Interoception: How the Brain Senses the Body

What is interoception?

Interoception is the way the nervous system senses, interprets and integrates signals that come from inside the body. It gives the brain a running map of the body's internal state, some of it conscious and much of it not [S1]. The signals include the heartbeat, breathing, hunger and fullness, temperature, and the state of the gut and other organs. A current research framework adds that interoception covers not only sensing these signals but also interpreting and regulating them [S2].

It is different from the senses that face outward. Sight, hearing and touch report on the world; this is called exteroception. Its stimuli — a light, a sound — are known and easy to control in an experiment, whereas the stimuli for interoception are often unknown and hard to control, which is one reason the field is difficult to study [S4]. Interoception is also usually kept apart from proprioception, the sense of where the limbs are and how they move, which comes from muscles and joints rather than from the organs.

How does it work?

Signals from the organs travel to the brain along nerves, including the vagus nerve, and along pathways through the spinal cord. Anatomical work has described an afferent (incoming) system in primates and in humans that represents the physiological condition of the whole body; its author suggested it might provide a foundation for feelings and self-awareness [S3]. Most of this traffic serves regulation: the autonomic nervous system uses it to adjust heart rate, breathing and digestion without any conscious effort.

Where in the brain?

The insula, a region of cortex folded deep inside the brain, is central to most accounts. A meta-analysis of brain-imaging studies in which people attended to their heartbeat found an extended network: the posterior insula on both sides, the claustrum, the precentral gyrus and the medial frontal gyrus; the posterior insula is presumed to be the main entry point for heart signals [S9]. The anterior insula has been proposed as the place where body signals are re-represented and become part of subjective feeling and awareness [S8]. That is a hypothesis about how feelings arise, not an established fact, and the imaging evidence for heartbeat attention comes from a modest number of studies.

The brain as a predictor — a model

Model, not settled fact. Predictive-processing accounts suggest that the brain does not simply wait for body signals. It predicts the state of the body and uses the incoming signals to correct those predictions. Seth's "interoceptive inference" proposes that feelings arise from the brain's predictive models of what is causing the body's signals [S10]. Barrett and Simmons proposed a related model in which interoceptive experience largely reflects predictions about the expected state of the body, constrained by signals from the organs [S11]. These models organise a great deal of research and make testable predictions, but they are theoretical frameworks; neither has been confirmed as the way the brain works.

How the heart and the brain influence each other moment to moment — including the neurovisceral integration model — is left to the planned page on heart–brain interaction.

How is it measured?

Three different things called "interoception"

Researchers separate three dimensions [S5]:

• Interoceptive accuracy — performance on an objective test, such as detecting your heartbeat.
• Interoceptive sensibility — what you report about your own body awareness in a questionnaire or interview.
• Interoceptive awareness — metacognition: whether your confidence matches your actual accuracy.

In a study of healthy adults, the three were distinct, and accuracy was only partly predicted by the other two [S5] (single study). Feeling that you are in tune with your body and being accurate about it are not the same thing.

Heartbeat counting

The most widely used test is the heartbeat counting task, often called the Schandry task. People silently count their heartbeats over short intervals without touching their pulse, and the count is compared with a recorded heartbeat [S6]. The task is simple, which explains its popularity, but it faces serious validity criticism:

• In a large sample, almost all errors were under-counts, the link between counted and actual beats was weak, and scores were tied to the person's heart rate even though they are not supposed to be [S6].
• When people were told to report only the beats they actually felt — not to estimate — scores dropped sharply; many people rely on what they know about their heart rate rather than on sensation [S7] (single study).

The critique has drawn published comments from other researchers, so the task's value is debated rather than settled. In practice, a counting score cannot be read as a clean measure of how well someone feels their heart.

Heartbeat discrimination

Discrimination tasks play tones that are either in time or out of time with the heartbeat, and the person judges which. A methods review argues that both counting and these forced-choice methods appear biased and of questionable validity, and recommends psychophysical methods that ask when in the heartbeat cycle a sensation occurs [S4].

Beyond the heart

The heart dominates research because it is easy to record. Breathing, the gut and other body systems are studied less, which is one of the gaps the field itself names [S1] [S2].

What affects it?

• Attention and strategy. Whether a person attends to sensation or estimates from knowledge changes the score on the most common test [S7].
• Beliefs and confidence. Self-reported body awareness and confidence can differ from measured accuracy [S5]; experienced meditators rated themselves as better at heartbeat detection without performing better [S17].
• The test itself. The same person can look different on a counting task and a discrimination task [S4] [S14].
• Mental-health conditions. Differences in interoception are reported in several conditions, as summarised below [S1].

What does the evidence show?

Established — it is a real, multi-part capacity. The nervous system continuously represents the body's internal state [S1] [S2], and accuracy, sensibility and awareness can be told apart [S5].

Debated — how to measure it. The heartbeat counting task is heavily influenced by non-interoceptive processes [S6] [S7], and the alternatives have their own problems [S4]. Many published findings rest on the counting task, so they inherit its weakness.

Context-dependent — the brain. The insula is a consistent part of the network engaged when people attend to their heartbeat [S9]; its exact role in feelings is a hypothesis [S8].

Debated — interoception and emotion. The idea that emotions are tied to the body is old. Influential theories, from the James–Lange tradition onward, hold that emotional feelings arise from changes within the body [S12]. Appraisal accounts add the brain's evaluation of those changes; Seth's model generalises them [S10]. Constructionist and predictive accounts, such as the one proposed by Barrett and Simmons, give predictions about the body a central role [S11]. These are competing theories. How much of an emotion comes from body signals, and how much from interpretation, is a contested question.

Context-dependent — mental health. Expert consensus counts disrupted interoception as a component of anxiety, mood, eating, addictive and somatic symptom disorders, while noting conceptual and methodological problems [S1]. The heartbeat-accuracy data are mixed:

• Anxiety. A pre-registered meta-analysis found no evidence of an association between cardiac interoceptive accuracy and anxiety [S13].
• Depression. A meta-analysis found a small negative correlation with counting-task accuracy, but none with discrimination tasks [S14].
• Anorexia nervosa. A meta-analysis found no significant difference from healthy controls, and noted the near-exclusive reliance on the counting task [S15].
• Brain imaging. Across several psychiatric conditions, imaging studies converge on altered activity in one part of the insula during interoceptive tasks [S16].

These are associations between groups. They do not show that altered interoception causes any condition, and they cannot be used to diagnose anyone.

"Meditators automatically feel their heartbeat better." The evidence does not support it. Experienced meditators did no better than non-meditators at heartbeat detection, although they rated their own performance higher and the task as easier [S17]. A later study that raised heart rate with a drug infusion again found no higher detection or accuracy, and its meta-analysis found little evidence of a difference [S18]. These studies tested heartbeat detection only; whether meditation changes other aspects of body awareness is a separate question.

Emerging — can it be trained? Training studies use external feedback that shows people when their heart actually beats:

• In one short experiment, feedback that matched the real heartbeat increased heartbeat-tracking accuracy more than feedback that did not match, a short mindfulness exercise or waiting [S19] (single session, healthy adults).
• In a small randomized trial, one session of heartbeat-perception training looked promising, but there were no significant improvements over time, and both groups showed similar descriptive trends [S20].

So external feedback is being studied as one possible route to better heartbeat detection; the evidence is emerging. Whether gains last, whether they reflect sensing rather than learning about one's heart rate, and whether they help mental health are not established. This heartbeat feedback is different from HRV biofeedback, which trains slow breathing and has its own evidence.

What it does not tell you

• A heartbeat score is not a measure of "body wisdom". Common tests capture a narrow skill and are distorted by estimation [S6] [S7].
• Feeling in tune is not the same as being accurate. Self-report and confidence often diverge from performance [S5] [S17].
• No diagnosis. Group differences in interoception do not identify a condition in one person [S1] [S13].
• No proven mental-health benefit. Better heartbeat detection has not been shown to improve anxiety or mood in controlled trials.
• More monitoring is not always better. For some people, especially those prone to anxiety, closer watching of the body can feed worry rather than calm it. The paradox of watching your body explains this loop and how to step out of it.
• When to see a doctor. Body sensations such as palpitations can have medical causes. See a doctor about palpitations with dizziness, and seek urgent care for chest pain, fainting or severe breathlessness.

Practical angles are covered in interoceptive precision, body awareness training, naming a feeling and meditation with measurable progress.

In ONDA

ONDA does not measure interoception and does not train it. ONDA measures heart rate from an Apple Watch, Apple Health or the iPhone camera [S21], and ONDA shows your pulse live during a practice, from the iPhone camera or an Apple Watch. A live coherence score, ONDA's own feedback metric rather than a clinical measure, needs an Apple Watch. Seeing your pulse on a screen is external information about the heart, not a test of how well you sense it. ONDA does not diagnose anything. See what ONDA measures.