Peter Attia MD#140 - Gerald Shulman, MD, PhD: Insulin resistance—molecular mechanisms and clinical implications
CHAPTERS
- 0:11 – 5:42
Episode Preview and Why Insulin Resistance Matters
Peter Attia introduces Dr. Gerald Shulman and his use of magnetic resonance spectroscopy to watch glucose and fat flux inside human cells. He warns that this is a very technical conversation and previews the closing discussion on metformin.
- •Shulman is a Yale professor, co-director of the Yale Diabetes Research Center and 2018 Banting Medal recipient
- •His NMR and mass spec methods show metabolic flux like a video rather than a blood-test snapshot
- •Understanding insulin resistance is framed as central to understanding longevity
- •Metformin discussion at the end is flagged as especially valuable
- 5:42 – 12:10
Shulman's Background and Insulin Resistance as a Root of Chronic Disease
Shulman describes his MD/PhD and endocrinology training and how exposure to his father's type 1 diabetes camp shaped his interest in metabolism. Attia argues that the continuum from hyperinsulinemia to type 2 diabetes underlies the major chronic diseases, and Shulman agrees, linking insulin resistance to heart disease, fatty liver and cancer.
- •Shulman's father was a diabetologist and he was exposed to type 1 diabetes at summer camp
- •He wanted to look inside the cell, which led him to NMR spectroscopy
- •Attia describes a continuum from hyperinsulinemia to NAFLD/NASH to type 2 diabetes underlying chronic disease
- •Shulman cites Reaven's work tying insulin resistance to atherosclerosis, fatty liver and cancer
- •Insulin infusion accelerated tumor growth in mouse models and insulin sensitizers slowed it
- 12:10 – 19:25
How NMR and Isotope Labeling Reveal Metabolism Inside Cells
At Attia's request, Shulman explains how tracers and NMR let researchers measure flux rather than static concentrations. He covers radiolabeled tracers, PET, and carbon-13 and phosphorus-31 NMR, which can track glucose into glycogen and measure intracellular intermediates non-invasively.
- •Flux (production versus uptake) matters more than static concentration
- •Traditional tracers use radioactive C-14 or tritium and measure turnover but not intracellular events
- •C-13 NMR tracks a labeled glucose carbon into glycogen without ionizing radiation
- •P-31 NMR sees ATP, phosphocreatine and glucose-6-phosphate in muscle
- •These tools let them ask where ingested carbohydrate goes in humans
- 19:25 – 25:35
Fate of Glucose: Healthy vs. Type 2 Diabetes
Attia sets up a comparison of a healthy person and one with type 2 diabetes after a high-carbohydrate meal. Shulman explains that most ingested glucose goes to glycogen in healthy people, while in diabetes the liver overproduces glucose and muscle uptake is blocked, and says it is easiest to start with pure insulin resistance.
- •In healthy people 80-90% of ingested glucose is stored as liver and muscle glycogen
- •In diabetes, gluconeogenesis is accelerated and peripheral glucose clearance is reduced
- •Beta cell failure can occur in established diabetes
- •Modest calorie restriction (1,200 kcal/day) can reverse these abnormalities by reducing ectopic lipid
- •Starting with young lean insulin-resistant people makes the progression easier to follow
- 25:35 – 31:37
Defining Insulin Resistance and Locating the Block at Glucose Transport
Shulman defines insulin resistance as insulin failing to do its normal work in muscle, liver and fat, noting it is likely present in about half of Western populations and missed by simple glucose tests. Using NMR in young lean volunteers, his group found that glycogen synthesis is down by half and that the block lies at glucose transport. Attia then reviews the normal insulin signaling pathway to GLUT4.
- •Insulin resistance means more insulin is needed to get normal effects in muscle, liver and fat
- •Young lean insulin-resistant volunteers have a 50% reduction in muscle glycogen synthesis
- •Measuring glucose and glucose-6-phosphate showed both reduced, pointing to transport rather than hexokinase or glycogen synthase
- •Attia outlines insulin receptor, IRS-1, PI3 kinase and GLUT4 translocation
- •The transport step is the target to fix, not the downstream enzymes
- 31:37 – 50:13
Lipid-Induced Muscle Insulin Resistance and the DAG–PKC Mechanism
Shulman shows slides explaining how fat inside muscle cells predicts insulin resistance and how lipid infusion causes it within hours in humans. He argues that diacylglycerol, not triglyceride, activates novel PKCs that block insulin signaling, supported by genetic tests in rodents.
- •Intralipid plus heparin raises fatty acids and induces insulin resistance in 3-4 hours via the same transport block
- •This contradicted Randle's proposed mechanism, since G6P fell rather than rose
- •Lipid infusion abolished insulin activation of PI3 kinase in human biopsies
- •DAGs activate PKC theta and epsilon when fatty acid flux exceeds oxidation or storage
- •Triglyceride is an inert storage form and a marker, not the culprit
- •Knockouts and manipulations of oxidation, fat entry and LPL support the mechanism
- 50:13 – 59:13
From Muscle Insulin Resistance to Fatty Liver and Dyslipidemia
Shulman explains that muscle insulin resistance diverts ingested carbohydrate to the liver, where portal hyperinsulinemia drives de novo lipogenesis. The data show raised liver triglyceride, higher plasma triglycerides and lower HDL in euglycemic young people, and Attia comments on clinical triglyceride thresholds.
- •Insulin-resistant young volunteers maintain normal glucose only through two-to-three-fold higher insulin
- •Portal vein insulin is about three times plasma levels
- •Liver triglyceride and de novo lipogenesis rise more than twofold after high-carbohydrate meals
- •Insulin turns on SREBP-1c, increasing DNL and VLDL export, raising triglycerides and lowering HDL
- •Attia and Shulman discuss the Hellerstein DNL data and the point that normal ranges may need resetting
- 59:13 – 1:07:32
How Exercise Bypasses Muscle Insulin Resistance and Reduces Liver Fat
Shulman describes studies in which exercise normalized muscle glycogen synthesis in insulin-resistant offspring, likely through AMPK-driven GLUT4 translocation independent of PI3 kinase. A single bout reduced de novo lipogenesis and liver triglyceride. Attia relates this to type 1 diabetes patients who exercise heavily.
- •Six weeks of stair-machine exercise normalized insulin-stimulated glycogen synthesis
- •A single 45-minute bout opens the glucose transport door
- •Shulman speculates acute effects come mainly from AMPK-induced GLUT4 translocation, with chronic exercise also lowering DAGs
- •A single bout of exercise reduced DNL and liver triglyceride
- •Attia describes type 1 patients using brisk walking to need very little insulin
- 1:07:32 – 1:21:41
Liver Insulin Resistance and the Evolutionary Case for Starvation Protection
Shulman explains that in humans muscle insulin resistance comes first and liver insulin resistance follows once fatty liver develops, unlike in rodents. He walks through how DAG–PKC epsilon phosphorylates a threonine on the insulin receptor, a site conserved from humans to fruit flies, and argues insulin resistance protects glucose supply for the brain during starvation.
- •In humans the order is muscle, then liver; rodents develop it in the reverse order
- •PKC epsilon phosphorylates a threonine in the insulin receptor catalytic domain
- •Mutating that threonine to alanine protects mice from lipid-induced hepatic insulin resistance despite equal liver fat
- •The specific sn-1,2 DAG isoform in the plasma membrane is what matters
- •Starved rats develop fatty liver and hepatic insulin resistance, and Mexican cavefish show a similar protective mutation
- •Shulman views insulin resistance as a starvation survival mechanism now harmful in overnutrition
- 1:21:41 – 1:30:58
Gluconeogenesis and How Insulin Really Regulates It
Attia and Shulman agree gluconeogenesis is essential for survival, though patients with glucose-6-phosphatase deficiency can survive with continuous feeding. Shulman challenges the textbook view that insulin suppresses it through FOXO transcription, citing evidence for acetyl-CoA allosteric control and an indirect effect through insulin's brake on fat-cell lipolysis.
- •Gluconeogenesis supplies glucose for the CNS overnight and during starvation
- •Knockout of AKT or FOXO does not prevent insulin from suppressing gluconeogenesis, and flux can be turned off in about five minutes
- •Liver samples from poorly controlled diabetics showed no correlation of PEPCK or G6Pase protein with glucose
- •Acetyl-CoA activating pyruvate carboxylase tracks gluconeogenic flux
- •Insulin controls glycogen synthesis directly in liver but gluconeogenesis mostly via suppression of peripheral lipolysis
- 1:30:58 – 1:41:21
Inflammation, Fat Tissue, and the Transition to Fasting Hyperglycemia
Shulman explains that insulin resistance can occur without inflammation, but moving to fasting hyperglycemia depends on adipose inflammation increasing lipolysis. This raises fatty acid and glycerol delivery to the liver, doubling acetyl-CoA and gluconeogenesis. Attia asks about fat-cell insulin sensitivity in obesity and visceral fat, and Shulman describes visceral fat as mostly a marker for liver fat, citing lipodystrophy.
- •Inflammation can be dissociated from insulin resistance, but the step to hyperglycemia depends on it
- •TNF-alpha and IL-6 from adipose macrophages boost lipolysis, raising acetyl-CoA and glycerol delivery to liver twofold
- •Obese people likely remain insulin resistant at the fat cell, with the DAG–PKC epsilon pathway also operating there
- •Visceral fat correlates with liver fat and is largely a marker
- •Lipodystrophy shows diabetes can occur without visceral fat when the liver is fatty
- •Shulman picks acetyl-CoA driving pyruvate carboxylase and DAG activating PKC epsilon as two key molecules
- 1:41:21 – 1:48:28
Treating Fatty Liver: Lifestyle, Surgery, GLP-1s and SGLT2 Inhibitors
Attia summarizes the nutrient-abundance framing and the limited treatment pipeline for NAFLD. Shulman explains why he sees the liver as the key target and as a thermodynamic problem of energy in versus out, then reviews diet, exercise, bariatric surgery, GLP-1 agonists and SGLT2 inhibitors.
- •Fatty liver is becoming a leading reason for liver transplant, and fatty donor livers cannot be used
- •Diet and exercise work but most patients cannot sustain them
- •Bariatric surgery effects appear to come from weight loss, as pair-fed weight loss gives the same results
- •GLP-1 agonists work mainly by reducing food intake and partially reverse fatty liver
- •SGLT2 inhibitors cause about 400 kcal/day urinary glucose loss but modest liver fat reduction
- 1:48:28 – 1:55:12
Liver-Targeted Mitochondrial Uncoupling: A New Pharmacologic Approach
Shulman describes his favorite approach: making liver mitochondria less efficient so they burn more fat. He tells the history of dinitrophenol and its toxicity, explains how liver targeting could widen the therapeutic window, and says preclinical results are promising, with human trials possibly first in lipodystrophy.
- •Mild uncoupling reverses fatty liver, NASH, fibrosis and diabetes in preclinical models
- •DNP was used for weight loss in the 1930s but caused fatal hyperthermia and was pulled
- •Liver-specific uncoupling avoids whole-body heat and widens the therapeutic window about a hundredfold
- •Approach lowers VLDL export, triglycerides and raises HDL, aiming to be heart-healthy
- •Still preclinical, with possible first use in lipodystrophy within a year or two
- 1:55:12 – 1:58:22
Dietary Approaches and Closing Thoughts on Insulin Resistance
Attia wraps up the technical discussion, stressing why understanding insulin resistance matters, and asks Shulman about carbohydrate or fructose restriction. Shulman says he tells patients to use whatever works and that the challenge is long-term adherence.
- •Attia argues chronic diseases should not be treated in silos without understanding insulin resistance
- •Attia reports carbohydrate restriction works well clinically for many insulin-resistant patients
- •Shulman advises patients to find a weight-loss approach that works for them
- •Maintaining weight loss over months and years is the main difficulty
- 1:58:22 – 2:08:41
Metformin's Mechanism and Its Use as a Longevity Agent
Shulman argues metformin lowers glucose mainly by inhibiting gluconeogenesis and that clinical plasma concentrations (roughly 30-50 micromolar) are far below those needed to inhibit complex I. He proposes it acts on mitochondrial glycerol-3-phosphate dehydrogenase, raising cytosolic redox and inhibiting gluconeogenesis only from lactate and glycerol. Both he and Attia are undecided on whether it helps healthy, insulin-sensitive people.
- •Metformin's glucose lowering is mainly through reduced gluconeogenesis, with greater effect in poorly controlled diabetes
- •Complex I inhibition needs millimolar concentrations, while patient levels are tens of micromolar
- •Shulman proposes inhibition of mitochondrial glycerol-3-phosphate dehydrogenase, raising NADH and cytosolic redox
- •Effect is substrate-dependent, blocking gluconeogenesis from lactate and glycerol but not alanine or DHAP
- •This could explain rare hypoglycemia and the lactic acidosis risk
- •Whether metformin benefits healthy people for aging is unknown to both speakers
