Method at a glance
- Modality: respiratory
- Measures: Anticipation and experience of an aversive inspiratory resistive load, and the gap between them — ‘interoceptive prediction accuracy’ (IPA). Distinct from respiratory detection tasks: the load is unmissable, and what is measured is the forecast of an aversive state rather than the perception of a faint one.
- Procedure: A visual cue signals an upcoming restricted-breathing condition. In the test phase the cue is followed by the load on some trials — after which the participant rates what they experienced — and by nothing on others, after which they rate what they had anticipated. Loads are produced by resistors or valves in the inspiratory limb of a two-way non-rebreathing circuit, so inspiration is restricted without affecting expiration. IPA = experienced magnitude minus anticipated magnitude, aggregated per participant.
- Also called: IBL, inspiratory breathing load, aversive inspiratory load, breathing load anticipation task, interoceptive prediction accuracy, IPA
Strengths
- A physically imposed, unambiguous perturbation. Unlike heartbeat tasks, the interoceptive signal is manufactured to order and is unmistakably present — so nothing here depends on faint-signal detection, and there is no route to a good score through beliefs about one’s own physiology.
- It measures the anticipatory limb, which almost no interoceptive task does. This is the paradigm that made Paulus & Stein’s ‘failure to anticipate interoceptive states’ into something with numbers attached.
- Transfers to the scanner. Walter et al. (2020) ran it in fMRI and recovered anterior insula and pre-SMA activation during prediction of future interoceptive states — the same regions engaged during craving down-regulation, which is the finding the paradigm is best known for.
- It travels outside interoception. Kruschwitz et al. (2019) linked IPA to self-control: participants better at predicting future interoceptive states were better at down-regulating craving via negative future thinking. Few interoceptive measures predict anything outside their own literature.
- Aversive-but-tolerable and quickly reversible, which is why it has been usable in anorexia-remitted (Berner et al. 2018), athlete (Paulus et al. 2012) and clinical samples.
Limitations
- The anticipation rating is retrospective and relief-contaminated. It is queried only after the moment the stimulus would have arrived, so it is a report of a forecast made under the knowledge that the forecast need not be endured. This is the defect Wilzok et al.’s ID paradigm was built to fix.
- One load level. With a single severity there is no stimulus-response function, so the relationship between load and sensation cannot be estimated and anticipation cannot be shown to be graded rather than categorical.
- IPA is a difference score between two self-reports, with all that implies for reliability and for shared scale-use variance.
- Respiratory loading is a loud stimulus and arguably closer to an effort judgement than to visceral perception — the same objection the wiki records against the FDT, and more forceful here because the load is suprathreshold by design.
- Voluntary breathing strategy is uncontrolled: participants can alter inspiratory effort, and the paradigm cannot distinguish that from altered perception. See Respiratory interoception on the channel’s exceptional voluntary accessibility.
- Anxiety about suffocation is a plausible mediator of everything the task measures, and is generally not measured alongside it.
Connections
The parent of Wilzok et al.’s two paradigms, and the wiki’s earliest instrument for interoceptive anticipation rather than interoceptive detection. Entered the wiki with the Wilzok et al. (2023) ingest, which is a critique of it as much as a descendant.
None of the primary IBL papers are in raw/; this page is held at one remove and should be read as such.
The lineage
| paper | what it added |
|---|---|
| Paulus, Flagan, Simmons et al. (2012), PLoS ONE 7:e29394 | The paradigm. Elite athletes under inspiratory load, looking for behavioural and neural signatures of optimal performance in extreme environments. |
| Kruschwitz, Kausch, Brovkin et al. (2019), Cognition 193:104028 | Interoceptive prediction accuracy (IPA) as a derived measure, and its link to self-control: better predictors of future interoceptive states were better at craving down-regulation. Also the finding that restricted breathing is experienced as worse than anticipated. |
| Walter, Kausch, Dorfschmidt et al. (2020), NeuroImage 215:116841 | The imaging replication: anterior insula and pre-SMA active both during prediction of aversive interoceptive states and during craving regulation. |
| Berner, Simmons, Wierenga et al. (2018), Psychol Med 48:142-154 | Clinical extension — altered interoceptive activation before, during and after aversive load in women remitted from anorexia nervosa. |
Why the paradigm matters to this wiki out of proportion to its size
Two reasons, both structural.
It is the instrument for the wiki’s most load-bearing clinical claim. Interoceptive psychopathology and computational psychosomatics both rest on the idea that disorders are mis-set predictions about the body. Martin P. Paulus’s page records the complaint that the forecast in question has never been written down; the IBL is where it was at least measured, and the fact that the wiki’s whole predictive-psychopathology edifice traces its empirical footing to one respiratory task with one load level is worth stating plainly.
Its most-cited result is not about interoception. The Kruschwitz/Walter finding that IPA predicts craving down-regulation, and that the same anterior-insula circuit serves both, is the paradigm’s calling card. That is a claim about self-control borrowing interoceptive machinery — which puts it beside somatic-marker arguments rather than beside the detection literature. Whether the shared activation reflects a shared computation or merely a busy region is the standing question about insula findings generally.
What it is not
It is not a respiratory detection task and should never be compared with the FDT or RRST as though it were. Those staircase the load down to threshold to find where perception fails. The IBL keeps the load plainly suprathreshold and asks what the person expected. Detection tasks measure sensitivity; this measures a forecast. Both use inspiratory resistance, which is exactly why the confusion is available.
Nor does it measure Forecasting (internal-state prediction) in the strict sense that page defines. What is learned is that a visual cue predicts an experimenter-imposed event. The body’s own dynamics do nothing. The distinction is the same one that disqualifies the BLT, and it disqualifies the IBL and its descendants for the same reason.