Interoception

You have five senses pointed outward and one pointed inward. Interoception is the inward one: the stream of information the brain receives about the state of the body — heart rate, breathing, gut, temperature, muscle tension, the thousand signals that never reach words.

It is the least discussed sense and arguably the most consequential, because emotion is partly built out of it. The tightness in the chest is not a metaphor for anxiety. It is one of the ingredients the brain uses to construct the experience it then labels as anxiety.

Which sets up the distinction this page is about. People assume the goal is to be better at reading the body — more tuned in, more aware. The research points somewhere more specific and more useful: what predicts trouble is not poor body-reading. It is the gap between how well you read your body and how well you believe you do.

Sarah Garfinkel and colleagues made the field tractable in 2015 by insisting these are separate, measurable and only loosely correlated:

• Interoceptive accuracy. How well you actually do on an objective task — traditionally counting your own heartbeats without taking a pulse, a method Rudolf Schandry introduced in 1981.
• Interoceptive sensibility. How tuned-in you believe you are. This is what a questionnaire measures. It is self-report, and it tracks accuracy loosely at best.
• Interoceptive awareness. Whether your confidence tracks your accuracy — whether you are more sure on the trials you get right. This is metacognition about the body, and it is the one that behaves most interestingly.

The important finding is that a person can be low on the first and high on the second. They feel intensely attuned to their body and are, in fact, reading it badly — and the mismatch is what shows up in anxiety. The signal is being amplified and misinterpreted at the same time. Khalsa's 2018 roadmap sets out how consistently that pattern appears across anxiety, panic, eating disorders and somatic conditions.

There is a mirror image too. People who score high on accuracy but rarely attend to it, and people who have very little access to the inward channel at all, for whom emotions arrive as unexplained physical events or as nothing.

Lisa Feldman Barrett and Simmons' predictive account is the current mainstream view, and it changes what "listening to your body" can mean.

The brain does not passively receive body signals and then feel something. It is continuously predicting what the body's state should be and using the incoming signal mostly to correct the prediction. What reaches awareness is largely the prediction, adjusted.

Two consequences follow, and both are recognisable.

Expectation shapes sensation. If you expect your heart to race, you will feel it race sooner and harder than the monitor justifies. This is not imagination; it is how the system is built. It is also the engine of a panic spiral: the prediction produces the sensation, which confirms the prediction.

Context supplies the label. The same physiological arousal can arrive as fear, excitement or anger depending on what the brain has decided it is about. Which is why the entry on the freeze response describes the same bodily machinery producing four quite different experiences, and why "am I anxious or am I excited" is a genuine question rather than a semantic one.

The heartbeat counting task has carried most of this literature, and it is not in great shape.

Zamariola and colleagues showed in 2018 that scores on it are heavily driven by under-counting — people who report fewer beats score as more accurate — and that the scores correlate with heart rate and with beliefs about one's own heart rate in ways that a pure accuracy measure should not. Other groups have pushed back and the argument is live, but the fair summary is that a single number from that task should not be treated as a stable trait about you.

So take the three-way distinction seriously and treat specific accuracy claims cautiously. That is roughly where the field itself sits, and it is why no reputable test hands you an interoception score. What a questionnaire can read is the sensibility dimension — how much you attend to and trust bodily signals — which is exactly what the Highly Sensitive Person test picks up on one side and the Stress Response test on the other.

If you feel everything and misread it. The move is not more attention to the body — that usually amplifies the loop. It is adding a second source: naming the situation, checking the context, asking what a person watching would say is happening. Interoceptive training that pairs the signal with feedback tends to work better than training that just turns the volume up.

If you feel nothing until it is a crisis. The gap is usually not in the body but in the vocabulary. The practical route in is granularity: distinguishing tired from hungry from flat from lonely, several times a day, in words. The entry on somatic symptoms of stress describes what happens when the signal has no route into language and takes a physical one instead.

If your body is the first to know. Some people's earliest reliable information about a situation is postural — the shoulders before the thought, the stomach before the opinion. That is worth treating as data rather than superstition, with one caveat: the body is reporting arousal, not truth. It can tell you something is happening. It cannot tell you what, and it is confidently wrong about why more often than people expect. The social battery is a good example of a bodily signal that is reliable about the cost and unreliable about the cause.

And the limit, stated plainly: chest pain, breathlessness, palpitations and dizziness are not a mystery to be solved by better self-knowledge. They are symptoms that get checked medically first. "It is probably interoception" is a conclusion available only after a doctor has ruled out the alternatives.