Evidence review
ApoB and LDL on low carb
What Ference, Sniderman, O’Neill/Raggi, and Norwitz actually support on ApoB, LDL-C, and LMHR — and where hard carnivore endpoints are still missing.
Auf einen Blick
- ApoB-containing lipoproteins are the more robust exposure marker in the cardiovascular literature than LDL-C alone.
- A low-carb, high-fat diet can lower, raise, or uncouple LDL-C and ApoB. Heterogeneity is the finding.
- LMHR is an observed phenotype with limited endpoint evidence. Neither 'harmless because lean' nor 'identical to familial hypercholesterolemia' is established.
- Strict carnivore has no hard-outcome RCTs of its own. Anyone who leaves that out is writing marketing.
How to read this review
The shorter companion, LDL on strict carnivore, is the starter card. This piece is about specific papers — what they actually carry, and what they do not.
Two mistakes sit on opposite sides of the same debate. One is seeing LDL-C or ApoB rise and treating the heart-attack risk as identical to that of a metabolically sick patient with the same numbers. The other is selling high LDL-C plus low triglycerides and high HDL as proven safe. Neither move is honest.
This is not a treatment recommendation
Statins, PCSK9 inhibitors, treatment targets, and stopping medication are clinical decisions. This text does not rebrand extreme values as “normal for carnivore.”
What the ApoB literature actually supports
Ference and colleagues (2017) pull genetic and interventional evidence together and conclude that LDL particles — more precisely, ApoB-containing lipoproteins — are causal in atherosclerosis. Risk scales with how high the exposure is and how long it lasts. Sniderman et al. (2019) make the same point from another angle: particle number (ApoB) is the relevant quantity, not just the cholesterol sitting inside the particles (LDL-C).
That is why a lab panel should include ApoB (or at least non-HDL-C) when the two disagree, not LDL-C alone.
What these reviews do not do: separately bless a lean, trained, low-carb person with isolated high LDL-C and very low triglycerides using hard endpoints. The causality arguments come from broad populations, Mendelian randomization, and lipid-lowering trials — not from carnivore cohorts.
What the low-carb / keto reviews actually support
O’Neill and Raggi (2020) and related overviews describe a common pattern on ketogenic diets: triglycerides often fall, HDL-C often rises, and LDL-C is heterogeneous — down, flat, or markedly up.
Weight loss and fewer ultra-processed carbohydrates change overall risk independently of isolated LDL-C. At the same time, a subset sees a sharp LDL-C rise. Pretending that phenomenon does not exist is unserious. Treating it as a mandatory complication of every carnivore practice is equally unserious. Carnivore is a subset of very low-carb eating, often with more dietary cholesterol and saturated fat than a mixed ketogenic diet. Transfer is plausible. It is not identity.
LMHR: what is published
Norwitz et al. (2022) describe the lean mass hyper-responder phenotype: typically high to very high LDL-C, high HDL-C, and very low triglycerides under carbohydrate restriction. It shows up most often in lean, active people. The Lipid Energy Model is a mechanistic hypothesis. It is not finished pathophysiology.
The phenotype is real enough to study. It is not a measurement fiction. Case series and cross-sections do not replace years of endpoint trials. Imaging surrogates are starting to appear and are being read contentiously. Being lean and athletic lowers many risks. It does not prove that decades of very high ApoB exposure are irrelevant. If you think you fit LMHR: document it. Do not celebrate it. Further workup belongs in clinic.
Guideline context, labeled honestly
The ESC/EAS dyslipidaemia guideline remains risk-stratified and LDL-C/ApoB-lowering in secondary prevention and high-risk primary prevention. It was not written for “carnivore patients.” Ignoring it because the community tells a different story is a choice — not evidence. Slapping it uncritically onto every metabolically healthy person with an isolated LDL rise is the same kind of mistake.
Practical synthesis
Measure ApoB (or non-HDL) when the numbers even raise a treatment question, not LDL-C alone. Read the pattern: triglycerides, HDL, glucose/HbA1c, blood pressure, smoking, family history, and Lp(a) if you can get it. Separate phenotypes. Say the sentence out loud: hard carnivore endpoints are missing. Extreme values and known coronary disease do not belong in self-normalization.
Further reading
Sources
- Ference et al. 2017 – Low-density lipoproteins cause atherosclerotic cardiovascular disease (PMID 28330828)
- Sniderman et al. 2019 – Apolipoprotein B particles and cardiovascular disease (PMID 30894319)
- O’Neill & Raggi 2020 – Review of ketogenic diets and the lipoprotein profile (PMID 31802496)
- Norwitz et al. 2022 – Elevated LDL-cholesterol with a carbohydrate-restricted diet: LMHR (PMID 35629964)
- Norwitz et al. 2021 – The Lipid Energy Model (discussion, PMID 35057469)
- Mach et al. 2020 – ESC/EAS Dyslipidaemia Guidelines (PMID 31504418) – guideline context, not a carnivore study
- Borén et al. 2020 – Low-density lipoproteins cause atherosclerotic cardiovascular disease: pathophysiological, genetic, and therapeutic insights (PMID 32052833)
- Nordestgaard & Langsted 2016 – Lipoprotein (a) as a cause of cardiovascular disease (PMID 27624320)
- Budoff et al. 2024 – Carbohydrate restriction-induced elevations in LDL-C and coronary plaque (PMID 39372369); surrogate, controversial
- Soto-Mota et al. 2025 – KETO-CTA one-year plaque (PMID 40192608); surrogate, not an event trial
This content is general information. It is not medical, dietetic, or diagnostic advice.
