An interoceptive illusion of effort induced by false heart-rate feedback
Iodice, Porciello, Bufalari, Barca & Pezzulo (2019), PNAS. The wiki’s first study to run the false-feedback-paradigm on a physiological variable rather than on emotional appraisal, and the source of the interoceptive-illusion concept: give someone false acoustic feedback that their heart is beating faster than it is, and the same physical effort feels harder. Corresponding/senior author giovanni-pezzulo (active inference); Tsakiris’s body-illusion programme is the intellectual parent, and the study is framed throughout as extending the rubber-hand-illusion literature inward.
The design in one paragraph
Eighteen fit young men cycled on an ergometer at hidden intensities (60–115 W) and rated perceived exertion on the Borg RPE scale. In a first (baseline) session with no sound, the experimenters recorded each man’s heart rate at each intensity and synthesised heartbeat sounds from it. In the second (experimental) session, each intensity was paired with congruent feedback (the heartbeat sound recorded at that intensity), faster feedback (recorded one level up), or slower feedback (one level down) — delivered through headphones, and believed by the participant to be his own live heart. The dependent variable is the gap between perceived effort and the effort actually being exerted, as a function of what the participant heard.
Why it belongs in this wiki
The paper is a clean, purpose-built demonstration of interoceptive-inference: if perceived effort is an estimate the brain forms by combining interoceptive signals with priors, then injecting a false (louder-heart) signal should shift the estimate — and it does, in the direction the theory predicts, without the body actually changing. It is evidence for the predictive pole of feedforward-vs-predictive-interoception that is behavioural and causal rather than anatomical (contrast Seth & Friston’s cytoarchitecture) or model-based-imaging (contrast Harrison et al.). See interoceptive-illusion for the construct and its cousins (thermal grill, cardio-visual RHI).
The asymmetry is the interesting part
Faster feedback inflated effort; slower feedback did nothing — and the authors backed the null with Bayes factors rather than resting on a failure to reject. Their reading: the interoceptive schema is risk-averse. Overestimating effort is cheap (you rest); underestimating it is dangerous (you overexert). A prior that resists being lowered by sensory evidence, while readily being raised, is precision-weighting with an asymmetric cost function — a top-down “concern for safety” limiting the plasticity of the estimate. This is a rare case in the wiki where a false-feedback effect has a shape that discriminates between a pure sensory-integration account (which predicts symmetry) and a cost-sensitive inferential one.
The asymmetry also earns the “illusion” label. If participants had simply been reading the effort off the sound, or counting the beats, or attending harder to the body because the feedback made it salient, slower feedback should have lowered RPE too. It did not — so the effect is not stimulus-following, not attention-amplification, and not entrainment (measured HR was flat, BF = 0.014).
The MAIA null, read carefully
The study’s a priori prediction — that people low in interoceptive awareness would lean harder on the false feedback and show a stronger illusion — came straight from Tsakiris et al. (2011) on the RHI and Tajadura-Jiménez & Tsakiris (2014) on the enfacement illusion, where low interoceptive sensitivity predicts greater malleability of the bodily self. It failed: MAIA Noticing and Body Listening did not correlate with illusion magnitude, and neither did four other trait questionnaires.
The wiki should hold the null precisely, because the measure matters. Tsakiris’s malleability result used an objective heartbeat task (the Mental Tracking Task); this study used the MAIA, a self-report sensibility instrument — and the two do not measure the same thing (Ventura-Bort shows sensibility itself splits in two; Banellis shows cardiac accuracy travels to no other channel). The authors say as much: they “did not observe a similar effect… using an explicit self-report measure,” and flag that objective accuracy remains to be tested against the illusion. So this is one more entry in the wiki’s long ledger of accuracy/sensibility non-correspondence, not evidence that interoceptive ability is irrelevant to the illusion. See is-more-interoceptive-awareness-better.
Relation to the wider false-feedback literature
The paradigm’s ancestor is Valins (1966), and its most-cited modern instance is Gray et al. (2007) — false cardiac feedback raising right-AIC responses and the emotional salience of neutral faces (already recorded on false-feedback-paradigm). Iodice et al. extend it from emotional appraisal to a homeostatic/effort variable, which is the move that makes it an interoception study rather than an emotion study. They also cite Herbert, Ulbrich & Schandry (2007): good heartbeat perceivers cover less distance and show smaller cardiac increases when free to self-pace — preliminary evidence that objective cardiac perception aids effort regulation, the outcome this study did not measure but its discussion points toward. See physical-activity-and-interoception.
Brakes
Small, homogeneous, all-male sample; one channel; the effect rides on a laboratory-induced belief that the sound is one’s heartbeat, so its ecological reach is unclear; and no downstream behaviour (pacing, effort discounting) was measured, so the functional consequences the authors hypothesise are untested. The illusion is robust and the controls (no entrainment, congruent = baseline, asymmetry) are unusually thorough for the claim’s size — but it is a proof of concept for a method as much as a result about effort.