Machine learning identifies microbiome and clinical predictors of sustained weight loss following prolonged fasting.

Machine learning predicts 12-week weight-loss response after prolonged fasting using baseline microbiome and clinical data, with key predictors including Faecalibacterium and LDL cholesterol.

• Why it matters: Understanding individual variability in fasting outcomes can optimize personalized interventions for metabolic health and weight management, addressing a significant knowledge gap.
• What they did: The study involved 38 healthy adults undergoing a 5-day fasting protocol, analyzing changes in body composition, microbiome, and metabolites, and applying machine learning models validated across three independent cohorts.
• The result: Baseline microbiome and clinical features accurately forecast long-term weight response, enabling tailored fasting strategies and advancing personalized metabolic health treatments.