Neuroticism is one of the most robust risk factors for addictive behaviors including food addiction (a key contributor to obesity), although the associated mechanisms are not well understood. A transdiagnostic approach was used to identify the neuroticism-related neuropsychological and gut metabolomic patterns associated with food addiction. Predictive modeling of neuroticism was implemented using multimodal features (23 clinical, 13,531 resting-state functional connectivity (rsFC), 336 gut metabolites) in 114 high body mass index (BMI ≥25 kg/m2) (cross-sectional) participants. Gradient boosting machine and logistic regression models were used to evaluate classification performance for food addiction. Neuroticism was significantly associated with food addiction (P < 0.001). Neuroticism-related features predicted food addiction with high performance (89% accuracy). Multimodal models performed better than single-modal models in predicting food addiction. Transdiagnostic alterations corresponded to rsFC involved in the emotion regulation, reward, and cognitive control and self-monitoring networks, and the metabolite 3-(4-hydroxyphenyl) propionate, as well as anxiety symptoms. Neuroticism moderated the relationship between BMI and food addiction. Neuroticism drives neuropsychological and gut microbial signatures implicated in dopamine synthesis and inflammation, anxiety, and food addiction. Such transdiagnostic models are essential in identifying mechanisms underlying food addiction in obesity, as it can help develop multiprong interventions to improve symptoms.
Keywords: Brain-gut-microbiome; Food addiction; Machine learning; Neuroticism; Omics.
Copyright © 2023 The Authors. Published by Elsevier Inc. All rights reserved.