Do we really have free will when it comes to eating? It’s a vexing question that is at the heart of why so many people find it so difficult to stick to a diet.

To get answers, one neuroscientist, Harvey J. Grill of the University of Pennsylvania, turned to rats and asked what would happen if he removed all of their brains except their brainstems. The brainstem controls basic functions like heart rate and breathing. But the animals could not smell, could not see, could not remember.

Would they know when they had consumed enough calories?

To find out, Dr. Grill dripped liquid food into their mouths.

“When they reached a stopping point, they allowed the food to drain out of their mouths,” he said.

Those studies, initiated decades ago, were a starting point for a body of research that has continually surprised scientists and driven home that how full animals feel has nothing to do with consciousness. The work has gained more relevance as scientists puzzle out how exactly the new drugs that cause weight loss, commonly called GLP-1s and including Ozempic, affect the brain’s eating-control systems.

The story that is emerging does not explain why some people get obese and others do not. Instead, it offers clues about what makes us start eating, and when we stop.

While most of the studies were in rodents, it defies belief to think that humans are somehow different, said Dr. Jeffrey Friedman, an obesity researcher at Rockefeller University in New York. Humans, he said, are subject to billions of years of evolution leading to elaborate neural pathways that control when to eat and when to stop eating.

As they have probed how eating is controlled, researchers learned that the brain is steadily getting signals that hint at how calorically dense a food is. There’s a certain amount of calories that the body needs, and these signals make sure the body gets them.

The process begins before a lab animal takes a single bite. Just the sight of food spurs neurons to anticipate whether a lot of calories will be packed into that food. The neurons respond more strongly to a food like peanut butter — loaded with calories — than to a low-calorie one like mouse chow.

The next control point occurs when the animal tastes the food: Neurons calculate the caloric density again from signals sent from the mouth to the brainstem.

Finally, when the food makes its way to the gut, a new set of signals to the brain lets the neurons again ascertain the caloric content.

And it is actually the calorie content that the gut assesses, as Zachary Knight, a neuroscientist at the University of California San Francisco, learned.

He saw this when he directly infused three types of food into the stomachs of mice. One infusion was of fatty food, another of carbohydrates and the third of protein. Each infusion had the same number of calories.

In each case, the message to the brain was the same: The neurons were signaling the amount of energy, in the form of calories, and not the source of the calories.

When the brain determines enough calories were consumed, neurons send a signal to stop eating.

Dr. Knight said these discoveries surprised him. He’d always thought that the signal to stop eating would be “a communication between the gut and the brain,” he said. There would be a sensation of having a full stomach and a deliberate decision to stop eating.

Using that reasoning, some dieters try to drink a big glass of water before a meal, or fill up on low-calorie foods, like celery.

But those tricks have not worked for most people because they don’t account for how the brain controls eating. In fact, Dr. Knight found that mice do not even send satiety signals to the brain when all they are getting is water.

It is true that people can decide to eat even when they are sated, or can decide not to eat when they are trying to lose weight. And, Dr. Grill said, in an intact brain — not just a brainstem — other areas of the brain also exert control.

But, Dr. Friedman said, in the end the brain’s controls typically override a person’s conscious decisions about whether they feel a need to eat. He said, by analogy, you can hold your breath — but only for so long. And you can suppress a cough — but only up to a point.

Scott Sternson, a neuroscientist with the University of California in San Diego and Howard Hughes Medical Institute, agreed.

“There is a very large proportion of appetite control that is automatic,” said Dr. Sternson, who is also a co-founder of a startup company, Penguin Bio, that is developing obesity treatments. People can decide to eat or not at a given moment. But, he added, maintaining that sort of control uses a lot of mental resources.

“Eventually, attention goes to other things and the automatic process will wind up dominating,” he said.

As they probed the brain’s eating-control systems, researchers were continually surprised.

They learned, for example, about the brain’s rapid response to just the sight of food.

Neuroscientists had found in mice a few thousand neurons in the hypothalamus, deep in the brain, that responded to hunger. But how are they regulated? They knew from previous studies that fasting turned these hunger neurons on and that the neurons were less active when an animal was well fed.

Their theory was that the neurons were responding to the body’s fat stores. When fat stores were low — as happens when an animal fasts, for example — levels of leptin, a hormone released from fat, also are low. That would turn the hunger neurons on. As an animal eats, its fat stores are replenished, leptin levels go up, and the neurons, it was assumed, would quiet down.

The whole system was thought to respond only slowly to the state of energy storage in the body.

But then three groups of researchers, independently led by Dr. Knight, Dr. Sternson and Mark Andermann of Beth Israel Deaconess Medical Center, examined the moment-to-moment activity of the hunger neurons.

They began with hungry mice. Their hunger neurons were firing rapidly, a sign the animals needed food.

The surprise happened when the investigators showed the animals food.

“Even before the first bite of food, the activity of those neurons shut off,” Dr. Knight said. “The neurons were making a prediction. The mouse looks at food. The mouse predicts how many calories it will eat.”

The more calorie-rich the food, the more neurons turn off.

“All three labs were shocked,” said Dr. Bradford B. Lowell, who worked with Dr. Andermann at Beth Israel Deaconess. “It was very unexpected.”

Dr. Lowell then asked what might happen if he deliberately turned off the hunger neurons even though the mice hadn’t had much to eat. Researchers can do this with genetic manipulations that mark neurons so they can turn them on and off with either a drug or with a blue light.

These mice would not eat for hours, even with food right in front of them.

Dr. Lowell and Dr. Sternson independently did the opposite experiment, turning the neurons on in mice that had just had a huge meal, the mouse equivalent of a Thanksgiving dinner. The animals were reclining, feeling stuffed.

But, said Dr. Andermann, who repeated the experiment, when they turned the hunger neurons on, “The mouse gets up and eats another 10 to 15 percent of its body weight.” He added, “The neurons are saying, ‘Just focus on food.’”

Researchers continue to be amazed by what they are finding — layers of controls in the brain that ensure eating is rigorously regulated. And hints of new ways to develop drugs to control eating.

One line of evidence was discovered by Amber Alhadeff, a neuroscientist at the Monell Chemical Senses Center and the University of Pennsylvania. She recently found two separate groups of neurons in the brainstem that respond to the GLP-1 obesity drugs.

One group of neurons signaled that the animals have had enough to eat. The other group caused the rodent equivalent of nausea. The current obesity drugs hit both groups of neurons, she reports, which may be a factor in the side effects many feel. She proposes that it might be possible to develop drugs that hit the satiety neurons but not the nausea ones.

Alexander Nectow, of Columbia University, has another surprise discovery. He identified a group of neurons in the brainstem that regulate how big a meal is desired, tracking each bite of food. “We don’t know how they do it,” he said.

“I’ve been studying this brainstem region for a decade and a half,” Dr. Nectow said, “but when we went and used all of our fancy tools, we found this population of neurons we had never studied.”

He’s now asking if the neurons could be targets for a class of weight loss drugs that could upstage the GLP-1s.

“That would be really amazing,” Dr. Nectow said.