Analysis of differentially expressed gut microbiota in short-stature children

WANG Wen; WANG Zhe; JIANG Xiulei; HU Yinghua; WANG Mingyue; MU Chunsun; WANG Chunxiang

doi:10.3969/j.issn.1002-266X.2024.32.006

Abstract：Objective To analyze the differences in gut microbiome expression between children with short stature and those with normal height,and to explore the relationship between changes in gut microbiome and short stature.Methods Sixty-six children with short stature were selected in the short stature group,and 56 healthy children were in-cluded in the control group.The diversity and abundance of gut microbiome were analyzed and compared between the two groups.Based on the differentially expressed gut microbiome between the two groups,a machine learning model was estab-lished using support vector machines,and the model's predictive performance for short stature was evaluated using receiver operating characteristic(ROC)curve and area under the curve(AUC).Results The proportions of the top 10 gut micro-biome at the phylum and genus levels were slightly different between the two groups,but the differences were not statistical-ly significant.The Shannon index and species richness index of gut microbiome in the short stature group were lower than those in the control group(both P<0.05).There were no statistically significant differences in Bray-Curtis or Jaccard dis-tances between the two groups.The relative abundances of Leptotrichiaceae,Paraprevotellaceae,Tannerellaceae,Eubac-teriumrectale,Agathobacter,Dietziaceae,Dietzia,and Faecalibacterium prausnitzii in the short stature group were higher than those in the control group,while the relative abundances of Helicobacte raceae,Clostridium sensu stricto,and Lach-nospiraceae NC2004 group were lower(all P<0.05).At the species level,cluster analysis of microbiome related to SDS revealed significant positive correlations among Bacteroides oleiciplenus,Alistipes,Actinomyces oris,and Ruminococ-cus.Based on the differentially expressed gut microbiome at the species level,an support vector machine prediction model was constructed.After parameter adjustment and optimization,the AUC of prediction model 1 for short stature was 0.856(95%CI:0.726 7-0.990 5);after further adjusting for other common factors that affected children's height(BMI,genetic height,premature birth,birth length,feeding method after birth,etc.),the AUC of prediction model 2 for short stature was 0.788(95%CI:0.636 1-0.939 7).In both models,the gut microbiome with greater influence on prediction included Prevotella histicola,Bacteroides oleiciplenus,Marvinbryantia,and Fusicatenibacter saccharivorans.Conclusions There are differences in gut microbiome expression between children with short stature and normal chil-dren.The prediction model based on differentially expressed gut microbiome shows good predictive performance for short stature.