By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.
Peter Attia MD#19 – Dave Feldman: stress testing the lipid energy model
Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.
CHAPTERS
- 0:04 – 11:41
Housekeeping and Peter's Post-Interview Verdict on Dave's Model
Peter Attia covers housekeeping, introduces Dave Feldman and his lean mass hyper-responder idea, and mentions Tom Dayspring's review of the transcript. He then lays out why he is not convinced by the model: mass balance, VLDL production, and the genetic natural experiments.
- •Dave is known for the cholesterol drop protocol and the lean mass hyper-responder concept
- •Peter's first objection is that the model does not explain mass balance for the extra LDL cholesterol and particles
- •His second objection is that insulin-sensitive people export fewer triglycerides in VLDL, with less ApoC3
- •His third objection is that about 2,000 genetic mutations, such as PCSK9 gain of function, are natural experiments that should not be ignored
- •Peter stresses probability over certainty and asks listeners to decide for themselves
- 11:41 – 18:03
Setting the Stage: Dave's Background and the Decision Problem
Peter welcomes Dave to San Diego and frames the conversation around the real clinical decisions he makes about lipids. Dave explains he is not a trained biochemist, describes how his software background shaped his view, and says he believes fatty acid energy use is the main influencer of LDL.
- •Peter must make real lipid decisions with incomplete information
- •Dave stresses he is not a formally trained biochemist or medical professional
- •He sees lipoproteins as a distributed network, using a software engineering lens
- •He claims he can move LDL-C up and down without drugs or supplements
- •He names energy metabolism, especially fatty acid utilization, as the main driver
- 18:03 – 30:27
What Moves LDL-P: Triglyceride, Synthesis, Absorption and Clearance
Peter reviews Dave's pre-keto lipid panel and explains the four things that move LDL particle number. He stresses sterol testing to separate synthesis from absorption, and Dave notes that dietary fat may travel with cholesterol.
- •Dave's pre-diet panel showed LDL-C 131, HDL 40, triglycerides 80
- •Peter says the four drivers are triglyceride burden, cholesterol synthesis, absorption and LDL clearance
- •Desmosterol serves as a marker of synthesis and phytosterols as markers of absorption
- •Dietary cholesterol plays only a trivial role in circulating cholesterol, Peter says
- •Clearance can only be inferred if the other three factors do not change
- 30:27 – 36:52
Dave's Story: Prediabetes, Keto and the Lipid Spike
Dave recounts a 6.1 A1c in 2015, going low-carb alongside his father and sister, and seeing his cholesterol skyrocket. He explains how this led him to study lipoproteins and to see them as an energy distribution network.
- •An A1c of 6.1 and a family history of type 2 diabetes prompted the diet
- •His father and sister also improved on the diet, and their LDL rose only modestly
- •His own total cholesterol reached 329 and LDL-C about 200 in November 2015
- •He saw lipoproteins as boats carrying energy, with apolipoproteins as routing metadata
- •Chylomicrons and VLDL deliver fat-based energy, while almost everything returns to the liver
- 36:52 – 47:19
Lipoprotein Lineages, ApoC3 and LDL Origins
Peter walks through the chylomicron, VLDL/IDL/LDL and HDL lineages and the role of reverse cholesterol transport. They discuss the large fraction of LDL secreted directly by the liver, the harmful effect of ApoC3 on residence time, and LDL half-life.
- •Peter says the overwhelming evidence points to LDL carrying out reverse cholesterol transport
- •About 40% of LDL comes directly from the liver, per the Sacks data Peter cites
- •ApoC3 increases residence time and tracks with insulin
- •LDL half-life is about a day under normal conditions, Peter says, versus a longer residence time
- •No clinical test shows which LDL came from VLDL
- 47:19 – 52:42
Defining Lean Mass Hyper-Responders
Dave defines the lean mass hyper-responder pattern and speculates it reflects heavy VLDL secretion for energy. The conversation touches on the Virta Health data and on whether PPAR variants or other SNPs explain the pattern.
- •The pattern is LDL-C of 200 or higher, HDL-C of 80 or higher, and triglycerides of 70 or lower
- •These people tend to be thin, fit and very low carb
- •No ApoE type or SNP has clearly been tied to the phenotype
- •Virta's participants are mostly type 2 diabetics, so they are not the ideal pool
- •Dave hypothesizes high VLDL secretion that leaves many LDL particles behind
- 52:42 – 1:00:24
How Atherosclerosis Develops: Retention, Oxidation and OxLDL
Peter explains why CAC and CIMT are late markers and describes how retained LDL in the arterial wall is oxidized and triggers inflammation. He and Dave also discuss the OxLDL assay and whether LDL has an antioxidant role.
- •CAC and CIMT gauge advanced disease and are not biomarkers
- •LDL enters the subendothelial space by diffusion and is more likely to leave than stay
- •Proteoglycan binding retains particles, which then oxidize and trigger inflammation
- •OxLDL measures a small subset of oxidized particles that escape the intima
- •Peter doubts the evidence that part of LDL's role is fighting oxidative stress
- 1:00:24 – 1:13:49
All-Cause Mortality, Lifetime Exposure and the Zero-LDL Question
Dave describes searching for evidence that lower LDL means living longer. Peter responds with power problems in all-cause mortality studies, the lifetime exposure idea illustrated by FOURIER and ODYSSEY, and why LDL is necessary but not sufficient for atherosclerosis, using a patient case with suppressed cholesterol synthesis.
- •Peter says all-cause mortality is rarely powered to be detected in lipid trials
- •FOURIER and ODYSSEY showed benefit in about 2.2 years, which Peter finds striking given lifetime exposure
- •Peter compares LDL to oxygen for fire, necessary but not sufficient
- •He does not endorse driving LDL to zero because of possible costs elsewhere
- •In one patient, unmeasurable desmosterol led him to move to a PCSK9 inhibitor and cut the statin
- 1:13:49 – 1:27:37
The Genetics Debate: Mendelian Randomization, FH and PCSK9
Dave presents his challenge for studies without genes or drugs showing high LDL, high HDL and low triglycerides with high disease. Peter argues that excluding genetic data throws away rich natural experiments, and they debate lipid malabsorption and endothelial health.
- •Dave's challenge asks for studies that rule out genes and drugs
- •Peter likens excluding genetics to looking for six-footers only in a kindergarten class
- •Dave's lipid malabsorption hypothesis says gene variants raising LDL may also impair cell uptake
- •Peter notes endothelial cells lack LDL receptors, so PCSK9 and FH effects act on the liver
- •Peter says FH is a phenotype with about 2,000 genetic causes, not one disease
- 1:27:37 – 1:41:39
Case Study: A Hyper-Responder Patient and Peter's Synthesis Hypothesis
Dave asks whether lean mass hyper-responders are at risk, and Peter walks through the labs of a recent patient with an LDL-P above 3,500. Peter argues the high desmosterol points to increased cholesterol synthesis, possibly driven by saturated fat or ketones.
- •The patient had total cholesterol 504, LDL-C 362, HDL-C 94 and LDL-P above 3,500
- •Peter views risk through three legs: lipoprotein burden, endothelial health and inflammation
- •Peter sees elevated desmosterol and absorption markers in these patients, suggesting more synthesis
- •He points to diabetic ketoacidosis cases with elevated cholesterol that normalizes once corrected
- •Dave suggests resistance training and receptor-mediated uptake in muscle may explain differences
- 1:41:39 – 1:55:44
Peter's Exercise Kinetics Experiment and Remnant Cholesterol
Peter shows his old kinetic experiment tracking cholesterol and triglyceride in lipoproteins around hard workouts. Dave then raises remnant cholesterol and Peter explains its limited usefulness, citing the Garvey particle-number data.
- •Peter's kinetic data showed triglyceride moving out of lipoproteins while cholesterol barely moved
- •Remnant cholesterol can be computed as non-HDL minus LDL, but it says little about pathology
- •Triglycerides divided by five is only a crude estimate of VLDL cholesterol
- •Garvey's data show VLDL-P rising only slightly while LDL-P rises far more with insulin resistance
- •Dave doubts this applies to fat-adapted people, calling it uncharted territory
- 1:55:44 – 2:14:27
Energy Flux and the Mass Balance Standoff
Peter and Dave discuss fat flux and the liver as the body's 'energostat', then clash over where an athlete's very high LDL cholesterol comes from. Peter presses on mass balance, and Dave concedes that increased synthesis plus recycling is involved.
- •They agree on fat flux: lipolysis equals de novo lipogenesis plus re-esterification
- •Peter calls the liver the body's energostat, borrowing the term from Mark Friedman
- •Peter says energy-driven VLDL turnover should give cholesterol-depleted LDL, not high LDL-C
- •Peter's mass balance argument is that a person with three times the LDL cholesterol must be making more
- •Dave concedes he is making more cholesterol but frames it as meeting demand plus recycling
- 2:14:27 – 2:26:24
Fasting, Fat Manipulation and the MUFA Experiment
The two compare how LDL responds to fasting, fat loading and carb swaps. Peter describes a patient who cut saturated fat in favor of MUFA and saw LDL-P fall dramatically, while Dave describes his own carb swap experiments around a roughly 90 g threshold.
- •Peter's LDL-C fell from 64 to 37 after a week-long fast, and he sees LDL drop under caloric or fat deprivation
- •Hyper-responders show triglycerides falling and desmosterol rising, Peter says
- •A young patient with an LDL-P above 3,500 returned to about 1,300 after swapping saturated fat for MUFA
- •Dave's LDL-C drops after about three days at roughly 90 g net carbs
- •Peter asks whether such effects are transient or only present at extreme conditions
- 2:26:24 – 2:55:00
Evidence, Skepticism and the Low-Carb Community
Peter explains why he thinks the LDL literature is of a different quality than the food-pyramid literature, and criticizes the low-carb community's dismissiveness. Dave describes his energy model, and they discuss necessity versus sufficiency of LDL and the ambulance analogy.
- •Peter distinguishes shoddy nutrition literature from the LDL literature, including Nobel-winning work
- •He says the low-carb mindset of rejecting mainstream information can spill over to lipids
- •Dave says his challenge came from looking at the energy delivery and support sides of lipoproteins
- •Both agree LDL is necessary but not sufficient for atherosclerosis
- •They debate whether Mendelian randomization can be dismissed, with the ambulance analogy
- 2:55:00 – 3:05:24
Data Sets, Stratification and HDL-C Limits
Dave says he wants large data sets such as MESA or Framingham Offspring stratified by HDL, triglycerides and LDL-P. Peter says HDL-C and triglycerides reveal little about lipid metabolism and that the Quebec study shows LDL-C matters less once ApoB is known.
- •Dave wants a regression on triglycerides, HDL and LDL-P without adjustments
- •Peter asks how he'd know risk is not increased without long follow-up
- •Dave cites two studies stratified by these markers, and Peter points to the Quebec Heart Study and ApoB
- •Peter says HDL-C reveals nothing about HDL function or lipid metabolism
- •Peter notes nothing in science is proven and the null hypothesis remains the lipid hypothesis
- 3:05:24 – 3:10:32
Closing Thoughts and Advice to Seek Lipid Experts
Peter suggests Dave engage lipidologists, for example at an NLA meeting, rather than only low-carb audiences. Dave agrees to seek voices who can challenge his model, and the two wrap up with notes on recording timing and the usual outro.
- •Peter says methotrexate and IL-1/IL-6 studies fit the multi-factor model
- •Dave says he is pursuing the data in order to try to disprove his model
- •Peter recommends going to an NLA meeting and offers to help connect Dave
- •Peter says he respects Dave's pursuit but disagrees with the model
- •The episode was recorded July 26th and released months later
