Dr. Scott Kanoski on How Food Shapes Memory and Cravings: GLP-1, the Vagus Nerve, Sugar and the Microbiome

High-Level Overview

Andrew Huberman talks with Dr. Scott Kanoski, a professor of human and evolutionary biology and neurobiology at the University of Southern California, whose lab studies how hunger and memory interact. The conversation starts with GLP-1, the gut hormone behind drugs such as Ozempic and Wegovy, and then widens into a bigger argument: eating is not a simple reflex run by the brainstem and hypothalamus. The hippocampus, the brain's memory hub, is deeply involved in when and how much we eat, and what we eat can in turn damage or protect it.

The tone is technical but relaxed and candid. Kanoski is careful throughout to separate rodent findings from human evidence, declines to hand out dietary rules, and admits where his field simply does not know the answer yet (whether the newer GLP-1 drugs reach the brain, which gut bacteria are good or bad, whether artificial sweeteners help or hurt). Huberman supplies the personal observations and the broader questions about obesity, processed food and dementia. The conversation moves from hormones and nerves to laboratory diet studies, then out to the microbiome, dopamine, food politics and a closing note on oxytocin.

The Discussion Explores:

• What GLP-1 is, how GLP-1 drugs suppress appetite through the vagus nerve, and the open question of whether they act in the brain
• Why the evidence on artificial and low-calorie sweeteners is mixed, and what Kanoski's own rat study found
• How the memory of a recent meal changes how hungry we feel later
• Rodent studies showing that sugar, junk food and heat-treated food early in life impair hippocampal memory into adulthood
• The vagus nerve to hippocampus pathway, acetylcholine release during meals, and vagal nerve stimulation as a possible treatment
• Why the hippocampus is needed to interpret hunger and fullness signals
• What can and cannot yet be said about the gut microbiome
• Dopamine, changing food preferences, the food addiction debate, the sugar lobby, and oxytocin's role in social eating

Deep Dive into Main Ideas

GLP-1 and what the drugs do

Kanoski explains that natural GLP-1 is an incretin hormone secreted from the distal intestines around eating: a little before the meal starts and more during it. A separate population of neurons in the brainstem also makes GLP-1 and sends it throughout the brain. Both sources suppress hunger, but he says they seem to operate independently. GLP-1 also helps regulate blood glucose by stimulating insulin secretion, which is why the first GLP-1 drugs were diabetes treatments before they became known for weight loss.

He agrees with Huberman that the drugs raise GLP-1 far above natural levels, and says most of what is known after more than a decade of use is upside. The main limitation is side effects, chiefly nausea and vomiting, which lead many people to stop. On the trend of microdosing, he says he is not aware of any data and notes that placebos are very effective. The conversation doubts the trials will be run, since they are expensive and drug companies have little incentive.

Because GLP-1 receptors are found well beyond appetite centers, including the brain's reward circuitry and the hippocampus, Kanoski sees possible applications for substance abuse and even memory disorders and dementia. The preclinical data support this and clinical data are only starting to come out. Whether the newer drugs actually get into the brain is, he says, controversial; studies using radio-labeled peptides by Elizabeth Rhea, Bill Banks and colleagues at the University of Washington suggest early-generation drugs readily entered the brain, while the newer ones are less clear.

How GLP-1 turns off hunger: the vagus nerve and food noise

Natural GLP-1 likely works locally. It is released from L cells in the intestines and acts on receptors sitting on nearby vagus nerve endings, and the vagus carries that signal to the brainstem. Kanoski describes meal ending as a symphony of signals, including stomach stretch and nutrients entering the intestines, in which GLP-1 is one instrument. In animals, cutting the sensory vagus stops injected native GLP-1 from reducing food intake, and dialing down the GLP-1 receptor on the vagus leads to what he believes are larger meals.

Neither man experiences the food noise that people on these drugs say goes quiet. Kanoski says it is not known whether that effect comes from the drugs reaching higher brain regions involved in memory and cognition or is simply a gut and vagus satiation mechanism.

Artificial sweeteners: mixed evidence

Asked about low-calorie sweeteners, Kanoski calls the topic very controversial and says there are no definitive answers. Sweet taste triggers an early insulin release that prepares the body for incoming sugar, and some animal work suggests that repeatedly breaking the link between sweet taste and calories can disturb appetite and glucose regulation. Human trials show effects in both directions on weight.

His own lab gave juvenile and adolescent rats a small daily amount of sweetener, within the FDA's acceptable daily intake, which they consumed voluntarily. As adults, off the sweeteners, the animals had some memory impairments, a larger insulin spike to infused glucose and changes in sweet taste receptor expression on the tongue. The effects were not large, and it did not matter whether the sweetener was stevia, Ace K or saccharin. He stresses that this is one rodent study, says "I don't think we have the evidence for recommendations at this point," and adds that he drinks Coke Zero regularly. Replacing habitual sugary drinks with artificially sweetened ones may help weight loss, but he sees no reliable trend line.

Memory decides how much we eat

"Our memory of eating occasions can have a powerful effect on how much we eat subsequently," Kanoski says. In a study by Suzanne Higgs and colleagues, people who were prompted to recall a pizza lunch a few hours later ate fewer cookies in a taste test. In work by Jeff Brunstrom and colleagues, a bowl of soup was secretly refilled or drained as people ate. Right afterward, hunger tracked what they had actually eaten; a couple of hours later it tracked what they had seen and believed they ate.

Kanoski's evolutionary reading is that finding food was once a critical event, so the systems that respond to nutrients also promote memory of what was eaten and where. On meal timing he sees large individual differences and no single recipe, and the conversation cites George Bray's view that the best diet is one that you can stick to.

Sugar, junk food and heat-treated food impair memory in rodents

In adult animals, a Western diet (about 35 percent of calories from fat and 15 percent from sugar, much of it simple sugar) produces hippocampal-dependent memory impairments, sometimes within a week. Kanoski calls the hippocampus "somewhat of a canary in the coal mine" because it is more vulnerable than other memory systems. In his lab's developmental models, rats given free access to a sugar solution, or a junk food diet of chocolate peanut butter cups and potato chips, showed memory deficits as adults that lasted even after a switch to healthy food.

A not yet peer-reviewed preprint from his lab isolates one form of processing: baking otherwise healthy food at high heat to drive the Maillard reaction, which creates advanced glycation end products. Animals raised on it had memory impairments. A population of bacteria was eliminated by the heat treatment, and reintroducing it reversed the impairments. In these spatial and object memory tests, control animals perform above chance while the diet groups perform at chance.

Diet, dementia, exercise and the vagus nerve

Kanoski says prospective human studies, which are correlations, link midlife dietary patterns and metabolic status, including visceral adiposity, with a higher risk of dementia later, and that a Western diet exacerbates symptoms in transgenic animal models of Alzheimer's. He agrees with Huberman that there is no one-to-one line from food to Alzheimer's, and thinks a healthy dietary pattern throughout life matters more than switching once there is a problem. In rats, exercise has been shown to reverse the diet-induced memory deficits, partly through changes in the brain.

Cutting the gut-innervating part of the vagus nerve gives animals hippocampal-dependent memory impairments. One small study in Taiwan found a positive correlation between vagotomy for stomach ulcers and later dementia, and Kanoski's group plans to look at the Swedish National Registry. His lab is testing a vagal nerve stimulation device, sutured to the stomach so that it fires only during eating, in a rat model of Alzheimer's; it has already rescued memory impairments caused by an unhealthy diet. The hippocampus also has many receptors for insulin, ghrelin and leptin, giving it a second, slower read on energy status.

GLP-1 drugs, nutrients and muscle

Huberman worries that people eating very little on GLP-1 drugs could become micronutrient deficient and lose muscle. Kanoski is less concerned, because a majority stop within two years and, to his knowledge, micronutrient deficiency has not shown up in the clinical trials.

He does say weight loss brings loss of both fat and lean mass, that the effects in lean people who do not need the drugs are unknown, and that "Simply taking a pill is not the optimal way to lose weight": the trials included nutrition and physical activity guidelines.

Acetylcholine and reading the body's signals

The vagal signal is not one message. Stretch receptors in the stomach and chemoreceptors in the intestines fire at different times, which Huberman sums up as "a song, not a gong." The route Kanoski's lab has traced runs from the brainstem to the medial septum, whose acetylcholine neurons project to the hippocampus. Acetylcholine rises while an animal eats, dips when it pauses, and ends the meal above where it started, like a staircase. The pattern appears with both healthy and high fat, high sugar food, but not with stevia alone, so he guesses it needs both taste and nutrients.

People with dense amnesia from medial temporal lobe damage will eat a meal, then a second and even a third when the experimenter brings one, because they do not remember having eaten. They are getting the gut feedback, Kanoski says, but do not know how to interpret it. He says there are analogous studies in rodents: a normal animal can learn to use its energy state as a cue, and an animal with a lesion to the hippocampus cannot. He adds that awareness of hunger versus fullness can itself be impaired by a high fat, high sugar diet.

The microbiome: early days

In the early-life sugar model, a bacterial population, Parabacteroides, was elevated and negatively correlated with memory. Enriching it in animals that had never consumed sugar produced memory impairments. In the heat-treated diet model, a lactic acid producing population was eliminated and restoring it rescued memory.

Still, Kanoski says no one can yet list good and bad bacteria: "We're not even close to that." The field barely existed before about 2012, and a population that looks harmful in one context can look helpful in another. He points to an FDA-approved fecal transplant for C. diff infection as a sign of promise. Huberman describes his own approach as feeding his gut bacteria with low-sugar fermented foods and fiber, plus sleep and managing stress.

Dopamine, cravings, food politics and oxytocin

The dopamine pathway from the midbrain to the nucleus accumbens responds to palatable food, but Kanoski notes it also responds to bland food in a hungry animal, or water in a thirsty one. Preferences are plastic: people who cut back on sugar come to find formerly enjoyable things too sweet, and training an animal to associate sucrose with malaise eliminates the dopamine response. On food addiction, he says one can make the argument using DSM criteria, but that does not make it the same thing as drug addiction, since everyone has to eat.

The two also discuss how more than 60 percent of calories many Americans consume come from ultra-processed foods and that over 40 percent of U.S. adults are obese. Kanoski says he had a few brushes with the sugar lobby early in his career, when a press release about his sugar study drew a public statement.

Oxytocin has long been known to reduce food intake, mainly by acting on the brainstem to make meals smaller. Kanoski's lab found a different role in the hippocampus. People and rodents eat more in company, especially familiar company, and in rodents that effect requires oxytocin signaling in the hippocampus, so there it promotes eating in a social context.

Conclusion

Kanoski's case is that appetite and memory are one intertwined system. The gut tells the brain about each meal through the vagus nerve and circulating hormones, the hippocampus uses that information to remember meals and interpret hunger, and diets high in sugar and saturated fat, at least in rodents and especially during development, degrade that same machinery in lasting ways. He will not prescribe a diet and repeatedly flags the limits of animal models, but he does note that human studies point the same way: excess sugar and excess saturated fat are not good for memory. Around that core he offers measured views on GLP-1 drugs, sweeteners and the microbiome, where much remains unknown, and some hopeful signs, from exercise reversing deficits in rats to food preferences that can be retrained.

This is a summary of "How Food Choices Shape Memory, Cognition & Cravings | Dr. Scott Kanoski" from Huberman Lab, giving you the GIST so you can decide if you want the REST. Listen to the full episode.