Alterations in Gut Microbiota Are Correlated With Serum Metabolites in Patients With Insomnia Disorder

Abstract

This study aimed to investigate insomnia-related alterations in gut microbiota and their association with serum metabolites. A total of 24 patients with insomnia disorder and 22 healthy controls were recruited. The fecal and serum samples were collected. The 16s rRNA sequencing and bioinformatics analysis were conducted to explore insomnia-related changes in the diversity, structure, and composition of the gut microbiota. UPLC-MS was performed to identify insomnia-related serum metabolites. Spearman correlation analysis was used to investigate the correlations between insomnia-related gut bacteria and the serum metabolites. Despite the nonsignificant changes in the diversity and structure of gut microbiota, insomnia disorder patients had significantly decreased family Bacteroidaceae, family Ruminococcaceae, and genus Bacteroides, along with significantly increased family Prevotellaceae and genus Prevotella, compared with healthy controls. Genus Gemmiger and genus Fusicatenibacter were dominant in patients with insomnia disorder, whereas genus Coprococcus, genus Oscillibacter, genus Clostridium XI, and family Peptostreptococcaceae were dominant in healthy controls. The UPLC-MS analysis identified 97 significantly decreased metabolites and 74 significantly increased metabolites in the serum samples of patients with insomnia disorder, compared with those of healthy controls. KEGG enrichment analysis revealed 1 significantly upregulated metabolic pathway and 16 downregulated metabolic pathways in patients with insomnia disorder. Furthermore, Spearman correlation analysis unveiled significant correlations among the altered bacteria genus and serum metabolites. Patients with insomnia disorder have differential gut microbiota and serum metabolic profiles compared with healthy controls. The alterations in gut microbiota were correlated with specific serum metabolites, suggesting that some serum metabolites might mediate gut microbiota-brain communication in the pathogenesis of insomnia disorder.

Keywords: gut microbiota; insomnia disorder; metabolic profile; microbiota-brain communication; serum metabolites.

Figure 1

Patients with insomnia showed differential gut microbial profiles compared with healthy controls. (A) α-diversity of Shannon index between insomnia and healthy control groups. *p < 0.05. (B) Plots of weighted and unweighted UniFrac PCoA of all samples. Orange dots: healthy controls; blue dots: insomnia patients. (C–F) Bar graphs of relative abundance of gut bacteria in healthy controls (A) and insomnia patients (B) at the levels of phylum, order, family, and genus. (G) Bar graphs of LEfSe analysis. Orange and blue bars represent the impact degree of enrichment of certain taxa in the healthy controls and insomnia patients, respectively. The LEfSe score threshold was 3 or -3. PCoA, principal coordinates analysis; LEfSe, linear-discriminant-analysis-effect-size; LDA, linear discriminant analysis; HC, Healthy control.

Figure 2

Patients with insomnia showed differential serum metabolic profiles compared with healthy controls. (A) A heatmap of metabolites with significantly altered concentrations between insomnia and healthy control groups. Red and blue represent increased and decreased metabolites, respectively, in patients with insomnia compared with those in healthy controls. (B) A dot plot of metabolites in the serum samples of the participants. Blue and red dots represent significantly decreased and increased metabolites in the serum samples of insomnia patients compared with those of healthy controls. Gray dots represent nonsignificantly changed metabolites.

Figure 3

Box plots (part I) of significantly changed metabolites identified in LC-MS analysis. Blue, patients with insomnia; Orange, healthy controls.

Figure 4

Box plots (part II) of significantly changed metabolites identified in LC-MS analysis. Blue, patients with insomnia; Orange, healthy controls.

Figure 5

Distinguishment between patients with insomnia and healthy controls using metabolic profiles. (A, B) Principal component analysis and orthogona projections to latent structures discriminant analysis (OPLS-DA) were performed to separate patients with insomnia and healthy controls using the metabolic profiles. Red: healthy controls. Blue: patients with insomnia. (C) The permutation test of OPLS-DA. (D) The metabolic pathway topology analysis. The horizontal axis indicates the −ln(p) values. The vertical axis indicates the impact values. PCA, principal component analysis; P1, 1st predictive principal component; O1, 1st orthogonal principal component.

Figure 6

Correlations between insomnia-related gut bacterial genus and serum metabolites. (A) A heat map of Spearman’s rank correlation coefficients and p-values of the correlation analysis. ⁺p < 0.05; *p < 0.01. (B) A bacteria-metabolite network. Orange dots represent the dominant gut bacteria in insomnia disorder. Blue dots represent the dominant gut bacteria in healthy control. The size of each dot reflects the impact value of each gut bacterial genus. The green dots represent metabolites.

Figure 7

Heat map of the Spearman correlation analysis between gut microbes and sleep/mood index. *P < 0.05; **P < 0.01. PSQI, Pittsburgh Sleep Quality Index; ISI, Insomnia Severity Index; HAMA, Hamilton Anxiety Rating Scale; HAMD, Hamilton Depression Rating Scale; SHAPS, Snaith-Hamilton Pleasure Scale.

Figure 8

Heat map of the Spearman correlation analysis between serum metabolites and sleep/mood index. *P < 0.05; **P < 0.01. PSQI, Pittsburgh Sleep Quality Index; ISI, Insomnia Severity Index; HAMA, Hamilton Anxiety Rating Scale; HAMD, Hamilton Depression Rating Scale; SHAPS, Snaith-Hamilton Pleasure Scale.