What Does Hunger Mean? New 2026 Research Invites Us to Rethink Interoception, Appetite and Eating

A stomach rumble seems straightforward.

Your stomach rumbles. You feel hungry. You eat.

Or perhaps you walk past a bakery and suddenly want something you had not been thinking about. You see a photograph of food when you have missed breakfast, and it becomes almost impossible not to notice. You eat something you really wanted, only to discover you don’t actually enjoy it very much.

We often talk about appetite as though there is a fairly simple chain running from the body’s need for food, to a hunger signal, to the brain noticing it, to eating.

Three papers brought together in the August 2026 issue of Appetite invite us to look again.

https://www.sciencedirect.com/journal/appetite/vol/223/suppl/C

Appetite describes itself as multidisciplinary research on eating and drinking. Its editorial subject areas span neuroscience, nutrition, physiology and metabolism, psychology and the senses, amongst other disciplines. Volume 223, dated 1 August 2026, contains an especially interesting cluster of papers exploring hunger, sensory cues, interoception, food motivation and interoception-based interventions. (ScienceDirect)

They are not three studies answering the same question.

That is precisely what makes reading them together so useful.

One asks whether the meaning of an internal bodily signal can itself contribute to hunger.

Another asks what happens to our responses to things we see and smell when we are actually hungry.

The third asks whether interventions that deliberately work with interoceptive experience can help people with eating disorders.

Together, they make the apparently simple question, “How do we know we are hungry?”, much more interesting.

Study one: When a stomach rumble becomes hunger

Stevenson and colleagues asked a rather clever question.

A rumbling stomach is associated with physiological activity in the gastrointestinal system. We commonly assume that the physiology produces the sensation, which then tells us that we are hungry.

But what if part of the power of that sensation comes from what we have learned it means?

The researchers recruited 130 participants. Sixty-five heard recordings of stomach rumbling during a task involving neutral, non-food pictures. Another 65 heard an electrical buzzing sound. Bowls of snack foods were available throughout the experiment.

The important part was not simply whether people heard the rumble.

Participants were also asked where they felt the sound had come from.

Among the 65 participants hearing stomach sounds, 37 experienced at least one occasion when the sound was perceived as coming from towards their own body. The researchers described this as illusory mislocalisation. The other 28 did not experience this illusion.

Then something striking happened.

The people who experienced the stomach rumble as if it belonged to their own body consumed around 50 per cent more energy than those who heard the same stomach sound but did not experience the illusion.

Their average intake was 1,199 kJ, compared with 767 kJ in the rumble without illusion group and 801 kJ in the electrical sound control group.

They also reported being hungrier, despite having eaten more during the experiment.

There was no significant difference between the groups in how long it had been since they had last eaten.

Perhaps even more interestingly, people who already reported experiencing stomach rumbling strongly when hungry were more likely to experience the illusion.

So, did the researchers create hunger without physiological hunger?

Not quite.

This study does not tell us that physiology is irrelevant to hunger, and it certainly does not suggest that hunger is simply “psychological”.

Physiological signalling remains an essential part of appetite regulation.

What the experiment does suggest is that physiology may not be the whole explanation.

A psychologically induced cue, when experienced as belonging to the body, was associated with both greater reported hunger and substantially greater eating.

The researchers interpret this as evidence consistent with a learning-based account of interoceptive hunger. Through experience, a bodily sensation may acquire meaning. A stomach rumble may come to mean something like, “Food will be rewarding now.”

That meaning may then contribute to what happens next.

This also fits with predictive accounts of interoception, where the brain is not understood as simply waiting passively for bodily messages to arrive. Previous experience contributes to the brain’s ongoing predictions about what is happening in the body.

That gives us a rather different picture of interoception.

• Sensation matters.
• But so might recognition.
• Expectation.
• Memory.
• Learning.
• Meaning.
• And context.

There are important cautions

The division between participants who did and did not experience the illusion was made after the experiment according to their localisation responses. The authors acknowledge this as a limitation.

They also consider whether participant expectations might have influenced behaviour. Their design attempted to reduce this possibility, but it cannot be completely excluded.

The participants were not a clinical population, and the study should not be generalised automatically to people with eating disorders, ARFID, neurodevelopmental differences or gastrointestinal conditions.

This is one experiment exploring a fascinating phenomenon. It does not establish a universal theory of hunger.

But it does make one familiar assumption much less comfortable: perhaps an interoceptive cue does not have a fixed meaning simply because it comes from the body.

Study two: Being hungry changes wanting more than liking

The second study comes at appetite from almost the opposite direction. Rather than creating an apparent bodily cue, Savva and colleagues deliberately changed participants’ actual hunger state and asked what happened to their responses to external sensory information.

Forty-three healthy adults completed two laboratory sessions.

• In one, they had fasted overnight.
• In the other, they ate breakfast until comfortably full before completing the experimental tasks.

Participants then encountered food and non-food cues presented either as pictures or as smells. Crucially, the researchers did not treat food reward as a single experience.

Participants rated two different things:

• how much they liked the stimulus,
• And how much they wanted the food.

That distinction produced one of the most interesting findings.

• When hungry, participants wanted food more.
• But they did not like it significantly more.

Across the food cues, hunger substantially increased wanting, while the effect on liking did not reach statistical significance. The effect of hunger was significantly greater for wanting than for liking.

That difference matters.

We often use phrases such as “I fancy that”, “I like that”, “I want that” and “I’m hungry” almost interchangeably.

Neuroscience and appetite research suggest they are not interchangeable processes.

• You can like something without particularly wanting it.
• You can strongly want something without its pleasantness having changed.
• And hunger appears capable of changing motivation towards food without necessarily changing how pleasant we judge that food to be.

That might sound academic until we recognise it in ordinary life.

• Perhaps you have eaten a satisfying meal and can still say that chocolate cake looks lovely without wanting to eat it.
• Or perhaps after a long day without lunch, an ordinary sandwich suddenly becomes something you urgently want even though your general opinion of sandwiches has not changed.

The research does not test those particular everyday examples, but the distinction between wanting and liking gives us a useful way of thinking about them.

And then the researchers got a surprise

The team had predicted that the sense of smell would be especially sensitive to hunger.

That seems intuitively plausible. Food smells can be extraordinarily evocative, and olfactory pathways are closely connected with neural systems involved in reward, memory and emotion.

But that was not what they found. Hunger altered wanting more strongly for food pictures than for food odours!

In exploratory analyses, wanting for food pictures increased markedly in the fasted state. There was initially a smaller increase for food odours, but that odour-specific effect was no longer statistically significant after individual differences in hunger and food craving were taken into account.

So another simple assumption becomes less certain. Different sensory routes into appetite may not be interchangeable. Seeing food and smelling food may contribute differently to the motivational experience.

The authors offer several possible explanations. Visual information may be easier to identify, describe and consciously evaluate. Odours are often harder to put into words and may carry learned associations with flavour and previous eating experiences in a different way.

Those explanations remain possibilities rather than established conclusions.

The experiment measured explicit ratings, not every unconscious motivational process that might be occurring.

The authors themselves suggest that future research should include implicit and physiological measures, particularly for olfaction.

Again, the value is not that one sense has “won”. Sensory integration matters!

It is that appetite appears to emerge through the integration of bodily state with different kinds of sensory information.

Study three: Can we intervene through interoception?

The third paper asks the question that often follows quickly in practice.

If altered interoceptive processing is associated with eating difficulties, can we improve outcomes by targeting interoception?

Diana and Barca conducted a systematic review of interoception-based interventions for anorexia nervosa, bulimia nervosa and binge eating disorder.

Their search identified 13 eligible studies involving 454 participants.

Nine examined mindfulness-based approaches, two explored interoceptive exposure, and two examined biofeedback.

At first glance, the findings appear encouraging.

Mindfulness-based interventions were frequently associated with reductions in eating disorder symptoms and improvements in areas such as emotional regulation and bodily awareness.

Early work involving interoceptive exposure and biofeedback also suggested possible benefits.

But this is where careful reading matters. The evidence base was very uneven.

Ten of the thirteen included studies were judged by the review authors to have a high, serious or critical risk of bias. Only two were rated at low risk, and both were case reports. One further case series received a moderate rating.

Interoceptive exposure was particularly underdeveloped. Across the two clinical reports identified by the reviewers, only five people had received the intervention.

There was another important problem.

Despite these being described as interoception-based interventions, many studies did not directly measure changes in specific dimensions of interoception.

Improved emotional regulation or reduced eating disorder symptoms were sometimes being used to infer that interoceptive functioning had changed.

Those are not necessarily the same thing.

The review therefore reaches a much more interesting conclusion than “interoception interventions work”.

• There is promise.
• There are plausible mechanisms.
• There are some encouraging clinical findings.
• But there is not yet enough rigorous evidence to know precisely which aspects of interoception are changing, for whom, through which intervention, or whether interoceptive change is actually the mechanism producing improvements.

Mindfulness, exposure and biofeedback are not the same intervention

This deserves particular attention.

The review groups three approaches under the broad heading of interoception-based intervention, but their proposed mechanisms are quite different.

Mindfulness-based approaches may alter how attention is directed towards bodily experience and how sensations are responded to.

Interoceptive exposure deliberately evokes uncomfortable bodily sensations and seeks to change avoidance, fear or distress associated with them.

Biofeedback makes physiological information available through external feedback and may support recognition or regulation of bodily states.

Calling all three “interoception interventions” should not make us assume they do the same thing.

This seems particularly important as interoception becomes more visible in clinical and educational practice.

Simply saying that an intervention “works on interoception” may no longer be precise enough.

Which interoceptive process?

• Detection?
• Attention?
• Interpretation?
• Confidence?
• Tolerance?
• Meaning?
• Prediction?
• Physiological regulation?

And how will we know that the intended process has actually changed?

What happens when we put all three papers together?

This is where the August Appetite issue becomes especially thought-provoking.

The Stevenson study starts with an external sound and asks what happens when it is experienced as an internal bodily event.

The Savva study starts with an internal metabolic state and asks how it changes our response to external sensory information.

The Diana and Barca review asks whether deliberately changing people’s relationship with bodily information might alter eating disorder outcomes.

These are separate studies. We should not combine them as though together they prove a single model.

But considered alongside one another, they invite a broader possibility.

Appetite may be better understood not as a one-way message travelling from stomach to brain, but as an ongoing interaction between bodily physiology, sensory information, previous experience, prediction, learning, motivation, environment and action.

Or, more simply:

body state ⇄ sensory world ⇄ learned meaning ⇄ wanting and liking ⇄ action.

That is our interpretation of the emerging picture, not a conclusion tested directly by any one of these papers.

And it makes interoception considerably more interesting than simply “noticing signals from inside the body”.

So, is a stomach rumble a hunger signal?

It can be.

But the Stevenson findings suggest that what the rumble has come to mean may also matter.

• For one person, stomach movement may be a highly recognisable precursor to eating.
• For another, hunger might be noticed first through concentration, shakiness, irritability, fatigue, headache, nausea, weakness or a change in thinking.
• For somebody else, reliable conscious hunger signals may be difficult to identify at all.

The study cannot tell us how these different experiences develop, but it gives us a reason not to assume that everybody’s body communicates hunger in the same language.

Does this mean hunger is learned rather than biological?

No. That would replace one oversimplification with another.

Hunger involves powerful physiological systems, including gastrointestinal, metabolic, endocrine and neural processes.

The Stevenson experiment suggests that learned meaning may contribute alongside physiology.

It does not erase physiology.

Nor should sensory or psychological interpretations of appetite replace appropriate medical, nutritional or mental health assessment when eating or appetite changes create concern.

Does seeing food actually make us hungry?

The Savva study does not quite answer that question.

Participants were already experimentally hungry or sated.

What changed was how much they reported wanting food in response to sensory cues.

The results suggest that our current bodily state alters the motivational significance of information outside us.

And this happened differently for pictures and smells.

So perhaps the more useful question is not whether appetite comes from “inside” or “outside”.

It may emerge through their interaction.

Is wanting food the same as enjoying food?

Apparently not.

This may be one of the most practically useful findings from the whole cluster.

Wanting and liking can move apart.

Hunger particularly influenced wanting in the Savva study.

That distinction could help us become more curious about eating experiences without assuming that motivation, pleasure, hunger and food preference are identical.

Should we simply teach people to listen to their body?

Perhaps that instruction is too small for what interoception actually involves. Listening assumes there is a clear message waiting to be heard.

• But what if the signal is ambiguous?
• What if several bodily sensations occur together?
• What if previous experiences have affected what a sensation predicts?
• What if gastrointestinal discomfort feels very similar to fullness?
• What if medication changes appetite?
• What if anxiety, pain, fatigue or illness changes internal sensations?
• What if someone knows they need nourishment before they consciously feel hungry?
• And what if a regular meal routine is actually an effective way of supporting their body rather than evidence that they have somehow failed to “listen” correctly?

These papers do not answer all those questions. They give us better reasons to ask them.

What about relationships?

This may be one of the most intriguing unanswered questions.

The Stevenson study did not test whether family interactions, feeding practices or professional responses shape later hunger perception. We should therefore be very careful not to claim that it did.

However, the authors connect their findings to earlier developmental work exploring similarities between parents and children in hunger sensations and caregiver responses to children’s interoceptive hunger and thirst cues. That makes relational experience an important area for future enquiry.

• What happens when a child says, “I’m hungry”, and is repeatedly told they cannot be?
• What happens when someone says they are full and is required to finish the meal?
• What happens when bodily reports are believed?
• Or doubted?
• Prompted?
• Corrected?
• Rushed?
• Supported?

None of those questions have been answered by this experiment.

But once interoceptive meaning is understood as potentially shaped by experience, relationships become difficult to leave out of the conversation.

Whose experience is missing?

This matters enormously before translating any of this research into practice.

The Stevenson study involved a non-clinical sample.

The Savva study involved healthy adults aged 18 to 45 and excluded people with a history of eating disorders, psychiatric or neurological conditions and several other factors.

The eating disorder systematic review was also heavily weighted towards adults and women. Across all 454 participants, the reviewers identified only 14 men, around three per cent of the total sample. Evidence involving younger people was limited.

The systematic review explicitly excluded ARFID because there were too few intervention studies within its chosen scope, although the authors recognise emerging interoception research in this population.

So these papers cannot tell us whether the same patterns occur in autistic people, people with ADHD, people with learning disabilities, people with dementia, people taking appetite-altering medication, people living with gastrointestinal or metabolic conditions, or children whose eating experiences differ substantially from those represented here.

Nor can they tell us enough about culture, food security, trauma, disability, communication differences, or the meaning of shared eating.

Those are not small omissions.

They are part of where the next research questions sit.

What might this change in practice?

Perhaps the challenge is not simply to help somebody “identify their hunger signal”.

It may be to become more curious about how hunger is experienced, interpreted and acted upon by this particular person, in this particular context.

This is where the PEAR TREE™ Lens can help us extend the thinking, not because these papers prove the framework, but because the research demonstrates why locating appetite entirely inside the person risks missing too much.

The Person brings physiology, interoceptive experience, sensory processing, health, medication, learning history, food preferences, communication and previous experience.

The Environment includes smells, visual food cues, noise, lighting, available foods, access, routines, culture, social expectations and predictability.

The Activity matters too. Eating breakfast before school is different from sharing a celebration meal, preparing food after work, eating during a short hospital break, managing lunch in a noisy dining hall or trying to notice hunger while deeply absorbed in something meaningful.

And the Relational Response asks what happens between people. Is the person’s report trusted? Are they hurried? Is food controlled by somebody else? Can they choose? Are internal signals explored with curiosity, or interpreted for them?

The papers did not test PEAR TREE™.

Rather, PEAR TREE™ gives us one way of holding the questions the evidence opens up, without prematurely reducing them to a single sensory explanation.

Questions worth carrying into everyday practice

• What does hunger actually feel like for this person, and are there several different signals rather than one?
• How does the person distinguish hunger from fullness, nausea, thirst, anxiety, pain, fatigue or gastrointestinal discomfort?
• Are those signals consistent, or do they change with time of day, health, stress, activity, medication or environment?
• What has the person learned about what particular bodily sensations mean, and where might that learning have come from?
• Which visual, olfactory, taste, social or contextual cues increase or reduce wanting?
• Does wanting a food correspond with liking it, enjoying it and feeling comfortable after eating it?
• What happens to bodily awareness when the person is deeply focused, overwhelmed, anxious, tired or busy?
• How do routines and relationships support or disrupt eating, nourishment, choice and bodily agency?
• What does the person themselves say is helpful, and what health, nutritional or specialist assessment may also be needed?

These are not questions designed to make eating more closely resemble somebody else’s idea of “normal”.

They are questions for understanding the experience together.

The bigger learning may be about interoception itself

Interoception is sometimes taught as though it were another sensory channel to add to a list.

• Vision.
• Hearing.
• Touch.
• Vestibular.
• Proprioception.
• Interoception.

But this cluster of research points towards something much more dynamic.

• A bodily sensation does not exist in isolation from previous experience.
• Metabolic state can change how external sensory information is valued.
• Different sensory modalities may influence motivation differently.
• Wanting and liking can be separate.
• Attention to bodily signals is not the same thing as accurately interpreting them.

And an intervention described as “interoceptive” cannot automatically be assumed to change interoceptive processing unless we actually measure that process.

This brings us back to sensory integration in its broadest neurological meaning!

The important question is not simply whether a signal was detected.It is what happens as information from the body and the world is integrated, interpreted and used in adaptive action.

And for practice, the question does not end with whether somebody can identify the sensation.

• What does being able to understand and respond to that experience enable them to do?
• Can they nourish themselves?
• Join a family meal?
• Manage a school day?
• Cook?
• Shop?
• Recognise when something feels wrong?
• Ask for food?
• Say no?
• Choose?
• Recover?
• Participate?

Perhaps we need a different question

Instead of beginning with:

“Can this person recognise when they are hungry?”

Perhaps we might begin with:

• “What does hunger feel like for you?”
• “What tells you that this sensation means you need food?”
• “What else is happening in your body and around you when you notice it?”
• “What makes the signal clearer or harder to recognise?”
• “What helps you respond to it?”
• “And what do you need from the people around you?”

The August 2026 issue of Appetite does not give us a new protocol for answering those questions. Something more useful may be happening:

• It reminds us that a human being is not simply a collection of sensory receptors reporting objectively to a brain.
• We are physiological, sensory, predictive, learning and relational beings, continually making meaning from what is happening within us and around us.
• Perhaps appetite is one particularly vivid place where we can see that integration happening.
• And perhaps the aim is not to teach everybody to experience hunger in the same way. It is to understand experience together, make space for reliable and personally meaningful ways of responding to bodily needs, protect agency and health, and support participation in the everyday occupations around food that matter.

References

Stevenson, R. J., Francis, H. M., Martin-Rivera, D., Wylie, F., Ho, V., & Saluja, S. (2026). Testing a learning based account of interoceptive hunger using an illusory induction. Appetite, 223, Article 108553. doi:10.1016/j.appet.2026.108553.

Savva, A., Gerlicher, A., Rey, L., Guitart-Masip, M., Ghaderi, A., Bulik, C. M., & Seubert, J. (2026). Reward responses to food stimuli across sensory modalities: Hunger modulates wanting differentially for pictures and odors. Appetite, 223, Article 108557. doi:10.1016/j.appet.2026.108557.

Diana, S. M., & Barca, L. (2026). Interoceptive based interventions for eating disorders: A systematic review. Appetite, 223, Article 108559. doi:10.1016/j.appet.2026.108559.

Suggested feature image caption: Appetite, Volume 223, 1 August 2026. Three papers in this issue invite a closer look at hunger, interoception, sensory cues, food motivation and what these emerging findings may mean for everyday participation.