Comprehensive analysis of gut microbiota of a healthy population and covariates affecting microbial variation in two large Japanese cohorts

Abstract

Background: Inter-individual variations in gut microbiota composition are observed even among healthy populations. The gut microbiota may exhibit a unique composition depending on the country of origin and race of individuals. To comprehensively understand the link between healthy gut microbiota and host state, it is beneficial to conduct large-scale cohort studies. The aim of the present study was to elucidate the integrated and non-redundant factors associated with gut microbiota composition within the Japanese population by 16S rRNA sequencing of fecal samples and questionnaire-based covariate analysis.

Results: A total of 1596 healthy Japanese individuals participated in this study via two independent cohorts, NIBIOHN cohort (n = 954) and MORINAGA cohort (n = 642). Gut microbiota composition was described and the interaction of these microorganisms with metadata parameters such as anthropometric measurements, bowel habits, medical history, and lifestyle were obtained. Thirteen genera, including Alistipes, Anaerostipes, Bacteroides, Bifidobacterium, Blautia, Eubacterium halli group, Faecalibacterium, Fusicatenibacter, Lachnoclostridium, Parabacteroides, Prevotella_9, Roseburia, and Subdoligranulum were predominant among the two cohorts. On the basis of univariate analysis for overall microbiome variation, 18 matching variables exhibited significant association in both cohorts. A stepwise redundancy analysis revealed that there were four common covariates, Bristol Stool Scale (BSS) scores, gender, age, and defecation frequency, displaying non-redundant association with gut microbial variance.

Conclusions: We conducted a comprehensive analysis of gut microbiota in healthy Japanese individuals, based on two independent cohorts, and obtained reliable evidence that questionnaire-based covariates such as frequency of bowel movement and specific dietary habit affects the microbial composition of the gut. To our knowledge, this was the first study to investigate integrated and non-redundant factors associated with gut microbiota among Japanese populations.

Keywords: 16S rRNA; Covariates; Gut microbiota; Japanese population; Large cohort.

Fig. 1

Gut microbiota distribution of the NIBIOHN cohort and MORINAGA cohort (a) Microbial community variation in each cohort represented by principal coordinates analysis (PCoA, genus-level JSD) and PAM clustering. Arrows indicated enterotype drivers; Bacteroides-enterotype (green), Prevotella-enterotype (blue), and Faecalibacterium-enterotype or Bifidobacterium-enterotype (red). b The estimated result of suitability of cluster number Calinski-Harabasz index (CH index) (c) The distribution of integrated data of two cohorts (d) The distribution of alpha diversity indices of NIBIOHN cohort (yellow) and MORINAGA cohort (blue) (e) The dominant genus among the two cohorts and their composition. **p<0.01, *p<0.05 (Wilcoxon rank sum test)

Fig. 2

Effect size of microbiome covariates (a) Effect size identified in the NIBOHN cohort (left) and the MORINAGA cohort (right). Factors are sorted according to their effect size and colored based on metadata category (Table S1) (b) PCoA-based on Bray-Curtis distance. Arrows show the ordination of 18 common covariates for overall microbiome community variation in the NIBIOHN cohort (top) and the MORINAGA cohort (bottom) (c) Cumulative effect size of non-redundant covariates. Microbial covariates selected by stepwise redundancy analysis in the NIBIOHN cohort (left) and the MORINAGA cohort (right). Common covariates in the two cohorts

Fig. 3

Correlation matrix heatmap between dominant bacteria and non-redundant covariates selected by stepwise redundancy analysis, as calculated by Spearmans rank correlation coefficients, in NIBIOHN cohort (top) and MORINAGA cohort (bottom)