Abstract
Anxiety is one of the most common and debilitating mental health disorders, and is related to changes in interoception (perception of bodily states). While anxiety is more prevalent in women than men, gender differences in interoception-anxiety associations are often overlooked. Here, we examined gender-specific relationships between anxiety and interoception in the breathing domain, utilising multicentre data pooled from four study sites (N = 175; 51% women). State anxiety scores were quantified via the Spielberger State-Trait Anxiety Inventory, and breathing-related interoceptive dimensions via an inspiratory load task to quantify sensitivity, decision bias, metacognitive bias (confidence in interoceptive decisions), and metacognitive insight (congruency between performance and confidence). Regression analyses revealed a significant negative relationship between state anxiety and metacognitive bias (β = -0.28; p = 0.01) and insight (β = -0.09; 95% highest density interval [HDI] in a hierarchical Bayesian regression = [-0.18, -0.004]) across the whole sample, while state anxiety did not relate to interoceptive sensitivity nor decision bias. While no mean interoceptive effects relating to gender were observed, the relationship between anxiety and metacognitive insight towards breathing was driven by women (women: β = -0.18; HDI = [-0.31, -0.05]; men: β = 0.02; HDI = [-0.12, 0.15]) with a significant interaction effect (β difference = -0.20; HDI = [-0.37, -0.01]), which did not hold for trait anxiety nor depression measures. In summary, state anxiety was associated with decreased metacognitive bias across all participants, while decreased interoceptive insight was only associated with anxiety in women but not men. Therefore, treatment programmes focusing on interoceptive metacognitive bias may be useful for all anxiety patients, while interoceptive insight might represent a specific treatment target for women with anxiety.