Abstract / Summary
Aims: Interoception has been proposed as a relevant process in alcohol use disorder (AUD), and studies using the heartbeat counting task (HCT) have often reported poorer performance in individuals with AUD than in healthy controls (HC). However, concerns about the construct validity of the HCT complicate interpretation of these findings. This systematic review and meta-analysis quantified the magnitude and consistency of group differences in HCT performance between adults with AUD and HC, without assuming that the task provides a process-pure measure of cardiac interoceptive accuracy.
Methods: Eligible studies compared adults diagnosed with AUD with HC and reported sufficient HCT data to compute standardized mean differences. PubMed/MEDLINE, PsycINFO, Embase, and Web of Science were searched up to 4 November 2025. Study quality was assessed using an adapted Newcastle-Ottawa Scale. Hedges' g values were pooled using a random-effects model. The review received no specific external funding and was preregistered in PROSPERO [ID = CRD420251173904].
Results: Six studies including 314 individuals with AUD and 324 HC were included. The pooled effect was moderate-to-large and negative, indicating lower HCT performance in AUD than HC, g = -0.79, 95% CI [-0.96, -0.63]. All included studies showed negative effect sizes, although the small evidence base limits certainty.
Conclusions: Lower HCT performance appears to represent a preliminary but consistent task-level behavioral difference in AUD. These findings should not be interpreted as direct evidence of reduced cardiac interoceptive accuracy, but they support further investigation of whether HCT performance relates to clinically relevant AUD features.