Association between gut microbiota and pan-dermatological diseases: a bidirectional Mendelian randomization research

Abstract

Background: Gut microbiota has been associated with dermatological problems in earlier observational studies. However, it is unclear whether gut microbiota has a causal function in dermatological diseases.

Methods: Thirteen dermatological diseases were the subject of bidirectional Mendelian randomization (MR) research aimed at identifying potential causal links between gut microbiota and these diseases. Summary statistics for the Genome-Wide Association Study (GWAS) of gut microbiota and dermatological diseases were obtained from public datasets. With the goal of evaluating the causal estimates, five acknowledged MR approaches were utilized along with multiple testing corrections, with inverse variance weighted (IVW) regression serving as the main methodology. Regarding the taxa that were causally linked with dermatological diseases in the forward MR analysis, reverse MR was performed. A series of sensitivity analyses were conducted to test the robustness of the causal estimates.

Results: The combined results of the five MR methods and sensitivity analysis showed 94 suggestive and five significant causal relationships. In particular, the genus Eubacterium_fissicatena_group increased the risk of developing psoriasis vulgaris (odds ratio [OR] = 1.32, p_FDR = 4.36 × 10^-3), family Bacteroidaceae (OR = 2.25, p_FDR = 4.39 × 10^-3), genus Allisonella (OR = 1.42, p_FDR = 1.29 × 10^-2), and genus Bacteroides (OR = 2.25, p_FDR = 1.29 × 10^-2) increased the risk of developing acne; and the genus Intestinibacter increased the risk of urticaria (OR = 1.30, p_FDR = 9.13 × 10^-3). A reverse MR study revealed insufficient evidence for a significant causal relationship. In addition, there was no discernible horizontal pleiotropy or heterogeneity.

Conclusion: This study provides novel insights into the causality of gut microbiota in dermatological diseases and therapeutic or preventive paradigms for cutaneous conditions.

Keywords: dermatological diseases; gut microbiota; inference; mendelian randomization analysis; therapeutics.

Figure 1

Overall workflow of the study and the premises of Mendelian randomization. GWAS, Genome-Wide Association Study; IVs, instrumental variables; SNP, single nucleotide polymorphisms.

Figure 2

Forest plots of the MR results for the five identified significant causal effects. The horizontal coordinate represents the odds ratio value, dots depict the point estimate of odds ratio, and horizontal bars depict the 95% confidence interval. Orange represents significant results, whereas blue represents non-significant results. The arrow in the figure indicates that the upper limit of the 95% confidence interval of the odds ratio value under the method exceeds the upper limit of the horizontal coordinate, which is what we did for the aesthetics of the image.

Figure 3

Scatter plots of estimates for significant associations between gut microbiota and dermatoses. Causal influence is represented by the slope value, which is equal to the b-value computed using the five methods. A positive slope indicates a risk factor for exposure.

Figure 4

Suggestive causality of the gut microbiota in dermatoses derived from the inverse variance weighted method. Estimates with p <0.05 were shown in purple or blue, while estimates with p ≥0.05 were shown in white or green.

Figure 5

Leave-one-out plots for significant associations between gut microbiota and dermatoses.