Almost nobody enjoys their first sip of black coffee. The bitterness is a flavor the brain is wired to treat with suspicion, yet millions of people end up looking forward to that same taste every morning. What changes along the way?

A team at the University of California, San Francisco (UCSF) has described a piece of the answer: a slow dopamine pathway that links the digestive system to the brain and helps teach us which flavors are worth wanting. The work was published on November 4, 2026 in Nature.

A second dopamine system, slower and quieter

Dopamine is a chemical messenger tied to motivation, reward and learning. Most research has focused on its fast signals, the bursts that appear within seconds of something rewarding, like the sight of a favorite meal.

The new study looks at a different timescale. This slower system responds to information coming from inside the body, so the brain can learn about the benefits of a food after it has been eaten. According to senior author Zachary Knight, professor of physiology at UCSF, this process largely happens outside our conscious awareness.

What the researchers did

Knight and then-graduate student James Grove delivered either sugar water or a milkshake-like liquid directly into the stomachs of mice, while recording dopamine activity across the brain. Putting the nutrients straight into the stomach separated their effects from the experience of tasting them.

The result was a strong, prolonged dopamine response in a region called the anterior basolateral amygdala (BLA). The amygdala is involved in emotional processing and in forming associations between experiences and their consequences. Dopamine release in this spot was already known, but what it did was not. Here it built up gradually after nutrients reached the gut and lasted for several minutes.

Switching the signal off, and on

Seeing a signal is not the same as showing it matters, so the team tested its role in two directions.

  • Blocking it: mice kept eating normally, but they could no longer learn that a new, unfamiliar flavor was linked to calories.
  • Activating it: after mice tasted a flavor with no calories, the researchers artificially stimulated dopamine fibers in the BLA. The animals then developed a preference for that flavor, even though it provided no nutrition.

Together, the experiments suggest this pathway helps the brain learn which flavors predict a nutritional payoff. Knight links this to coffee: once the brain connects the taste with caffeine’s stimulating effects, he says, “that coffee starts to taste a lot better.”

Hints of the same system in humans

Collaborators looked for a similar pattern in people. Yale psychologist Dana Small reanalyzed data from an earlier PET imaging study of 13 men who drank either a milkshake or a drink without calories. Dopamine release after the nutrients reached the gut was concentrated in the left side of the BLA. A separate fMRI study led by Marc Tittgemeyer examined 38 adults after they consumed glucose and found increased activity in the same region.

One surprise: in both mice and humans, the response was considerably stronger in the left BLA than in the right. Brain circuits for essential functions like feeding are generally expected to be fairly symmetrical, and Knight’s lab is now investigating why this one is not.

What this does and does not show

  • Coffee is an illustration, not the experiment. The core experiments used sugar water and a milkshake-like liquid in mice. The caffeine explanation is Knight’s way of connecting the findings to a familiar example.
  • The human evidence is indirect. It comes from a reanalysis of an earlier PET dataset (13 men) and an fMRI study (38 adults). fMRI tracks activity linked to blood flow, not dopamine release itself.
  • GLP-1 drugs were not tested. Knight suggests that some effects of these medications could involve dampening the gut-triggered dopamine signal, making foods gradually less rewarding. This is a hypothesis, not a demonstrated effect.

Why it matters

Many approaches to changing eating behavior aim to reduce appetite or increase the feeling of fullness. A mapped pathway for how the brain learns food preferences opens a different possibility: eventually influencing what we come to like, not just how much we want to eat. Knight puts it as fine-tuning the steps between nutrient detection in the gut and learning in the brain, which might subtly shift our preferences for different foods. For now, that remains a long way from any treatment.

It also offers a tidy way to think about everyday tastes. Our love of a bitter morning cup, or of calorie-rich foods, may depend partly on a quiet conversation between gut and brain that continues long after the first bite.

Sources

  • Press release: UCSF, “How your brain learned to love coffee, chocolate, fat, and sugar” – EurekAlert
  • Paper: Grove J.C.R. et al., “The neural basis of flavour-nutrient learning,” Nature, 2026 – nature.com (DOI: 10.1038/s41586-026-11229-5)