A predictive regression model of the obesity-related inflammatory status based on gut microbiota composition

Background and aim Fecal microbiome disturbances are linked to different human diseases. In the case of obesity, gut microbiota seems to play a role in the development of low-grade inflammation. The purpose of the present study was to identify specific bacterial families and genera associated with a...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:INTERNATIONAL JOURNAL OF OBESITY 2021-10, Vol.45 (10), p.2261-2268
Hauptverfasser: Aranaz, Paula, Ramos-Lopez, Omar, Cuevas-Sierra, Amanda, Martinez, J. Alfredo, Milagro, Fermin I., Riezu-Boj, Jose I.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Background and aim Fecal microbiome disturbances are linked to different human diseases. In the case of obesity, gut microbiota seems to play a role in the development of low-grade inflammation. The purpose of the present study was to identify specific bacterial families and genera associated with an increased obesity-related inflammatory status, which would allow to build a regression model for the prediction of the inflammatory status of obese and overweight subjects based on fecal microorganisms. Methods A total of 361 volunteers from the Obekit trial (65 normal-weight, 110 overweight, and 186 obese) were classified according to four variables: waist/hip ratio (≥0.86 for women and ≥1.00 for men), leptin/adiponectin ratio (LAR, ≥3.0 for women and ≥1.4 for men), and plasma C-reactive protein (≥2 mg/L) and TNF levels (≥0.85 pg/mL). An inflammation score was designed to classify individuals in low (those subjects who did exceed the threshold value in 0 or 1 variable) or high inflammatory index (those subjects who did exceed the threshold value in 2 or more variables). Fecal 16 S rRNA sequencing was performed for all participants, and differential abundance analyses for family and genera were performed using the MicrobiomeAnalyst web-based platform. Results Methanobacteriaceae, Christensenellaceae, Coriobacteriaceae, Bifidobacteriaceae, Catabacteriaceae, and Dehalobacteriaceae families, and Methanobrevibacter, Eggerthella, Gemmiger, Anaerostipes, and Collinsella genera were significantly overrepresented in subjects with low inflammatory index. Conversely, Carnobacteriaceae, Veillonellaceae, Pasteurellaceae, Prevotellaceae and Enterobacteriaceae families, and Granulicatella, Veillonella, Haemophilus, Dialister Parabacteroides, Prevotella, Shigella, and Allisonella genera were more abundant in subjects with a high inflammatory index. A regression model adjusted by BMI, sex, and age and including the families Coriobacteriaceae and Prevotellaceae and the genus Veillonella was developed. Conclusion A microbiota-based regression model was able to predict the obesity-related inflammatory status (area under the ROC curve = 0.8570 ± 0.0092 Harrell’s optimism-correction) and could be useful in the precision management of inflammobesity.
ISSN:0307-0565
1476-5497
DOI:10.1038/s41366-021-00904-4