LMHR: Why Your Cholesterol Went Up on Keto (And Why It Might Be Fine)

By Dr. Matt Altman, MD — Emergency Medicine Physician & Functional Medicine Practitioner

Reading time: 14 minutes | Last updated: February 2026

---

You started keto. You feel amazing — more energy, clearer thinking, lost weight, inflammation down. Then you get blood work done and your doctor nearly falls out of their chair.

Total cholesterol: 320. LDL: 220.

"You need a statin immediately," they say. "This diet is killing you."

But is it? Let's talk about the Lean Mass Hyper-Responder (LMHR) phenotype — one of the most fascinating and controversial topics in lipidology right now.

Disclaimer: This content is for educational purposes only and is not medical advice. Elevated LDL on any diet should be discussed with your healthcare provider. Lab Sage provides educational analysis to help you have more informed conversations with your doctor.

---

What Is an LMHR?

The term "Lean Mass Hyper-Responder" was coined by Dave Feldman and the Citizen Science Foundation. An LMHR is someone who, on a low-carb or ketogenic diet, develops a specific lipid pattern:

• LDL-C ≥ 200 mg/dL
• HDL-C ≥ 80 mg/dL
• Triglycerides ≤ 70 mg/dL

This triad — high LDL, high HDL, and very low triglycerides — occurs predominantly in lean, metabolically healthy individuals. It's estimated to affect somewhere between 5-25% of people who adopt a strict low-carbohydrate diet, with higher rates in leaner individuals.

The Key Distinction

This is NOT the same as someone with obesity, insulin resistance, and high LDL. That pattern — high LDL, low HDL, high triglycerides — is a well-established cardiovascular risk profile.

The LMHR pattern is metabolically the opposite:

• Excellent insulin sensitivity
• Very low triglycerides (a marker of good metabolic health)
• High HDL (protective)
• Lean body composition
• No markers of metabolic syndrome

---

Why Does This Happen? The Lipid Energy Model

The prevailing hypothesis, proposed by Dave Feldman and increasingly supported by research, is the Lipid Energy Model.

Here's the simplified version:

When you're lean and eating very low carb:

1. Your body runs primarily on fat for fuel (ketosis)
2. Your liver needs to distribute a LOT of energy to your muscles and tissues — via lipoproteins
3. More VLDL particles are produced to deliver triglycerides as fuel
4. After VLDL delivers its triglyceride cargo, the remaining particle becomes an LDL particle
5. Higher throughput of VLDL → more LDL particles in circulation

In other words: the elevated LDL may be a traffic phenomenon — more delivery trucks on the road because there's more energy to deliver — rather than a sign of atherosclerosis.

The Critical Variable: Lean Mass

The leaner you are, the more pronounced this effect tends to be. Why? Because lean individuals have less body fat to draw from passively. Their liver has to work harder to actively distribute dietary fat energy to tissues, requiring more lipoprotein "delivery trucks."

This is why the same ketogenic diet might cause minimal LDL changes in someone with higher body fat (who can release stored fat locally) but dramatic LDL increases in someone who is already lean.

---

What Does the Research Say?

This is where we need to be honest: the research is still early, and this is genuinely uncertain territory.

Evidence suggesting LMHR may not carry the same risk:

1. The Lipid Energy Model Study (Feldman et al.) Multiple controlled experiments have demonstrated that LDL in hyper-responders can be dramatically manipulated (up or down by 100+ points) simply by changing dietary fat intake over 3-5 days. This kind of acute responsiveness is inconsistent with the traditional atherosclerosis model, which involves LDL particles spending extended time in the arterial wall.

2. Favorable metabolic markers LMHRs consistently show:

• Very low triglycerides (typically < 70)
• Very high HDL (typically > 80)
• Low fasting insulin
• Low hs-CRP (low inflammation)
• TG:HDL ratio often < 1.0 (excellent)
• Low or absent coronary artery calcium (in many case reports)

3. The Framingham Paradox In the original Framingham data, individuals with high LDL but also high HDL and low triglycerides had relatively low cardiovascular event rates. The LDL risk was most pronounced when accompanied by low HDL and high triglycerides.

4. The "Pattern A" likelihood With very low triglycerides, LDL particles are more likely to be large and buoyant (Pattern A) rather than small and dense (Pattern B). This has been confirmed in multiple NMR analyses of LMHR individuals.

Evidence suggesting caution:

1. The Coronary Artery Calcium (CAC) question The LMHR Study (an ongoing clinical trial) is specifically looking at CAC scores in LMHRs. Preliminary data is being analyzed. Until this data is published, we cannot definitively say that sustained very high LDL is benign.

2. Mendelian randomization studies Genetic studies consistently show that lifelong LDL exposure increases cardiovascular risk. The counter-argument is that genetically elevated LDL is biologically different from diet-responsive LDL elevation in metabolically healthy individuals — but this remains debated.

3. ApoB concerns If ApoB is elevated alongside LDL-C (which it often is in LMHRs), some lipidologists argue the particle count itself confers risk regardless of the metabolic context.

4. Duration of exposure Even if LMHR LDL particles are less atherogenic per-particle, sustained exposure over decades is an unknown. Most LMHR data covers years, not decades.

---

What Should You Actually Do?

If you're on keto and your LDL has skyrocketed, here's a practical framework:

Step 1: Confirm the LMHR Pattern

Check all three criteria:

• LDL-C ≥ 200
• HDL-C ≥ 80
• Triglycerides ≤ 70

If you have high LDL but your HDL is low and/or triglycerides are high — that's NOT the LMHR phenotype. That's a different (and more concerning) lipid pattern that needs attention.

Step 2: Get Advanced Testing

A standard lipid panel is not enough to assess risk in this context. Request:

• ApoB — measures actual atherogenic particle count
• Lp(a) — genetic risk factor (test once)
• LDL Particle Number and Size (NMR LipoProfile) — are they large-buoyant or small-dense?
• hs-CRP — systemic inflammation
• Fasting insulin and HOMA-IR — confirm insulin sensitivity
• Coronary Artery Calcium (CAC) Score — the most direct measure of actual atherosclerosis

Step 3: Assess Your Full Metabolic Picture

If you're an LMHR, your metabolic picture likely looks like this:

• ✅ Fasting insulin < 5 μIU/mL
• ✅ TG:HDL ratio < 1.0
• ✅ hs-CRP < 0.5 mg/L
• ✅ Fasting glucose < 90 mg/dL
• ✅ Normal or zero CAC score
• ✅ Normal blood pressure
• ✅ Lean body composition

If ALL of these are true alongside elevated LDL, you're in a genuinely gray area of medicine. The traditional risk calculators (Framingham, ASCVD) were not validated on this population and almost certainly overestimate risk.

Step 4: Make an Informed Decision

Options include:

1. Stay the course, monitor closely. Get advanced lipid testing + CAC every 1-2 years. If CAC progresses, reassess.

2. Modify the diet. Adding some carbohydrates back (even 75-100g/day) or reducing saturated fat intake (swap butter/coconut oil for olive oil/avocado) often reduces LDL significantly while maintaining ketosis benefits.

3. Consider medication. If your ApoB is very high, Lp(a) is elevated, or CAC is progressing — a statin or other lipid-lowering therapy may be appropriate regardless of diet. This is a personal risk-benefit calculation to make with your doctor.

---

The "Just Add Carbs" Test

Here's a useful experiment your doctor probably won't suggest:

If you want to know whether your elevated LDL is diet-responsive (LMHR pattern) or structural (genetic/metabolic):

1. Reintroduce 100-150g of carbs per day for 2-3 weeks
2. Retest your lipid panel

If LDL drops dramatically (often by 50-100+ points), it confirms the diet-responsive LMHR mechanism. If it stays elevated regardless, there may be an underlying factor (like familial hypercholesterolemia or elevated Lp(a)) worth investigating.

---

What the Conventional Medical System Gets Wrong

The current medical paradigm treats all LDL elevation the same way: statin. But context matters enormously:

Treating both columns the same is bad medicine. Period.

---

The Bottom Line

The LMHR phenotype is real, reproducible, and currently being studied in rigorous clinical trials. If you fit this pattern:

1. Don't panic — but don't ignore it either
2. Get advanced testing — ApoB, Lp(a), NMR, hs-CRP, CAC
3. Look at the full metabolic picture — not just one number
4. Find a doctor who understands lipidology in the context of low-carb diets
5. Monitor regularly — this is a watch-and-measure situation, not ignore-or-medicate

The science is evolving. The honest answer is: we don't have 30-year outcome data on LMHRs yet. What we do have suggests this pattern is metabolically different from traditional dyslipidemia — but "metabolically different" isn't the same as "proven safe."

Make informed decisions. Get the right tests. And don't let anyone reduce your cardiovascular health to a single LDL number.

---

How Lab Sage Helps

Lab Sage doesn't just flag your LDL as "high" and tell you to take a statin. We analyze your complete lipid picture — TG:HDL ratio, metabolic context, and functional medicine ranges — to give you a nuanced interpretation that helps you have a smarter conversation with your doctor.

Upload your labs at labsage.ai and see your results through a functional medicine lens.

---

Dr. Matt Altman is an emergency medicine physician and functional medicine practitioner. Lab Sage was built to bridge the gap between conventional lab interpretation and functional medicine analysis.

This article is for educational purposes only and does not constitute medical advice. Decisions about cholesterol management, medication, and dietary interventions should be made in consultation with your healthcare provider.