Cardiac Interoception Measures Challenged by New Computational Model

New Computational Model Questions Cardiac Interoception Tasks

A September 2026 study published via OpenAlex (W7212226397) introduces a Bayesian computational framework that challenges the validity of traditional cardiac interoception tasks. Cardiac interoception, the ability to sense internal bodily states such as heartbeats, is often assessed by comparing subjective heart rate estimates with physiological measurements. However, the study demonstrates that such tasks may conflate true interoceptive sensitivity with participants' prior beliefs, potentially undermining their utility as endpoints in both basic and clinical research.

The authors analyzed the largest Heart Rate Discrimination dataset to date (n=549) and found that only 43% of participants exhibited behaviors minimally compatible with genuine interoceptive processing. The remainder appeared to rely primarily on expectations rather than actual physiological signals, suggesting that existing measures may not accurately reflect interoceptive abilities.

Mechanism: Bayesian Updating and Belief Formation

The study formalizes interoception as a process of weighted Bayesian updating, in which perceptual beliefs about bodily states are continuously adjusted based on the interplay of prior expectations and incoming physiological evidence. This approach recognizes that the brain integrates both afferent (body-to-brain) signals and cognitive expectations to form a subjective sense of bodily state.

Unlike traditional tasks that treat interoceptive accuracy as a direct readout of bodily awareness, the new model quantifies cardiac interoceptive sensitivity—the extent to which beliefs are updated in response to naturally occurring heart rate variability. This computational metric distinguishes between true sensitivity to internal signals and the influence of prior beliefs, offering a more nuanced understanding of interoceptive processes. Notably, the study highlights that many participants show weak sensitivity, indicating that belief updating from physiological signals is far from universal in healthy populations.

Implications for Psychedelic Research and Clinical Trials

The findings carry important implications for psychedelic research and clinical trials that employ interoceptive measures as endpoints or mechanistic readouts. Many studies investigating the effects of psychedelics on mind-body interactions use cardiac interoception tasks to quantify changes in bodily awareness or self-perception. If these tasks are confounded by prior beliefs, their validity as mechanistic or clinical endpoints is called into question.

Researchers designing future trials may need to adopt more rigorous or alternative approaches to measuring interoception, incorporating computational models that account for belief updating. This is particularly relevant for studies aiming to link subjective effects of psychedelics—such as altered body awareness or ego dissolution—to objective physiological processes. The new framework also provides a foundation for embodied psychophysics, enabling the study of bodily signals that cannot be experimentally controlled, such as spontaneous heart rate fluctuations, alongside experimentally manipulated external signals.

One non-obvious implication is that regulatory agencies and trial sponsors may need to revisit protocols that rely on conventional interoceptive endpoints, especially in early-phase mechanistic studies. The computational approach could facilitate more precise stratification of participants based on genuine interoceptive sensitivity, potentially reducing noise and improving the interpretability of trial results.

Risks, Limitations, and Unknowns

The direct answer is that the new computational model exposes substantial limitations in traditional cardiac interoception tasks, but its own generalizability and clinical utility remain unproven. While the Bayesian updating framework provides a theoretically robust alternative, it requires large datasets and sophisticated modeling, which may not be feasible for all research teams or trial settings.

Moreover, the study focused on healthy participants; its findings may not generalize to clinical populations, such as individuals with anxiety, depression, or those undergoing psychedelic therapy. The extent to which computational measures of interoceptive sensitivity predict clinical outcomes or subjective experiences during psychedelic sessions is still unknown. There is also a risk that over-reliance on computational models could obscure other relevant dimensions of interoception, such as emotional or social context, which are often salient in psychedelic experiences.

Looking Forward: Rethinking Interoception in Psychedelic Science

The introduction of a Bayesian computational framework for belief updating marks a pivotal moment for mind-body research, with direct consequences for the design and interpretation of psychedelic trials. As the field moves toward more mechanistically precise endpoints, researchers and sponsors must critically evaluate the validity of their interoceptive measures and consider adopting computational approaches that disentangle belief from physiological signal.

Future studies should validate the new model in diverse populations and clinical contexts, and explore its predictive value for therapeutic outcomes. For now, the study serves as a reminder that the tools used to measure subjective experience are themselves subject to cognitive biases and methodological limitations—a consideration that is especially salient in the rapidly evolving field of psychedelic science.

How we research: Reviewed by Dr. Alex J. Harper, PhD (neuroscience), on 2026-09-13. Sources include the original study (OpenAlex W7212226397) and recent FDA trial guidance.