The development and validation of the Addiction-like Eating Behaviour Scale.

BACKGROUND: Overeating and obesity are frequently attributed to an addiction to food. However, there is currently a lack of evidence to support the idea that certain foods contain any specific addictive substance. An alternative approach is to focus on dimensions of observable behaviour, which may underpin a behavioural addiction to eating. To facilitate this, it is necessary to develop a tool to quantify addiction-like eating behaviour, which is not based on the clinical criteria for substance dependence. The current study provides initial validation of the Addiction-like Eating Behaviour Scale (AEBS).

METHODS: English speaking male and female participants (N=511) from a community sample completed the AEBS, alongside a range of other health- and eating-related questionnaires including the Yale Food Addiction Scale (YFAS) and Binge Eating Scale (BES). Participants also provided their height and weight to enable calculation of body mass index (BMI). Finally, to assess test-retest reliability, an additional 70 participants completed the AEBS twice, 2 weeks apart.

RESULTS: Principle components analysis revealed that a two-factor structure best accounted for the data. Factor 1 consisted of items that referred to appetitive drive, whereas factor two consisted of items that referred to dietary control practices. Both subscales demonstrated good internal reliability and test-retest reliability, and a confirmatory factor analysis confirmed the two-factor scale structure. AEBS scores correlated positively with body mass index (BMI) (P<0.001) and other self-report measures of overeating. Importantly, the AEBS significantly predicted variance in BMI above that accounted for by both the YFAS and BES (P=0.027).

CONCLUSIONS: The AEBS provides a valid and reliable tool to quantify the behavioural features of a potential 'eating addiction'. In doing so, the AEBS overcomes many limitations associated with applying substance-dependence criteria to eating.