The Minnesota Effect: What Starvation Science Teaches Us About Dieting and Human Behavior

In 1944, 36 healthy young men volunteered to go hungry.

They were conscientious objectors to military service during World War II, and they wanted to contribute to the relief effort in another way. Across Europe, millions of people were starving. Scientists knew that food could keep famine victims alive, but they knew much less about how to rehabilitate a person after prolonged deprivation.

Physiologist Ancel Keys and his colleagues at the University of Minnesota designed an experiment to find out. The volunteers would first eat normally under controlled conditions. Then, for six months, researchers would cut their food intake until they had lost roughly a quarter of their body weight. A rehabilitation period would follow.

The experiment became one of the most important studies of human starvation ever conducted. Its findings were eventually published in the two-volume Biology of Human Starvation, and later reviews still describe the Minnesota study as a key reference on the physical and psychological effects of severe dietary deprivation (Magowska, 2020; Dalle Grave, Pasqualoni, & Marchesini, 2011).

The study was not a modern weight-loss trial. Its participants were lean men, its restriction was severe, and its purpose was to study famine. Applying every observation directly to an ordinary diet would be a mistake.

Still, it exposed something that diet culture often prefers to ignore: when food becomes scarce, the body and mind do not sit quietly and accept the new arrangement. They adapt. Hunger intensifies. Energy expenditure falls. Attention narrows around food. Mood and social behavior change. Once food becomes freely available again, eating may surge past ordinary levels.

These reactions are sometimes called the Minnesota effect. The phrase does not refer to a formal diagnosis or a single biological mechanism. It is shorthand for the cluster of metabolic, psychological, and behavioral responses associated with prolonged energy restriction—and for the rebound that can follow it.

The deepest lesson is not that weight loss is impossible. It is that human behavior cannot be separated from human biology.

What happened during semistarvation

The men did not stop eating. They entered a state of semistarvation: enough food to continue functioning, but too little to maintain their previous body weight and ordinary physical state. Their prescribed meals were paired with substantial weekly physical activity.

As deprivation continued, food became psychologically enormous. The men talked about it, read about it, collected recipes, stretched meals out, and developed rituals around eating. Other interests lost some of their pull. Their moods and relationships changed, too. Depression, irritability, fatigue, social withdrawal, and reduced sexual interest became part of the experience.

That change in personality matters.

A simplistic account of dieting treats hunger as a sensation that disciplined people can ignore. The Minnesota observations suggest something broader. Sustained deprivation can reorganize attention, motivation, and emotion. The person who becomes obsessed with food after months of restriction may not have suddenly become shallow, weak, or greedy. Their brain has started treating food as an urgent problem.

Later research on dietary restraint has developed this point. Jane Polivy's review of food restriction concluded that starvation and self-imposed dieting can produce similar psychological consequences and may lead to binge eating when food becomes available (Polivy, 1996). Polivy and C. Peter Herman argued that dieting can play a causal role in binge eating, rather than always being a harmless response to overeating (Polivy & Herman, 1985).

This does not mean every calorie deficit causes a binge or that all forms of restraint are identical. It means that restriction is not psychologically neutral. The degree, duration, and rigidity of a diet matter.

The body begins to economize.

Weight loss lowers energy expenditure for an obvious reason: a smaller body generally costs less energy to maintain and move. But deprivation can also provoke an additional energy-saving response often called adaptive thermogenesis.

In plain English, the body becomes economical.

This response makes sense from an evolutionary perspective. When food supplies collapse, an organism that burns energy at its old rate is at a disadvantage. Reducing energy expenditure helps stretch limited reserves. Analyses of starvation physiology describe depression of basal metabolism as a survival adaptation, not a defect (Øristland, 1990).

For a person trying to lose weight, that adaptation creates a frustrating asymmetry. The behavior that produced early weight loss may produce less loss later. The body is smaller, expends less energy, and appetite pushes in the opposite direction. Weight loss slows and may plateau.

That plateau is often interpreted as proof that the dieter has stopped trying. Sometimes adherence does change, of course. Yet physiology predicts a slowdown even when a person continues making a real effort. Energy-balance research shows that a fixed reduction in intake typically causes weight to decline gradually toward a new equilibrium rather than producing the same amount of loss forever (Hall et al., 2012; Hall, 2012).

So the familiar straight-line promise—cut a fixed number of calories every day and lose a fixed amount every week indefinitely—is physiologically naïve. Bodies are dynamic systems, not bank accounts with legs.

Appetite may be the stronger opponent.

Metabolic adaptation receives a lot of attention, partly because it sounds mechanical and measurable. Yet increased appetite may be the more powerful force.

In a 2016 analysis, David Polidori, Kevin Hall, and colleagues quantified how strongly appetite responded to weight loss. Their model indicated that feedback through increased food intake opposed continued weight loss and helped create the familiar plateau (Polidori et al., 2016). The key point is easy to miss: someone may not merely want more food in a casual sense. Their regulatory biology may be applying persistent pressure to eat.

Hormonal research supports that interpretation. In one study, 50 adults with overweight or obesity followed a 10-week very-low-energy diet and lost an average of 13.5 kilograms. Weight loss altered several appetite-regulating hormones. Leptin, peptide YY, cholecystokinin, insulin, and amylin fell, while ghrelin and some other signals rose. Participants also reported greater hunger.

A year after the initial weight loss, many of those hormonal changes—and the increase in hunger—were still detectable (Sumithran et al., 2011).

This helps explain one of dieting's most demoralizing experiences. During the active phase, there is a project: rules to follow, progress to track, and a visible reward on the scale. Maintenance asks the person to keep doing much of the work after the rewarding part has slowed down, while appetite remains elevated.

That is not a short test of willpower. It is open-ended resistance against a recurring biological signal.

Why rebound eating makes sense.

When unrestricted food became available after semistarvation, some Minnesota participants ate enormous amounts. Fullness did not always switch off their desire to keep eating. Their bodies were repairing a serious energy deficit, and the drive to restore depleted tissue did not disappear the instant a meal entered the stomach.

Later interpretations of the experiment have focused on hyperphagia, meaning unusually high food intake after deprivation, and preferential catch-up fat, in which fat stores may recover faster than lean tissue. A 2023 systematic review and meta-analysis discussing dieting and subsequent weight gain notes that adaptive thermogenesis and hyperphagia can drive a rapid rebound of fat mass after restriction (Schiavo et al., 2023).

The result can look irrational from the outside. A person finishes a punishing diet, begins eating more, and regains weight quickly. They may temporarily rise above their previous weight or regain proportionally more fat than lean tissue. The common moral is that they let themselves go."

The biological account is less emotionally satisfying, but more scientifically useful. Restriction generated compensatory pressures. Once the restraint weakened or food access changed, those pressures shaped behavior.

Again, ordinary dieting is not identical to the Minnesota protocol. The volunteers underwent severe, prolonged deprivation under unusual experimental conditions. We should not pretend that skipping dessert recreates a famine study. The value of the experiment lies in demonstrating the direction of the response: the larger and longer the energy shortage, the harder the organism is likely to defend against it.

What the Minnesota effect does—and does not—prove

The Minnesota experiment is sometimes used to support claims it cannot establish.

It does not prove that every diet permanently damages metabolism. It does not show that all lost weight must return. It does not tell us that intentional weight loss is always harmful, or that exercise and nutrition have no effect on health. Its participants were a small and highly selected group of young, lean men living under controlled conditions in the 1940s. The intervention was far more severe than most medically supervised weight-management programs.

It also cannot tell us how women, older adults, children, or people with obesity would respond under the same conditions. There was no diverse, population-representative sample. By current standards, both the sample and the ethical design have serious limits. The experiment does establish that severe energy restriction itself can produce a striking collection of symptoms. Food obsession, emotional distress, lethargy, withdrawal, and intense post-restriction eating need not reveal a preexisting character flaw. They can emerge from deprivation.

Modern obesity research reaches a parallel, more carefully qualified conclusion: losing weight is possible through many approaches, but keeping it off is usually harder because biological, behavioral, and environmental pressures favor regain. Hall Kahan's view describes a common pattern of early loss, plateau, and progressive regain, and argues that obesity often requires continuing care rather than a one-time intervention (Hall & Kahan, 2018).

The difference between "weight loss is impossible" and "weight loss requires continuing support" is huge. The first is fatalistic. The second is realistic.

Lesson one: extreme restriction can create the behavior it is meant to control

Strict diets are often built on a theory of human nature: if people face enough rules, they will become more disciplined. Sometimes the opposite happens.

A rigid diet increases food's mental importance. It creates forbidden categories. Hunger makes those foods more salient, and the effort required to resist them consumes attention. When a dieter breaks a rule, they may interpret one deviation as total failure. One cookie becomes evidence that the day is ruined, so the rest of the package follows.

Biology and cognition reinforce each other. Deprivation increases appetite; rigid thinking turns an ordinary lapse into a crisis. Shame then encourages another round of severe restriction, beginning the cycle again.

Research on dietary restraint does not imply that planning meals or setting limits is inherently dangerous. A key difference exists between flexible structure and brittle control. A person who usually eats in a certain way but can adapt to travel, celebrations, and hunger cues is not doing the same thing as someone who treats any deviation as moral collapse.

A sustainable eating pattern has to survive contact with real life.

Lesson two: regain is not proof that the original loss was fake

People often imagine weight regain as a conscious return to old habits. That can happen, but it is not the whole story.

After weight loss, the same meal may feel less satisfying. Hunger may arrive earlier. The person may burn fewer calories at rest and during movement because their body is smaller and more efficient. Tiny, mostly unconscious shifts in portion size or activity can accumulate. Nobody has to make a dramatic decision.

The Minnesota effect teaches us to look for feedback loops rather than villains. Weight loss changes physiology; changed physiology influences eating and movement; those behaviors influence weight. The person inside the loop experiences it as an exhausting series of daily choices, but the pattern is bigger than any one choice.

This is why short-term success says surprisingly little about long-term maintenance. Don't judge an intervention only by what happens during its most intensive phase. The better question is what it requires after six months, two years, or ten years—and whether people can reasonably keep doing it.

Lesson three: hunger changes the mind, not just the stomach

We like to imagine that our values and personalities sit above our physical state. Hunger exposes how fragile that distinction can be.

Under prolonged deprivation, attention shifts toward food. Patience may shorten. Social life can become less rewarding. Work that once felt meaningful may become harder to sustain. A person can know that their thoughts are disproportionate and still be unable to stop them.

This principle extends beyond dieting. Sleep deprivation narrows attention. Chronic stress changes risk perception. Loneliness affects what people notice in social interactions. Money scarcity can consume mental bandwidth much like food scarcity.

The broader behavioral lesson is simple: before explaining conduct as character, examine the conditions surrounding it.

A person who seems impulsive may be depleted. Someone who cannot concentrate may be hungry. Someone who repeatedly "chooses badly" may be responding predictably to a system that keeps creating scarcity and then demanding perfect self-control.

Context does not remove responsibility. It tells us where intervention is likely to work.

Lesson four: willpower is a costly strategy

Self-control works best when it does not have to work very often.

A diet that requires someone to resist strong hunger every waking hour is poorly designed, even if resistance remains theoretically possible. It is like constructing a bridge that stays upright only when workers constantly hold it in place.

This insight changes the practical goal. Instead of asking, "How can I become tougher?" a dieter can ask, "How can I reduce the amount of toughness this plan requires?"

That might mean pursuing a smaller calorie deficit, eating enough protein and fiber to support satiety, building meals around foods that are filling for their energy content, preserving enjoyable foods in reasonable amounts, and adjusting the plan when hunger becomes intrusive. It can also mean improving sleep, changing the home food environment, planning for predictable high-risk situations, and seeking medical help when appropriate.

None of those strategies defeats biology. They reduce the frequency and intensity of the fight.

Lesson five: treatment should match a chronic problem

Many weight-loss programs are sold as temporary challenges. Follow the plan for eight or twelve weeks, reach the target, and graduate.

The body does not recognize graduation.

If appetite remains elevated and energy requirements remain reduced, ending all support at goal weight is a strange treatment model. It withdraws the intervention just as the maintenance problem becomes clear.

Hall and Kahan argue that long-term weight management needs ongoing clinical attention and maintenance-specific counseling (Hall & Kahan, 2018). That does not mean every person needs lifelong intensive treatment. It means the plan for maintaining change deserves as much thought as the plan for producing it.

For some people, continuing support may involve dietary counseling, regular self-monitoring, and structured physical activity. Others may benefit from anti-obesity medication, bariatric surgery, or treatment for binge eating and other eating disorders. The proper choice depends on medical history, weight-related risks, preferences, access, and the presence of disordered eating. A qualified clinician should make the choice, not a historical starvation experiment.

What a more humane approach to weight loss looks like

A humane approach begins by dropping the idea that suffering validates a diet. Constant hunger is not proof that a plan is working well. It may show the plan is too aggressive to last.

It also separates health from spectacle. Rapid scale changes are dramatic, but preserving muscle, improving metabolic markers, sleeping better, moving more easily, and building a stable eating pattern all matter. A slower intervention that survives five years may accomplish more than a severe one that produces a striking before-and-after photograph followed by rebound.

Expectations matter too. Weight-loss curves are not usually straight. Plateaus are predictable. Day-to-day scale changes include water, glycogen, and digestive contents, not just changes in fat. Planning for these realities can prevent normal fluctuation from being treated as failure.

Most of all, a humane approach refuses to turn biology into blame. People remain capable of choice, but they make choices under pressure. Hunger, hormones, energy expenditure, food availability, stress, medication, income, sleep, and culture all help determine how difficult a given choice feels"

"Just eat less" describes part of the energy equation. It does not describe the experience of doing it.

The larger lesson about human behavior

The Minnesota experiment is often remembered as a study of food, but it is also a study of explanation.

Before deprivation, researchers selected the volunteers partly for their physical and psychological health. During semistarvation, they developed behaviors and symptoms that could easily have been mistaken for fixed personality traits if the cause had been hidden. Once we know the context, those same behaviors look different.

That pattern appears everywhere.

We admire disciplined people without noticing the environments that make discipline easier. We condemn impulsive people without measuring the scarcity they face. We treat productivity, patience, and restraint as private virtues, even though sleep, stress, nutrition, safety, and social support affect each.

The Minnesota effect warns against explanations that begin and end with character. Behavior is real, and choices matter. But behavior is also adaptive. Put a healthy person under enough deprivation and their attention, emotions, and priorities will change.

That is not weakness. It is one way a human organism stays alive.

The useful question is therefore not whether willpower matters. Of course it does. The useful question is how much strain a person's biology and environment are placing on it—and whether a better-designed system could make the desired behavior easier.

For dieting, that means replacing punishment with sustainability, short campaigns with maintenance plans, and moral judgment with physiological realism. For human behavior more broadly, it means becoming slower to blame and quicker to inspect the conditions producing what we see.

The men in Minnesota volunteered to teach scientists how starvation works. Eight decades later, their experience still teaches a quieter lesson: if a plan requires people to behave as though they do not have bodies, the problem may be the plan.