No gender difference in cardiac interoceptive accuracy: Potential psychophysiological contributors in heartbeat counting task

Abstract

Gender differences in interoceptive awareness-awareness of internal bodily signals such as heartbeat perception-have been suggested, with some findings indicating behaviourally reduced but subjectively enhanced awareness in women, though these findings are still contentious. This study aimed to comprehensively examine gender differences in three aspects of interoceptive awareness: behavioural accuracy, subjective confidence, and relationship between them (i.e., metacognition). We used a modified heartbeat counting task that prohibited estimation strategies and increased the number of trials up to 20. Using data from 74 healthy young adults (39 women and 35 men), we evaluated gender differences and practice effects for each measure via Bayesian linear mixed models, controlling for individual heart rate and trial duration on a trial-by-trial basis. Contrary to previous research, the results revealed no reduced interoceptive accuracy in women; instead, higher interoceptive accuracy score was associated with shorter trial durations and lower heart rates regardless of gender. Moreover, women exhibited underconfidence about their performance, and therefore lower metacognition scores, compared to men. Trial repetitions moderated women's lowered metacognition but did not affect accuracy or confidence. These findings highlight potential physiological and psychological confounding factors in the heartbeat counting task, such as heart rate and reporting style, and emphasise several cautions for studying gender differences in interoceptive awareness.

Keywords: Gender difference; Heartbeat counting task; Interoceptive accuracy; Interoceptive awareness; Metacognition.

Fig. 1

Gender differences in interoceptive awareness. A) Parameter estimates for each factor of the Multidimensional Assessment of Interoceptive Awareness, contrasted between women and men, are displayed as a forest plot. Values greater than zero indicate higher scores in women compared to men, and vice versa. All 95% credible intervals estimated via Bayesian independent samples t-tests contain zero, suggesting no gender difference in interoceptive sensibility assessed by the questionnaire. B), C), and D) Individual scores for interoceptive accuracy, confidence, and metacognition, respectively, are plotted by gender and trial session. Each point represents an individual score within the session, with group means and 95% confidence intervals depicted

Fig. 2

Forest plots of gender and practice effects in the heartbeat counting task, estimated via Bayesian Linear Mixed Model analyses. Marginal means for four contrasts of interest were defined: gender differences (men vs. women) over four trial sessions (W < M); practice effects (latter two sessions vs. initial two sessions) in women (W), in men (M), and in both women and men (W&M). Each parameter was estimated by Markov Chain Monte Carlo sampling. The models included trial duration, heart rate, and the interaction between gender and session as fixed effects, with random intercepts and random slopes for each variable modeled as random effects. (A) All 95% credible intervals for contrasted estimates of marginal means in interoceptive accuracy contained zero, indicating no gender difference or practice effect in either men or women. (B) The 95% credible intervals for gender differences did not contain zero, suggesting general underconfidence in women. (C) The 95% credible intervals for gender differences, practice effects in women, and practice effects in the whole sample did not contain zero, though the lower limit of gender difference was close to zero (0.013)