The relationship between self-reported interoception and depression: A systematic review and multilevel meta-analysis

Background: Emerging research highlights interoceptive dysfunction as a potential mechanism underlying depression. Whereas physiological measures of interoception (e.g., cardiac accuracy tasks) have shown inconsistent associations with depression, subjective self-report measures may provide greater insight. However, no previous meta-analysis has synthesised evidence on self-reported interoception and depression. Methods: Following PRISMA guidelines, a systematic search of PubMed and Scopus identified 44 studies (N = 18,377) meeting our inclusion criteria. A three-level meta-analysis was conducted to account for nested effect sizes. Primary analyses focused on the Multidimensional Assessment of Interoceptive Awareness (MAIA), with secondary analyses including other self-report tools. Moderator analyses tested study design, depression and interoception measures, age, gender, and publication year. Study quality was evaluated using the AXIS tool. The review was preregistered on OSF (https://osf.io/vq62p). Results: Self-reported interoception was significantly negatively associated with depression (r = -0.17, 95% CI [-0.23, -0.11], p < .001), particularly in regulatory and metacognitive domains (attention regulation, not-worrying, self-regulation, trusting). Effect sizes were larger in between-group designs than in within-group designs. Depression measure moderated the strength of association, but age, gender, and publication year did not. Conclusions: Depression is reliably linked to self-reported interoceptive dysfunction, particularly in domains that involve the metacognitive evaluation and regulation of bodily states. Our findings reinforce the view of interoception as a multidimensional construct and identifies specific interoceptive capacities, such as attention regulation, self-regulation, and trust in bodily signals, as promising targets for psychological interventions aimed at reducing depression.