Body-wandering
When the mind is left alone, some of what it does is think about the body. That is the whole claim, and the reason it needed a name is that the mind-wandering literature had no slot for it: self-generated thought has been characterized along self-reference, temporal orientation (past/future), social content and affective tone, and interoceptive content was neither probed nor theorized.
Banellis et al. (2026) added eight body probes to the standard MDES battery, gave it to 536 people after a 14-minute resting scan, and found that the resulting dimension is stable across individuals (split-half ICCs 0.88–0.94), affectively distinctive, physiologically distinctive, and neurally distinctive.
The three signatures
Affective. Body-wandering is negative. Thoughts about heart, bladder, skin and stomach correlate with reduced positive and increased negative affect; thoughts about the past, the future, other people and one’s own focus correlate with the opposite. Exploratory factor analysis recovers this as a single bipolar axis — visceral-and-unpleasant against social-and-pleasant — explaining 9% of variance and reproducing at r = 0.92 across split halves.
That polarity is the page’s most theoretically loaded fact. It says body-oriented and socially-oriented thought compete at the level of individual disposition. A person whose resting stream is full of their stomach is not a person whose resting stream is full of other people.
Physiological. Body-wandering tracks high autonomic arousal: elevated heart rate, reduced parasympathetic HRV (RMSSD), with convergent patterns in respiration and electrogastrography. Cognitively-oriented mind-wandering tracks the low-arousal state. So the two poles of the affective axis are also two poles of an arousal axis — which is core-affect’s valence/arousal pair falling out of a thought-content questionnaire.
Neural. Not the default-mode-network. The connectivity mode associated with body-wandering is centred on thalamus and striatum coupling to primary somatosensory and motor cortex, with right thalamus as hub, plus insula and cingulate. Canonical mind-wandering is described as perceptual decoupling; body-wandering involves sensorimotor circuitry rather than disengaging from it, and its topography bridges the principal unimodal-to-transmodal cortical gradient.
The clinical paradox
Body-wandering feels bad and goes with fewer symptoms. Heart-, bladder- and skin-related thought correlates negatively with both ADHD (ASRS) and depression (MDI) scores; depression’s positive correlate is mental time travel instead.
The authors’ proposed reading, offered as speculation: spontaneous attention to bodily sensation reflects preserved engagement with ongoing sensory input, and that engagement is protective relative to thought decoupled from the current bodily state. Unpleasantness at rest is not the same as maladaptive cognition.
This is the same shape the wiki already holds from a completely different literature. Farb et al. (2011): elaborative self-referential reactivity predicts relapse, sensory reactivity protects, felt sadness predicts nothing. Farb et al. (2010): trained participants keep the insula online under a sad film where controls deactivate it, and more insula recruitment tracks lower depression. decentering and cognitive-reactivity carry the same distinction from the clinical side.
What Banellis et al. add is that this is not only a trained or clinical phenomenon: the elaborative/sensory split shows up as ordinary dispositional variation in an unselected sample at rest, without any intervention, and the sensory pole is genuinely interoceptive rather than visual — which the Farb thread has wanted and mostly not had.
Held with the obvious caution: rs ≈ 0.10–0.21, cross-sectional, and a sample with little clinical depression in it.
What kind of construct is this?
Body-wandering is self-reported attention to the body, and therefore sits on the sensibility side of the interoceptive-taxonomy — near MAIA-style noticing, far from any perceptual accuracy measure. Nobody in this study was asked to detect anything.
That placement is not incidental, because the same research group’s other 2026 paper found that the interoceptive quantity behaving like a person-level trait is confidence while performance is organ-specific. Body-wandering shows exactly the trait-like stability that finding predicts for a sensibility-side measure — and inherits exactly the warning: it cannot be assumed to track any ability to actually perceive the organs it is about.
An open question worth flagging rather than answering: is high body-wandering the same disposition as high MAIA noticing, or a different one? The attachment-anxious profile is high-noticing and worse off; body-wandering is high-noticing and better off. Either these are different constructs, or the difference is between vigilant monitoring for threat and undirected drift toward the body. Nothing measured here separates them.
Two cautions on the construct itself
The scanner is a strange body to wander in. Lying motionless in a narrow tube with the instruction to have used the bathroom first is not a neutral bodily context, and the bladder item is one of the strongest correlates of everything on this page. The authors control for it statistically and the patterns survive, but the ecological question — is body-wandering in a scanner the same phenomenon as body-wandering on a bus? — is untested.
Which body depends on the analysis. Unsupervised factor analysis picks skin, heart, bladder and stomach. Brain-constrained CCA picks stomach, arousal, breath and movement. Same data, same sample, different organs foregrounded — a small worked example of the wiki’s standing worry about generalizing from whichever interoceptive channel a method happens to favour. See is-interoception-domain-general.
The stomach, met from the other direction
The organ the brain-constrained CCA foregrounds most strongly (stomach, loading 0.58) is also the one channel with an established neural coupling signature — the gastric-network, a set of unimodal sensory and motor regions phase-locked to the stomach’s ~0.05 Hz rhythm. Body-wandering’s own connectivity mode is thalamo-somatomotor and unimodal. Two 2020s literatures, working from opposite ends (thought content vs. organ rhythm), arriving in overlapping cortex.
Whether that is convergence or coincidence is untested and cheap to test, and one result constrains what such a test could show. Levakov et al. (2023) found gastric–brain coupling has no test–retest reliability in individuals, while the EGG signal itself retests at r = .74. So a study correlating body-wandering with the gastric signal is viable; one correlating it with gastric–brain coupling is bounded by a reliability near zero before it starts. See is-brain-body-coupling-a-reliable-individual-difference.