Home Andrew Huberman Notes
Andrew Huberman · 2024-06-17 · 2h 18m

The Science of Hunger & Medications to Combat Obesity | Dr. Zachary Knight

A UCSF neuroscientist explains how the brain controls hunger, thirst and salt, and why GLP-1 drugs like Ozempic actually work.

The Science of Hunger & Medications to Combat Obesity | Dr. Zachary Knight
The guest

Dr. Zachary Knight — Professor of physiology at UC San Francisco and a Howard Hughes Medical Institute investigator whose lab studies homeostasis, mapping the brain circuits that drive hunger, thirst and body-temperature regulation.

The gist

Andrew Huberman and Dr. Zachary Knight take a deep dive into the modern neuroscience of appetite. They cover the brain's two-system control of eating (a short-term brain-stem 'meal size' system and a long-term hypothalamic 'body fat' system), the discovery of leptin, and surprising findings from Knight's lab that hunger (AgRP) neurons predict how much you'll eat before you take a bite. The conversation then turns to the science and history of GLP-1 drugs (Ozempic, Mounjaro and beyond), explaining how they were discovered, why they work, their unexpected health benefits, and the next generation of dual- and triple-agonist obesity drugs. They also explore dopamine's true role in eating (wanting vs. liking, learning rather than pleasure), the genetics of obesity, ultra-processed foods, and the tightly regulated systems controlling thirst and salt appetite.

Big reveals

  • Knight's lab found AgRP hunger neurons shut off within seconds of seeing food — before the first bite — predicting how much the animal will eat.
  • Amgen/Amgen-era leptin was hyped to 'cure obesity' but failed in obese people because they are leptin-resistant, not leptin-deficient.
  • Body weight is ~80% heritable — one of the most heritable traits known, second only to height.
  • DPP-4 inhibitors raise natural GLP-1 threefold yet cause no weight loss, proving GLP-1's natural job isn't body-weight control.
  • Ozempic's cardiac benefits appeared before patients lost weight, hinting at anti-inflammatory effects independent of weight loss.
  • Eli Lilly's triple-agonist (GLP-1 + GIP + glucagon) drove ~25% weight loss in a phase 2 trial and people were still losing — approaching bariatric-surgery territory.
  • Stimulating thirst neurons is deeply aversive (mice work hard to stop it), while hunger neurons mainly make food more attractive rather than feeling unpleasant.
  • Old dogma said only the long-term hypothalamic system could change body weight; GLP-1 drugs disproved it by hammering the short-term brain-stem system 24/7.

Things worth remembering

  • A 'decerebrate' rat with 80% of its brain removed can still regulate the size of a single meal using only its brain stem.
  • Leptin, cloned in 1994, is secreted by fat in direct proportion to body-fat reserves, giving the brain a readout of energy stores.
  • To AgRP neurons, 120 calories of olive oil and 120 calories of chicken are equally satiating — they track calories, not macronutrients.
  • For every ~2 lbs of weight lost, energy expenditure drops ~30 calories/day and hunger rises ~100 calories/day, explaining why weight is regained.
  • GLP-1 has a natural half-life of just ~2 minutes; engineering pushed it from 2 min to 2 hrs to 13 hrs to 7 days (semaglutide).
  • The first GLP-1 drug, exenatide, came from the venom of the Gila monster, a lizard with long gaps between meals.
  • Ozempic-class drugs raise GLP-1 a thousand- to ten-thousand-fold — levels no food or supplement can ever replicate.
  • In New Zealand studies, wild rabbits drink zero water nine months a year, getting all their hydration from food.
  • The brain detects a mere 1% rise in blood osmolality as thirst; a 20% rise can land you in the hospital.
  • Water leaves the stomach almost instantly, but high-fat, high-calorie food can take hours, regulated by gut feedback on gastric emptying.