What If the Food Is Designed to Be Hard to Stop Eating?

When people struggle to control their eating, we tend to look for an explanation inside the person. Maybe they lack discipline. Maybe they need more motivation. Maybe they haven’t learned to listen to their fullness cues. We rarely give the same scrutiny to the food itself.

That’s a strange omission, because foods differ enormously in the demands they place on human appetite regulation. Some foods require substantial chewing and take time to consume. Some contain relatively little energy for their volume, while others concentrate hundreds of calories into a few bites. Some contain combinations of fat, carbohydrate, sugar and sodium commonly found in minimally processed foods. Others combine and concentrate those properties in ways that are relatively unusual in nature.

None of this means there is something inherently wrong with processed food, or that particular foods make us powerless. It does mean that eating behavior can’t be understood entirely by studying the person doing the eating. We also have to study the properties of what they’re eating.

When a normal response meets an exaggerated stimulus

There is an old concept from animal behavior research that provides a useful way of thinking about this problem: the supernormal stimulus.

Ethologists, including Nobel Prize-winning researcher Nikolaas Tinbergen, discovered that animals sometimes respond more strongly to exaggerated versions of biologically important stimuli than they do to the naturally occurring versions. Certain birds, for example, could be induced to direct their behavior toward artificial objects that exaggerated important visual characteristics of their natural environment.

The concept doesn’t translate perfectly to human eating, and nutrition researchers generally don’t describe foods as “supernormal stimuli.” Instead, they study overlapping concepts such as food reward, hyper-palatable foods, hedonic eating, energy density, eating rate and ultra-processed foods. Taken together, though, this research raises a fascinating question: What happens when modern technology becomes extremely good at concentrating and combining the signals that naturally make food attractive to us?

Our attraction to those signals isn’t a defect. Humans didn’t evolve to be indifferent to food. Sweetness can signal readily available carbohydrate. Fat provides concentrated energy and essential fatty acids. Sodium is physiologically necessary. Aroma helps us anticipate what we’re about to eat, while taste and texture provide information as we eat it. The nervous system also learns from what happens after nutrients enter the body, gradually associating the sensory properties of foods with their nutritional consequences.

These systems exist because finding and consuming food is necessary for survival. Modern food production, however, can take the properties those systems respond to and alter them in ways that weren’t previously possible. We can concentrate nutrients, combine them in unusual proportions, manipulate texture, remove physical structures that slow consumption, increase energy density, intensify flavors and reproduce a highly specific sensory experience millions of times.

The result is a food environment very different from the one in which human appetite regulation developed.

Scientists can actually measure hyper-palatability

“Junk food” isn’t a particularly useful scientific category. It tells us something about how a culture judges a food but very little about what properties of that food might affect eating behavior.

Tera Fazzino and colleagues took a different approach. In research published in Obesity, they developed quantitative criteria for identifying what they termed hyper-palatable foods. Rather than deciding subjectively which foods seemed tempting, they looked for specific combinations of nutrients that appeared repeatedly in highly palatable foods.

Three clusters emerged: foods combining substantial fat and sodium, foods combining substantial fat and simple sugars, and foods combining substantial carbohydrate and sodium. When the researchers applied their definitions to the U.S. food supply, they found that a substantial proportion of available foods met at least one of these criteria.

That doesn’t establish that every food meeting the definition causes overeating, and “hyper-palatable” shouldn’t become another moral category for sorting foods into good and bad. What the research gives us is something much more useful: a way to move beyond the assumption that people simply like “junk food” too much and begin identifying specific properties of foods that may influence eating behavior.

One of those properties has produced particularly interesting experimental results.

What happens when fat and carbohydrate arrive together?

In 2018, Alexandra DiFeliceantonio, Dana Small and colleagues compared people’s responses to foods containing primarily fat, primarily carbohydrate, or substantial amounts of both. The foods were matched for energy density, portion size and other characteristics, allowing the researchers to investigate whether the nutrient combination itself affected reward.

Participants were willing to pay more for foods containing both fat and carbohydrate than would be predicted by their responses to foods containing either nutrient predominantly. Brain imaging also showed greater recruitment of regions involved in reward valuation. The researchers described the effect as supra-additive: the combination appeared to have a reward value greater than would be expected from simply adding the responses to fat and carbohydrate separately.

There is an intriguing evolutionary context for this finding. Foods naturally containing high concentrations of both fat and carbohydrate are relatively uncommon, while the combination is routine in the modern food environment. Cookies, donuts, chocolate, pizza, French fries and pastries can all deliver substantial fat and refined carbohydrate together.

The researchers proposed that signals generated by fat and carbohydrate may travel through partly distinct physiological pathways involved in communicating nutritional value to the brain. Combining the two could therefore create a particularly powerful signal of food value. The exact mechanisms remain an active area of research, so it would be premature to claim that fat and carbohydrate somehow “hijack” the brain. What the experiment does demonstrate is important enough without that exaggeration: the nutrient architecture of food can influence its reward value beyond its calorie content alone.

And nutrient composition is only one part of the story.

Food controls how quickly we can eat it

Consider an apple and applesauce. The nutritional differences can be relatively modest, but the physical experience of eating them is not. One requires biting, chewing and repeated oral processing. The other can be swallowed much more rapidly.

We often talk about eating quickly as though it were simply a personal habit: Slow down. Chew your food. Pay attention.

But eating rate isn’t determined solely by the eater. The physical structure of the food helps determine how quickly a human being can consume it.

That matters because appetite regulation unfolds over time. As food is eaten, mechanical and chemical signals develop throughout the gastrointestinal tract. Nutrients reach the intestine. Hormonal and neural signals communicate information about what is arriving. Those processes occur while the meal is still happening.

Researchers have therefore begun manipulating food texture and eating rate experimentally. Controlled feeding studies show that changing how quickly foods can be consumed can produce meaningful differences in spontaneous energy intake, even when other characteristics of the foods are carefully controlled.

Now add another variable: energy density. A food that can be consumed quickly and delivers a large amount of energy per bite creates a very different physiological situation from a bulky, low-energy-density food that requires substantial oral processing.

Then add portion size, palatability, sensory variety and convenience. At that point, looking exclusively at the eater’s “self-control” starts to seem like a remarkably incomplete way of explaining what happened.

The 500-calorie experiment

One of the most important demonstrations of this problem came from Kevin Hall and colleagues at the National Institutes of Health. In a tightly controlled inpatient randomized trial, participants lived at the NIH Clinical Center and were provided either an ultra-processed or an unprocessed diet for two weeks before switching to the other condition. They were allowed to eat as much or as little of the provided food as they wanted.

When participants were eating the ultra-processed diet, they spontaneously consumed about 500 more calories per day than they did on the unprocessed diet. They gained weight during the ultra-processed condition and lost weight during the unprocessed condition.

The study did not establish that “ultra-processing” itself is a single mechanism that causes overeating. In fact, subsequent research has been trying to determine which characteristics of those foods produced the difference. Eating rate, energy density, texture, palatability and specific nutrient combinations are all part of that continuing investigation.

But Hall’s experiment demonstrates something that should fundamentally change how we talk about eating behavior: when researchers changed the food environment while observing the same people under controlled conditions, the amount those people chose to eat changed dramatically.

The person wasn’t the only variable.

Maybe we have been asking the wrong question

For decades, much of our cultural response to overeating has begun with some version of the same question: Why can’t people control themselves around food?

That question already contains an assumption about where the problem resides.

A better scientific question is: What happens when a normal human appetitive system encounters an environment filled with foods that concentrate, combine, intensify and accelerate the signals that system evolved to respond to?

That doesn’t mean humans have no agency. It doesn’t mean cookies are addictive drugs, processed foods are inherently harmful, or people are incapable of choosing what they eat. It means behavior emerges from an interaction between a person and an environment. If we want to understand the behavior, both sides of that interaction deserve investigation.

This distinction matters clinically, too. If someone repeatedly struggles around a particular food, telling them to develop more self-control gives them exactly one variable to work with: themselves. Understanding the food environment gives them many more. They can notice which foods leave them satisfied and which seem to intensify appetite, how eating speed affects their experience, whether particular combinations change satisfaction, how portion and packaging influence them, and what happens when their environment changes.

The goal isn’t to become afraid of food. It is to understand the forces acting on us well enough that we can make decisions without interpreting every predictable human response as a personal failure.

Over the rest of this series, we’ll take those forces apart one at a time. We’ll look at why fat and carbohydrate together appear to have unusual reward value, how food texture controls eating speed, why energy density matters, how portion size changes consumption even when hunger hasn’t changed, and what happens when all of those properties are combined in the modern food environment.

Because before we ask people to get better at controlling their appetite, we should probably understand what we’re asking their appetite to contend with.

References

DiFeliceantonio AG, Coppin G, Rigoux L, et al. Supra-additive effects of combining fat and carbohydrate on food reward. Cell Metabolism. 2018;28(1):33–44.e3.

Fazzino TL, Rohde K, Sullivan DK. Hyper-palatable foods: development of a quantitative definition and application to the US food system database. Obesity. 2019;27(11):1761–1768.

Hall KD, Ayuketah A, Brychta R, et al. Ultra-processed diets cause excess calorie intake and weight gain: an inpatient randomized controlled trial of ad libitum food intake. Cell Metabolism. 2019;30(1):67–77.e3.

Tinbergen N. The Study of Instinct. Oxford University Press; 1951.

Related reading

Educational information only. This article is not a diagnosis or individualized medical or nutrition treatment. New, severe, persistent, or concerning symptoms should be evaluated by an appropriate healthcare professional.