Gut microbiota and postpartum depression: a Mendelian randomization study

Background Increasing evidence suggests a correlation between intestinal microbiota and the gut-brain axis; however, the causal relationship between gut microbiota and postpartum depression (PPD) remains unclear. Methods In this study, a two-sample Mendelian randomization (MR) design was employed to analyze the GWAS data of gut microorganisms from the Mibiogen database and PPD data from the UK biobank. Various statistical methods, including inverse variance weighted, MR-Egger, weighted median, weighted model, and MR-PRESSO, were utilized to investigate the causal relationship between gut microbiota and PPD. Additionally, sensitivity analysis was conducted to assess the robustness of the findings. Results Through MR analysis, it was found that phylum Actinobacteria (P=0.014, OR=0.971, 95% CI=0.948-0.994) and genus Holdemanella (P=0.023, OR=0.979, 95% CI=0.961-0.997) have protective effects on PPD, while the other two unknown genera, genus Unknown Ids 2001 (P=0.025, OR=0.972,95% CI=0.947-0.996), and genus Unknown Ids 2755 (P=0.012, OR=0.977, 95% CI=0.959-0.995) also has a protective effect on PPD. The sensitivity analysis results indicate that there is no heterogeneity or horizontal pleiotropy. Conclusion This study has identified a causal association between Actinomycetota, Holdemanella, and PDD through MR analysis. These findings offer significant contributions to the development of personalized treatment approaches for PPD, encompassing interventions such as dietary modifications or microbiome interventions.


Introduction
Postpartum depression (PPD) is a non-psychotic depressive episode that begins or continues into the postpartum period (1).Postpartum depression is the most common complication of childbirth, affecting women and mother-infant relationships and cognitive and emotional problems in children, with serious consequences for mothers, families, and children (2).Nearly 20% of patients with depression relapse within 20 years of their initial onset, and most people develop suicidal thoughts, with 4% -5% dying from depression-related suicide (3,4).Not only does it significantly impact the mother itself, but it also affects the quality of life of a family.
Recent studies have shown that the gut microbiota (GM) plays an important physiological role in maintaining gastrointestinal, hormonal, immune, and neural homeostasis (5).The concept of the "microbiota-gut-brain (MGB) axis" has been developed to understand the impact of the gut-brain axis on human homeostasis, particularly in the field of psychiatry (6).There is a close relationship between depression and the microbiota, as recent research suggests that the gut microbiota may have a significant impact on the onset and development of depression.Animal experiments indicate that the gut microbiota can influence brain function and potentially affect behavior.For example, oral administration of Lactobacillus can reduce the expression of proinflammatory cytokines and increase the levels of BDNF in the hippocampus, leading to anti-anxiety and antidepressant effects in mice (5).
Evidence from human studies indicates that the gut microbiota of individuals with depression differs significantly from that of healthy individuals, including changes in the abundance of specific bacterial genera and alterations in the overall microbial community structure.Some microbial families have been found to be positively associated with anxiety and depressive symptoms, while others may help alleviate depressive symptoms (7)(8)(9).Additionally, the relative abundance of certain bacterial taxa, such as the Firmicutes phylum, appears to be more representative in major depressive disorder (MDD) (6,10).This association is likely mediated through mechanisms such as regulating inflammation, influencing neurotransmitter synthesis and metabolism, and modulating the gut-brain axis signaling.Overall, the research suggests that the gut microbiota may have a profound impact on the molecular pathways involved in the occurrence and development of anxiety and depression-related behaviors, despite the differences between human and murine microbiomes (11).
Mendelian randomization (MR) is a new approach to exploring the causal relationship between gut microbiota and PPD by constructing working exposure variables using genetic variation to assess the causal relationship between exposure and outcome (12).Due to the random assignment of genes, the influence of other confounding factors is also avoided (13).The Mibiogen database is a bioinformatics platform that can be used for multi-omics data analysis and interactive visualization (14), based on which numerous authors have explored the causal relationship between gut flora and a variety of diseases, including eclampsia (15), adverse pregnancy outcomes (16) and ischemic Stroke (17).
In this study, a two-sample MR Analysis was performed using pooled statistics from genome-wide Association Studies (GWAS) from MiBioGen and the UK biobank consortium to explore specific gut microbiota causally associated with PPD.

Data sources
We obtained summary statistics of genome-wide association studies of the gut microbiota in mibiogen (18), and the MiBioGen study coordinated 2021S rRNA gene sequencing profiles and genome-wide genotyping data from 18,473 individuals (25 cohorts) and is the largest, multi-ethnic, genome-wide metaanalysis of the gut microbiome to date (19).This study included 211 taxa: 9 phylum, 16 orders, 20 families, 35 families, 131 genera, and 7738 participants of European ancestry, as determined by 16S ribosomal RNA gene sequencing (18).Data for PPD were obtained from the UK biobank, containing 4834 patients and 33173 controls from the European population, containing a total of 11,982,120 SNPs (20).All the people are European.

Genetic variants selection criteria
Based on the screening criteria from previous literature, we chose a stringent threshold of P<1×10 -5 to select instrumental variables (IVs) for our analysis.This threshold ensures that only genetic variants with a very low probability of being associated with the outcome are included as IVs, reducing the likelihood of including SNPs with weak or spurious associations.
Additionally, to ensure the independence of each IV, we applied a threshold of r 2 <0.001 within a window size of 10,000 kb.This step aimed to mitigate the effects of linkage disequilibrium (LD), a phenomenon where genetic variants close to each other on the chromosome are inherited together.By trimming IVs that are in high LD with each other, we aimed to reduce redundancy and remove SNPs that are essentially providing the same information.This helps in ensuring that the selected IVs are truly independent and provide unique information for the analysis.
Furthermore, we removed "echo SNPs" which are SNPs that are redundant due to LD and do not provide additional information beyond the already included SNPs.We also excluded SNPs that were not present in the results from the IVs, ensuring that all SNPs used in the analysis had valid and reliable data available for the research.
By applying these stringent criteria, we aimed to ensure that the selected IVs were robust, independent, and unlikely to be influenced by LD, thus enhancing the quality and reliability of our instrumental variable analysis.

MR analysis
The IVW method is an extension of the Wald ratio estimator based on meta-analysis principles (21).The random effects model with inverse variance weight was selected as the main MR method.For the flora with causality in IVW (p<0.05),four additional methods were selected as supplements (MR Egger, weighted median, simple model, and Weighting pattern).In addition, we conducted a sensitivity analysis of the results.Firstly, we used the MR Egger interception test and the MR PRESSO global test to detect horizontal pleiotropy (22,23).We reported the heterogeneity of the Wald estimator using the Cochrane Q statistic (24).In addition, a retention analysis was conducted to evaluate the robustness of the results.
All analyses in this study were conducted based on R software (version 4.2.1).The "TwoSampleMR" R package and the "MRPRESSO" R package were used for our MR research.

Results
According to the selection criteria of IVs, a total of 2044 SNPs were used as IVs for 5 levels and 211 sets, including 9 phylum, 16 classes, 20 orders, 35 families, and 119 bacterial genera.
We used IVW testing and MR Egger regression to test the Qstatistic results and did not find any heterogeneity in the results.We used MR Egger regression to detect the presence of horizontal pleiotropy in Genus (Table 2).Holdemanella (P=0.04), but no other results showed the presence of horizontal pleiotropy.At the same time, we used the MR-PRESSO algorithm for detection and did not find the existence of horizontal pleiotropy.The forest diagram of causal effects using a single SNP shows that their association with mental illness/ traits is not very significant, and sensitivity analysis indicates that there is no single SNP driving causal association signal (Figure 2).
In addition, the MR Steiger directionality test showed that the variance explained by the included bacterial exposure SNP was greater than the mental outcome, indicating a true causal correlation in the direction.

Discussion
In this study, the causal relationship between four bacterial features in the gut microbiota genome-wide association study (GWAS) and postpartum depression (PPD) was demonstrated through Mendelian randomization (MR) analysis.This research is not only significant in understanding the role of the gut microbiota in postpartum depression, but also provides new evidence for the "microbiota-gut-brain (MGB) axis" concept.The gut-brain axis is involved in the shared genetic basis of gastrointestinal and mental disorders, a notion which has been confirmed through comprehensive genomic range analysis (25).
This study identified a causal relationship between four bacterial genera and postpartum depression (PPD), allowing for in-depth exploration of the impact of these microbial changes on function and metabolism.Notably, the research on Actinobacteria Scatter plots for the causal association between gut microbiota and PPD. and Holdemanella genera is particularly intriguing.Evidence of the protective role of Actinobacteria as a key member of the gut microbiota against depression continues to accumulate, as demonstrated by Tian et al.'s analysis of the gut microbiota in normal mice and those with PPD, which revealed higher abundance o f A c t i n o b a c t e r i a ( i n c l u d i n g B i fi d o b a c t e r i u m a n d Corynebacterium) in the normal group compared to the PPD group (26).Previous research has also indicated an association between gut microbiota imbalance and certain mental disorders such as anxiety and depression.Jiang's high-throughput sequencing analysis of 46 depressed patients and 30 healthy controls showed significantly higher abundance of Actinobacteria and Firmicutes in the healthy control group at the phylum level compared to the depressed patients (27).Moreover, previous MR studies have indicated a protective effect of Actinobacteria against major depressive disorder (MDD) incidence (OR 0.88, 95% CI 0.87-0.9)(28).Therefore, bacteria within the Actinobacteria phylum may indirectly influence the onset and development of mental disorders by affecting the balance of the gut microbiota.In fact, Actinobacteria are producers of many important antibiotics (29), including penicillin, tetracycline, and erythromycin, and their increased abundance may compete with pathogens for nutrients and adhesion sites, thereby inhibiting pathogen colonization and growth, and contributing to the maintenance of gut microbiota balance.Additionally, some Actinobacteria may modulate the host's immune system, contributing to immune response regulation and maintaining gut immune system balance.However, further research and exploration are needed to elucidate the specific mechanisms and effects of Actinobacteria in alleviating postpartum depression.
Research on the Holdemanella genus also suggests its potential beneficial impact in reducing the risk of postpartum depression (PPD).In a study on post-stroke depression (PSD), researchers analyzed fecal samples from 232 patients with acute ischemic stroke using 16S rRNA sequencing.The samples were assessed using the Hamilton Depression Rating Scale (HAMD-3).The results indicated a significant decrease in the abundance of Holdemanella genus in PSD patients, and a negative correlation between the abundance of Holdemanella genus and HAMD scores, suggesting a potential beneficial impact of Holdemanella genus in reducing the risk of PSD (30).Furthermore, Jiang's study observed lower abundance of Firmicutes in the gut microbiota of depressed patients compared to healthy controls (7).Additionally, several studies consistently indicate that increasing the levels of the Holdemanella genus is beneficial in reducing the incidence of depression (31,32).It is worth mentioning that depressed patients often have lower levels of omega-3 fatty acids (33).While there is no consensus on whether supplementing omega-3 alone can effectively alleviate depression, it has been observed that consuming omega-3-rich fish may be associated with increased abundance of Holdemanella genus (27).Could increasing the abundance of Holdemanella genus and reducing the risk of postpartum depression be achieved through omega-3 supplementation?This is purely speculative, but it also provides a new perspective on the role of dietary intervention in preventing postpartum depression.Leave-one-out analysis for the causal association between gut microbiota and PPD.
This study identified a causal relationship between four bacterial features in the gut microbiota genome-wide association study (GWAS) and postpartum depression (PPD) through Mendelian randomization (MR) analysis.Additionally, it suggests that the Actinobacteria and Holdemanella genera may have a potential beneficial impact in reducing the risk of depression.Adjusting the abundance of these microorganisms in the gut microbiota may help improve symptoms of certain mental disorders, providing important evidence for understanding the role of the gut microbiota in postpartum depression.
However, this study also has some limitations.Firstly, the GWAS meta-analysis of the gut microbiota included male and female participants.Even though genetic variants located on the sex chromosomes were excluded from the analysis and adjustments for gender were made, it may still introduce bias (18).Moreover, the majority of the data is from individuals of European descent, potentially introducing interference from racial differences.Therefore, in future studies, we hope to conduct detailed subgroup analyses targeting specific populations to explore the influence of gender on the relationship between the gut microbiota and PDD.Additionally, we aim to conduct in-depth research on specific gut microbiota to understand their association with PDD, and further explore the mechanisms of specific microbiota in PDD through metagenomic analysis and functional experiments, deepening our understanding of the relationship between the gut microbiota and PDD, and providing a scientific basis for more precise intervention measures in the future.

TABLE 1
Causal estimations of gut microbiota on postpartum depression in the MR analysis.

TABLE 2
Heterogeneity test and horizontal pleiotropy test of gut microbiota on postpartum depression.