Emotional eating-consuming food in response to feelings rather than hunger-is a significant, frequently enough overlooked factor in the growing global obesity crisis. A new study unveiled at the 2026 Pacific Symposium on Biocomputing demonstrates a potential leap forward in identifying individuals predisposed to this behavior, moving beyond traditional, time-consuming psychological assessments. Researchers have developed an artificial intelligence model capable of estimating emotional eating risk based on a person’s genetic makeup, opening the door to more targeted and effective interventions.
Emotional eating, a complex pattern of food intake driven by feelings rather than physiological hunger, plays a significant role in the development and persistence of obesity. Identifying specific patterns of disordered eating, like emotional eating, is crucial for tailoring effective weight management strategies.
A new study presented at the 2026 Pacific Symposium on Biocomputing in Hawaii details a novel approach to identifying individuals prone to emotional eating using artificial intelligence and genetic information. Traditionally, clinicians have relied on lengthy questionnaires to assess this behavior, a process that can be time-consuming for both patients and providers.
Researchers have developed a predictive model that estimates the risk of emotional eating by analyzing a person’s genetic makeup. This could allow for targeted lifestyle interventions and potential pharmacological treatments without the need for extensive psychological evaluations. In the study, emotional eating was defined as a high score on a hospital anxiety and depression scale, assessed within a cohort of nearly 4,000 individuals participating in the Omni Exome study.
The research demonstrates that artificial intelligence can accurately pinpoint individuals at risk for eating driven by emotional factors, potentially streamlining the diagnostic process. By combining genetic risk scores with clinical and behavioral data, the model captures relevant biological mechanisms involved in appetite regulation and eating behavior. This approach could significantly improve how patients with obesity are categorized and treated, paving the way for more personalized interventions.
Genetic analysis involved genome-wide association studies, followed by the calculation of genetic risk scores based on variations near selected genes. Artificial intelligence algorithms assisted in feature selection, and the final models integrated genetic scores with demographic and anthropometric variables like age, sex, and height. The models’ performance was evaluated through accuracy analyses on independent groups, demonstrating an increased ability to predict emotional eating risk.
“Emotional eating is a real contributor to weight gain and is often overlooked or misunderstood in current obesity care. These preliminary findings reinforce the direction the field is heading: a precision approach to obesity that recognizes the biological, behavioral, and emotional components of the disease,” said Andres Acosta, MD, PhD, co-founder of Phenomix Sciences and an obesity expert at Mayo Clinic, Minnesota, in an article published in Medscape.
The study found that a simplified model, based on a limited number of genes, could accurately predict the risk of emotional eating. Genes playing a major role included HTR2A and TPH2, involved in serotonin synthesis and signaling, as well as DRD2 and ANKK1, essential components of the dopamine reward system. This suggests a consistent neurobiological basis for emotional eating behaviors.
Responses from the Three Factor Eating Questionnaire also added predictive value, particularly elements related to feelings of loneliness and emotional control. Integrating this data allowed for better separation of high-risk individuals, with the genetic risk score demonstrating a stronger linear relationship with emotional eating behavior compared to traditional demographic factors.
From a clinical perspective, the findings suggest that a genetic risk score for emotional eating expands the capabilities of standard behavioral assessments. This tool could help researchers test the relationship between behavioral phenotypes and treatment response in observational studies, and prioritize therapeutic strategies based on dominant mechanisms, such as hunger driven by the gut-brain axis or reward circuits.
In conclusion, using AI-assisted genetic risk scores offers a new way to identify those at risk for emotional eating and personalize obesity management interventions. This strategy supports the transition to precision medicine, where treatments are tailored based on individual biological and behavioral mechanisms, with the potential to significantly improve long-term clinical outcomes.
Material drafted with the assistance of AI, verified and edited by the Raportul de Garda team, comprised of medical specialists.
Citește și:
