How Food Choices Shape Memory, Cognition & Cravings | Dr. Scott Kanoski

The episode opens with an account of early research on how diet affects neurocognition and memory. The first model studied was a high-sugar one: animals had ad libitum access to a sugar solution supplying roughly the same percentage of calories as a soft drink does for people, while otherwise eating healthy food and drinking water.

In this model, the animals showed hippocampal-dependent memory deficits when they were older, and the deficits appeared to persist later in life even after the sugar was no longer being consumed. The microbiome is raised as a possible factor: the sugar model produced changes in gut bacteria that "may have a functional connection to the memory impairments." The wording is tentative and presents this as a possibility, not an established mechanism.

A second line of work used what is described as a total junk food diet: chocolate peanut butter cups, ruffled potato chips, and a rodent high-fat, high-sugar diet. These animals also showed memory impairments that lasted even after they were switched to a healthy diet. All of these findings come from animal models.

Andrew Huberman, who introduces himself as a professor of neurobiology and ophthalmology at Stanford School of Medicine, introduces the guest, Dr. Scott Kanoski, a professor of human and evolutionary biology and neurobiology at the University of Southern California. According to the introduction, Kanoski's lab studies how hunger and memory interact, how the mind and body affect appetite, body weight regulation, food craving and choice, and how different nutrients affect brain development and later brain function.

The introduction lists GLP drugs among the planned topics: how they work and how they affect hunger, longevity, cognition and drives in general. It frames the conversation around interactions between brain and body involving hormones, neurotransmitters, memory and experience. Huberman also states that the podcast is separate from his teaching and research roles at Stanford and that the episode includes sponsors.

The first question sets up the context. Hunger, feeding and satiety are everyday experiences, and the obesity epidemic "seems to be partially mitigated by the GLP drugs." The discussion starts with what glucagon-like peptide 1 is in its natural form, because these drugs are so widely used or debated. As the question puts it, people are taking them, coming off them, thinking about them, objecting to them, or know someone on them.

GLP-1 is described as an incretin hormone secreted from the distal intestines around food ingestion. A little is secreted before a meal starts and more during the meal, and it enters circulation. A separate source exists in the brainstem: a population of neurons in the nucleus tractus solitarius makes GLP-1 and sends GLP-1 signals extensively throughout the brain.

The peripheral and central GLP-1 systems have very similar effects, including hunger suppression, but they appear to work independently. There "doesn't seem to be a strong connection" between the peripheral physiological system and the GLP-1-producing neurons in the brain, although both suppress appetite.

GLP-1 also helps regulate blood glucose by stimulating insulin secretion when blood glucose is high. That is why GLP-1-based drugs were first used to treat diabetes, not obesity. Because they also suppress appetite and food intake, which leads to weight loss, they are now probably better known for their weight-loss effects.

The discussion confirms that drugs such as Ozempic and Wegovy raise GLP-1 activity far beyond natural levels. The question describes the increase as thousandfold, and the answer agrees: "That's correct. Super physiological for sure."

Asked about downsides and upsides, the answer is that most of what is known is positive. People have taken GLP-1-based drugs for over a decade, and clinical trials support their efficacy. Side effects exist, and the most common are nausea and emesis (vomiting). These lead many people to stop: they tolerate the effects for a while and then discontinue treatment.

So if there's any primary limitation of these drugs, it's trying to mitigate those side effects and to have more long-term efficacy beyond the initial weight loss.

The conversation turns to reports of people taking far less than their prescribed dose. The term "microdosing" is borrowed from the psychedelics community. People do this to reduce side effects while keeping some appetite suppression, and possibly to lower the cost, although many drugs are moving to generic forms. Asked whether any animal or human data exist on this practice, the answer is: "I'm not aware of any data on that."

The answer raises placebo effects as a possible part of what people experience. It sets a clear standard for evidence: "I'm old-fashioned, and I like to see the large clinical trials," meaning randomized controlled clinical trials, before believing something works. No such studies of microdosing exist yet.

The discussion also suggests such trials may never happen. Drug companies have little incentive, because people taking and buying less is not a good business strategy, and the studies are so expensive that nobody is likely to be running them now or in the near future. The conclusion: "So we're kind of left with anecdotal reports in that zone."

Anecdotes about benefits unrelated to weight loss are raised next. The question describes people of healthy body weight who take micro or full doses for claimed benefits to cognition and energy, and a friend with chronic fatigue syndrome who microdoses under a physician's care. She likes the effects but not the appetite suppression, because she is already very lean and close to underweight. These are presented as claims and personal anecdotes, not data.

The biological basis for such effects is that GLP-1 receptors are not limited to brain regions that control appetite, food intake and glucose regulation. They are densely expressed throughout the brain, including the reward circuitry of the mesostriatal pathway: the midbrain VTA, the striatum and the nucleus accumbens. The hippocampus, described as a favorite brain region, also has many GLP-1 receptors.

Based on this receptor distribution, the discussion sees possible uses for these drugs or other GLP-1 receptor agonists in substance use disorders and potentially in memory disorders and dementia. The evidence is uneven. Plenty of preclinical data support the idea, but clinical data are still being collected and are "just ongoing and starting to come out now."

Whether GLP-1 receptor agonists help with anything beyond appetite depends partly on whether they get into the brain, which is called "somewhat of a controversial topic" in the field. For the first generation of GLP-1 drugs, from 2005 to 2012, preclinical rodent evidence suggested that much of their effect on food intake worked through the brain. For newer drugs like Ozempic and Wegovy, "it's less clear."

The methods have limits. Some studies, which are exclusively preclinical, tag the peptides with fluorescent molecules and track where they accumulate. Much of the signal clusters in the circumventricular organs, where no blood-brain barrier shields the brain from circulation. This supports one school of thought that the drugs act mainly there. But fluorescent molecules fade and brain tissue produces its own autofluorescent signal, so these studies could miss something.

A better but more expensive and complicated method radiolabels the peptides, for example with carbon-14, to determine whether and where they enter the brain. Elizabeth Rhea, Bill Banks and colleagues at the University of Washington are named as starting this work. Early results suggest that some early-generation GLP-1 drugs readily enter the brain, while for some newer ones the picture is less clear.

Maybe it takes a little bit longer, but the jury's still out.

Where in the brain the drugs end up is also described as something without a clear understanding at this point.

After the sponsor segments, the discussion moves toward reward, reinforcement and memory. The point is that these topics apply to food as well as to drug-seeking: what we find rewarding and what we crave are closely related, and remembering which foods will soothe or satisfy us seems to be a strong factor.

The question then sets aside brain entry and asks what people actually feel on a full dose or microdose. Is it satiety, does that mean fullness, and what is the "food noise" that users say goes quiet, biologically? The questioner says they have never taken one of these drugs.

The answer describes a likely paracrine mechanism for the body's natural GLP-1. The vagus nerve innervates the GI tract and has GLP-1 receptors on its terminals, where it contacts the gut. GLP-1 released from L cells in the intestines acts immediately on these nearby vagal receptors. This is offered as one route by which both natural GLP-1 and the drugs "may be producing this feeling of satiation."

The vagus nerve carries many within-meal signals that help us stop eating as we become full, a process called satiation. GLP-1 is described as a big player in satiation, working "not exclusively through the vagus nerve, but probably largely through the vagus nerve." The hedge leaves room for other pathways.

The section opens partway through an explanation of how GLP-1 acts locally, "through this paracrine mode of action," to send a signal to the brainstem. The brainstem is described as the first site in the brain that the vagus nerve informs that the animal has eaten enough. By this point in a meal the stomach is starting to expand and nutrients are entering the intestines.

GLP-1 is not presented as the single switch for fullness. Many signals contribute. Some travel through the vagus nerve and some do not, and together they eventually produce the feeling of fullness and the end of a meal.

And GLP-1 is one instrument in that symphony.

The conversation then asks whether GLP-1 also switches off the desire for food. People taking these drugs often report less "food noise." Some of them exercise hard and some do not, and that difference is set aside for later. The common report of reduced food noise implies there was noise before, though the questioner says they have not asked those people whether they were aware of it at the time. The questioner tries to picture what food noise is for someone who has never taken the drugs. Their own wanting depends on the food: blueberries always appeal, while bread or steak depends on when they last ate, the time of day and similar factors.

The response also comes from someone who does not experience food noise and finds the idea somewhat strange. They have heard the same anecdotal reports that GLP-1-based drugs "really, really canceled the food noise." The mechanism is not known. The effect could come from receptor binding in higher-order brain regions involved in memory and cognition, such as the hippocampus and prefrontal cortex. It could also be essentially a gut-vagus satiation mechanism.

Again, I think that's one of the big gaps in this field is understanding where the drugs are acting.

The point is left as an open research question: "It's a lot to uncover in that area still." The food-noise reports stay in the category of anecdote, and no mechanism is claimed for them.

Asked for evidence that the vagus nerve is crucial for GLP-1's appetite-suppressing effect, the answer describes lesion experiments. If the sensory vagus nerve running from the GI tract to the brain is selectively lesioned, injected GLP-1 no longer reduces food intake.

This finding is limited to native GLP-1, which has a short half-life. Long-acting GLP-1 drugs are described as "a different story" because they stay in circulation longer and can get into the brain. So the vagal-dependence result should not be read as a direct account of how the drug versions work.