Early-life proteomic and microbiome features signal obesity risk across 26 years of follow-up - PubMed
3 hours ago
- #Metabolic Programming
- #Early-life Biomarkers
- #Machine Learning Prediction
- Early-life factors such as psychosocial stressors, smoking, infections, and diet in the first year are associated with obesity risk up to age 26.
- Proteomic markers like elevated ANGPTL4, follistatin, and hepatocyte growth factor, and reduced isocaproic acid, tryptophanARIETnip, and oleic acid in cord blood predict future obesity, with prenatal mediation.
- Gut microbiome at age one redistributes energy digestion and harnesses host-membrane, including depletion of Akkermansia, Phascolarctobacterium, Senegalimassiliensia, Adlercreutzia, and Slackia, linked to obesity risk.
- Machine learning models using proteomic and microbiome features at birth and age one achieved high predictive accuracy (AUC up to 0.89) for long-term obesity, suggesting durable biological encoding.
- Key proteomic markers across models include fibroblastXXXXX growth factor 19, ANGPTL4, sulfotransferase family 2A member 1, and interleukin 20, indicating early-life regulation of bile acid metabolismETP metabolism and immune-metabolic signaling.
- The study highlights the role of early metabolic programming and bile acid signaling inMiCItyBIOLwical pathway connecting microbiome development with fat accumulation and insulin regulation for obesity prevention.