Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

We theorize that the perception of bodily sensations (i.e., interoception) depends not only on sensory input, but also on learned expectations – or predictions. From this point of view, peripheral physiological input is neither necessary, nor sufficient to experience bodily sensations and symptoms. In some cases, the perception of the internal body can be determined strongly by learned expectations about stimulation, possibly leading to interoceptive illusions (i.e., medically unexplained symptoms). There is evidence that predictions, conceptualized as categories, impact interoception. In this study we examine whether threatening categorical predictions have more impact on interoception. First, participants learned that inspiratory stimuli belonged to a ‘low’ or ‘high’ intensity category. Afterwards, participants were asked to classify the resistances into the correct category in both a threatening and safe phase (within-subjects), and were asked to rate fear, intensity, and unpleasantness. In the threatening phase, participants were told they could get up to ten painful electrocutaneous stimuli following resistances belonging to a specific category (‘low’ or ‘high’, between-subjects) (threat-of-shock manipulation; Schmitz & Grillon, 2012). No electrocutaneous stimuli were administered in the safe phase. We hypothesized that threat influences (1) between-category discrimination, (2) the threshold to classify resistances in the ‘low’, and the ‘high’ category (i.e., subjective category border), and (3) perceived sensations (fear, intensity, and unpleasantness). Prior results indicating the impact of categorical predictions on perceived sensations were replicated (Petersen, Schroijen, Mölders, Zenker, & Van den Bergh, 2014). The threat manipulation was successful in increasing anxiety and fear, but did not change the categorical predictions, nor their impact on perceived sensations. An explanation could be that too little prediction error was created and therefore we need more learning trials, especially in healthy participants. Implications for future research are discussed.