Scientists pinpoint memory-related neurons that drive eating behavior

A new study published in Nature Metabolism has uncovered a surprising mechanism in the brain that helps explain why people may overeat, even when they’re not hungry. Researchers identified specific neurons in the hippocampus—a brain region known for memory—that store detailed information about the location and type of foods rich in fat and sugar. When activated, these neurons increase food consumption, while silencing them can prevent weight gain in mice. The findings suggest that memory, not just hunger or pleasure, plays a powerful role in driving what and how much we eat.

The research was led by Guillaume de Lartigue, whose team set out to investigate how food-related memories influence dietary behavior. While scientists have long known that humans and animals can remember where to find palatable food, the direct role of memory circuits in overeating had not been clearly demonstrated. In modern environments where calorie-dense foods are readily available and food cues are constant, the researchers suspected that memory systems in the brain might be overactive and contributing to obesity.

The hippocampus is best known for its role in forming spatial and episodic memories—those tied to specific times, places, or experiences. Previous studies have hinted at its involvement in regulating food intake. For instance, lesions in this region increase eating and body weight in animals, and people with hippocampal damage may eat repeated meals without remembering the last one. The research team hypothesized that some hippocampal neurons might encode not just general satiety or hunger signals, but specific memories about fat or sugar consumption.

“We were interested in how memory influences eating behavior. While it’s well known that hunger and food pleasure drive eating, people often eat simply because an external cue (like the smell of fresh bread or the sight of a slice of cake) triggers the urge, even when they’re not hungry,” said de Lartigue, an associate professor at Monell Chemical Senses Center and the Perelman School of Medicine at the University of Pennsylvania.

“Sometimes, just remembering that there’s a chocolate bar in your desk drawer can make you want to eat it. That led us to ask: can memory alone be enough to drive eating? And if so, how does the brain store and retrieve those food-related memories?”

To test this, the researchers used genetically modified mice that allowed them to “tag” active neurons in the dorsal hippocampus when the animals consumed fat or sugar. These neurons responded to post-ingestive nutrient signals delivered directly to the gut and were activated through a gut-brain communication pathway involving the vagus nerve. They found that fat and sugar each triggered a distinct group of neurons, and these two populations were largely non-overlapping. Fat-responsive neurons were concentrated in areas of the hippocampus associated with motivation, while sugar-responsive neurons were linked to spatial memory.

The researchers then manipulated these neurons in different ways. When they activated the sugar-responsive neurons, mice showed increased preference for and consumption of sugary solutions. Activation of fat-responsive neurons had a similar effect on fat intake. Conversely, silencing these neurons had the opposite outcome. Mice with deactivated sugar-memory neurons forgot where sugar was located, consumed less of it, and showed reduced weight gain—even when eating high-fat, high-sugar diets. Mice with silenced fat-memory neurons consumed less fat and showed less motivation to work for fatty rewards.

In further experiments, the team demonstrated that sugar-responsive neurons functioned like a “memory trace” for food location. When these neurons were stimulated, mice were better at remembering where they had previously found sugar. But when the neurons were removed, this spatial memory was impaired, and sugar consumption dropped. These effects were specific to food-related memories—general memory functions remained intact. Meanwhile, fat-responsive neurons appeared to enhance the motivational pull of fat. Silencing them reduced the animals’ willingness to expend effort for fat rewards, while stimulating them increased that motivation.

“What surprised us most was how specific these memories are. The brain doesn’t just store a general memory of food, it forms distinct memory traces for fat and sugar, using different sets of neurons. Even more surprising, these two groups of neurons affect behavior differently: sugar memories help the brain remember where to find sugar, while fat memories increase the motivation to go get it.”

In addition to influencing immediate food choices, these neurons also shaped long-term feeding behavior. Mice with deactivated sugar-memory neurons ate fewer meals during the inactive light phase of their day and resisted weight gain over several weeks, despite having unrestricted access to calorie-dense food. This pattern mimics some of the benefits of time-restricted eating, suggesting that impairing memory for food cues may help reduce snacking and prevent overeating.

“Our study shows that the brain not only remembers where high-calorie foods like sugar and fat were found, but that those memories can actually drive you to seek them out, even without hunger. We discovered specific neurons in the hippocampus (the brain’s memory center) that store these food memories, and when we activated or deleted them in mice, it changed how much and what they ate. In other words, memory itself can make you want to eat.”

The researchers argue that their findings open up new avenues for understanding and treating obesity. While much attention has focused on appetite hormones and reward pathways in the brain, this study highlights memory as a key driver of eating behavior. In environments where high-calorie food is easy to access and advertisements constantly remind us of past enjoyable meals, memory circuits may be overstimulated, encouraging people to eat even when they’re not hungry.

“Memory systems in the hippocampus evolved to help animals locate and remember food sources critical for survival,” said first author Mingxin Yang, a University of Pennsylvania doctoral student in the de Lartigue lab. “In modern environments, where food is abundant and cues are everywhere, these memory circuits may drive overeating, contributing to obesity.”

The study has some limitations. It was conducted in mice, and while the hippocampus performs similar functions in humans, more research is needed to confirm that these specific neuronal populations exist and operate the same way in people. Additionally, the researchers focused on isolated nutrients—sugar and fat—rather than real-world complex meals. This reductionist approach helps pinpoint mechanisms but may not capture the full complexity of human eating behavior.

Looking ahead, the researchers hope to investigate whether these memory-driven eating patterns can be modified or suppressed. If specific brain circuits that encode fat or sugar memories can be selectively turned down, it could lead to new strategies for reducing unhealthy eating habits.

“We’re now exploring whether these memory-driven eating behaviors can be modified, which could open new doors for helping people who struggle with overeating or obesity. We are looking for ways of selectively increasing or decreasing the gain on the activity of these neurons to precisely influence what we eat.”

“One of the most exciting parts of this work is that it challenges the traditional view that only hunger, or pleasure, drive eating. We show that memory can be its own trigger, and that the hippocampus plays a central role in that process. It helps explain why we sometimes eat even when we don’t need to—and why high-calorie foods are especially hard to resist.”

The study, “Separate orexigenic hippocampal ensembles shape dietary choice by enhancing contextual memory and motivation,” was authored by Mingxin Yang, Arashdeep Singh, Alan de Araujo, Molly McDougle, Hillary Ellis, Léa Décarie-Spain, Scott E. Kanoski, and Guillaume de Lartigue.