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

Summary

Andrew Huberman interviews neuroscientist Dr. Scott Kanoski about how food, hunger, memory, and the gut–brain connection interact. A recurring theme is that eating is not controlled by a single “hunger center”: signals from the gut, hormones, memory, reward, and context all contribute.

GLP-1 and appetite

• The body naturally releases GLP-1 from the intestines around meals. GLP-1 is also produced by neurons in the brainstem. Both sources can suppress appetite, though they appear to act largely independently.
• GLP-1 also helps regulate blood glucose and insulin. GLP-1-based medicines were first used for diabetes and became widely known for their effects on appetite and weight.
• Nausea and vomiting are common side effects and can lead people to stop treatment. Kanoski says there is not yet robust clinical-trial evidence for “microdosing” these drugs.
• GLP-1 receptors occur in brain regions involved in reward and memory, including the hippocampus. This raises the possibility of effects beyond appetite, but clinical evidence for such benefits is still emerging. Researchers also continue to debate how much newer drugs enter the brain and where they act.
• A likely route for natural GLP-1’s meal-terminating effects is through receptors on the vagus nerve. The vagus carries information about stomach expansion and nutrients to the brainstem as part of a broader set of signals that contribute to satiation.

Sweeteners: evidence remains mixed

• Sweet taste can trigger an anticipatory insulin response before calories reach the gut. Animal studies suggest that repeatedly separating sweet taste from its usual caloric consequences may affect appetite and glucose regulation.
• In one lab study, juvenile and adolescent rodents consumed daily amounts of stevia, acesulfame-K, or saccharin within the stated acceptable daily intake range. As adults, the animals showed memory and glucose-regulation changes, as well as altered sweet-taste receptor expression.
• These findings are preclinical and do not establish that artificial or low-calorie sweeteners cause the same effects in people. Human evidence on weight outcomes is mixed, and Kanoski does not consider it strong enough to support a clear recommendation either for or against them.

Memory influences eating—and eating can affect memory

• Remembering a recent meal can influence later intake. In a human study discussed by Kanoski, participants who were prompted to recall a pizza lunch later ate fewer cookies than those who were not prompted to remember it.
• Another study manipulated how much soup participants thought they had eaten using a concealed pump. Immediately afterward, reported hunger reflected actual consumption; later, hunger ratings were more influenced by the amount participants believed they had eaten.
• Kanoski argues that systems involved in appetite and satiation may also support memory. Remembering what food was found, where it was obtained, and what followed from eating it would have been useful for survival.

Diet and cognitive function

• In Kanoski’s rodent research, diets high in sugar, fat, and simple carbohydrates can impair hippocampus-dependent memory. The hippocampus appears especially vulnerable compared with some other kinds of memory.
• In developmental models, excess sugar or a junk-food diet was associated with memory deficits that persisted into adulthood, even after the animals returned to healthy food. Gut-microbiome changes may contribute, though the precise mechanisms are not settled.
• Kanoski also describes unpublished, preprint-stage work in which researchers heat-treated otherwise healthy food to increase compounds called advanced glycation end-products (AGEs). Rodents exposed during development later showed memory impairments. Reintroducing a particular bacterial population reversed the deficit in that model. These results require peer review and replication.
• Human studies are less controlled than rodent studies because dietary reporting is imprecise. Still, some human research has linked habitual sugar and saturated-fat intake with performance on hippocampus-dependent memory tasks.
• The discussion emphasizes that these results do not prove a direct, one-to-one connection between a particular food and dementia. Observational studies link midlife dietary and metabolic patterns with later dementia risk, while animal models suggest unhealthy diets can worsen Alzheimer’s-like pathology. Kanoski stresses the importance of long-term patterns rather than assuming that dietary changes after symptoms begin will reverse disease.

The gut–brain connection

• The vagus nerve is one route by which the gut communicates with the brain, but it does not carry one simple “I have eaten” message. Different sensory neurons respond to stomach stretching, nutrients, and other inputs, creating a varied pattern of signals.
• In Kanoski’s model, vagal information reaches the hippocampus through a pathway involving the brainstem and medial septum. The medial septum’s acetylcholine-producing neurons then signal to the hippocampus.
• In rodent recordings, hippocampal acetylcholine rises and falls during eating, with its baseline ending higher after a meal. Kanoski proposes that this may help encode the meal and its context. Sweet taste alone was not enough to produce the same pattern in the lab’s experiments; the combination of taste and post-ingestive nutrients may matter.
• The hippocampus may also help interpret internal bodily signals. People with severe amnesia can still experience fullness, yet may not use that feeling to guide eating appropriately. Kanoski describes this as difficulty interpreting internal signals, rather than necessarily failing to sense them.

Microbiome findings and limits

• Kanoski’s lab has used several approaches to test whether gut bacteria relate to diet-associated memory changes: measuring bacterial populations, correlating them with memory performance, manipulating populations, and testing whether adding or restoring bacteria changes outcomes.
• In one sugar-related rodent model, a bacterial population increased and was associated with worse memory performance; enriching it in a separate experiment also produced impairments.
• In the heat-treated-food preprint, a different bacterial population was reduced, and restoring it improved memory performance.
• These examples do not identify universal “good” or “bad” bacteria. The microbiome is complex, context-dependent, and a relatively young research field; the discussion cautions against treating diversity or any single bacterial species as a complete answer.
• The speakers note that microbiome recovery is not necessarily immediate. They also discuss fecal microbiota treatment as an established intervention for certain cases of C. difficile infection, while emphasizing that this does not mean microbiome treatments are broadly solved.

Exercise, eating patterns, and practical implications

• Kanoski says exercise can reverse some diet-associated memory deficits in rodent studies, likely through both metabolic and brain-related changes.
• There is no single meal schedule that works for everyone. People differ in whether they eat several meals, graze, or eat less frequently; Huberman and Kanoski suggest that individuals may need to notice how meal timing and size affect their own focus and energy.
• Kanoski’s broad, cautious takeaway is that excess sugar and saturated fat are not favorable for cognition, while avoiding overly rigid prescriptions is important. He characterizes a sustainable eating pattern as more useful than searching for a single universally “best” diet.
• For people taking appetite-suppressing drugs, the speakers raise the importance of nutrition and physical activity alongside weight loss. Eating less does not automatically ensure adequate nutrient intake, and clinical-trial protocols have included dietary and activity guidance.
• The discussion also notes that animal findings cannot be assumed to apply directly to people, and that many questions about long-term effects—especially in people without obesity—remain unanswered.

Reward, cravings, and food choice

• Dopamine-related reward circuitry responds not only to palatable foods but also to things the body needs, such as water during thirst or sodium when depleted. The animal’s physiological state influences how rewarding a food is.
• Food preferences can change. People who reduce sugar may find very sweet foods less appealing, while foods they eat regularly can become more desirable. The speakers distinguish this plasticity from claiming that food is equivalent to addictive drugs.
• Kanoski describes conditioned taste-aversion experiments in animals as an extreme example of how learned associations can change a food’s reward response. He avoids taking a definitive position on “food addiction,” noting that some diagnostic criteria may appear to fit, but food differs from substances because eating is necessary.

Oxytocin and social eating

• Oxytocin can reduce food intake, particularly by reducing meal size through brainstem pathways.
• Kanoski’s lab also found that oxytocin signaling in the hippocampus is needed for a social-eating effect in rodents: animals ate more in the company of familiar others than when alone or with unfamiliar animals.
• Thus, oxytocin’s effects on eating may depend on the brain region and social context.

Speakers and sources featured

Speakers

• Andrew Huberman — podcast host and interviewer.
• Dr. Scott Kanoski — guest neuroscientist.

Research and researchers discussed

• Studies and work associated with Suzanne Higgs, Jeff Brunstrom, Wolfgang Langhans, Bill Banks, Elizabeth Rhea, Dick Stevenson, George Bray, Paul Rosen, Harvey Grill, and Michael Kilgard.
• Kanoski’s laboratory research on diet, memory, the microbiome, vagal signaling, acetylcholine, and oxytocin.
• Michael Pollan and work associated with Sonnenberg are mentioned in passing.

Sponsor messages read by Huberman

• Our Place
• Helix Sleep
• AG1
• Element
• Function