Gut microbiota and metabolic health among overweight and obese individuals

Obesity prevalence and related diseases are increasing globally, but cardiometabolic risk among individuals with obesity varies according to metabolic health. A subset, termed metabolically healthy obesity (MHO), displays lower systemic inflammation and favorable immune and liver profiles, though many may transition to metabolically unhealthy obesity (MUO) over time. Prior studies link gut microbiota to obesity and metabolic diseases, with obesity often associated with reduced microbial diversity and altered composition. Despite evidence connecting microbiota to type 2 diabetes, hypertension, and dyslipidemia, no prior work specifically compared gut microbiomes between metabolically healthy and unhealthy overweight/obese individuals. The study’s objective was to evaluate gut microbiota diversity, composition, and predicted function by metabolic health status among overweight and obese adults to clarify microbial features associated with metabolic health within obesity.

MHO individuals often are younger, physically active, and have lower ectopic/visceral fat and inflammation, yet 33–47.6% transition to MUO over 5–10 years; nevertheless, MHO shows lower risks of CVD and metabolic disease than MUO, and longer metabolic syndrome duration associates linearly with CVD. Gut microbiota contribute to obesity and metabolic disease via immune and dietary interactions, with obesity generally showing reduced microbiome diversity and altered energy harvest compared with lean individuals. Microbiota have been implicated in obesity comorbidities, including type 2 diabetes, hypertension, and dyslipidemia, and modulation of microbiota can ameliorate metabolic syndrome components, suggesting specific microbial compositions underlie metabolic phenotypes.

Design and participants: Cross-sectional analysis within the Kangbuk Samsung Health Study. Of 1463 adults (23–78 years) providing fecal samples in June–September 2014, exclusions were applied for missing data (n=18), BMI<23 (n=616), recent antibiotics (6 weeks; n=55) or probiotics (4 weeks; n=19), CVD history (n=24), malignancy (n=52), or <2000 sequences (n=19), yielding 747 overweight/obese participants. Ethical approval was obtained (Kangbuk Samsung Hospital IRB 2013-01-245-12) with informed consent. Group definitions: Overweight and obese were defined by Asian-specific criteria (overweight BMI 23.0–24.9 kg/m²; obese BMI ≥30 kg/m²). Metabolic health status followed NCEP-ATP III criteria: metabolically unhealthy (MU) if ≥1 component present (fasting glucose ≥100 mg/dL or medication; BP ≥130/85 mmHg or medication; triglycerides ≥150 mg/dL; low HDL-C <40 mg/dL men/<50 mg/dL women; waist >102 cm men/>88 cm women). Metabolically healthy (MH) had none of these abnormalities. Data collection: Self-administered questionnaires captured medical history, medications, and behaviors; trained staff performed physical/ultrasound measures; biochemical assays were obtained from fasting blood. Microbiome profiling: Fecal samples were frozen at −20 °C immediately and stored at −70 °C within 24 h. DNA was extracted (MO BIO PowerSoil) within one month. V3–V4 16S rRNA gene regions were amplified with indexed fusion primers and sequenced on Illumina MiSeq. QIIME2 (v2019.7) with DADA2 performed quality control, chimera removal, and ASV inference (100% OTUs). Taxonomy was assigned using a pre-trained naive Bayes classifier against Greengenes 99% OTUs (13_8). Features present in only one sample were filtered out. Diversity analyses: Feature tables were rarefied to 2019 sequences/sample. Alpha diversity indices included observed ASVs, Shannon index, Pielou’s evenness, and Faith’s phylogenetic diversity; group differences tested with Kruskal–Wallis. Beta diversity used unweighted/weighted UniFrac and non-phylogenetic Bray–Curtis and Jaccard distances; group differences tested by pairwise PERMANOVA (999 permutations). Plots used ggplot2. Differential abundance: MaAsLin generalized linear models evaluated associations between taxa and metabolic status: Model 1 unadjusted, Model 2 adjusted for age, sex, and BMI; FDR correction applied (q<0.05 significant). LEfSe identified taxa discriminating groups (LDA>3, p<0.05). Functional inference: PICRUSt2 (v2.2.0-b) predicted EC gene families and MetaCyc pathways from ASVs via HMMER, EPA-NG, and GAPPA. STAMP visualized pathway differences (Welch’s t test; BH FDR q<0.05). Sex-stratified analyses and comparisons with a metabolically healthy non-obese control group (supplementary) were also conducted.

Participants: 747 overweight/obese adults; MH n=317 (42.4%), MU n=430 (57.6%). MU were older and had higher BMI and adverse metabolic indicators; smoking and nutritional intake did not differ significantly. Alpha diversity: MU showed significantly lower diversity than MH in observed ASVs (p=3.63×10⁻³), Faith’s PD (p≈1.96×10⁻⁴), and Shannon index (p=1.03×10⁻³); Pielou’s evenness was not different (p=0.51). Beta diversity: Significant group differences for unweighted UniFrac (PERMANOVA Pseudo-F=3.815, p=0.001), Bray–Curtis (Pseudo-F=1.603, p=0.004), and Jaccard (Pseudo-F=1.635, p=0.001); weighted UniFrac not significant (p=0.067). PCoA did not clearly separate groups due to interindividual variation. Controls and sex: MU had lower alpha diversity than metabolically healthy non-obese (MHN) and MH; no difference between MHN and MH. Females had higher alpha diversity than males; sex-stratified analyses confirmed lower diversity in MU. Differential taxa (adjusted for age, sex, BMI): Enriched in MU—Fusobacteria and lower taxa including Fusobacteriaceae and Fusobacterium (e.g., f_Fusobacteriaceae q=0.020; g_Fusobacterium q=0.024). Enriched in MH—Ruminococcaceae (q=0.034), Oscillospira (coef≈−0.022 for MU vs MH; q=0.015), Clostridium (q=0.021), Odoribacteraceae (q=0.045), Coriobacteriaceae (q=0.031), and Leuconostocaceae (q=0.048). Actinobacteria and Bifidobacteriaceae lost significance after adjustment. LEfSe corroborated enrichment of Fusobacteria in MU and Oscillospira/Clostridium/Ruminococcaceae in MH. Functional prediction (PICRUSt2): MU showed enrichment of vitamin and nucleotide biosynthesis pathways (e.g., cob(II)yrinate a,c-diamide biosynthesis I; preQ0 biosynthesis; 6-hydroxymethyl-dihydropterin diphosphate biosynthesis III; thiamin diphosphate superpathway; pyrimidine and purine pathways) (q<0.05). MH showed higher L-lysine biosynthesis and glycogen biosynthesis I (from ADP-D-Glucose). Overall: Metabolically unhealthy status in overweight/obesity is associated with reduced gut microbial richness and distinct community composition, with inflammation-associated Fusobacteria enriched in MU, and putative SCFA/butyrate-associated taxa (Oscillospira, Clostridium, Ruminococcaceae) and Coriobacteriaceae enriched in MH.

The study demonstrates that metabolic health among overweight/obese adults is reflected in gut microbiota diversity and composition. Lower alpha diversity and distinct beta diversity in MU align with prior links between reduced gut richness and adverse metabolic traits (higher TG, lower HDL-C). Enrichment of Oscillospira, Ruminococcaceae, and butyrate-producing Clostridium in MH supports a role for SCFAs in maintaining metabolic homeostasis, improving insulin sensitivity, and modulating inflammation. Higher Coriobacteriaceae in MH is consistent with potential benefits in bile acid, steroid, and lipid metabolism. Conversely, enrichment of Fusobacteria in MU aligns with inflammation-related dysbiosis observed in T2D and intestinal inflammation, suggesting a microbiota–inflammation axis contributing to metabolic unhealth independent of BMI. No significant difference in the Firmicutes/Bacteroidetes ratio between MH and MU underscores that broad phylum-level ratios may be less informative than specific taxa. Functional predictions, with increased nucleotide and certain vitamin biosynthesis in MU and higher lysine and glycogen biosynthesis in MH, are consistent with metabolic pathway alterations, though inference is limited by 16S-based predictions. Collectively, findings indicate that microbiome features may help differentiate metabolically healthy from unhealthy phenotypes within obesity and could inform strategies to preserve or restore metabolic health.

Metabolically healthy and unhealthy overweight/obese adults exhibit significant differences in gut microbial diversity, composition, and predicted functions. MU status is associated with reduced alpha diversity and enrichment of Fusobacteria, whereas MH status is associated with higher abundance of Oscillospira, Clostridium, Ruminococcaceae, Coriobacteriaceae, and Leuconostocaceae. These microbial patterns may play a role in maintaining metabolic health in obesity. Future longitudinal and multi-omic studies are needed to elucidate causal mechanisms and to evaluate microbiome-targeted interventions for preventing metabolic deterioration in obese populations.

Cross-sectional, single time point design precludes causal inference and does not capture transitions between MHO and MUO. Study population comprised Korean adults, limiting generalizability to other ethnicities and lifestyles. Microbiome profiling used 16S rRNA sequencing, limiting taxonomic resolution and providing only inferred functional profiles without direct metagenomic, transcriptomic, proteomic, or metabolomic validation.