Item

Abstract

Rising obesity prevalence urges the need to understand unhealthy food decisions and potential modifiers. We aim to identify how psychological factors such as wanting and hunger modify food memory, and how prefrontal-temporal fiber connections contribute to food memory processing.
The study was designed as a randomized controlled cross-over intervention in 60 healthy, naive-eating, overweight adults (20f). Data for this pre-registered cross-sectional analysis (osf.io/2z4dn) was drawn from all time points of the larger study (intervention as confounder of no interest). Each testing day included 3T brain MRI with two tasks and diffusion-weighted imaging (DWI). Firstly, food vs. non-food wanting was assessed followed by a recognition memory task incl. lure discrimination. After quality assessment, 57 participants (181 data sets) entered behavioral analyses and 55 (176 data sets) were suitable for DWI analyses. We conducted advanced fiber tracking to assess the microstructure of the uncinate fasciculus (UF). Normalized quantitative anisotropy (nQA) served as outcome measure. We performed the statistical analysis with Bayesian mixed modeling.
At the behavioral level, recognition (d’) and lure discrimination (LDI) was better for food than non-food. Stronger single item wanting predicted higher response correctness. However, better food memory was not explained by individual hunger or categorized wanting. Neither nQA of the UF nor age nor body composition determined memory performance.
Individual food wanting predicted food memory in this sample of young, overweight adults. This might help to improve weight-loss interventions. To better understand reward-related determinants of food memory, we currently investigate reward-related brain activity during encoding and recognition.