, B-R) |
Adapted from ACC/AHA 2026 [e1172–e1228]
- High Lp(a) (≥125 nmol/L): control all other risk factors early; with ASCVD not at goal, add a PCSK9 mAb. Cascade-test relatives. [1 · e1168, e1234]
Statin intolerance and safety
- Assess muscle symptoms for secondary causes; discuss the higher risk of stopping; rechallenge with a lower dose, another statin or intermittent dosing; use bempedoic acid, ezetimibe or a PCSK9 mAb when goals aren't met. [1 · e1235]
- New diabetes on a statin: continue the statin. Stable chronic liver disease (including MASLD): statins are reasonable. [1, A; 2a, B-R · e1238]
- Pregnancy: stop statins 1–2 months before trying to conceive unless high risk (FH or ASCVD may continue after discussion); bile acid sequestrants are reasonable in pregnancy; fibrates (after the first trimester) or omega-3 for TG ≥500. [e1220]
Clinical pearls
- Goals are back: <100 (moderate risk), <70 (high risk or ASCVD), <55 (very-high-risk ASCVD).
- PREVENT replaces the Pooled Cohort Equations; thresholds 3%, 5%, 10%.
- Lp(a) once in every adult.
- CAC 0 can defer statins; CAC ≥100 means treat.
- Add-on order: ezetimibe, then PCSK9 mAb (or bempedoic acid, inclisiran).
Classes and levels of evidence
How ACC/AHA 2026 grades its recommendations. Adapted from ACC/AHA 2026 Table 3 [e1162]. Hover over any class or level cited on this page to see what it means.
| Class | Strength | Meaning | Wording used |
|---|---|---|---|
| Class 1 | Strong | Benefit >>> risk | is recommended; is indicated, useful, effective or beneficial; should be performed or administered |
| Class 2a | Moderate | Benefit >> risk | is reasonable; can be useful, effective or beneficial |
| Class 2b | Weak | Benefit ≥ risk | may or might be reasonable; may or might be considered; usefulness is unknown, unclear or uncertain |
| Class 3: No benefit | Moderate | Benefit = risk (generally used only with level A or B) | is not recommended; is not indicated, useful or effective; should not be performed or administered |
| Class 3: Harm | Strong | Risk > benefit | potentially harmful; causes harm; associated with excess morbidity or mortality; should not be performed or administered |
| Level | Meaning |
|---|---|
| Level A | High-quality evidence from more than one RCT, meta-analyses of high-quality RCTs, or one or more RCTs corroborated by high-quality registry studies |
| Level B-R (randomised) | Moderate-quality evidence from one or more RCTs, or meta-analyses of moderate-quality RCTs |
| Level B-NR (non-randomised) | Moderate-quality evidence from one or more well-designed, well-executed non-randomised, observational or registry studies, or meta-analyses of such studies |
| Level C-LD (limited data) | Randomised or non-randomised observational or registry studies with limitations of design or execution, meta-analyses of such studies, or physiological or mechanistic studies in humans |
| Level C-EO (expert opinion) | Consensus of expert opinion based on clinical experience |
Class (strength) and level (quality) are judged independently, so any class can pair with any level. A level C recommendation isn't necessarily weak: many important clinical questions can't be tested in trials, yet there may be clear clinical consensus.