Why Your LDL Exploded on Keto, And Why 'Just Ignore It' Is Bad Advice

Dr. Feld doesn't do small talk, so the pause on the phone was the first thing Marcus noticed.

Your LDL is 258. Your triglycerides are 58. Your HDL is 92. Your fasting glucose is 88.

Marcus wrote the numbers down and said he would call back. Then he went looking for what they meant, and walked into a fight.

The keto corner of the internet told him to relax. LDL doesn't matter. It's triglycerides and HDL that matter. Your doctor is stuck in 1985. The medical corner told him his diet was taking years off his life and he should stop before it did damage. Neither side asked a single question about his labs, his family history, or his life. Both wanted him to stop thinking and pick a team. It's the same blame-first reflex that makes people quit low-carb.

This is the piece Marcus wishes he had found that night. It won't tell you to ignore your labs. It won't tell you to panic, either. It'll give you the biology, name the pattern he's in, and hand you better questions than a comment section will.

The pattern that scares people

To see why Dr. Feld sounded worried, look at the shape of the change, not just the number.

On a well-run low-carb diet, most people move in a familiar, reassuring direction. Triglycerides fall, often sharply. HDL climbs. Fasting blood sugar and insulin improve. Blood pressure and weight come down. It's the pattern the low-carb world talks about endlessly, and it's real.

Marcus was the picture of it. Forty-four, lean, four lifting sessions a week, a bike commute even when it's raining sideways. He had never once had a doctor's appointment end with a serious face. He had cut the carbs nine months earlier, mostly because a friend at the gym would not stop talking about how much better he felt, and he had kept his two Friday IPAs, three if the week had earned it.

He also got something stranger layered on top. His LDL didn't drift up a little. It rocketed, straight into territory a cardiologist would flag in someone who had already had a heart attack. And it climbed while everything else was getting better.

That mismatch is what scrambles everybody's head. If your health is improving on almost every metric, how can the one number most tied to heart disease be running the wrong way? Either something is wrong with the diet, or something is wrong with the number.

The honest answer is that we don't fully know. The people most certain they do are usually selling something.

It has a name, and Marcus fits it

This isn't a forum rumour. Researchers have described a reproducible phenotype, a pattern of how certain people answer carbohydrate restriction. They call it the lean mass hyper-responder, or LMHR.

The working definition, laid out in the Journal of Clinical Lipidology and in survey research on low-carb dieters, is a triad: an LDL cholesterol of 200 mg/dL or higher, an HDL of 80 or higher, and triglycerides of 70 or lower, all while eating low-carb and carrying relatively little body fat.

Read that again, because the second two numbers matter as much as the first. A triglyceride-to-HDL ratio that low is one of the best markers we have of good metabolic health. It's the mirror image of the classic metabolic-syndrome picture, where triglycerides run high, HDL runs low, and LDL is only mildly elevated. The LMHR is that picture flipped. The LDL climb seems to depend on the carbohydrate restriction itself, and not on weight gain or poor health.

Marcus's numbers, 258 and 92 and 58, read like a definition rather than a description. The type who lands here tends to be lean, active, often younger, often someone who was metabolically healthy before they changed anything. The survey research found LMHR respondents had a lower average BMI than other low-carb dieters, and LDL numbers similar to everyone else before they started. Their baseline was fine. The diet moved the needle.

That matters, because it kills the lazy explanation. This isn't a man who was always headed for trouble. It's a man whose body answers fat-adaptation in an unusual way.

The lab pattern is real, it's documented, and it is not you doing keto wrong.

The part you can't wave away

Here's where the keto cheerleaders lose Marcus, and where he thinks they do real harm.

In the general population, LDL cholesterol is not a fringe theory or a debunked idea. Decades of evidence back the view that LDL is a causal and cumulative driver of atherosclerosis, the plaque that narrows arteries over years. That evidence includes genetic studies, randomized trials, and the plain fact that lowering LDL reduces heart attacks. It isn't marketing. It's one of the sturdier findings in preventive cardiology.

None of that felt abstract to Marcus. The night Dr. Feld called, the kids were at their mother's, so the flat was quiet, and quiet is when the divorce still finds him. There was a beer open and a block of dark chocolate he had been picking at, and a number in the hundreds that was not supposed to be there. He could not decide the evidence didn't apply to him just because he felt good.

Feeling good doesn't get you off that hook. If your LDL sits at 250 for a decade, you're running a long experiment on your own arteries. Low triglycerides don't cancel it out. Anyone who tells you LDL is meaningless isn't being straight with you, and the honest version of this conversation starts by admitting that.

So Marcus sits inside a real tension. The same number that predicts risk across a whole population shows up in people with an otherwise ideal metabolic profile. That's the question worth ten arguments: is the LDL itself doing the damage, or the context it travels in?

The other side of the ledger

Now the part the medical world tends to skip, because it's harder to fit into a clean warning.

The case against blind alarm rests on a couple of things. First, LMHR labs look excellent everywhere except LDL. Low triglycerides, high HDL, normal blood sugar, normal inflammation. Second, there's early imaging data.

In 2024, researchers reported in JACC: Advances on a group of metabolically healthy people with carbohydrate-restriction-induced LDL elevations averaging around 272 mg/dL over roughly 4.7 years. Compared with a matched population cohort carrying much lower LDL, they did not have a greater burden of coronary plaque. That's reassuring, and it's the strongest card the don't-panic side holds.

Notice what that study can't tell you, though. Most of those participants were younger and leaner than the average person walking into a cardiology clinic. A few years of imaging is not thirty years of follow-up. A group that hasn't shown extra plaque yet is not a group that never will. Marcus is 44, not 70. His arteries have decades left in which to disagree with a short study.

Reassuring data and missing data are not the same thing. Only one of them is a reason to stop looking.

The honest gap in the science

Nobody has run the study that would settle this.

A 2025 review in the Journal of the American Heart Association makes the point plainly. It notes that while LDL is a well-established causal risk factor in the general population, the long-term consequences of ketogenic-diet-induced LDL elevation in lean, healthy people haven't been studied directly, and coronary imaging in this specific group is limited. We have a plausible mechanism, a real phenotype, some short-term reassurance, and one large open question.

There's a darker data point worth knowing too. Clinicians have published case series of people who developed extreme LDL elevations after starting a ketogenic diet. Cholesterol fell again once they stopped. Those cases show that in at least some people the response is severe enough to deserve real medical attention rather than a shrug and a slogan. That doesn't make keto dangerous. It does mean "hyper-responder" means hyper, and a number in the hundreds should start a conversation rather than end one.

What Marcus did with his labs

Here's the framework he used. It rests on one idea: don't panic, don't deny, and get better information than a single number on a standard panel.

Priya came with him to that appointment. She's a veterinarian, which he insists is practically a doctor, and by then she had heard enough of both internet camps to want a real opinion in the room. She also describes Marcus as the healthiest man she's ever dated and the worst person to take to brunch.

First he asked Dr. Feld for the advanced lipid panel. That means ApoB, which counts the actual number of atherogenic particles in your blood and predicts risk better than LDL-C alone, and Lp(a), a genetically set particle that raises baseline risk and is almost never checked even once. Marcus's ApoB came back on the high side. His Lp(a) was normal, real good news he would never have known without asking.

Then he asked whether a coronary artery calcium (CAC) score made sense for him. It did. The scan measures calcified plaque in your arteries. His score came back zero, which means there is no detectable calcified plaque right now. Not a guarantee. Not a receipt that says he's fine forever. Real information from his own body, though, instead of a slogan borrowed from either camp.

He also started treating the numbers as a trend rather than a snapshot. One alarming reading after eight weeks is information. The same reading holding steady for three years is different information. Marcus put a repeat lab on the calendar for six months out, then another after that. He's still eating low-carb. He still has his Friday beers. He is not cured, and he would be the first to say so.

What to ask your doctor

If you're staring at a scary number and don't know how to open the conversation, borrow Marcus's script. You don't need to be an expert to ask good questions.

Ask: "I eat a very low-carb diet and my LDL went up a lot. Can we look beyond the standard panel? Can we check my ApoB and my Lp(a)?" Ask whether a CAC score is worth doing in your situation. Ask what your overall risk looks like given your blood pressure, family history, and the rest of your numbers, not just this one line. And ask the direct question: "If this were your body, what would you do?"

Then listen. A good doctor will engage with the LMHR pattern instead of dismissing it, and will tell you honestly when a number is high enough that the plan needs to change. If yours won't have that discussion at all, it's fair to seek a second opinion, ideally from someone who reasons about particle burden and long-term risk rather than reaching straight for a prescription.

Where this leaves Marcus

He didn't get a verdict, which is the honest ending. He got a better picture.

You are not imagining the trade-off, and you are not necessarily in danger. The LMHR pattern is real, documented, and the early plaque imaging is reassuring. Marcus didn't let anyone scare him off a way of eating that works for him, based on one number and no context.

He also didn't let anyone tell him to wave it away. LDL being causal in the general population isn't something you get to opt out of just by feeling well. So he got the sharper picture: ApoB, Lp(a), a CAC score. He's tracking it across years with a doctor who treats him like an adult. He'll go back in six months, and six months after that. If something changes, he'll know, because he's watching instead of arguing.

Your bloodwork isn't a verdict, and it isn't noise. It's a question, and it's still open.

Sources

• Norwitz et al., "Elevated LDL-cholesterol levels among lean mass hyper-responders on ketogenic diets," Journal of Clinical Lipidology, doi.org/10.1016/j.jacl.2022.10.010
• "Elevated LDL Cholesterol with a Carbohydrate-Restricted Diet" (LMHR survey and defining triad), doi.org/10.1093/cdn/nzab144
• "Carbohydrate Restriction-Induced Elevations in LDL Cholesterol and Coronary Plaque," JACC: Advances, jacc.org
• "Ketogenic Diets and Low-Density Lipoprotein Cholesterol," Journal of the American Heart Association, ahajournals.org
• "Profound Elevation in LDL Cholesterol Level Following a Ketogenic Diet" (case series), PMC9402960
• "Lean mass hyper-responder", PubMed 36351849