Should You Be Worried About High LDL On Carnivore?

What you are getting yourself into: what actually happens to your blood fats when you run on meat, who decided what counts as high and when they moved the line, why the damaged particle matters more than the particle count, the lean mass hyper responder question answered honestly, and the exact conversation to have across the desk when the printout comes out.

The printout comes across the desk turned to face you, which is never a good sign.

One number has a flag beside it.

Your LDL is up, and not politely. Three months of carnivore. Waist down two notches, lifts climbing, sleeping through the night for the first time in a decade. None of that is on the sheet. The sheet has one red number, and the red number says the fat is winning.

So you are now in the strangest position modern medicine has to offer. You feel better than you have in years, and you are being told you are getting sicker, on the strength of a single figure, measured against a reference range that was never built for anyone eating the way you now eat.

Read the whole panel. Not the one red number. All of it.

If your LDL went up while your triglycerides went down and your HDL climbed, the diet is doing exactly what it was hired to do. That is a fat-burning metabolism working, not a disease starting. The number that deserves your attention is not the LDL count. It is the ratio of triglycerides to HDL, and on carnivore it almost always moves the right way.

The argument has changed

For years the answer to "where are the studies on carnivore?" was that there weren't any, and both sides were arguing from proxy. That has changed. There are carnivore-specific papers now, a review pulling them together, and case reports doing the rounds with the diet's name in the headline.

Your relatives have started sending them. This is the reply.

What the carnivore evidence actually is

The entire published human literature on the carnivore diet is nine studies. Not nine trials. Nine studies, and the word is doing heavy lifting.

Two are meal plans typed into nutrition software. Nobody ate anything. Researchers built a hypothetical carnivore week on a computer and scored its nutrients against the recommended intakes.

Two are social media surveys, where people who follow carnivore accounts filled in a questionnaire. The larger had 2,029 in it, and the blood results came from the fraction who had numbers to hand.

Four are case reports or small series. One person here, seven there.

One is an exploratory study with fasting bloods taken before and during. Twenty-four people.

No randomised trials. No cohorts. Not one study anywhere measuring a heart attack, a stroke or a death. Nobody has followed a group of carnivores long enough to count them.

Both sides live with that. Nobody has shown carnivore causes heart attacks. Nobody has shown it doesn't. Every confident sentence you have read in either direction is somebody extrapolating.

The 2026 review

In January 2026 a team at Münster published a review in Nutrients, and it now sits behind most of the "science says carnivore is dangerous" posts on your timeline. Read what it says rather than what it is being used to say.

Those nine studies are its entire evidence base. Its findings on blood fats, in its own numbers.

LDL and total cholesterol go up. In the twenty-four-person German study, LDL went from 157 to 256 mg/dL, 4.1 to 6.6 mmol/L, and total cholesterol from 224 to 305, 5.8 to 7.9. Every dataset with before-and-after bloods found the same direction. That is real, and it is what this article exists to deal with.

Triglycerides and HDL improve. In the 2,029-person survey, triglycerides fell from 83 to 68 mg/dL and HDL rose from 58 to 68, both significant. That is the ratio this page tells you to watch, moving the right way, in the sceptics' own data.

CRP, the inflammation marker, fell rather than rose: 1.0 to 0.7 mg/dL. GGT, the liver marker, fell alongside it.

And the finding nobody quotes. In the German study, the participants who arrived with metabolic disease improved their HbA1c and their triglycerides. The ones who arrived already healthy saw both go up. Same diet, opposite results, sorted entirely by who walked in the door.

The review's conclusion is that the diet carries substantial risk of cardiovascular disease and that long-term adherence cannot be recommended. That conclusion rests on no carnivore outcome data, because there is none. It is assembled by importing the general epidemiology of red meat and saturated fat, the same food-questionnaire work this article deals with elsewhere, and applying it to a diet nobody has followed to an endpoint.

The authors say as much themselves, and they get the credit for saying it plainly. Not one study in their review reported a clinical endpoint. The evidence is very limited: small samples, short durations, no control groups. They grade the whole body of it at the bottom of their own scale.

That is the strongest paper anyone will send you this year. A review of nine studies, not one of which measured a single cardiac event.

Concede the LDL rise. It is real, you should know your number, and the article has never said otherwise. Then refuse the swap the review invites you to make, because "LDL went up in twenty-four people" is not "carnivore causes heart disease", and a recommendation with no outcome data under it is an opinion with a reference list.

What carnivore actually does to your blood fats

Three things happen to a blood panel in the first months. Two of them never make it into the conversation.

Triglycerides fall, often sharply. HDL rises. And LDL goes up, sometimes a little, sometimes enormously, with the biggest rises in exactly the people you would least expect to be in trouble.

The lean. The fit. The fully fat-adapted.

One table, so you can see the whole board at once.

Every reference range on that printout was built in populations eating carbohydrate every few hours. Not one of them was measured in people running on fat.

On sugar, a climbing LDL is a warning light.

In ketosis, it is a delivery fleet scaled to the job it has been given.

Same number. Different machine.

Why LDL rises when you run on fat

LDL is not a poison. It is a courier.

Fat does not dissolve in blood, so it travels by boat. The liver loads triglycerides into VLDL particles and sends them out to feed muscle and organs. As a VLDL drops its cargo it shrinks, and what is left of the boat is an LDL particle, still carrying cholesterol for cell walls, hormones and repair.

Now change the fuel. On carnivore, fat is not a garnish on a carbohydrate diet. It is the entire energy economy. More fat moving means more boats on the water. A fat-adapted body, with low insulin and empty glycogen, turns those VLDLs over fast and efficiently. Triglycerides drop. HDL climbs. And LDL, the emptied courier fleet, sits higher because the whole port is busier.

Which is precisely why the leanest and most active people see the biggest rises. Less fat sitting in storage means more fuel in transit.

Even a fast can double LDL in lean, healthy subjects while leaving obese subjects untouched. Very hard to explain if LDL is a disease marker. Very easy to explain if it is a fuel logistics marker.

Who decided what counts as high

The line your number is being measured against has moved. Repeatedly. Always in the same direction.

For years the goal for most people sat around 130. Then the American panel that writes the cholesterol rulebook revised the guidance downwards. High-risk patients were now to be medicated in the 100 to 129 band, where drug treatment had previously been optional, and a new optional target of 70 was introduced for the highest risk of all (Grundy et al., Circulation, 2004).

Nobody got healthier on the day the line moved. Millions of people simply woke up as patients.

Then the disclosure. Eight of the nine authors of that update had financial relationships with statin manufacturers, and not one of those relationships was declared when the guidance was published. It took outside researchers to dig them out afterwards, and the petition for an independent review went nowhere (BMJ, 2004).

You do not need to believe in a conspiracy to notice the shape of that.

A treatment exists that lowers one number. The definition of sickness is then lowered until enough people qualify for the treatment. The men who moved the line were, by their own belated disclosures, not neutral about the answer.

Hold that beside the reference range problem from the table above and the appointment starts to look rather different. You are being graded against a line drawn by interested parties, on a scale calibrated in carb-fed bodies, while the two numbers that actually track metabolic damage sit unread in the same column.

The damage is not the count

Arteries do not fur up because too many couriers are on the road.

Plaque is a repair job that goes wrong. The artery wall gets damaged, LDL arrives as part of the cleanup, and under the wrong conditions it gets stuck, oxidises, and builds.

The wrong conditions have names.

Glycation, which is sugar chemically caramelising the particle until its receptors no longer recognise it. And oxidation, which is the particle being chemically damaged, a job for which industrially processed seed oils are singularly well qualified.

A damaged particle lingers, embeds and calcifies. A healthy one delivers and goes home.

Which is why the population data keeps refusing to behave. Higher LDL tracking with lower all-cause mortality in older adults. Saturated fat showing little effect on LDL in the obese, and, in one of the more inconvenient findings on record, an association with slower progression of arterial plaque rather than faster.

None of it fits the leaflet. All of it fits the damaged-particle model.

And it explains the great failed experiment. Half the world swapped butter for seed oils. LDL duly fell. The heart disease tide did not turn. Fewer couriers, identical damage, because the damage was never about how many couriers were on the road.

The case reports

Three case reports are doing the rounds, and they point in three different directions, which is the most useful thing about them.

The famous one is JAMA Cardiology, January 2025. A man in his 40s, eight months into what was reported as a carnivore diet, turns up with yellow cholesterol deposits on his palms, soles and elbows, and a total cholesterol over 1,000 mg/dL, 25 mmol/L.

Three details before you share it in either direction. His intake was six to nine pounds of cheese a day, sticks of butter, and extra fat worked into his daily hamburgers. His total cholesterol was already 210 to 300 before he started. He had not had a heart attack.

It is a case report of skin deposits produced by an extreme lipid level on an extreme intake, and at that number I would be sat in a cardiologist's office too. It is not a heart attack, and it is not what a plate of steak and eggs does to blood.

The second is the closest thing to a carnivore heart attack in the literature, and it is not carnivore. A 2022 report describes a 38-year-old man four weeks into a very low carbohydrate ketogenic diet, with chest pain and raised troponins. Catheterisation found his arteries clean. The diagnosis was a type 2 myocardial infarction, the supply and demand kind rather than the blocked artery kind, and the authors could do nothing with it beyond noting the timing.

The third runs the other way. A 2026 report describes a man in his 30s who went ketogenic to manage ulcerative colitis and watched his LDL climb from 95 to 574 mg/dL, his total cholesterol to 705, 18 mmol/L, with HDL at 124 and triglycerides at 34. Seven years at that level.

Then a CT angiogram: zero measurable plaque, soft or calcified, in any coronary vessel.

It is written up by the same researcher behind the retracted imaging study this page discusses further down, which is worth knowing and changes nothing about what it is, a single man with a scan.

One man at 1,000 with cholesterol coming out through his skin. One man at 700 for seven years with clean arteries. The distance between those two is everything this article says about context. If case reports settled anything, both sides would have declared victory years ago.

And the claim that keeps circulating, answered flat: there is no published case report of a heart attack on a carnivore diet.

There are ketogenic ones, and the closest is a 2024 conference abstract: a man of 62 with a family history of early heart attacks, seven months into keto for weight loss, who arrested with a fully blocked artery and an LDL of 542 mg/dL, and who had eaten strictly carnivorous for the five weeks before it.

Have that one in full before somebody hands it to you in pieces. It is an abstract rather than a peer-reviewed report, and one man with a family history settles nothing, in either direction. A peer-reviewed carnivore case report of a heart attack still does not exist. If that changes, this page changes with it.

ApoB, and the argument you will actually get from a cardiologist

The sharper version of the LDL argument has moved on from LDL cholesterol, and the number a good cardiologist reaches for now is ApoB.

Every particle capable of carrying cholesterol into an artery wall, LDL included, carries exactly one molecule of apolipoprotein B. Measure ApoB and you have counted the particles rather than weighed their cargo. In the mainstream risk model that count is the best single predictor of atherosclerosis on offer: more particles in circulation, more collisions with the wall, more chances for one to lodge and oxidise.

On carnivore, ApoB usually rises along with LDL. So when the conversation moves from "your LDL is high" to "your ApoB is high", you have met the stronger form of the argument, and the particle quality points made elsewhere on this page do not answer it, because ApoB is a claim about quantity.

Where the counter-argument actually sits, and it is not that ApoB is meaningless: the price list was built on populations eating a mixed diet, a great many of them insulin resistant, and nobody has checked whether the same particle count carries the same risk in a lean, insulin-sensitive person running on fat.

That is the exact question the retracted imaging study claimed to have answered. It stands unanswered. Test it, know it, and hold it open with real weight on both sides.

The lean mass hyper responder question

If you are lean, fit, long-adapted, and your LDL has gone somewhere your GP has never seen before, 6, 7, 8 mmol/L, while your HDL sits high and your triglycerides are on the floor, you have a name. Lean mass hyper responder. It is the courier model in its purest form. Maximum fat flux, minimum storage.

And the study everyone in this corner of the internet was quoting at their doctor has been retracted.

It followed a hundred of these exact people for a year with CT imaging and reported that LDL and ApoB did not predict plaque progression. It was held up as the answer. Critics pointed out that plaque progressed anyway, that the preregistered primary outcome had been sidelined in favour of a secondary one, and that there was no control group to speak of. The authors and the editors then withdrew the paper together, agreeing the methodology problems were too large to fix with a correction (retraction notice, JACC: Advances, 2026).

I am not going to quote a retracted paper at you, and I am not going to pretend it never existed either. What it means is that the strongest piece of reassurance this phenotype had is off the table, and the honest position is that nobody knows yet.

What I can tell you without a single study is this. A lean mass hyper responder with low triglycerides, high HDL, low inflammation and a clean glucose panel is in a categorically different room from a man with the same LDL, a forty inch waist and an HbA1c heading north.

The tragedy of the single-number appointment is that it cannot tell those two men apart.

That is not proof of safety. It is a reason to go and photograph the artery rather than keep arguing about the courier, which is what the scan further down is for.

The statin conversation

At some point the prescription pad comes out. So be precise about what is on offer.

Measured as postponement of death inside the trials’ own running time, statins bought a median of 3.2 days in people who had never had a heart event, and 4.1 days in people who had (Kristensen et al., BMJ Open, 2015).

Days.

And be fair about the counter, because this site does not get to complain about selective citation and then do it. A later re-analysis, modelling the gap between the survival curves rather than measuring it off the page, put the figures higher, roughly 10 days and 17 days (Hansen et al., Journal of General Internal Medicine, 2019). Take the generous version. It is still days and weeks rather than years, and it is still measured inside trials that ran a handful of years apiece.

Against that you weigh the muscle aches, the documented glucose disruption, and a daily tablet with your name on it for the rest of your life.

If you have already had a heart attack, the calculation genuinely does change. The event data in secondary prevention is stronger than the survival arithmetic makes it sound, and that conversation belongs with your cardiologist rather than with a fitness coach on the internet. I am not your doctor.

What I am is someone pointing out that “your LDL is high, here is a statin” skips every question that actually matters. High in what context. Damaged by what diet. Alongside what triglycerides. Inside what metabolism.

Lowering LDL with a pill while the glycation and the oxidation carry on is bailing the boat and ignoring the hole.

LDL-C against LDL-P: weighed or counted

One more pair of letters, because it decides how alarming your LDL number actually is. LDL-C weighs the cholesterol sitting inside your LDL particles. LDL-P counts the particles, measured directly by NMR. ApoB is the count's close cousin.

The two usually move together. The information lives in the times they don't. The best data on this is MESA, 6,814 people with no cardiovascular disease at entry, followed five and a half years. Where the cholesterol weight and the particle count disagreed, the events tracked the particle count.

People with high LDL-C and a low particle count did fine. People with unremarkable LDL-C and a high particle count, the classic insulin resistant pattern of many small dense particles, did worst of the lot.

That second group walks out of a standard cholesterol panel with a clean bill and stays sick, which tells you exactly what a standard panel is worth on its own.

The pattern low carb typically produces, low triglycerides and high HDL, goes with larger particles, meaning the same LDL-C is riding in fewer of them. That is the better direction to be discordant in.

It is also a thing you measure rather than assume, because some people on carnivore run a high count and a heavy cargo at the same time. An NMR panel or an ApoB test settles which of those you are for the price of a blood draw.

The numbers that actually deserve your attention

In the order I would look at them.

One. Triglycerides against HDL. The single most informative pairing on a standard NHS panel, and it costs nothing extra. It is also the pairing that tracks metabolic dysfunction most closely.

Two. HbA1c and fasting glucose. The glycation input. Carnivore usually puts these right, and when they are right, the sugar-damage pathway is shut.

Three. ApoB, once a year, or after any large change to how you eat. It is the number the strongest version of the opposing argument runs on, and either result is useful to you. A moderate ApoB largely retires the argument. A high one tells you precisely which open question you are carrying.

If your LDL is very high and you want the full picture, an NMR particle panel (LDL-P) tells you whether that cholesterol is riding in many particles or few.

Four. Waist, blood pressure, resting pulse. Free, boring, and collectively worth more than the lipid panel.

Five. A coronary artery calcium scan, if you want the arterial answer rather than another argument about couriers. It looks at the artery itself. A zero is the most reassuring result in this entire subject, whatever your LDL says. It is not routinely offered on the NHS, it is available privately for roughly the price of a decent pair of trainers, and it converts an argument about proxies into a photograph of the thing itself.

It is a one-off dose of radiation, typically around one to one and a half millisieverts on a standard protocol and lower on modern low-dose ones. A mammogram or two, and comfortably less than a British year of sky and ground.

Lipids are one of seven panels your GP will run, and carnivore moves numbers on the other six as well. The blood tests article walks all seven before you sit down.

What to say across the desk

You do not win this appointment by lecturing your GP about lipoproteins.

You win it by asking for the numbers that answer the actual question.

Ask what your triglycerides did. Ask what your HDL did. Ask for your HbA1c. If all three moved the right way, say so, out loud, and then ask the only question that matters. Given the whole panel, not the one number, what exactly has got worse?

Be firm and be polite. Your GP is not the villain here. They are reading the sheet they were handed, against lines they were handed, with ten minutes on the clock.

The lines are the problem.

And if the conversation truly stalls, ask about the calcium scan and settle it with a photograph.

Just because fat can clog your kitchen sink does not mean it can do the same to your arteries.

Sinks do not have a liver.

What nobody knows yet

Three things are unknown, and this page is stronger for putting them in writing.

Nobody knows the long-term cardiovascular outcome of a carnivore diet. No cohort, no trial, no registry. Anyone selling you certainty here, for the diet or against it, is selling.

Nobody knows whether the lean mass hyper-responder phenotype is protective, neutral or dangerous. The one imaging study was retracted in May 2026, and the detail that never survives the retelling is who asked for it.

The authors did. They went to the journal after publication, having found that the plaque analysis was not blinded and that one of their own co-authors was chief medical officer at the company running it. Until an independent group does it properly, that question stays open.

Nobody knows whether ApoB carries the same risk per particle in a lean, fat-adapted person as in the mixed-diet populations the risk model was built on. Arguments both ways, data neither.

What is known: LDL and ApoB rise on carnivore, most of all in the leanest people. Triglycerides, HDL, inflammation and glucose markers generally improve. Both halves of that sit in the sceptics' own review.

That is the map. Anything drawn on it in confident ink is decoration.

Key Takeaways

• LDL usually rises on carnivore, and rises most in the leanest and fittest. That is fuel logistics, not disease.
• Triglycerides down plus HDL up is the pattern of a working fat metabolism, and it is the pattern carnivore reliably produces.
• The danger was never the count. It is glycation and oxidation of the particles, driven by sugar and seed oils, both of which carnivore removes.
• The thresholds your number is judged against were moved by a panel whose ties to the companies selling the remedy were not disclosed at the time.
• The lean mass hyper responder question is genuinely open. The study that appeared to settle it has been retracted, and this site is not going to pretend otherwise.
• Judge your heart on the trig:HDL ratio, the glucose panel, and if you want certainty, a calcium scan. Not on one red number.

FAQ

Does the carnivore diet raise cholesterol?

Usually, yes. LDL in particular tends to rise, and rises most in lean, active people. Triglycerides typically fall and HDL typically rises at the same time, which is the pattern of a healthy fat-burning metabolism rather than a warning sign.

Is high LDL dangerous on a low carb diet?

The count alone tells you very little. Arterial damage is driven by glycated and oxidised LDL particles, products of high blood sugar and industrial seed oils. A carnivore eater with low triglycerides, high HDL and a normal HbA1c has removed both drivers, whatever the LDL count reads.

What is a lean mass hyper responder?

A lean, fit, carb-restricted person whose panel shows very high LDL, high HDL and low triglycerides together. The pattern comes from a high rate of fat transport in a body with little stored fat. Research on the phenotype is young, and the largest imaging study of it was retracted by its own authors and editors over methodology, so the honest answer is that the question is still open.

Should I take a statin if my LDL is high on carnivore?

That is a decision for you and your doctor, not a website. What the trials show is that in people with no history of heart disease, statins postponed death by a median of a few days within the trial period, and by a couple of weeks on a more generous later re-analysis. If you have existing heart disease the calculation is different. Either way, insist the decision is made on your whole panel rather than one number.

Which cholesterol numbers matter most?

The triglyceride to HDL ratio first, then HbA1c and fasting glucose. Those track the actual damage pathways. A coronary artery calcium scan answers the arterial question directly if you want it settled.

Will my LDL come back down if I quit carnivore?

Almost certainly, because you will have re-fuelled on carbohydrate and reduced the fat traffic the LDL was carrying. The question is what you buy with that lower number. The triglycerides, glucose and waistline that came down on carnivore tend to go back up with it.

Should I get a calcium score?

If the LDL argument is causing you or your doctor real anxiety, it is the cleanest way to end it. It images calcified plaque in the arteries themselves. A score of zero means very low risk over the following decade, whatever the lipid panel says.

Are there studies showing carnivore causes heart attacks?

No. As of 2026 there are nine published human studies on the carnivore diet and not one measured heart attacks, strokes or deaths. The 2026 review being circulated as evidence against the diet reached its conclusion by importing general red meat epidemiology, while stating in its own discussion that no carnivore study has ever reported a clinical endpoint. LDL does rise, and that deserves proper monitoring. "Raises LDL in small short studies" and "causes heart attacks" are different claims.

What about the man whose hands turned yellow?

A real JAMA Cardiology case from 2025. Total cholesterol over 1,000 mg/dL and cholesterol deposits in his skin, on a reported daily intake of six to nine pounds of cheese, sticks of butter and extra fat in his hamburgers, with a starting cholesterol already between 210 and 300. No heart attack. It shows an extreme intake can produce a genuine lipid emergency, which nobody disputes. There is also a 2026 case report of the opposite: seven years at 705 mg/dL with zero plaque on imaging. Case reports raise questions and cannot settle them.

Should I test ApoB?

Yes. It is the strongest number the mainstream case runs on, it is cheap, and either answer is worth having. A moderate result largely closes the argument. A high result tells you what you are actually carrying, and makes imaging, a calcium score for instance, more worth doing.

---

The Carnivore Fat Loss Series

This is part of an ongoing series on how carnivore strips body fat without counting calories. See the full hub →

• 7 Fixes For When You're Hungry On Carnivore
• Why Carnivore Outperforms Other Diets For Shedding Pounds
• What Nutrients Are Lacking On A Carnivore Diet?
• What I’ve Changed My Mind On After Four Years Of Carnivore

---

If you want to put together your own metabolic revival, follow me on Twitter or Instagram, or find out what the coaching programme involves below.