The association between obesity severity and food reward in adolescents with obesity: a one-stage individual participant data meta-analysis

Abstract

Background: Food reward and cue reactivity have been linked prospectively to problematic eating behaviours and excess weight gain in adults and children. However, evidence to date in support of an association between degree of adiposity and food reward is tenuous. A non-linear relationship between reward sensitivity and obesity degree has been previously proposed, suggesting a peak is reached in mild obesity and decreases in more severe obesity in a quadratic fashion.

Objective: To investigate and characterise in detail the relationship between obesity severity, body composition, and explicit and implicit food reward in adolescents with obesity.

Methods: Data from seven clinical trials in adolescents with obesity were aggregated and analysed in an independent participant data meta-analysis. Linear and curvilinear relationships between the degree of obesity and explicit and implicit reward for sweet and high fat foods were tested in fasted and fed states with BMI-z score as a continuous and discrete predictor using clinically recognised partitions.

Results: Although positive associations between obesity severity and preference for high-fat (i.e. energy dense) foods were observed when fasted, none reached significance in either analysis. Conversely, adiposity was reliably associated with lower reward for sweet, particularly when measured as implicit wanting (p = 0.012, ηp² = 0.06), independent of metabolic state. However, this significant association was only observed in the linear model. Fat distribution was consistently associated with explicit and implicit preference for high-fat foods.

Conclusions: A limited relationship was demonstrated between obesity severity and food reward in adolescents, although a lower preference for sweet could be a signal of severe obesity in a linear trend. Obesity is likely a heterogenous condition associated with multiple potential phenotypes, which metrics of body composition may help define.

Clinical trial registrations: NCT02925572: https://classic.

Clinicaltrials: gov/ct2/show/NCT02925572 . NCT03807609: https://classic.

Clinicaltrials: gov/ct2/show/NCT03807609 . NCT03742622: https://classic.

Clinicaltrials: gov/ct2/show/NCT03742622 . NCT03967782: https://classic.

Clinicaltrials: gov/ct2/show/NCT03967782 . NCT03968458: https://classic.

Clinicaltrials: gov/ct2/show/NCT03968458 . NCT04739189: https://classic.

Clinicaltrials: gov/ct2/show/NCT04739189 . NCT05365685: https://www.

Clinicaltrials: gov/study/NCT05365685?tab=history .

Keywords: Adolescent; Body composition; Food preference; Morbid obesity; Obesity; Reward.