Peter Attia MD337- Insulin resistance masterclass: The full body impact of metabolic dysfunction, treatment & more
CHAPTERS
- 0:00 – 5:59
Cahill, starvation studies, and DeFronzo's path into metabolism
After framing metabolic disease as a foundational fourth horseman, Peter Attia asks Ralph DeFronzo about his background. DeFronzo credits his mentor Professor Cahill, and they discuss the forty-day starvation study and how much energy the body stores.
- •Attia frames metabolic disease, from hyperinsulinemia to type 2 diabetes, as amplifying risk for the other major causes of death
- •DeFronzo describes decades of NIDDK funding and his formative time with Professor Cahill
- •The Cahill fasting study showed the brain still relied on some glucose even in prolonged starvation
- •DeFronzo says fat stores let obese people fast for months, but muscle breakdown becomes the problem
- 5:59 – 10:05
What insulin resistance means and why it is tissue-specific
DeFronzo defines insulin resistance by comparing responses to infused insulin in lean, obese and diabetic people. He and Attia stress that the term is nonspecific because insulin acts on muscle, liver, fat and protein metabolism.
- •Insulin releases after a meal and regulates glucose uptake, fat release and protein metabolism
- •Obese people are resistant in muscle glucose uptake and diabetics are even more resistant
- •Insulin keeps fat in fat cells, but in obesity this action fails
- •Insulin resistance should be specified by tissue and by metabolic process
- 10:05 – 25:13
How the euglycemic clamp works and what it showed
DeFronzo explains why he developed the insulin clamp and walks through the procedure with Attia. They cover what happens to glucose under clamp conditions and the liver's basal glucose output.
- •Older approaches used oral glucose tolerance tests and swapped numerator and denominator, which DeFronzo found unsound
- •The clamp holds insulin steady while glucose is infused to keep blood sugar constant
- •Insulin shuts down liver glucose output, drives muscle uptake (80-90%), and suppresses lipolysis
- •Fat responds to small insulin rises, the liver needs about 50 µU/mL, and muscle keeps responding up to 200
- •The liver produces about 2 mg/kg/min of glucose at baseline
- 25:13 – 29:15
Insulin resistance in the brain, beta cells, and the heart
DeFronzo broadens insulin resistance to the brain and says the two big concepts in type 2 diabetes are insulin resistance and impaired beta cell function. He links the defective insulin signaling pathway, which also drives nitric oxide production, to cardiovascular disease.
- •Insulin receptors exist in the brain, and brain insulin resistance may relate to Alzheimer's, though DeFronzo is unsure about calling it type 3 diabetes
- •Insulin-resistant people with healthy beta cells compensate and don't become diabetic
- •The same signaling pathway activates nitric oxide synthase, so defects promote cardiovascular disease
- •Most diabetics already have cardiovascular disease at diagnosis if examined carefully
- 29:15 – 37:52
Hyperinsulinemia as a cause of insulin resistance
DeFronzo and Attia discuss the harms of high insulin doses compared with a physiologic total of about 35 units per day. DeFronzo describes a study in which a mild insulin infusion made lean young volunteers as insulin resistant as type 2 diabetics within 48-72 hours.
- •Physiologic insulin secretion is about 35 units a day, while many patients take 100 or more
- •DeFronzo cautions that people who need insulin should still get it
- •Raising fasting insulin from 8 to 20 induced severe insulin resistance within 48-72 hours
- •Hyperinsulinemia downregulates insulin signaling, and the liver removing half of secreted insulin protects the periphery
- 37:52 – 47:49
The genetics of insulin resistance and type 2 diabetes
DeFronzo says little is known about the genetic basis of insulin resistance, since GWAS findings are mostly associations. He argues diabetes is too heterogeneous a phenotype and describes a muscle-insulin-resistance phenotyping approach with epigenetic findings.
- •Lipodystrophy and PPAR gamma are among the few clearer causal links
- •Diabetes is heterogeneous, unlike single-mutation diseases such as sickle cell anemia
- •DeFronzo's group uses clamp-defined muscle insulin resistance with biopsies, finding chromatin opening
- •He argues for combination therapy from the start, and says the ADA first acknowledged it in 2022
- 47:49 – 58:44
The ominous octet: fat cell, gut, and alpha cell
DeFronzo explains how his triumvirate of beta cell, muscle and liver became the ominous octet. He covers the fat cell turning from friend to foe, incretin resistance in the gut, and excess glucagon from alpha cells.
- •Expanded fat cells resist insulin's anti-lipolytic effect, raising free fatty acids and causing lipotoxicity
- •Incretins GLP-1 and GIP account for about 70% of meal-driven insulin, but beta cells become resistant to them
- •GLP-1 agonists overcome this resistance with a pharmacologic dose
- •Alpha cells overproduce glucagon, driving gluconeogenesis and fasting hyperglycemia
- 58:44 – 1:08:39
The kidney and SGLT2 inhibitors
DeFronzo describes how the diabetic kidney upregulates SGLT2 and holds on to glucose. He tells the story of developing SGLT2 inhibitors, which lower glucose and improve insulin sensitivity and beta cell function.
- •SGLT2 reabsorbs about 90% of filtered glucose and is upregulated in diabetes
- •DeFronzo's work led to dapagliflozin and the first human proof of glucotoxicity
- •Fourteen days of dapagliflozin improved insulin sensitivity by 35% and beta cell function
- •Concerns about kidney damage were eased by familial renal glucosuria, and the drugs protect kidney and heart
- 1:08:39 – 1:15:33
The brain, overeating, and lipotoxicity
DeFronzo covers the eighth player, the brain, which becomes resistant to satiety hormones such as GLP-1, amylin and leptin. He then describes lipotoxicity experiments, including acipimox, and the question of mitochondrial dysfunction.
- •Brain areas controlling appetite show gray matter shrinkage and abnormal glucose uptake during insulin clamps in obese people
- •Infusing free fatty acids induces severe insulin resistance in hours
- •Lowering FFA with acipimox improved muscle insulin sensitivity and mitochondrial ATP generation about 50%
- •Mitochondrial dysfunction is partly secondary to lipotoxicity, though still debated
- 1:15:33 – 1:26:06
Pioglitazone: the true insulin sensitizer and its misconceptions
DeFronzo defends pioglitazone, calling it the only true insulin sensitizer. He covers the insulin signaling pathway, Shulman's DGAT work, weight gain and fluid retention, and cardiac data from a 2017 study.
- •DeFronzo says more weight gain on pioglitazone tracks with better A1c, insulin sensitivity and lipids because fat is redistributed
- •The insulin signaling cascade runs from receptor to IRS-1, PI3 kinase and Akt, and is also tied to nitric oxide
- •Pioglitazone caused fluid retention through vasodilation, but improved myocardial insulin sensitivity by 75% and cardiac function
- •He calls it the best NASH drug
- 1:26:06 – 1:31:39
Metformin: not an insulin sensitizer
DeFronzo argues metformin works by lowering hepatic glucose production rather than sensitizing muscle. He says it cannot enter muscle because the transporter is absent there, and explains the lactate effect.
- •Clamp studies never showed metformin improves insulin sensitivity
- •Metformin enters cells via the organic cation transporter, absent in skeletal and cardiac muscle
- •Labeled metformin accumulates in liver, kidney and bladder, not muscle
- •Weight loss from GI effects probably led to the insulin-sensitizer belief
- •He still considers it a good, cheap fourth drug
- 1:31:39 – 1:45:29
EDICT triple therapy versus the ADA stepwise approach
DeFronzo describes the EDICT study, comparing early triple therapy with the ADA stepwise approach. Triple therapy yielded far better control and improved insulin sensitivity and beta cell function, and he cites GRADE, ADOPT and UKPDS as showing failure of stepwise therapy.
- •EDICT used metformin, exenatide and pioglitazone versus metformin, sulfonylurea and insulin in 315 people
- •At six years, 29% of the ADA-approach group had failed, versus about 70% of the triple-therapy group reaching A1c below 6.5
- •The ADA approach gave no improvement in insulin sensitivity
- •GRADE, ADOPT and UKPDS showed add-ons to metformin failing over time
- •Most of the old triple regimen is cheap, but patients resist pioglitazone's weight gain
- 1:45:29 – 1:56:11
Choosing drugs: GLP-1s, the Qatar study, and rethinking treatment
DeFronzo ranks drugs: a newer GLP-1 first, pioglitazone second, SGLT2 inhibitor third. He discusses glucagon in retatrutide, caps on benefit, and the Qatar study, where exenatide plus pioglitazone beat split-mixed insulin in poorly controlled patients.
- •DeFronzo would choose a GLP-1 agonist if allowed only one drug, adding pioglitazone and then an SGLT2 inhibitor
- •GLP-1 drugs also protect beta cells, a benefit he thinks gets overlooked
- •Cardiovascular reduction is about 20% across GLP-1 trials despite greater weight loss with newer drugs
- •In the Qatar study, A1c averaged 6.1 on exenatide plus pioglitazone versus 7.1 on insulin at three years
- 1:56:11 – 2:03:33
Hyperglycemic clamps, C-peptide, and measuring beta cell function
DeFronzo describes a three-step hyperglycemic clamp used to look for genes tied to beta cell failure with glucose, GLP-1 and amino acid stimuli. He explains why C-peptide is a better measure of insulin secretion than insulin.
- •Sequential stimuli help isolate defects in glucose, GLP-1 and amino acid responses
- •C-peptide is secreted one-to-one with insulin but isn't taken up by the liver
- •Insulin degradation is impaired in diabetes, so insulin levels mislead
- •Beta cell function should be expressed per rise in glucose, relative to insulin resistance
- 2:03:33 – 2:15:53
GLP-1 safety, muscle loss, and myostatin inhibitors
DeFronzo calls current GLP-1 drugs quite safe and takes up the controversy over muscle loss, citing bypass data in which function improved. He and Attia discuss myostatin inhibitors and the need for better functional endpoints than lean mass or the six-minute walk.
- •Lean body mass is an imperfect measure and his group is measuring muscle by MRI
- •In a bypass study, relative strength, walking distance and chair rises improved despite lean mass loss
- •Myostatin inhibitors are in phase two, and DeFronzo suggests insulin sensitivity could be a functional benefit
- •Attia criticizes the six-minute walk test and prefers strength tests
- 2:15:53 – 2:34:41
Childhood obesity, the obesity epidemic, and brain circuitry
DeFronzo describes type 2 diabetes in adolescents, who respond poorly to drugs, and the dilemma of treating obesity with lifelong medication. He attributes the epidemic to environmental stimuli changing brain neurocircuitry, plus genetic predisposition shown by lean Hispanic children of diabetic parents.
- •Childhood obesity predicts adult obesity, and teens with diabetes are very insulin resistant and hard to control
- •Half of patients stop GLP-1 drugs within a year due to cost, side effects or wanting to eat
- •Processed foods, calorie density and inactivity change hedonic brain circuits
- •Lean offspring of diabetic parents were insulin resistant with signaling defects starting at IRS-1
- 2:34:41 – 2:42:17
Multiple defects within muscle and the OGTT as a practical test
DeFronzo discusses disagreements with Shulman over GLUT4, hexokinase and IRS-1, and says defects in different tissues can each reproduce diabetes. Attia then raises the OGTT, and DeFronzo interprets sample cases and highlights the one-hour glucose above 155.
- •DeFronzo believes the signaling defect, impaired transport and phosphorylation all contribute, so one drug is unlikely to fix everything
- •A defect in fat, liver, muscle, beta cell or brain can reproduce diabetes
- •A high 30-minute insulin with later hypoglycemia signals a prediabetic insulin-resistant state despite normal A1c
- •A one-hour glucose above 155 is a strong predictor of future diabetes, and low 30-minute insulin is another
- •DeFronzo says Attia's target numbers may be overly aggressive but meeting them is probably safe
