Children’s gut microbiota predicts the efficacy of obesity treatment

Background & Objective Responses to dietary interventions may vary depending on baseline gut microbiota composition. This study aimed to determine whether baseline gut microbiota diversity and composition predict the effectiveness of childhood obesity interventions.Methods Anthropometry, triglycerides, HDL-cholesterol, HOMA-IR, and systolic and diastolic blood pressure (SBP, DBP) were evaluated and standardised in 41 children with obesity (8–14yrs). Faecal samples were collected at baseline and after one year. Intervention success was defined by improvements in metabolic risk score (MetScore) or BMI z-score. Associations between baseline microbiota features (diversity and composition) and intervention success were evaluated using Spearman’s correlation and linear regression models. Gut microbiota composition and differential abundance were analyzed using ANCOM-BC2. Exploratory biomarker discovery was analyzed using LEfSe, and predictive modelling using a Random Forest (RF) classifier. Receiver operating characteristic (ROC) curve analysis was used to determine a Simpson index cut-off.Results Higher baseline Shannon and Simpson indices, and greater abundances of Faecalibacterium and Eubacterium coprostanoligenes group, were associated with greater improvements in MetScore. Faecalibacterium was the most influential feature with the highest importance in the RF model, which achieved an AUC of 0.876 for MetScore and 0.873 for BMI z-score improvement. Eighty-four features differed between MetScore response groups (FDR