• rheumatoid arthritis
• disease activity
• inflammation

We thank Liu et al for their interest in our paper and for providing their thoughts through correspondence.1 We agree that gut microbiome studies are promising to identify novel insights into the sexual dimorphism in rheumatoid arthritis (RA). We reported that the genus of Gardnerella and the species of Gardnerella vaginalis increased in the gut microbiome of patients with RA.2 G. vaginalis is known as a representative causal bacterium of vaginosis. In our study, G. vaginalis was detected in both male and female samples, and there were no significant gender differences in their relative abundance (P=0.41). Thus, the increment of G. vaginalis was less likely to be contamination from the female genital organ in perineum. We further performed case-control association tests of G. vaginalis stratified by gender with age, sequencing groups and the top two principal components as covariates (figure 1A). We found that the effect size was larger in female samples (beta=1.14, P=4.1×10^-4) than in male samples (beta=0.596, P=0.22). This result suggests that the increment of Gardnerella in the gut microbiome of patients with RA is specific to the female samples.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Characteristics of the relative abundance of Gardnerella vaginalis in RA samples. (A) Boxplots of the relative abundance of G. vaginalis in RA and control samples. The y-axes indicate the relative abundance of G. vaginalis in a logarithmic scale. The left, centre and right boxplots are for all samples, only female samples and only male samples, respectively. The lower and upper hinges of the boxes indicate the first and third quartiles, respectively. The horizontal lines within the boxes indicate median levels. (B) Correlation of the relative abundance of G. vaginalis with that of Prevotella spp. The x-axes of the left, centre and right figures indicate the relative abundance of the genus Prevotella, the total abundance of the five Prevotella species with significant RA-control discrepancy (i.e., P. denticola, P. marshii, P. disiens, P. corporis and P. amnii) and the relative abundance of P. amnii in a logarithmic scale, respectively. The y-axes indicate the relative abundance of G. vaginalis in a logarithmic scale. RA, rheumatoid arthritis.

Gardnerella has been detected not only in vaginitis but also in a variety of infections, such as hip arthritis and joint infections.3 However, their biological and pathological roles in gut microbiome have been elusive. G. vaginalis was reported to have a symbiotic positive relationship with Prevotella bivia.4 Considering the association of Prevotella with RA aetiology,5–7 Liu et al proposed that G. vaginalis affects RA aetiology through the symbiotic proliferation of Prevotella. In our study, there was no significant positive correlation between the relative abundance of the genus Prevotella and that of G. vaginalis (r=0.054, P=0.47; figure 1B). However, when we focused on the total abundance of the five Prevotella species with significant RA-control discrepancy (i.e., P. denticola, P. marshii, P. disiens, P. corporis and P. amnii), significant positive correlation was found (r=0.20, P=0.024). Among the five RA-associated species, P. amnii had nominally significant positive correlation (r=0.19, P=0.035) with G. vaginalis, while the others did not (P>0.062). There was no significant correlation between P. bivia and G. vaginalis either (P=0.84). These results demonstrated that Gardnerella and RA-associated Prevotella had a symbiotic relationship in gut microbiome. Further studies are required to reveal the role of the G. vaginalis in the gut microbiome of patients with RA.

Liu et al also mentioned that female patients with RA confer more severe inflammation profiles (i.e., 28-joint Disease Activity Score (DAS28)) than male patients.8 Assessing whether Gardnerella is responsible for the sexual difference of RA severity, we performed association tests between the relative abundance of G. vaginalis and DAS28-CRP, but no significant association was found (P=0.91, r=−0.013). This result indicates that Gardnerella in gut microbiome may not be related to RA severity in female samples.

We hoped that researches focusing on G. vaginalis would lead to further clarification of aetiology of RA.