From infancy and diet to GLP-1 therapies and AI, researchers are piecing together how shifting microbial ecosystems and their chemical signals intersect with weight regulation throughout life.
Review: Microbiome and Metabolomics in Obesity: Advances in Understanding and Interventions Across the Lifespan. Image Credit: Vicente Fernandez Garcia / Shutterstock
In a recent review published in the journal Obesity Reviews, researchers synthesized the latest multi-omics evidence exploring the reciprocal interactions between the human gut microbiome and host energy regulation in obesity.
The review mapped physiologically relevant mechanistic links across the human lifespan (infancy to old age). Additionally, it assessed the potential effects of incretin-based therapies on gut ecology and surveyed artificial intelligence (AI) integration frameworks for multi-omics analysis in obesity research.
Review findings emphasized that obesity is increasingly understood to involve complex, bidirectional host-microbiome pathways rather than isolated dysbiotic states. Microbial metabolites can form an extended endocrine network that influences host appetite and insulin sensitivity and modulates pathways involved in low-grade inflammation.
The authors concluded that, when combined with AI-driven multi-omics integration, functional pathway profiling could help advance future precision nutrition and targeted pharmacotherapies, although standardized methods, longitudinal cohorts, and causal studies will be needed for clinical translation.
Background
In 2022, public health records estimated that 890 million adults were living with obesity, while roughly one in eight people worldwide met clinical criteria for the disease. These estimates underscored the growing global burden of obesity and highlighted the need for improving clinical understanding of the disease and its mechanistic underpinnings.
Obesity was historically considered a condition caused by simple energy imbalances. However, a landmark germ-free mouse experiment provided preclinical evidence that gut microbiota can influence adiposity. When fecal microbiota from human twin pairs discordant for obesity were transferred into germ-free mice, animals receiving microbiota from obese donors gained more weight and accumulated greater fat mass than those colonized with microbiota from their lean co-twins. The experiment implicated microbe-derived functions in energy harvest and fat storage, emphasizing the microbiome's importance as a research target for obesity-centric interventions.
Subsequent biochemical research has elucidated that the gut microbiota uses small-molecule metabolites to communicate with enteroendocrine and gut-brain circuits, thereby modulating the host's appetite and insulin sensitivity. Despite evidence for this bidirectional communication network, reviews indicate that obesity research often examines microbial taxonomy and metabolic shifts in isolation.
About the review
The present review aimed to address this knowledge gap and inform future personalized obesity interventions by consolidating current mechanistic, lifespan, and pharmacological evidence within the framework of the European BETTER4U consortium.
The review's evidence base comprised published microbiome and multi-omics studies spanning four developmental life stages: 1. Infancy and early life, 2. Childhood and adolescence, 3. Adulthood, and 4. Older age. The review did not generate or analyze a new participant dataset.
The infancy section focused on the impacts of delivery modes and human milk oligosaccharides (HMOs). Similarly, the childhood and adulthood sections examined the roles of pubertal endocrine shifts and lifestyle behaviors, respectively. Finally, the older-age section considered age-related microbiome and metabolic changes associated with obesity, insulin resistance, and cardiometabolic risk, and also considered sarcopenia within the broader lifespan framework.
The authors also examined the physiological effects of incretin-based pharmacotherapies (e.g., semaglutide and tirzepatide) on the gut microbiome and metabolomic profiles in obesity, and the potential utility of advanced computational strategies for multi-omics integration using machine learning and deep learning approaches.
Review findings
Evidence synthesized in the review indicated that diet-microbiome interactions can influence host metabolism via enteroendocrine cells (EECs) and vagal afferents projecting to the hypothalamus and brainstem.
Mechanistically, the microbial fermentation of dietary fiber generates SCFAs, which in turn bind to free fatty acid receptors (FFAR2/3) on EECs. This binding then stimulates the EECs to produce appetite-suppressing hormones GLP-1 and peptide YY (PYY). Concurrently, secondary bile acids can activate TGR5 receptors, thereby influencing GLP-1 secretion and glycemic control.
The paper also reported that Westernized dietary patterns are associated with fewer beneficial fermenters, altered bile acid signaling, greater endotoxin load, and amino acid fermentation products implicated in insulin resistance. Separately, LPS from Gram-negative bacteria can activate toll-like receptor-mediated inflammatory pathways, while pro-inflammatory stimuli may also influence the hypothalamic-pituitary-adrenal axis. Together, these mechanisms provide plausible links between diet, microbiome function, metabolic inflammation, and obesity.
Lifespan evaluations of infant data revealed that cesarean delivery is associated with reduced vertical (mother-to-offspring) transmission of Bifidobacterium and Bacteroides, along with microbial and metabolic changes that may have consequences for later obesity risk. Breastfeeding, however, provides bioactive HMOs that are selectively utilized by beneficial gut bacteria, particularly Bifidobacterium, supporting immune maturation and gut barrier integrity.
Obesity throughout childhood and adulthood has frequently been associated with the depletion of Akkermansia muciniphila and butyrate-producing taxa (such as Faecalibacterium prausnitzii and Roseburia). The review further concluded that the historical Firmicutes/Bacteroidetes ratio is not a robust biomarker in humans because of inconsistent findings, methodological sensitivity, and functional heterogeneity within bacterial phyla. Differences in sampling, sequencing methods, diet, medication use, geography, and population characteristics also continue to limit reproducibility across microbiome and metabolomics studies.
Finally, preclinical models evaluating GLP-1 receptor agonists have reported microbiome changes involving Akkermansia and SCFA producers. However, human causal evidence remains limited, and it is unclear how much microbiome remodeling reflects direct drug effects rather than reduced food intake, weight loss, or metabolic improvement. In one human metabolomics study comparing liraglutide with bariatric surgery, most initially observed metabolite changes lost statistical significance after adjustment for weight loss, suggesting that much of the metabolic remodeling may reflect weight reduction rather than a treatment-specific effect.
Conclusions
This review emphasizes that obesity involves a reciprocal host-microbiome-metabolome system and identifies diet as one of the most tractable ways to influence this axis. It supports a shift from conventional descriptive taxonomic counts to functional pathway profiling, with Mediterranean-style, fiber-rich, and fermented-food dietary patterns highlighted as particularly relevant to microbiome function and metabolic health.
The authors posit that while incretin therapies offer potent metabolic control, translation into personalized care will require harmonized methods, diverse longitudinal cohorts, causal study designs, and explainable AI models to distinguish direct treatment effects from changes associated with weight loss and to identify clinically useful responder profiles.
Gut bacteria may send different biological signals in Parkinson’s disease