Abstract Interoception - the physiological sense of our inner bodies - has risen to the forefront of psychological and psychiatric research. Much of this research utilizes tasks that attempt to measure the ability to accurately detect cardiac signals. Unfortunately, these approaches are confounded by well-known issues limiting their validity and interpretation. At the core of this controversy is the role of subjective beliefs about the heart rate in confounding measures of interoceptive accuracy. Here, we recast these beliefs as an important part of the causal machinery of interoception, and offer a novel psychophysical “heart rate discrimination“ method to estimate their accuracy and precision. By applying this task in 223 healthy participants, we demonstrate that cardiac interoceptive beliefs are more biased, less precise, and are associated with poorer metacognitive insight relative to an exteroceptive control condition. Our task, provided as an open-source python package, offers a robust approach to quantifying cardiac beliefs. Highlights Current interoception tasks conflate cardiac beliefs with accuracy. We introduce a Bayesian method for estimating cardiac belief accuracy and precision. Individuals underestimate their heart rate by -7 BPM (95% CI [-8.6 -5.3]) on average. Cardiac beliefs are associated with reduced precision and metacognitive insight. The task and modelling tools are provided in the Python Cardioception Package.

The heart rate discrimination task: a psychophysical method to estimate the accuracy and precision of interoceptive beliefs