#337 - Insulin resistance masterclass: The full body impact of metabolic dysfunction and prevention, diagnosis, and treatment | Ralph DeFronzo, M.D.

Episode Summary

Executive Summary: Peter Attia interviews Dr. Ralph DeFranzo on the organ-specific biology of insulin resistance and type 2 diabetes. The discussion explains why diabetes is a multi-organ, heterogeneous disease, how euglycemic clamps reveal tissue-level defects, and why combination therapy targeting beta cell failure, hepatic glucose output, lipolysis, and appetite outperforms stepwise treatment. DeFranzo also details the roles of GLP-1s, SGLT2 inhibitors, pioglitazone, and metformin.

Main Topics: Defining insulin resistance across organs (Priority: 5/5): DeFranzo explains that insulin resistance is not a single phenomenon but varies by tissue and pathway: liver, muscle, fat, beta cell, gut incretins, alpha cell, kidney, and brain. Euglycemic clamp and metabolic phenotyping (Priority: 5/5): He describes the clamp as the gold standard for quantifying insulin sensitivity, showing how it isolates insulin action while keeping glucose constant and revealing organ-specific responses. The ominous octet and type 2 diabetes pathophysiology (Priority: 5/5): The conversation expands the classic model of diabetes from a triumvirate to eight interlocking defects, emphasizing beta-cell failure, insulin resistance, lipotoxicity, glucagon excess, renal glucose reabsorption, and brain appetite signaling. Drug mechanisms and precision combination therapy (Priority: 5/5): They compare GLP-1 receptor agonists, SGLT2 inhibitors, pioglitazone, and metformin, arguing that no single drug addresses the full pathophysiology and that early combination therapy is superior. Hyperinsulinemia, lipotoxicity, and cardiovascular risk (Priority: 4/5): DeFranzo argues that chronic hyperinsulinemia can worsen insulin resistance and may be atherogenic, while elevated free fatty acids drive multi-organ metabolic injury. Diagnosis beyond HbA1c (Priority: 4/5): The discussion emphasizes OGTT patterns, one-hour glucose, C-peptide, and clamp-derived measures as more informative than HbA1c alone for identifying early or hidden dysglycemia. Obesity, brain circuitry, and future therapies (Priority: 3/5): They discuss how processed foods and obesogenic environments may alter brain reward circuitry, and consider future therapies including muscle-preserving agents and possibly next-generation anti-obesity drugs.

Key Arguments: Insulin resistance is organ- and pathway-specific; liver, muscle, fat, beta cell, kidney, gut, alpha cell, and brain can each show distinct defects. The euglycemic clamp is the gold standard because it keeps glucose constant while precisely measuring tissue response to insulin. Type 2 diabetes is heterogeneous, so treating it as one disease with one drug is mechanistically unsound. Hyperinsulinemia can itself induce insulin resistance by down-regulating insulin signaling. Free fatty acids are toxic to insulin signaling, beta-cell function, and mitochondrial function, making lipolysis a key therapeutic target. GLP-1/GIP defects are mainly at the beta cell, not because the gut fails to secrete the hormones. SGLT2 inhibition works by forcing glycosuria and can improve glucotoxicity, beta-cell function, and insulin sensitivity despite acting only in the kidney. Pioglitazone is presented as the only true insulin sensitizer, with benefits extending to heart, liver, muscle, triglycerides, and blood pressure. Metformin lowers hepatic glucose output but is not a true insulin sensitizer in muscle and does not enter skeletal muscle. Combination therapy started early outperforms stepwise escalation and should be matched to the patient’s phenotype. HbA1c alone can miss important risk; OGTT timing, especially one-hour glucose, can identify people likely to progress to diabetes. The brain likely contributes to overeating and obesity through altered hedonic circuitry and resistance to anorectic signals like leptin, amylin, and GLP-1.

Data Points: Years in metabolic disease research: 53+ years - DeFranzo says he has been funded continuously as an NIDDK investigator for 53 years. Major organ defects in the ominous octet: 8 - He explains the expanded model of type 2 diabetes pathophysiology. Original triumvirate: 3 defects - Beta-cell failure, muscle insulin resistance, and liver insulin resistance. Glucose production by liver at baseline: 2 mg/kg/min - Basal hepatic glucose output in a fasting person. Hepatic glucose output in a 100 kg person: 200 mg/min - Converted from 2 mg/kg/min to absolute output. Approximate glucose output per hour: 12 g/hour - Derived from basal hepatic glucose production. Insulin rise needed to suppress lipolysis: 10 microU/mL increase - A small insulin increase can completely inhibit adipose lipolysis. Insulin rise needed to suppress liver glucose output: ~50 microU/mL - Higher insulin is needed to shut down hepatic glucose production. Insulin level for maximal muscle glucose uptake: ~200 microU/mL - Even at 100 microU/mL, muscle uptake is not fully saturated. Muscle glucose uptake distribution: 80-90% - Most insulin-stimulated glucose disposal goes to muscle. Adipocyte glucose uptake distribution: ~10% - A smaller share of insulin-stimulated glucose disposal goes to fat cells. Type 2 diabetes prevalence at first diagnosis with CVD: 10-15% - Attia notes DeFranzo’s claim that many patients already have clinically significant cardiovascular disease. Physiologic daily insulin secretion: 35 units/day - DeFranzo says normal beta cells secrete about this much per day. Insulin resistance induced in healthy volunteers: 48-72 hours - Low-dose insulin infusion rapidly induced insulin resistance in lean subjects. Insulin in hyperinsulinemia experiment: 8 to 20 microU/mL - Healthy volunteers were raised from fasting insulin of 8 to 20 with infusion. Weight reduction in SGLT2 study: 14 days - Early dapagliflozin human studies showed metabolic improvement within two weeks. Insulin sensitivity improvement with dapagliflozin: 35% - Reported improvement after 14 days in human studies. SGLT2 glucose excretion: 40-60 g/day to 120 g/day - Range depends on kidney function and filtered glucose load. GLP-1/GIP contribution to post-meal insulin: ~70% - He says incretins account for most meal-stimulated insulin release. Older low-dose insulin clamp trial participants: 18-25 years old - Lean healthy volunteers in an insulin resistance induction study. Cardiovascular event reduction with GLP-1 trials: ~20% - He notes this is fairly uniform across trials, old and new agents. Pioglitazone dose-related weight gain: ~2 to 2.5 kg - Typical gain at 15-30 mg over a year. Pioglitazone improvement in myocardial insulin sensitivity: 75% - Measured with PET and fluorodeoxyglucose in a 2017 study. ADA treat-to-fail benchmark: A1C 6.5% - Target used in the comparison of treatment strategies. EDIC/early combination outcome: 70% below A1C 6.5% at 6 years - Triple therapy outperformed ADA stepwise treatment. ADA sequential-treatment outcome: 29% above A1C 6.5% at 6 years - A large fraction failed standard escalation. Qatar study starting A1C: ~10% - Patients were poorly controlled on prior therapy before randomized treatment. Qatar symptomatic subgroup starting A1C: 12.2% - One subgroup had severe symptomatic hyperglycemia. Qatar symptomatic subgroup outcome: 6.1% - After 3 years on exenatide plus pioglitazone. Mixed-split insulin outcome in Qatar study: 7.1% - After 3 years, good but inferior to exenatide plus pioglitazone. Predictive OGTT one-hour glucose threshold: >155 mg/dL - Strong predictor of future type 2 diabetes. Current obesity prevalence in the U.S.: 42% - Raised as a public health challenge in selecting who to treat.

Pivotal Quotes: "Insulin resistance is a very important concept, but you all have to be a little bit more specific about what aspect you want to address." — Dr. Ralph DeFranzo: He is defining why insulin resistance cannot be treated or measured as a single uniform entity. "If you have eight problems, I'm sure are going to be more to be found, and I can give you a few more if you want. But if you have eight problems, why in the world do you think one drug is going to correct eight problems?" — Dr. Ralph DeFranzo: He explains why type 2 diabetes requires combination therapy targeting multiple defects. "Metformin cannot get into skeletal muscle. It does not exist in skeletal muscle. It does not exist in cardiac muscle. So metformin cannot get into these tissues." — Dr. Ralph DeFranzo: He argues against the common belief that metformin is a muscle insulin sensitizer.

Implications: The episode argues for earlier, phenotype-driven combination therapy in diabetes, using mechanistic testing and multi-drug regimens rather than stepwise escalation. It also suggests future care may better target appetite, lipotoxicity, and tissue-specific insulin resistance.