Respiratory Interoception and Pathological Illness Anxiety: Disentangling Bias.

OBJECTIVE: Biased interoception decoupled from physiology might be relevant in the etiology of pathological illness anxiety (PIA). Empirical evidence for interoceptive deviations in illness anxiety is scarce but potentially informative to optimize treatments. We hypothesized that persons with PIA differ fundamentally in the classification of bodily sensations from those without PIA. METHODS: In a respiratory categorization task, participants breathed into a pulmonary training device. Inspiration effort was varied by eight resistive loads. The lower/higher four loads were introduced as belonging to arbitrary categories "A"/"B," respectively. Participants memorized respiratory sensations in a first experimental block and were asked to label the resistances in a second block. We calculated the sensitivity of resistance classification according to category and response bias in terms of categorical misclassification. Data of 39 participants with PIA and 35 controls were compared with regard to sensitivity and response bias by group, resistive load, and their interaction in a multiple regression. RESULTS: With similar sensitivity, patients more often labeled loads above the categorical border erroneously as belonging to category A, thus underestimating their resistance ( ß = -0.06, p = .001; η2 = 0.02). CONCLUSIONS: Individuals with PIA showed a systematic "wait and see" approach. Altered respiroception in PIA might stem from biased perception during training phase, the recognition phase, biased memory, or a combination of these. Its exact characteristics remain unknown, and future research must address the challenge of developing reliable and valid paradigms accounting for the variability of interoceptive biases. REGISTRATION: This work was preregistered on OSF ( https://osf.io/9shcw ).

Symptom Perception in Pathological Illness Anxiety: Tactile Sensitivity and Bias.

OBJECTIVE: Symptom perception in pathological illness anxiety (PIA) might be biased so that somatic signals are overreported. In the somatic signal detection task (SSDT), performance in detecting weak tactile stimuli gives information on overreporting or underreporting of stimuli. This task has not yet been applied in PIA. METHODS: Participants with PIA (n = 44) and healthy controls (n = 40) underwent two versions of the SSDT in randomized order. In the original version, tactile and auxiliary light-emitting diode (LED) stimuli were each presented in half of the trials. In the adapted version, illness or neutral words were presented alongside tactile stimuli. Participants also conducted a heartbeat mental tracking task. RESULTS: We found significantly higher sensitivity and a more liberal response bias in LED versus no-LED trials, but no significant differences between word types. An interaction effect showed a more pronounced increase of sensitivity from no LED to LED trials in participants with PIA when compared with the adapted SSDT and control group (F(1,76) = 5.34, p = .024, η2 = 0.066). Heartbeat perception scores did not differ between groups (BF01 of 3.63). CONCLUSIONS: The increase in sensitivity from no LED to LED trials in participants with PIA suggests stronger multisensory integration. Low sensitivity in the adapted SSDT indicates that attentional resources were exhausted by processing word stimuli. Word effects on response bias might have carried over to the original SSDT when the word version was presented first, compromising group effects regarding bias. TRIAL REGISTRATION: The study was preregistered on OSF (https://osf.io/sna5v/).