Assessing the relationship between gut microbiota and hyperprolactinemia: A bidirectional two-sample Mendelian randomization study.

Hyperprolactinemia is a prevalent endocrine disorder of the hypothalamic-pituitary axis, with prolactinoma being one of its clinical causes. [ 1 ] In a recent epidemiological survey of hyperprolactinemic patients in Saudi, the majority of cases were found to occur in women (86.8%) and were more frequently observed in patients aged 21 to 40 years. [ 2 ] Typical symptoms of hyperprolactinemia include hypogonadism, infertility, galactorrhea, osteopenia, and tumor mass effects. [ 1 ] Patients with this condition may experience a significant reduction in their quality of life due to severe breast symptoms, breastfeeding discomfort, menstrual changes, and even male erectile dysfunction. [ 3 ] Moreover, hyperprolactinemia has been reported to be associated with endometriosis in infertile women. [ 4 ] Currently, dopamine agonists and surgery are the treatments for hyperprolactinemia. [ 5 ] However, due to the complex nature of this endocrine disease, it is imperative to identify its pathogenesis and develop more effective and minimally invasive treatment options. The pathophysiological mechanism of hyperprolactinemia is multifactorial; however, the current understanding of its underlying causes remains incomplete. Studies have shown that there is a strong genetic component in the risk of hyperprolactinemia. [ 6 ] The genetic mechanism of hyperprolactinemia may not be a simple single gene mutation, but multiple gene changes, such as prolactin receptor mutations and polymorphisms. [ 7 ] Gut microbiota may influence the occurrence and progression of diseases by producing certain metabolites and regulating disease-related genes, thereby causing changes in endocrine system signaling pathways. [ 8 ] Recent evidence suggests that the pathogenesis of hyperprolactinemia may be closely related to the gut microbiota. [ 9 ] In an animal study, the researchers found that severe hyperprolactinemia can exacerbate metabolic imbalances induced by a high-fat diet, [ 10 ] which may be related to changes in the gut microbiota. [ 11 ] Therefore, gut microbes may directly or indirectly influence the occurrence and development of hyperprolactinemia. To date, there have been no studies establishing a link between the gut microbiota and hyperprolactinemia. Therefore, it remains unclear whether a causal relationship exists. Mendelian randomization is an established epidemiological method that combines observational data to directly estimate causal effects. [ 12 ] It has become a valuable tool for analyzing potential causal relationships between the gut microbiota and disease risk genes. [ 13 – 17 ] In order to investigate the potential causal relationship between the gut microbiota and hyperprolactinemia, as well as identify specific classifications of pathogenic bacteria, we conducted a Mendelian randomization (MR) Analysis using pooled data from genome-wide association studies.

Conceptualization: Teng Qi, Fuqing Ji. Formal analysis: Teng Qi, Zhenping Ouyang. Funding acquisition: Fuqing Ji. Methodology: Teng Qi, Lanhui Zhang. Software: Teng Qi. Visualization: Yujie Hu, Xiaorui Zhou, Sujin Wang. Writing – original draft: Xiaorui Zhou, Miao Wei, Sujin Wang. Writing – review & editing: Lanhui Zhang, Xulong Zhu, Mingkun Zhang, Chuanlin Zhao.

The overall process design of this study is shown in Figure 1 . We obtained summary statistics for human gut microbiome and abdominal aortic aneurysm from a recent meta-analysis of genome-wide association studies (GWAS). [ 18 ] The meta-analysis of human gut microbiome included 18,340 participants from 24 different cohorts. More information about this study can be found elsewhere. [ 18 ] In summary, the study combined data from various countries, such as United States of America, Canada, Germany, Denmark, Netherlands, Belgium, Sweden, Finland, and the United Kingdom. It integrated 16S rRNA gene sequencing profiles with genotyping data and performed association analyses while adjusting for factors such as age, sex, technical covariates, and genetic principal components. [ 18 ] The data for hyperprolactinemia were derived from samples of Finland ancestry, including 1099 disease samples and 395,289 control samples. [ 19 ] Since our study utilized publicly available summary data, there was no need for additional ethics approval or participant consent. The detailed information of the data sources used in this MR study can be found in Table 1 . Details of the genome-wide association studies and datasets used in our analyses. The study design of the associations of gut microbiota and hyperprolactinemia. MR = Mendelian randomization, SNP = single-nucleotide polymorphism. Initially, we excluded 15 unidentified bacterial traits from our analysis, leaving us with a remaining set of 194 bacterial traits that span 9 phyla, 14 classes, 20 orders, 32 families, and 119 genera. Subsequently, we applied a significance threshold of P <  1.0 × 10 −5 to select instrumental variables (IVs). To ensure genetic independence between loci, we utilized a linkage disequilibrium (LD) threshold of R 2   < 0.001 and clumping distance = 10,000 kb using the “TwoSampleMR” package on the 1000 Genomes EUR data. For each associated trait, we retained single-nucleotide polymorphisms (SNPs) with the lowest P -value for subsequent clumping analysis alongside the other 194 bacterial traits. This process resulted in a total of 2515 independent SNPs that were found to be associated with these bacterial traits. In the reverse MR analysis for hyperprolactinemia, we used the threshold for P  < 1 × 10 −5 , as previously described in Table 2 . [ 20 , 21 ] Extracting pertinent information such as the effect allele, effect size (including β -value), standard error, and P -value for each SNP allowed us to calculate the proportion of variation explained ( R 2 ) and F -statistics to evaluate instrument strength. These calculations followed the equation: Effect estimates of the associations of hyperprolactinemia with 9 types of gut microbiota in the reverse MR analyses. where “MAF” represents the minor allele frequency of IVs used, “n” denotes the sample size, and “k” signifies the number of IVs employed. [ 20 , 21 ] We employed various methods to evaluate the potential causal links between the gut microbiota and hyperprolactinemia. These approaches consisted of the fixed/random-effects inverse-variance weighted (IVW) method, weighted-median method, MR-Egger regression, and MR pleiotropy residual sum and outlier (MR-PRESSO) test. The IVW method was our primary analysis due to its accurate effect estimates and widespread usage in MR analysis. [ 22 – 24 ] Initially, individual SNPs were utilized with the Wald estimator and Delta method to calculate ratio estimates. These estimates were then combined to obtain the main causal estimate. [ 25 ] Cochran’s Q test was used to assess heterogeneity among selected SNPs. If heterogeneity was detected ( P <  .05), we applied the random-effects IVW method; otherwise, we utilized the fixed-effects IVW method. [ 26 ] To ensure robust associations considering valid instruments and potential pleiotropic effects, sensitivity analysis were conducted. First, we implemented the weighted median method, which provided reliable causal effect estimates even when valid instruments are limited. [ 27 ] This approach yields valid results even if <50% of information is derived from invalid instruments. [ 27 ] Second, we performed MR-Egger regression to examine potential horizontal pleiotropy by evaluating the P -value of the intercept (<.05 indicates possible horizontal pleiotropy of SNPs). [ 28 ] Last, we employed the MR-PRESSO test to identify outliers among SNPs through a global test of heterogeneity. After identifying and removing potential outliers, a corrected association result was obtained. [ 29 ] To further explore potential directional pleiotropy, each SNP used as IVs was assessed for their associations with secondary phenotypes using GWAS Catalog ( http://www.ebi.ac.uk/gwas , last accessed on August 23, 2023). Subsequently, MR analysis weas repeated after excluding SNPs associated with other phenotypes. The associations between human gut microbiota and the risk of hyperprolactinemia were reported as odds ratios (ORs) with 95% confidence intervals (CIs). The reverse MR analysis was carried out only when there was unanimous support from all MR methods regarding the link between the gut microbiota and hyperprolactinemia. All MR analysis were conducted using R version 4.2.2 ( https://www.r-project.org/ ) with the “Mendelian Randomization,” “TwoSampleMR,” and “MR-PRESSO” R packages.

Simply put, we observed suggestive evidence that 9 bacterial traits were associated with the risk of hyperprolactinemia using the IVW approach (Fig. 2 ). The MR results of the associations between all 194 bacterial traits and the risk of hyperprolactinemia are presented in Table S1, Supplemental Digital Content, https://links.lww.com/MD/Q466 , and the IVs used for these 9 bacterial traits are listed in Table S2, Supplemental Digital Content, https://links.lww.com/MD/Q466 . Forest plot of the associations between 9 genetically determined bacterial traits and the risk of hyperprolactinemia. CI = confidence interval, OR = odds ratio, SNP = single-nucleotide polymorphism. The IVW method and MR-Egger regression results revealed that certain bacterial groups, namely Family Acidaminococcaceae (Fig. 3 A, OR: 0.632; 95% CI: 0.405 to 0.988; P  = .044), Genus Eubacterium ruminantium group (Fig. 3 B, OR: 0.681; 95% CI: 0.531 to 0.873; P  = .002), Genus Anaerofilum (Fig. 3 D, OR: 0.686; 95% CI: 0.522 to 0.901; P  = .007), and Genus Erysipelotrichaceae UCG003 (Fig. 3 F, OR: 0.595; 95% CI: 0.416 to 0.852; P  = .005), may serve as potential protective bacterial communities, exhibiting a negative association with hyperprolactinemia. Conversely, Genus Ruminococcus gauvreaui group (Fig. 3 C, OR: 1.589; 95% CI: 1.057 to 2.389; P  = .026), Genus Eisenbergiella (Fig. 3 E, OR: 1.333; 95% CI: 1.017 to 1.747; P  = .037), and Genus Ruminococcaceae UCG014 (Fig. 3 F, OR: 1.3986; 95% CI: 1.00 to 1.954; P  = .05) may act as potential pathogenic bacterial communities, showing a positive association with hyperprolactinemia. Scatter plot of the associations of genetic variants with 9 bacterial traits and the risk of hyperprolactinemia. (A) Family Acidaminococcaceaeid ; (B) Genus Eubacterium ruminantium group ; (C) Genus Ruminococcus gauvreauii group ; (D) Genus Anaerofilum ; (E) Genus Eisenbergiella ; (F) Genus Erysipelotrichaceae UCG003 ; (G) Genus Ruminococcaceae UCG014 ; (H) Family Bacteroidales S24.7 group ; (I) Genus Peptococcus . MR = Mendelian randomization, SNP = single-nucleotide polymorphism. However, it should be noted that the impact of Family Bacteroidales S24.7 group (Fig. 3 H, OR: 0.685; 95% CI: 0.483 to 0.970; P  = .033) and Genus Peptococcus (Fig. 3 I, OR: 0.781; 95% CI: 0.612 to 0.998; P  = .048) on hyperprolactinemia appeared to be contradictory in the IVW analysis and MR-Egger regression. Further foundational research is needed to validate their influence on hyperprolactinemia. To further explore the impact of potential directional pleiotropy on the estimates of causal effects, we utilized the GWAS Catalog to identify single-nucleotide polymorphisms (SNPs) associated with the 9 bacterial traits under investigation. We found that 9 SNPs were associated with other traits, as shown in Table S3, Supplemental Digital Content, https://links.lww.com/MD/Q466 . The leave-one-out method was employed to observe the minimal changes in results (Fig. 4 ), confirming the stability of the associations between various factors and the increased risk of hyperprolactinemia. “Leave-one-out” sensitivity analysis for 9 types of gut microbiota on hyperprolactinemia. (A) Family Acidaminococcaceaeid ; (B) Genus Eubacterium ruminantium group ; (C) Genus Ruminococcus gauvreauii group ; (D) Genus Anaerofilum ; (E) Genus Eisenbergiella ; (F) Genus Erysipelotrichaceae UCG003 ; (G) Genus Ruminococcaceae UCG014 ; (H) Family Bacteroidales S24.7 group ; (I) Genus Peptococcus . MR = Mendelian randomization. Subsequently, we conducted reverse MR analyses to evaluate the potential reverse associations between the 9 bacterial traits and hyperprolactinemia. We employed the IVW method for this purpose. However, we did not detect any statistically significant associations between hyperprolactinemia and any of these 9 bacterial traits. These findings remained consistent across sensitivity analysis, as presented in Table 2 .

The human gut microbiota exhibits a high sensitivity to pathophysiological changes within the body, making it crucial to investigate specific alterations in order to comprehend disease pathogenesis and enable accurate diagnoses. [ 30 ] Such susceptibility to change can potentially lead to modifications in the microbiome, consequently contributing to the development of various diseases in humans. [ 31 ] This 2-sample MR study identified 9 bacterial taxa that displayed potential associations with an increased prevalence of hyperprolactinemia. These taxa include the Family Acidaminococcaceaeid and Bacteroidales S24.7 group , as well as the genus Ruminococcus gauvreauii group, Eisenbergiella, Ruminococcaceae UCG014, Eubacterium ruminantium group, Anaerofilum, Erysipelotrichaceae UCG003 , and Peptococcus . However, further sensitivity analysis utilizing different MR methods and restricted IV sets revealed that specifically, 7 bacterial taxa, Ruminococcus gauvreauii group, Eisenbergiella, Ruminococcaceae UCG014, Acidaminococcaceaeid, Eubacterium ruminantium group, Anaerofilum , and Erysipelotrichaceae UCG003 , were associated with the risk of hyperprolactinemia. In this study, a positive correlation was observed between hyperprolactinemia and 3 types of bacteria: Genus Ruminococcus gauvreauii group , Genus Eisenbergiella , and Genus Ruminococcaceae UCG014 . [ 32 ] Genus Ruminococcus gauvreauii group in human disease-related research is less, only Djawad Radjabzadeh and others a cohort study on the gut microbiota of depression in a report. Previous studies have shown that hyperprolactinemia is closely related to depression. [ 1 , 33 ] Patients with hyperprolactinemia usually have poor quality of life and personality characteristics such as anxiety and depression. [ 34 ] So combining with the research of Genus Ruminococcus gauvreauii group relevant results suggest that Genus Ruminococcus gauvreauii group may through a series of pathophysiological mechanisms result in increased prolactin and cause depression. As a new microbiome found in human feces collected after bariatric surgery, [ 35 ] The Genus Eisenbergiella has been reported to be associated with multiple sclerosis, rheumatoid arthritis and high blood pressure. [ 36 – 38 ] Therefore, we speculate that Genus Eisenbergiella may influence the occurrence and development of hyperprolactinemia through certain metabolites through inflammatory, lipid metabolism, and immune pathways.Genus Ruminococcaceae UCG014 has been suggested to be associated with asthma in a previous MR study. [ 39 ] Further inquiry into the gutMGene database revealed that metabolites of Genus Ruminococcaceae UCG014 include indolepropionic acid, creatine, goandeoxycholic acid, 4-acetylaminobutyric acid, aquaposin, N(6) -methyllysine, n-acetyl-D-mannosamine, n-isovalerylglycine, etc. Its effect on hyperprolactinemia needs more in-depth basic research to clarify. [ 40 ] After conducting our study, we discovered a negative correlation between hyperprolactinemia and the Family Acidaminococcaceae , Genus eubacterium raminantium group , Genus Anaerofilum , and Genus Erysipelotrichaceae UCG003 . MR studies have linked Family Acidaminococcaceae to obstructive sleep apnea (OSA) disease severity and asthma. [ 41 ] Further research has shown that these bacteria can ferment carbohydrates to produce metabolites such as butyric acid and propionic acid. [ 42 ] It is conceivable that alterations in these metabolites by Family Acidaminococcaceae may modulate the body’s state, thereby contributing to the onset of diseases. The genus Eubacterium ruminantium has the capacity to generate short-chain fatty acids and bile metabolites, influencing the intestinal microenvironment. These metabolites play a crucial role in modulating the body’s inflammatory immune response, regulating blood glucose and cholesterol levels, and preserving the integrity of the intestinal barrier. [ 43 ] Previous studies have confirmed that increased abundance of the Genus Anaerofilum has a positive effect on patients with depression. [ 44 ] In conjunction with the findings of this study, the genus Anaerofilum could serve as a foundation for understanding the mechanistic link between depression and hyperprolactinemia.The association between Erysipelotrichaceae UCG003 and cancer has been confirmed. The abundance of Erysipelotrichaceae UCG003 is negatively correlated with glycerophospholipid metabolism which could be an essential mechanism for regulating metabolism and tumor development in vivo. [ 45 ] Additionally, Erysipelotrichaceae UCG003 is one of the main bacteria that produce butyrate, which affects intestinal flora and improves gastrointestinal function. [ 45 ] In our research, we utilized genetic epidemiological methods to establish a causal relationship between the gut microbiota and hyperprolactinemia. Our IV analysis ensured that our results were less likely to be affected by weak instrument bias, as indicated by the F -statistic of our IVs exceeding the threshold of > 10. Furthermore, we conducted a reverse MR analysis to exclude the possibility of reverse causality. Moving forward, investigating the causal association between gut microbiota and disease development should be the focus of future research. Our MR analysis offers valuable insights into selecting specific gut bacteria for studying the role of gut microbiota in hyperprolactinemia pathogenesis. Nevertheless, our study does have limitations. First, we only analyzed bacterial taxa at the genus level, rather than at a more detailed level, such as species or strain. Second, the majority of participants in our GWAS were of European descent, which may limit the generalizability of our findings to other ethnic groups. Third, we relaxed the threshold for selecting IVs for gut microbiota ( P  < 1.0 × 10 −5 ) to obtain a sufficient number of IVs. Additionally, the effects of the bacterial traits we identified were relatively weak, and there were no other independent GWASs with an adequate sample size to validate our findings. Finally, due to the lack of specific information on the subtypes and gender of hyperprolactinemia, further research is needed when this information is available.

Our systematic analysis provides evidence supporting a potential causal relationship between specific gut microbiota taxa and the risk of hyperprolactinemia. Further research is needed to elucidate the mechanisms through which gut microbiota influence and expedite the development and progression of hyperprolactinemia.