Differences in bacterial taxa between treatment-naive patients with major depressive disorder and non-affected controls may be related to a proinflammatory profile

Abstract

Background: Major depressive disorder (MDD) is characterized by sadness and anhedonia, but also physical symptoms such as changes in appetite and weight. Gut microbiota has been hypothesized to be involved in MDD through gut-brain axis signaling. Moreover, antidepressants display antibacterial properties in the gastrointestinal tract. The aim of this study was to compare the gut microbiota and systemic inflammatory profile of young patients with MDD before and after initiation of antidepressant treatment and/or psychotherapy in comparison with a non-depressed control group (nonMDD).

Methods: Fecal and blood samples were collected at baseline and at follow-up after four and twelve weeks, respectively. Patients started treatment immediately after collection of the baseline samples. The gut microbiota was characterized by 16 S rRNA gene sequencing targeting the hypervariable V4 region. Plasma levels of 49 unique immune markers were assessed using Mesoscale.

Results: In total, 27 MDD patients and 32 nonMDD controls were included in the study. The gut microbiota in the baseline samples of MDD versus nonMDD participants did not differ regarding α- or β-diversity. However, there was a higher relative abundance of the genera Ruminococcus gnavus group, and a lower relative abundance of the genera Desulfovibrio, Tyzzerella, Megamonas, Olsenella, Gordonibacter, Allisonella and Rothia in the MDD group compared to the nonMDD group. In the MDD group, there was an increase in the genera Rothia, Desulfovibrio, Gordinobacteer and Lactobacillus, while genera belonging to the Firmicutes phylum were found depleted at twelve weeks follow-up compared to baseline. In the MDD group, IL-7, IL-8 and IL-17b levels were elevated compared to the nonMDD group at baseline. Furthermore, MDI score in the MDD group was found to correlate with Bray-Curtis dissimilarity at baseline, and several inflammatory markers at both baseline and after initiation of antidepressant treatment.

Conclusion: Several bacterial taxa differed between the MDD group and the nonMDD group at baseline and changed in relative abundance during antidepressant treatment and/or psychotherapy. The MDD group was furthermore found to have a pro-inflammatory profile compared to the nonMDD group at baseline. Further studies are required to investigate the gut microbiota and pro-inflammatory profile of patients with MDD.

Keywords: Depression; Gut-Brain Axis; Major depressive disorder; Microbiome; Microbiota.

Fig. 1

CONSORT flow diagram displaying the process of recruitment and adherence to the study

Fig. 2

α-diversity in untreated patients with MDD and compared to healthy individuals (nonMDD). Number of observed amplicon sequence variants (ASVs) (A, D, G), Faith’s phylogenetic diversity (B, E, H) and Shannon diversity index (C, F, I) as compared between MDD and nonMDD (A, B, C), in-between samples of MDD collected over time (D, E, F) and in-between samples of nonMDD collected over time (G, H, I)

Fig. 3

Gut microbiota β-diversity. Bray-Curtis dissimilarity (A, D, G), weighted UniFrac (B, E, H) and unweighted UniFrac (C, F, I) as compared between MDD and nonMDD (A, B, C), within the MDD group with samples collected at baseline and at twelve weeks (D, E, F) and within the nonMDD group with samples collected at baseline and at twelve weeks (G, H, I)

Fig. 4

ANCOM-BC analysis of individual bacterial taxa changes. Bar plots display the log fold change in relative bacterial taxa compared between the MDD group versus the nonMDD group at baseline (A), between the MDD group at baseline versus 12 weeks follow-up (B), and between the nonMDD group at baseline versus 12 weeks follow-up (C)

Fig. 5

Plasma inflammatory markers significantly different between the MDD and nonMDD groups. The red dashed line represents the log2 signal intensity of the buffer control. Measurements with a signal below this cutoff line were included in the statistical analyses

Fig. 6

Correlation matrix of microbiota markers and LBP compared to plasma immune markers. Correlations were tested using Spearman’s correlation. Red boxes indicate positive correlations (>0), while blue boxes indicate negative correlations (<0). Blank boxes indicate that correlations did not reach significance

Fig. 7

Correlation analysis between MDI score, gut microbiota measures and inflammatory biomarkers. Correlation was tested using Spearman correlation at baseline and after 12 weeks. Only those correlations that reached significance are depicted. The red dashes represent the log2 signal intensity of the buffer control. Measurements with a signal below this cutoff line were included in the statistical analyses