Lipids & statin decisions

The 2026 guideline prefers Martin/Hopkins or Sampson/NIH over Friedewald and over direct LDL-C assays (other than beta-quantification) COR 1. Non-HDL-C should be reported with every panel COR 1. Nonfasting panels are fine; repeat fasting if nonfasting TG is ≥400 mg/dL.

Martin/Hopkins (mg/dL)

Sampson/NIH (mg/dL)

Friedewald (mg/dL)

• Friedewald underestimates LDL-C when TG ≥150 mg/dL, LDL-C <70 mg/dL or non-HDL-C <100 mg/dL, and cannot be used at TG ≥400 mg/dL.
• The original Martin/Hopkins table was validated below TG 400 mg/dL; an extended version exists for 400–799 mg/dL (not implemented here). Sampson/NIH is validated to TG 800 mg/dL.
• Above TG 800 mg/dL no calculated LDL-C is reported here: use non-HDL-C, apoB, or beta-quantification.
• SI entry is converted with 1 mmol/L cholesterol = 38.67 mg/dL and 1 mmol/L TG = 88.57 mg/dL before calculation.

Median TG:VLDL-C ratio by stratum (Martin 2013). Rows: triglycerides (mg/dL). Columns: non-HDL-C (mg/dL).

Sex-specific, race-free equations for adults 30–79 without CVD, predicting 10- and 30-year risk of ASCVD (nonfatal MI, CHD death, fatal or nonfatal stroke), heart failure and total CVD. The guideline uses 10-year PREVENT-ASCVD for lipid decisions and 30-year ASCVD in ages 30–59.

Terms: age (per 10 y from 55; 30-year models add age²), non-HDL-C and HDL-C in mmol/L (×0.02586; centered at 3.5 and 1.3, HDL per 0.3), SBP spline at 110 mmHg (per 20), eGFR spline at 60 (per −15), BMI spline at 30 (per 5; heart failure only), diabetes, current smoking, BP treatment, statin, and interactions (treated SBP, statin × non-HDL-C, age × non-HDL-C, HDL-C, SBP, diabetes, smoking, BMI, eGFR). Optional UACR (ln) and HbA1c (centered at 5.3%, separate terms with and without diabetes) switch to the augmented models; entering both uses the full model with social deprivation index set to missing.

Valid input ranges: TC 130–320 mg/dL, HDL-C 20–100 mg/dL, SBP 90–180 mmHg, eGFR 15–140, BMI 18.5–39.9, UACR 0.1–25,000 mg/g, HbA1c 4.5–15%. Values outside these ranges are not clamped — PREVENT is not calculated. Categories use the unrounded 10-year value (shown to two decimals when rounding would cross 3, 5 or 10%).

Validation of this implementation: all 768 coefficients match the preventr package (from Table S12 of Khan 2024), and 24 reference profiles from preventr's test suite (base, UACR, HbA1c and full models; both sexes; statin, smoking, eGFR, BMI and age variations) reproduce every published 10- and 30-year output to 0.1%.

Indicated regardless of PREVENT category (ages 40–75): diabetes, CKD stage 3–4 (moderate-intensity statin ± ezetimibe), and HIV on stable antiretroviral therapy COR 1. Untreated LDL-C <70 and non-HDL-C <100 mg/dL without other risk factors: starting LLT is unlikely to reduce risk COR 3.

Risk enhancers (Table 13): premature ASCVD in a parent or sibling; higher-risk ancestry; high polygenic risk; chronic inflammatory disease; Lp(a) ≥125 nmol/L (≥50 mg/dL); hsCRP ≥2 mg/L on more than one occasion; TG persistently ≥175 mg/dL nonfasting or ≥150 fasting; CKM syndrome; LDL-C persistently 160–189, non-HDL-C 190–219 or apoB ≥120 mg/dL; reproductive risk markers.

Very high risk: ≥2 major ASCVD events, or 1 major event plus ≥2 high-risk conditions. Major events: ACS within 12 months, prior MI, ischemic stroke, symptomatic PAD. High-risk conditions: age >65, coronary revascularization, current smoking, diabetes, heart failure, hypertension, LDL-C >100 mg/dL despite maximally tolerated statin + ezetimibe. Most patients with ASCVD qualify.

ASCVD with elevated Lp(a) not at goal on maximally tolerated statin: add a PCSK9 mAb with proven cardiovascular benefit COR 1. Inclisiran has no completed outcome trial and is a second-line PCSK9 inhibitor.

When to order: intermediate risk, or selected borderline risk, when the treatment decision is uncertain COR 1; intermediate or high risk when the intensity of therapy is uncertain 2a. Incidental CAC on a noncardiac CT should be factored into the decision COR 1. Not useful to “de-risk” severe hypercholesterolemia/FH, diabetes over 40, current smokers or a strong family history of premature ASCVD, and not for monitoring treatment.

Incidental CAC on noncardiac CT: moderate-to-severe → high-intensity statin, goal <70 / <100; mild → moderate-intensity, goal <100 / <130 2a.

• LDL-C ≥190 mg/dL: exclude and treat secondary causes, then maximally tolerated statin COR 1.
• Not at goal on maximally tolerated statin: add ezetimibe, a PCSK9 mAb and/or bempedoic acid COR 1. Goal <100 / <130 without HeFH, additional risk factors or coronary calcium; <70 / <100 with any of these; <55 / <85 with clinical ASCVD.
• LDL-C ≥100 mg/dL despite maximally tolerated statin ± ezetimibe: inclisiran reasonable 2a.
• Do not use PREVENT or other population risk tools in HeFH COR 3.
• Panel-based genetic testing for FH when LDL-C ≥190 without a secondary cause 2a; cascade-test relatives.
• HoFH on statin, ezetimibe and PCSK9 mAb with LDL-C ≥100: evinacumab may be reasonable 2b.

• Age 40–75: moderate-intensity statin, goal <100 / <130 COR 1. Multiple ASCVD risk factors: high-intensity, goal <70 / <100 2a. PREVENT ≥10% and above goal: adding ezetimibe or a PCSK9 mAb may be reasonable 2b.
• Statin-attributed side effects: ezetimibe and/or bempedoic acid or a PCSK9 mAb COR 1.
• Age 20–39 with T2D ≥10 y or T1D ≥20 y, albuminuria ≥30 mg/g, eGFR <60, retinopathy, neuropathy or ABI <0.9: moderate-intensity statin may be reasonable 2b. From age 30, PREVENT also applies.
• Over 75 with life expectancy ≥2.5 years: moderate-intensity statin may be reasonable after discussion 2b.
• On statin with LDL-C <100, TG 150–499 and additional risk factors: icosapent ethyl may be considered 2b.

• Persistent TG ≥150 mg/dL: treat secondary causes and start with lifestyle (low added sugar, less alcohol and saturated fat, exercise, 5–10% weight loss) COR 1.
• Age 40–75 without ASCVD or diabetes, TG 150–499: use PREVENT to guide the statin discussion COR 1. Statins remain first-line for ASCVD risk when TG <1000.
• ASCVD with LDL-C ≥55 and non-HDL-C ≥85 on maximally tolerated statin and TG 150–999: intensify LDL-lowering first COR 1.
• Age ≥50 with ASCVD, or diabetes + ≥1 risk factor, TG 150–499 and LDL-C <100 on maximally tolerated statin: icosapent ethyl may be reasonable 2b.
• TG 500–999, especially ≥1000 despite diet: a fibrate (fenofibrate, not gemfibrozil, with a statin) or prescription omega-3 is reasonable to lower pancreatitis risk 2a.
• Fasting TG ≥1000: refer to a registered dietitian nutritionist COR 1; familial chylomicronemia syndrome: olezarsen with diet COR 1.

• Measure Lp(a) at least once in every adult COR 1. Relative ASCVD risk vs the median: 1.4× at 125 nmol/L (50 mg/dL), 2× at 250 (100), 3× at 350 (150), 4× at 430 (180). Do not convert nmol/L to mg/dL — the units are not interchangeable.
• Elevated Lp(a): early, tight control of all modifiable risk factors COR 1. Consider lipid-specialist referral at ≥200 nmol/L (≥75 mg/dL).
• ApoB on LLT (ASCVD, CKM syndrome, type 2 diabetes, elevated TG) once LDL-C/non-HDL-C goals are met, to decide on further intensification 2a.
• Lipid panel every 5 years from age 19; universal screening once at 9–11 years.
• Dietary supplements are not recommended to lower LDL-C or TG COR 3.
• Refer to a lipid specialist: suspected or confirmed FH, premature ASCVD (<40 y), not reaching ≥50% reduction and goal on statin + nonstatin therapy, side effects on ≥2 statins, Lp(a) ≥200 nmol/L, diabetes with dyslipidemia under 40.

Educational — verify doses, interactions and renal adjustment against the product label and local protocol. Simvastatin 80 mg is not recommended (myopathy).