Causal relationship between gut microbiome and sex hormone-binding globulin: A bidirectional two-sample Mendelian randomization study.

Currently, there is a variety of evidence linking the gut microbiota to changes in sex hormones. In contrast, the causal relationship between SHBG, a carrier of sex hormones, and the gut microbiota is unclear. Bidirectional two-sample Mendelian randomization (MR) analysis was used to detect the causal effect between SHBG and the gut microbiome. Summary statistics of genome-wide association studies (GWASs) for the gut microbiome and SHBG were obtained from public datasets. Inverse-variance weighting (IVW), weighted median, weighted mode, MR-Egger and simple mode methods were used to operate the MR analysis. F-statistics and sensitivity analyses performed to evaluate bias and reliability. When we set gut microbiome as exposure and SHBG as outcome, we identified nine causal relationships. In males, Coprobacter (PIVW = 2.01 × 10-6 ), Ruminococcus2 (PIVW = 3.40 × 10-5 ), Barnesiella (PIVW = 2.79 × 10-2 ), Actinobacteria (PIVW = 3.25 × 10-2 ) and Eubacterium fissicatena groups (PIVW = 3.64 × 10-2 ) were associated with lower SHBG levels; Alphaproteobacteria (PIVW = 1.61 × 10-2 ) is associated with higher SHBG levels. In females, Lachnoclostridium (PIVW = 9.75 × 10-3 ) and Defluviitaleaceae UCG011 (PIVW = 3.67 × 10-2 ) were associated with higher SHBG levels; Victivallaceae (PIVW = 2.23 × 10-2 ) was associated with lower SHBG levels. According to the results of reverse MR analysis, three significant causal effect of SHBG was found on gut microbiota. In males, Dorea (PIVW = 4.17 × 10-2 ) and Clostridiales (PIVW = 4.36 × 10-2 ) were associated with higher SHBG levels. In females, Lachnoclostridium (PIVW = 7.44 × 10-4 ) was associated with higherr SHBG levels. No signifcant heterogeneity of instrumental variables or horizontal pleiotropy was found in bidirectional two-sample MR analysis. This study may provide new insights into the causal relationship between the gut microbiome and sex hormone-binding protein levels, as well as new treatment and prevention strategies for diseases such as abnormal changes in sex hormones.

Yan Z ，Zheng Z ，Xia T ，Ni Z ，Dou Y ，Liu X ... -  《-》

[Genetic Causation Analysis of Hyperandrogenemia Testing Indicators and Preeclampsia].

Some epidemiological studies have shown that pregnant women who develop preeclampsia (PE) have elevated levels of testosterone in their maternal plasma compared to women with normal blood pressure during pregnancy, revealing a potential association between hyperandrogenism in women and PE. To explore the causal relationship between hyperandrogenism and PE, this study selected total testosterone (TT), bioavailable testosterone (BIOT), and sex hormone binding globulin (SHBG) as exposure factors and PE and chronic hypertension with superimposed PE as disease outcomes. Two-sample Mendelian randomization (MR) analyses were used to genetically dissect the causal relationships between the three exposure factors (TT, BIOT, and SHBG) and the outcomes of PE and chronic hypertension with superimposed PE. Two independent genome-wide association study (GWAS) databases were used for the two-sample MR analysis. In the GWAS data of female participants from the UK Biobank cohort, single nucleotide polymorphisms (SNPs) associated with TT, BIOT, and SHBG were analyzed, involving 230454, 188507, and 188908 samples, respectively. GWAS data on PE and chronic hypertension with superimposed PE from the Finnish database were used to calculate SNP, involving 3556 PE cases and 114735 controls, as well as 38 cases of chronic hypertension with superimposed PE and 114735 controls. To meet the assumptions of instrumental relevance and independence in MR analysis, SNPs associated with exposure were identified at the genome-wide level (P<5.0×10-8), and those in linkage disequilibrium interference were excluded based on clustering thresholds of R 2<0.001 and an allele distance greater than 10000 kb. Known confounding factors, including previous PE, chronic kidney disease, chronic hypertension, diabetes, systemic lupus erythematosus, or antiphospholipid syndrome, were also identified and the relevant SNPs were removed. Finally, we extracted the outcome data based on the exposure-related SNPs in the outcome GWAS, integrating exposure and outcome data, and removing palindromic sequences. Five genetic causal analysis methods, including inverse variance-weighted method (IVW), MR-Egger regression, weighted median method, simple mode method, and weighted mode method, were used to infer causal relationships. In the IVW, it was assumed that the selected SNPs satisfied the three assumptions and provided the most ideal estimate of the effect. IVW was consequently used as the primary analysis method in this study. Considering the potential heterogeneity among the instrumental variables, random-effects IVW was used for MR analysis. The results were interpreted using odds ratios (OR) and the corresponding 95% confidence interval (CI) to explain the impact of exposure factors on PE and chronic hypertension with superimposed PE. If the CI did not include 1 and had a P value less than 0.05, the difference was considered statistically significant. Sensitivity analysis was conducted to assess heterogeneity and pleiotropy. Heterogeneity was examined using Cochran's Q test, and pleiotropy was assessed using MR-Egger intercept analysis. Additionally, leave-one-out analysis was conducted to examine whether individual SNPs were driving the causal associations. To further validate the findings, MR analyses were performed using the same methods and outcome variables, but with different exposure factors, including waist-to-hip ratio adjusted for BMI (WHRadjBMI) and 25-hydroxyvitamin D levels, with MR results for WHRadjBMI and PE serving as the positive controls and MR results for 25-hydroxyvitamin D levels and PE as the negative controls. According to the criteria for selecting genetic instrumental variables, 186, 127, and 262 SNPs were identified as genetic instrumental variables significantly associated with testosterone indicators TT, BIOT, and SHBG. MR analysis did not find a causal relationship between the TT, BIOT, and SHBG levels and the risk of developing PE and chronic hypertension with superimposed PE. The IVW method predicted that genetically predicted TT (OR [95% CI]=1.018 [0.897-1.156], P=0.78), BIOT (OR [95% CI]=1.11 [0.874-1.408], P=0.392), and SHBG (OR [95% CI]=0.855 [0.659-1.109], P=0.239) were not associated with PE. Similarly, genetically predicted TT (OR [95% CI]=1.222 [0.548-2.722], P=0.624), BIOT (OR [95% CI]=1.066 [0.242-4.695], P=0.933), and SHBG (OR [95% CI]=0.529 [0.119-2.343], P=0.402) were not significantly associated with chronic hypertension with superimposed PE. Additionally, MR analysis using the MR-Egger method, weighted median method, simple mode method, and weighted mode method yielded consistent results, indicating no significant causal relationship between elevated testosterone levels and PE or chronic hypertension with superimposed PE. Heterogeneity was observed for SHBG in the analysis with PE (Cochran's Q test, P=0.01), and pleiotropy was detected for BIOT in the analysis with PE (MR-Egger intercept analysis, P=0.014), suggesting that the instrumental variables did not affect PE through BIOT. Other instrumental variables did not show significant heterogeneity or pleiotropy. Leave-one-out analysis confirmed that the results of the MR analysis were not driven by individual instrumental variables. Consistent with previous MR studies, the results of the control MR analyses using WHRadjBMI and 25-hydroxyvitamin D levels supported the accuracy of the MR analysis approach and the methods used in this study. The MR analysis results suggest that current genetic evidence does not support a causal relationship between TT, BIOT, and SHBG levels and the development of PE and chronic hypertension with superimposed PE. This study suggests that elevated testosterone may be a risk factor for PE but not a direct cause.

Lin C ，Chen J ，Zhao X 《-》

Exploring reciprocal causation: bidirectional mendelian randomization study of gut microbiota composition and thyroid cancer.

While an association between gut microbiota composition and thyroid cancer (TC) has been observed, the directionality and causality of this relationship remain unclear. We conducted a bidirectional two-sample Mendelian randomization (MR) analysis to investigate the causal effect between gut microbiota composition and TC. Gut microbiota data were derived from a diverse population encompassing various ethnicities (n = 18,340 samples), while TC data were sourced from an European population (n = 218,792 samples). Instrumental variables, represented by single nucleotide polymorphisms (SNPs), were employed to assess the causal relationship using multiple MR methods, including inverse-variance weighting (IVW), weighted median, weighted mode, MR-Egger, and simple mode. F-statistics and sensitivity analyses were performed to evaluate the robustness of the findings. Our investigation identified a comprehensive set of 2934 instrumental variables significantly linked to gut microbiota composition (p < 1 × 10-5). The analysis illuminated notable candidates within the phylum Euryarchaeota, including families Christensenellaceae and Victivallaceae, and genera Methanobrevibacter, Ruminococcus2, and Subdoligranulum, which emerged as potential risk factors for TC. On the other hand, a protective influence against TC was attributed to class Betaproteobacteria, family FamilyXI, and genera Anaerofilum, Odoribacter, and Sutterella, alongside order Burkholderiales. Further enhancing our insights, the integration of 7 instrumental variables from TC data (p < 1 × 10-5) disclosed the regulatory potential of one family and five genera. Notably, the genus Coprobacter innocuum group (p = 0.012, OR = 0.944) exhibited the highest probability of regulation. Our meticulous analyses remained free from significant bias, heterogeneity, or horizontal pleiotropy concerns. Through a bidirectional two-sample Mendelian randomization approach, we elucidated a potential bidirectional causal relationship between gut microbiota composition and TC. Specific microbial taxa were associated with an increased risk or conferred protection against TC. These findings advance our understanding of the complex interplay between the gut microbiota and TC pathogenesis, offering new insights into the therapeutic potential of modulating the gut microbiota for managing TC.

Zhou J ，Zhang X ，Xie Z ，Li Z ... -  《-》

Investigating causal relationships between the gut microbiota and inflammatory skin diseases: A Mendelian randomization study.

Numerous inflammatory skin diseases are associated with the gut microbiota. Studies of the association between gut microbiota and inflammatory skin diseases have yielded conflicting results owing to confounding factors, and the causal relationship between them remains undetermined. Two-sample Mendelian randomization (MR) was used to examine the association between gut microbiota and four common inflammatory skin diseases: acne, psoriasis, urticaria and atopic dermatitis. The summary statistics of the gut microbiota from the largest available genome-wide association study meta-analysis (n = 13,266) conducted by the MiBioGen consortium along with the summary statistics of the four diseases were obtained from the FinnGen consortium. Causal relationships were assessed using the inverse variance weighted (IVW), weighted median, MR-Egger and maximum likelihood methods, and several sensitivity analyses were performed to ensure the accuracy of the results. Finally, reverse and multivariable MR analyses were performed to verify the robustness of the results. We found causal associations of Bacteroidaceae [odds ratio (OR), 2.25; 95% confidence interval (CI), 1.48-3.42; pivw = 0.0001], Allisonella (OR, 1.42; 95% CI, 1.18-1.70; pivw = 0.0002) and Bacteroides (OR, 2.25; 95% CI, 1.48-3.42; pivw = 0.0001) with acne, the Eubacterium fissicatena group with psoriasis (OR, 1.22; 95% CI, 1.10-1.35; pivw = 0.0002) and Intestinibacter with urticaria (OR, 1.28; 95% CI, 1.13-1.45; pivw = 0.0001). These results were corrected for a false discovery rate. Sensitivity analyses were performed to validate the robustness of the associations and reverse MR confirmed that the results were not influenced by the reverse effect. Our study revealed that some gut microbiota are risk factors for inflammatory skin diseases, providing new information on potential therapeutic targets. Additionally, a possible association with the gut-skin axis was confirmed. Further research is required to elucidate the mechanisms underlying these relationships.

Gu Y ，Zhang W ，Zhao W ，Zeng X ... -  《-》

Causality Investigation between Gut Microbiota, Derived Metabolites, and Obstructive Sleep Apnea: A Bidirectional Mendelian Randomization Study.

Yan W ，Jiang M ，Hu W ，Zhan X ，Liu Y ，Zhou J ，Ji J ，Wang S ，Tai J ... -  《Nutrients》