Obesity Abstracts

Background: Contrary to the idea that compensatory increases in food reward occur after weight loss (WL), a systematic review has shown that liking (L) and implicit wanting (W) decrease after WL interventions. However, there is a large individual variability in WL-induced changes in food reward, with potential implications for weight regain. The aim of this multivariate analysis was to summarize the changes in L&W during WL at the individual level.

Method: 30 women with overweight and obesity were assigned to either a continuous (25% energy restriction) or intermittent energy restriction (ad-libitum day alternating with 75% energy restriction day) until ³5% WL within 12 weeks. Before an ad-libitum meal, at week 0 and after WL had been achieved, the Leeds Food Preference Questionnaire (LFPQ) was used to assesses L&W according to 4 categories of food (high-fat savoury, low-fat savoury, high-fat sweet and low-fat sweet). A principal component analysis (PCA) and cluster analysis were performed on the changes in 8 variables of food reward from the LFPQ.

Results: L (e.g. mean change in HFSW=−7.26, P=0.029) but not W decreased after a WL of −6.40% (P<2e-16). PCA revealed that changes in W were distinct from L as were responses to sweet versus savoury but not high-fat versus low-fat food categories. Cluster analysis identified three main patterns of WL-induced changes in food reward: (1) increase in L&W; (2) decrease in sweet and increase in savoury categories and vice versa; (3) decrease in L&W.

Conclusion: There is no ‘one size fits all’ effect of WL on food reward, and some individuals may show increases, decreases, or both, in their food reward depending on sweet or savoury taste. Liking and implicit wanting are methodologically and statistically distinct and individuals may respond differently according to these dimensions. These exploratory results should be replicated to investigate whether it could contribute in explaining why some people regain weight after WL.

Keywords: Food reward, weight loss, individual-level

Disclosures: None