Interoceptive individual differences have garnered interest because of their relationship with mental health. One type of individual difference that has received little attention is variability in the sensation/s that are understood to mean a particular interoceptive state, something that may be especially relevant for hunger. We examined if interoceptive hunger is multidimensional and idiosyncratic, if it is reliable, and if it is linked to dysfunctional eating and beliefs about the causes of hunger. Participants completed a survey just before a main meal, with most retested around 1 month later. We found that interoceptive hunger has 11 dimensions, and while people differ considerably in their combinations of interoceptive hungers, these represent only 4% of all possible permutations. Hunger reports were reliable. We found relationships between variability in hunger interoception and dysfunctional eating, especially for uncontrolled eating. We also found that hunger beliefs were in some cases strongly related to aspects of hunger interoception. The implications of these findings are discussed.

Hunger is often reported when people experience certain internal sensations (e.g., fatigue) or when they anticipate that a food will be good to eat. The latter results from associative learning, while the former was thought to signal an energy deficit. However, energy-deficit models of hunger are not well supported, so if interoceptive hungers are not "fuel gauges," what are they? We examined an alternate perspective, where internal states signaling hunger, which are quite diverse, are learned during childhood. A basic prediction from this idea is offspring-caregiver similarity, which should be evident if caregivers teach their child the meaning of internal hunger cues. We tested 111 university student offspring-primary caregiver pairs, by having them complete a survey about their internal hunger states, alongside other information that may moderate this relationship (i.e., gender, body mass index, eating attitudes, and beliefs about hunger). We observed substantial similarity between offspring-caregiver pairs (Cohen's ds from 0.33 to 1.55), with the main moderator being beliefs about an energyneeds model of hunger, which tended to increase similarity. We discuss whether these findings may also reflect heritable influences, the form that any learning might take, and the implications for child feeding practices.