Multi-Population Cohorts Meta-Analysis Of Gut Microbiota Associated With BMI

Human gut microbiota is a complex microbial community in the human body,and the total genetic information of microorganisms in the gut constitutes the human gut microbiome.More and more studies have shown that the homeostasis and changes of gut microbes were closely related to the health status of hosts.Obesity is a common disease worldwide,and its incidence is gradually increasing.Obesity is often accompanied by a variety of complications,such as diabetes,fatty liver,etc.The sharp increase of the obesity incidence has brought heavy health and economic burden to the society.With the development of high-throughput sequencing technology and the increasing number of gut microbiome sequencing data,many studies have begun to focus on exploring the relationship between gut microbiota and obesity.However,due to the limited sample size,differences in method design or inter-individual heterogeneity,the research results of independent studies are often inconsistent,which greatly increases the difficulty of translating the research results of gut microbiome markers into practical application to improve the current situation of obesity.This study utilized the strategy of comprehensive re-analysis and meta-analysis of multiple metagenomic datasets,aiming to improve the statistical efficiency of the study,enhance the credibility of the results,and to identify gut microbiota associated with BMI across populations.This study comprehensively re-analyzed the shotgun metagenomic sequencing data of 4255 healthy subjects in 14 studies worldwide,which were splitted into 16 data sets by nationality in the analysis.We explored the association of BMI with the diversity of intestinal flora,species co-abundance network,species abundance and functional pathways.Discriminant analysis and machine learning methods were also used to identify feature bacteria of obese individuals.We found that there was no difference in the dominant gut microbiota,genera,and species between the normal weight group(NW group,BMI<25kg/m~2)and the overweight group(OW,BMI ≥ 25kg/m~2).In addition,changes in the abundance ratios of Firmicutes to Bacteroidetes were only detected in three data sets,and the direction of changes was inconsistent among the three studies.Meta-analysis results also showed that the ratio of Firmicutes to Bacteroidetes was not different between the two BMI groups.As for the analysis of the micobiota diversity,we detected significant differences of the alpha diversity indices of Chao1,Simpson and/or Fisher(p<0.05)between the NW/OW groups in 4 data sets.There was statistically significant difference in beta diversity among the NW/OW groups in 5 datasets(p<0.05).Through metagenomic-wide association analysis(MWAS)with multivariable linear regression model in individual datasets and the following meta-analysis with random-effect model on 16 datasets,we found that the abundances of 26 bacterial species were associated with BMI(p.adjust<0.05),of which 2 were positively associated.Association analysis of the functional pathways with BMI and meta-analysis were also performed,and the results showed that 48 pathways were associated with BMI(p.adjust<0.05).Through LEf Se discriminant analysis,we identified 69 characteristic features(FDR<0.05,|LDA score|>2.0)of gut microbiota for distinguishing normal weight individuals and obese individuals(OB,BMI ≥ 30 kg/m~2);The prediction model consisting of 50 gut microbiota trained through machine learning using the LASSO＿ll model has a predictive power of AUC value of 0.772 for predicting obese individuals.Briefly,in this study,BMI was correlated with alpha and beta diversity of gut microbiota in multiple groups of study data,and alpha diversity decreased in high BMI group.Through metagenome-BMI association analysis and meta-analysis of multi-population cohorts,multiple gut microbiota and functional pathways such as s＿Parabacteroides＿distasonis,s＿Intestinimonas＿butyriciproducens and PWY-7357 were identified.Differences in the topology of gut microbial co-abundance network between the NW group and the OW group were identified.Through LEf Se analysis and LASSO＿ll machine learning training identification,s＿Roseburia＿faecis,s＿Bacteroides＿plebeius and other characteristic gut microbiota which can be used to indicate obesity were obtained.This study laid more basis for the development of gut microbiota-based interventions for obesity in the future.