Appetite

Abstract

Appetite, which the Medical Subject Headings thesaurus classifies under psychophysiology, is the desire to eat: a psychological state that selects when, what, and how much food is consumed. It is distinguished from hunger, the physiological drive arising from energy depletion, because appetite can be aroused by the sight, smell, or memory of palatable food in the complete absence of energy need. Modern accounts treat appetite as the output of two interacting control systems: a homeostatic system that tracks the body's energy stores through hormonal signals to the brain, and a hedonic system that assigns reward value to food and can override energy signals. Layered over both is cognitive control, in which memory, attention, learned restraint, and environmental cues shape eating. This article develops each system, their integration, and the measurable regularities they produce.

Keywords: appetite, hunger, satiety, hedonic eating, energy homeostasis

Appetite is the desire for food, a motivational state that regulates food intake and is set apart from hunger, the physiological drive that follows energy depletion (Rogers & Brunstrom, 2016). The distinction matters because the two dissociate routinely: a person who has just finished a large meal, and is by any energy measure sated, will nonetheless report an appetite for a familiar dessert, while a genuinely energy-depleted person may have no appetite when ill or anxious. Contemporary research frames appetite as the joint product of homeostatic signalling from the body's energy stores and hedonic signalling from the brain's reward circuitry, with cognitive and environmental factors modulating both (Berthoud et al., 2017). Understanding appetite therefore means understanding how a biological regulator and a reward system are reconciled inside a cognitive agent.

What Appetite Is

Appetite is best defined by contrast with hunger. Hunger is a physiological drive state produced by energy deficit; appetite is a psychological desire that guides food selection and can be present or absent regardless of that deficit (Rogers & Brunstrom, 2016). The clearest evidence for the separation is that appetite is food-specific and cue-driven: the desire is not for calories in the abstract but for a particular food, and it can be switched on by the smell of baking or the sight of a menu with no change in energy status. This is why appetite, not hunger, is the variable that most directly governs everyday eating in an environment where food is abundant and energy deficit is rare.

Because appetite is a private motivational state, it is measured indirectly. Researchers infer it from visual-analogue ratings of desire to eat and fullness, from the amount eaten at a test meal, and from the latency to begin eating, each with characteristic strengths and biases (Blundell et al., 2010). A recurring methodological lesson is that subjective appetite ratings and actual food intake correlate only modestly, so a complete account must measure behaviour as well as report. The regulation of appetite unfolds across a cycle: satiation is the process that brings an episode of eating to an end and determines meal size, whereas satiety is the state of suppressed appetite between meals that determines when eating resumes (Blundell et al., 2010). Table 1 summarizes the principal peripheral signals that act on this cycle.

Note. Tonic adiposity signals report the size of energy stores to the brain over hours to days (Woods et al., 1998; Morton et al., 2006); episodic gut signals act meal by meal (Cummings et al., 2001).

Types of Appetite

In the Medical Subject Headings thesaurus, Appetite is a descriptor filed under psychophysiology, and it has one narrower descriptor: Appetite Regulation, the physiological and psychological processes that govern the onset, maintenance, and termination of the desire to eat. Appetite Regulation does not yet have its own article on this site, so it is named here rather than linked. The parent-and-child relation is a classification device for indexing the biomedical literature; it is not a claim that appetite decomposes exhaustively into these categories, and the psychological distinctions that organize the rest of this article — homeostatic versus hedonic versus cognitive control — cut across the MeSH tree rather than mirroring it.

It is more useful for a cognitive account to distinguish appetite by the system that generates it. Homeostatic appetite arises from energy need and is corrected by feedback about energy stores. Hedonic appetite, sometimes called hedonic hunger, arises from the anticipated pleasure of eating and can drive consumption of palatable food when no energy is required (Berthoud et al., 2017). Cognitively controlled appetite reflects deliberate influences such as dietary restraint, and learned influences such as the memory of what and when one last ate (Higgs, 2016). These are not mutually exclusive categories but overlapping contributions to a single behaviour, which is why the same meal can be initiated by hunger, sustained by pleasure, and ended by a cognitive rule.

Homeostatic Control: Signals From the Body

The oldest scientific account of appetite located it in the periphery. Cannon and Washburn argued that the sensation of hunger was produced by contractions of the empty stomach, which they recorded with a swallowed balloon and correlated with reported hunger pangs (Cannon & Washburn, 1912). The peripheral theory was incomplete — people whose stomachs are removed still report hunger and regulate intake — but it established the enduring idea that appetite is driven by signals reporting the state of the body. The modern form of that idea is the adiposity-signal model: hormones circulating in proportion to the body's fat stores act on the brain to adjust food intake and defend a stable body weight over the long term (Woods et al., 1998).

Between Cannon's peripheral account and those modern signals lay two mid-century homeostatic theories that framed the field. Mayer's glucostatic theory proposed that short-term fluctuations in glucose availability, sensed by central receptors, trigger and terminate individual meals (Mayer, 1953). Kennedy's lipostatic hypothesis proposed the complementary long-term control: a signal proportional to the body's fat stores acts on the hypothalamus to hold adiposity roughly constant (Kennedy, 1953). The lipostatic hypothesis was a prediction in search of a molecule, and the adiposity-signal model is its direct descendant — leptin is the circulating fat-store signal that Kennedy's theory required but could not name.

The decisive discovery was leptin. Friedman and colleagues used positional cloning to identify the mouse obese gene and its product, a hormone secreted by adipose tissue that signals the size of energy stores to the brain; animals lacking it eat voraciously and become massively obese (Zhang et al., 1994). Leptin and insulin together constitute the slow, tonic arm of homeostatic control, informing the hypothalamus of how much energy is banked (Morton et al., 2006). The fast, episodic arm is dominated by ghrelin, the only known circulating hormone that stimulates appetite: Cummings and colleagues showed that plasma ghrelin rises sharply just before each expected meal and falls after eating, exactly the profile expected of a meal-initiation signal (Cummings et al., 2001). Administering ghrelin to humans increases hunger ratings and food intake, confirming a causal role (Wren et al., 2001). The first demonstration shows how these tonic and episodic signals combine across a day.

Homeostatic signals across a day

Leptin is a tonic adiposity signal: it rises with body fat and suppresses appetite over the long term. Ghrelin is an episodic gut signal that climbs before each expected meal and falls after eating. Net appetite drive is the ghrelin signal minus the tonic leptin suppression.

Leptin suppression: 0.40 · ghrelin peaks per day: 3 · mean net appetite drive: 0.21. Higher adiposity raises tonic leptin and lowers the net drive; each meal resets the pre-meal ghrelin rise.

Hedonic Control: Liking and Wanting

Homeostasis cannot explain why people eat when they are not energy-depleted, or why some foods are almost irresistible. That is the province of the hedonic system, which assigns reward value to food. Its central contribution to the psychology of appetite is the dissociation of reward into two components that feel like one. Berridge and Robinson argued that the brain separates liking — the hedonic pleasure a food actually delivers — from wanting — the motivational pull that makes a food attractive and worth pursuing, which they termed incentive salience (Berridge & Robinson, 1998). The two are normally correlated but are generated by different neural systems and can be pulled apart: dopamine manipulations that raise wanting need not raise liking at all.

The dissociation reframes overeating and craving. A food cue can generate intense wanting — an urgent appetite — for a food whose consumption produces only ordinary pleasure, so that the desire outstrips the enjoyment (Berridge, 2009). This helps explain why craving is a poor guide to satisfaction, and why appetite for highly palatable, energy-dense food can persist against a person's own stated goals. The neural picture that has emerged is not of a separate hedonic centre but of reward and homeostatic circuits that overlap and interact, with hypothalamic energy signals modulating the mesolimbic reward pathway and vice versa (Rossi & Stuber, 2018). The second demonstration lets the reader dissociate liking from wanting by manipulating a reward signal.

Liking versus wanting

Food reward is not one thing. Liking is the pleasure a food actually delivers; wanting is the motivational pull that makes it worth pursuing. Palatability sets liking. Raising the dopamine signal scales wanting alone, so the desire for a food can outrun the enjoyment it gives.

Liking 60 · wanting 57 · wanting-to-liking ratio 0.95. Wanting is at or below liking; desire tracks enjoyment.

Cognitive and External Control

Appetite in humans is powerfully shaped by cognition and environment. Schachter's externality hypothesis proposed that eating in some individuals is governed less by internal state than by external cues such as the sight of food, the time on the clock, and the palatability of what is offered (Schachter, 1968). The idea was later refined rather than discarded: external cue sensitivity is now understood as one dimension of appetite that interacts with reward and restraint rather than a simple trait of obesity. A related programme showed that deliberate restraint can backfire. Herman and Mack found that dieters who restrain their intake will, once that restraint is broken by a forced high-calorie preload, eat more than non-dieters — the counter-regulatory pattern of disinhibited eating (Herman & Mack, 1975).

Memory and attention are equally consequential. Higgs showed that the memory of a recent meal suppresses later intake: people distracted while eating, or unable to recall what they ate, eat more at a subsequent occasion, so that appetite depends on cognitive representations of eating and not only on physiological state (Higgs, 2016). Two robust environmental effects illustrate how easily intake is moved without any change in reported fullness. Sensory-specific satiety is the decline in the pleasantness of a food as it is eaten, specific to that food, so that variety in a meal increases total consumption (Rolls et al., 1981). The portion-size effect is the finding that larger served portions increase intake, because people consume a roughly constant proportion of what is presented rather than compensating for the extra amount (Rolls et al., 2002). Both are amplified by food cue reactivity, the conditioned appetitive response to food cues that predicts eating and weight gain (Boswell & Kober, 2016). The third demonstration models the portion-size effect.

The portion-size effect

Served a larger portion, people eat more rather than compensating for it. A simple model captures this: intake is a roughly constant proportion p of the mass served, so energy intake is E = p · S · d, with S the mass served and d the energy density. Portion and density multiply.

Energy intake E = 0.90 × 500 × 1.5 = 675.0 kcal. At the baseline worked-example value of 675 kcal.

Worked Example

Consider the portion-size effect with a simple, empirically motivated model: people eat a roughly constant proportion p of the food served, so energy intake is E = p x S x d, where S is the mass served in grams and d is the food's energy density in kilocalories per gram (Rolls et al., 2002). Take a baseline serving of S = 500 g of a food at d = 1.5 kcal/g, with a consumed proportion p = 0.90. The predicted intake is E1 = 0.90 x 500 x 1.5 = 675 kcal.

Now increase the portion by half, to S = 750 g, leaving the food and the person unchanged. Because intake tracks the proportion served rather than energy need, E2 = 0.90 x 750 x 1.5 = 1012.5 kcal — an extra 337.5 kcal, a 50 percent rise, typically consumed with no reported change in fullness. This is the portion-size effect: the served amount, not an internal set-point, sets intake over a wide range. Energy density multiplies in the same way. Returning to the 500 g portion but lowering density to d = 1.0 kcal/g gives E3 = 0.90 x 500 x 1.0 = 450 kcal, a 225 kcal reduction from baseline. Because portion and density enter the model as a product, a large portion of a low-density food can deliver fewer calories than a small portion of a dense one, which is why manipulating energy density is an effective way to hold appetite constant while reducing intake (Blundell et al., 2010). The third demonstration reproduces each of these values.

Discussion

The study of appetite has moved from single-cause theories toward an integrated account. Peripheral, homeostatic, hedonic, and cognitive explanations were once rivals; each captured part of the phenomenon and failed to explain the rest. The gastric theory could not survive gastrectomy; pure homeostasis could not explain eating without need; pure hedonics could not explain the defence of body weight; pure externality could not explain individual differences in restraint. The contemporary synthesis treats these as interacting controllers rather than competing theories, in which slow adiposity signals set the gain on fast meal-related signals, and both are modulated by the reward value of food and by cognitive representations of eating (Morton et al., 2006; Berthoud et al., 2017).

That integration carries a practical lesson that mirrors the modern food environment. In conditions of scarcity, homeostatic control suffices, because appetite and energy need are aligned. In conditions of abundance, where palatable, energy-dense food is continuously available and cued, hedonic and external control can systematically override the homeostatic defence of body weight (Berthoud et al., 2017). This reframes overeating not as a failure of a single regulator but as the predictable output of a multi-system controller placed in an environment unlike the one it evolved to manage — and it locates the leverage for intervention in the hedonic and cognitive systems as much as in the homeostatic one (Higgs, 2016).

Current Directions

The most active current question is how homeostatic and hedonic signals are integrated at the level of neural circuits. High-resolution recording and circuit-manipulation methods have shown that the classical separation of a hypothalamic energy centre from a mesolimbic reward centre is too clean: the two are anatomically and functionally overlapping, with defined cell populations that carry both energy-state and reward information, so feeding is controlled by a distributed network rather than by two independent modules (Rossi & Stuber, 2018). This work is reframing obesity as a disorder of that integrated network rather than of any single signal, and it is shifting the search for intervention toward the points where energy and reward signals converge (Berthoud et al., 2017).

A second direction concerns the cognitive control of eating. Evidence that memory for recent meals and attention at the time of eating causally affect later appetite has opened a line of work on cognitive and behavioural interventions — attentive eating, memory cues, and expectation manipulations — that target intake without pharmacology (Higgs, 2016). A third re-examines whether human appetite defends energy balance at all precisely: analyses suggest that people compensate for energy expenditure weakly and asymmetrically, defending more vigorously against energy deficit than against surplus, which would help explain why weight is easier to gain than to lose in an environment of abundance (Rogers & Brunstrom, 2016). Across all three, food cue reactivity is emerging as a measurable predictor of who is most vulnerable to an obesogenic environment (Boswell & Kober, 2016).

Common Misconceptions

Appetite and hunger are the same thing.

They dissociate routinely. Hunger is a physiological drive from energy depletion; appetite is a psychological desire that can be aroused by palatable food with no energy need, and can be absent despite genuine deficit (Rogers & Brunstrom, 2016).

Wanting a food and enjoying it are one process.

Reward splits into wanting (incentive salience) and liking (hedonic pleasure), which are generated by different neural systems and can be dissociated, so a strong appetite need not track how much a food is actually enjoyed (Berridge, 2009).

Dietary restraint reliably reduces eating.

Restraint can invert. Once a restrained eater's control is disinhibited by a preload or emotional stress, they may eat more than an unrestrained eater — the counter-regulatory pattern of disinhibited eating (Herman & Mack, 1975).

Glossary

Adiposity Signal.

A hormone, such as leptin or insulin, that circulates in proportion to body fat and informs the brain of long-term energy stores.

Appetite Regulation.

The physiological and psychological processes that govern the onset, maintenance, and termination of the desire to eat.

Appetite.

The psychological desire to eat that selects when, what, and how much food is consumed, distinct from the drive of hunger.

Ghrelin.

A stomach-derived hormone that stimulates appetite; its plasma level rises before expected meals and falls after eating.

Glucostatic Theory.

Mayer's hypothesis that short-term changes in glucose availability, sensed centrally, trigger and terminate individual meals.

Hedonic Hunger.

Appetite driven by the anticipated pleasure of eating palatable food rather than by energy need.

Homeostatic Regulation.

The control of food intake by feedback signals that report the body's energy stores and defend a stable body weight.

Hunger.

The physiological drive state produced by energy depletion, distinguished from the psychological desire of appetite.

Incentive Salience.

The motivational property that makes a food cue attractive and worth pursuing; the wanting component of reward.

Leptin.

A hormone secreted by adipose tissue that signals the size of energy stores to the brain and suppresses appetite over the long term.

Liking.

The hedonic pleasure actually produced by consuming a food, dissociable from the wanting that motivates its pursuit.

Lipostatic Hypothesis.

Kennedy's proposal that a signal proportional to body-fat stores acts on the hypothalamus to hold adiposity constant; the conceptual forerunner of the adiposity-signal model and of leptin.

Portion-Size Effect.

The finding that larger served portions increase intake because people eat a roughly constant proportion of what is presented.

Restrained Eating.

Deliberate cognitive limitation of food intake, which can invert into overeating when the restraint is disinhibited.

Satiation.

The process that brings an episode of eating to an end and thereby determines meal size.

Satiety.

The state of suppressed appetite between meals that determines the interval before eating resumes.

Sensory-Specific Satiety.

The decline in the pleasantness of a food as it is eaten, specific to that food, so that variety increases total intake.

Wanting.

The motivational pull toward a food, generated by incentive salience and dissociable from the liking it delivers.

Key Researchers

Kent C. Berridge (b. contemporary). Psychologist and neuroscientist at the University of Michigan; he drew the liking-versus-wanting distinction and the incentive-salience account of food reward. Wikipedia - Wikidata - Google Scholar

Hans-Rudolf Berthoud (b. contemporary). Neurobiologist at Pennington Biomedical Research Center; he advanced the integration of homeostatic and hedonic control through brain-gut signalling. ORCID

Walter B. Cannon (1871-1945). Physiologist at Harvard Medical School; with A. L. Washburn he advanced the peripheral, gastric-contraction theory of hunger. Wikipedia - Wikidata

David E. Cummings (b. contemporary). Endocrinologist at the University of Washington; he showed that plasma ghrelin rises before meals and falls after eating, implicating it in meal initiation. Faculty Page

Jeffrey M. Friedman (b. contemporary). Molecular geneticist at The Rockefeller University and HHMI; he discovered leptin through positional cloning of the mouse obese gene. Wikipedia - Wikidata - Faculty Page

Suzanne Higgs (b. contemporary). Psychologist at the University of Birmingham; she established memory and attention as determinants of subsequent eating and the cognitive control of appetite. ORCID - Wikipedia - Google Scholar

Barbara J. Rolls (b. contemporary). Nutritional scientist at Pennsylvania State University; she established sensory-specific satiety and the portion-size effect on energy intake. Google Scholar - Faculty Page

Michael W. Schwartz (b. contemporary). Physician-scientist at the University of Washington; he co-developed the central negative-feedback model of food intake and body-weight regulation. ORCID - Wikipedia - Faculty Page

Stephen C. Woods (b. contemporary). Psychologist at the University of Cincinnati; he framed the adiposity-signal model of energy homeostasis, with insulin and leptin as feedback to the brain. Google Scholar - Faculty Page

Frequently Asked Questions

What is the difference between appetite and hunger?
Hunger is a physiological drive produced by energy depletion, while appetite is the psychological desire to eat a particular food, which can be present without energy need and absent despite it (Rogers & Brunstrom, 2016).

What controls appetite in the body?
A homeostatic system uses hormonal signals of energy state: slow adiposity signals such as leptin and insulin report fat stores, while the stomach hormone ghrelin rises before meals to stimulate eating (Woods et al., 1998; Cummings et al., 2001).

What is leptin?
Leptin is a hormone secreted by fat tissue that signals the size of energy stores to the brain; its discovery through cloning of the obese gene showed that appetite is regulated by feedback from the body's fat (Zhang et al., 1994).

What is the difference between liking and wanting?
Liking is the pleasure a food actually delivers, and wanting is the motivational pull to obtain it; the two are produced by different brain systems and can be pulled apart, so appetite can outrun enjoyment (Berridge & Robinson, 1998).

Why do bigger portions make people eat more?
Because people consume a roughly constant proportion of what is served rather than compensating for the extra amount, so a larger portion raises intake with little change in reported fullness (Rolls et al., 2002).

What is sensory-specific satiety?
It is the decline in a food's pleasantness as it is eaten, specific to that food, so offering variety at a meal increases the total amount consumed (Rolls et al., 1981).

Does memory affect appetite?
Yes. The memory of a recent meal suppresses later intake, so people who eat while distracted, or who cannot recall what they ate, tend to eat more afterward (Higgs, 2016).

Are homeostatic and hedonic appetite separate systems in the brain?
Increasingly the evidence says no: energy-state and reward signals are carried by overlapping, interacting circuits rather than two independent modules, so feeding is governed by a distributed network (Rossi & Stuber, 2018).

References