Abstract
Objective
The relationship between autoimmune diseases (ADs) and male reproductive health is easily neglected. Epidemiological studies indicated that multiple ADs may adversely affect male reproductive health. This study aimed to explore the causal relationship between ADs and erectile dysfunction (ED), abnormal spermatozoa, and sex hormones using Mendelian randomization (MR) analysis.
Methods
We selected 12 common ADs as exposures, with outcomes including ED, abnormal spermatozoa, and sex hormones. The selection of instrumental variables was based on summary statistics from European GWAS studies. A combination of multiple methods was performed for MR analysis of the above data, with the inverse-variance weighted method being the main method. We also conducted several sensitivity analyses to verify the reliability of the results.
Results
Inflammatory bowel disease (IBD) and Crohn’s disease (CD) were associated with an increased risk of ED (IBD, OR: 1.101, 95% CI: 1.016-1.194, P = 0.019; CD, OR: 1.086, 95% CI: 1.012-1.165, P = 0.022). Asthma was positively causally related to abnormal spermatozoa (OR: 1.270, 95% CI: 1.045-1.545, P = 0.017), but negatively causally related to total testosterone (β: -0.042, 95% CI: -0.057 - -0.027, P < 0.001). Detection of sensitivity analysis supported findings related to ED and abnormal spermatozoa.
Conclusion
Current genetic evidence suggested a potential causal effect of IBD and CD on the risk of ED. Furthermore, genetic liability to asthma was associated with abnormal spermatozoa and decreased total testosterone levels. Our results contributed to targeted prevention and management of these conditions, ultimately promoting male reproductive health.
Keywords
1. Introduction
Sexual health is not only important for the physical and mental well-being of individuals and their partners, but also contributes to public health development and social stability. 1 As a common male sexual disorder, erectile dysfunction (ED) refers to the inability of patients to achieve a satisfactory sexual intercourse experience. According to reports, the incidence of ED among middle-aged and elderly men is 52% in the United States, 30% in Europe, and 63% in Asia.2,3 As another global problem, infertility is plaguing about 8-12% of childbearing couples, and about 50% of these cases are related to male factors. 4 Male infertility has a serious negative impact on population growth and is more likely to be a predictor of poor health and early death. 5 Normal sperm count, motility, and morphology are conducive to successful conception. Sperm abnormalities could reduce the ability of sperm to penetrate cervical mucus and zona pellucida, and increase the risk of miscarriage. 6 Therefore, maintaining male reproductive health is particularly important, which would improve patients’ quality of life and reduce the resulting social medical burden.
While sex hormones play a fundamental regulatory role in male reproductive health, this status is actively influenced by various systemic factors. Specifically, immunoregulation along with its mediated inflammatory responses and cytokine profiles could affect sperm production, sex hormone secretion and penile erection, playing a key role in maintaining reproductive homeostasis.7,8 In recent years, studies have shown that autoimmune diseases (ADs) are also associated with male reproductive health and may impair erectile function, semen quality, and sex hormones.9–12 ADs, such as Crohn’s disease (CD), ulcerative colitis (UC), are chronic inflammatory conditions that are caused by immune-mediated destruction of self-tissue. In addition to generalized inflammation, the distinct pathogenesis of ADs includes disease-specific immune insults (such as anti-sperm antibodies), treatment-induced gonadotoxicity, chronic psychological distress, and more, which uniquely compound the risk of male reproductive impairment. However, the current evidence is mostly derived from observational studies. These conclusions may be biased due to potential confounders and reverse causation. Whether there is a causal relationship between ADs and male reproductive diseases still needs further verification.
Mendelian randomization (MR) analysis utilizes genetic variation as instrumental variables (IVs) to assess the causal relationship between exposure and outcome. Precisely because its design is similar to that of randomized controlled trials, MR analysis can circumvent the limitations of observational studies to a certain extent and has been widely used in causality research. 13 Therefore, this study aimed to use MR analysis to explore whether ADs are causally related to ED, abnormal spermatozoa, and sex hormones.
2. Methods
2.1 Study design
In this two-sample MR analysis, single nucleotide polymorphisms (SNPs) were used as IVs and should satisfy three key assumptions: the relevance assumption, independence assumption, and exclusion restriction assumption. The key assumptions and flow chart were shown in Figure 1. This study strictly followed the STROBE-MR guidelines. Ethical approval and informed consent were obtained for all original studies included in this article, so no additional ethical approval is required. The study design of two-sample MR analysis. (A) Three major assumptions of MR analysis. Partly created with BioRender.com. (B) Flow chart of MR analysis. MR, Mendelian randomization; IVs, instrumental variables; SNPs, single nucleotide polymorphisms; SHBG, sex hormone-binding globulin; TT, total testosterone; BT, bioavailable testosterone.
2.2 Data resources
This study considered 12 common ADs, including ankylosing spondylitis, asthma, celiac disease (CeD), eczema, primary sclerosing cholangitis, psoriasis, rheumatoid arthritis, systemic lupus erythematosus, inflammatory bowel disease (IBD), CD, UC and type 1 diabetes. It is worth noting that this study, to maintain consistency with previous research, refers to asthma as an AD, actually meaning an immune-mediated disease.14,15 Sex hormones in this study referred to sex hormone-binding globulin (SHBG), total testosterone (TT), and bioavailable testosterone (BT). All summary statistics on exposures and outcomes were obtained from the IEU OpenGWAS project (https://gwas.mrcieu.ac.uk/). We preferred to adopt publicly downloadable GWAS data with larger samples and more recent dates for proper MR analysis. 14 All participants included in these studies were of European ancestry, and diagnostic criteria for diseases were documented in the original publications. More detailed information about the GWAS cohorts was shown in Supplementary Table S1 and S2.
2.3 IVs selection
We performed a series of rigorous screening on the summary data from the above-mentioned GWAS studies and obtained suitable SNPs as IVs. Firstly, only SNPs significantly associated with exposure were extracted (P < 5×10-8). Secondly, a clustering process was performed to avoid the impact of linkage disequilibrium under the following conditions (R2<0.001 and window size = 10,000 kb). Thirdly, r2 values for all SNPs were pooled to estimate the proportion of total exposure variance explained by IVs. The criterion of F > 10 was adopted to avoid the bias of weak IVs. 16 Fourthly, we used the phenoscanner database (https://www.phenoscanner.medschl.cam.ac.uk/) to remove SNPs related to possible confounding factors in IVs. Besides, the harmonization process was conducted across the exposure and outcome data sets.
2.4 Statistical analysis
We selected the inverse-variance weighted (IVW) method as the main method in the MR analysis, and also conducted verification by other methods, including MR-Egger, weighted median, weighted mode, and simple mode. Subsequently, a series of sensitivity analyses were conducted to test the robustness of the results. Cochran’s Q test based on IVW and MR-Egger was able to detect heterogeneity. The intercept calculated by MR-Egger and the results of MR-PRESSO analysis (the number of distributions defaulted to 1000) could reflect the existence of pleiotropy. In addition, MR-PRESSO analysis was also used to detect and exclude outliers. Leave-one-out analysis was applied to analyze whether a single SNP had a biasing effect on the whole. Visualizations included scatter plots and funnel plots based on the above analysis. Finally, we utilized online tools (https://sb452.shinyapps.io/power/) to calculate statistical power.
In this study, we utilized the “TwoSampleMR (version 0.5.7)” and “MRPRESSO (version 1.0)” packages in R software (version 4.1.0) to perform two-sample MR analysis. The difference was considered statistically significant when P < 0.05.
3. Results
3.1 Baseline characteristics
Overall, the number of SNPs used as IVs ranged from 7 to 146 (Supplementary Table S3-S5). Due to the lack of certain data in GWAS studies, we were unable to calculate r2 and F statistics in ankylosing spondylitis, psoriasis, rheumatoid arthritis, and systemic lupus erythematosus. Nonetheless, given that all instrumental variables were rigorously selected at the genome-wide significance level (P < 5×10-8), combined with the massive sample sizes of the included cohorts, the potential bias introduced by weak instruments is expected to be critically low. In the remaining GWAS cohorts, except for one SNP in CD with the F value less than 10 (rs25491056, F = 8.324), all other calculable F values were greater than 10.
3.2 The causal effect of ADs on ED
As shown in Figure 2, different ADs had different causal effects on ED. According to the results of IVW, IBD and CD were positively correlated with ED (IBD, OR: 1.101, 95% CI: 1.016-1.194, P = 0.019; CD, OR: 1.086, 95% CI: 1.012-1.165, P = 0.022). The analysis of MR Egger, weighted median, and weighted mode between IBD and ED also produced positive results. The statistical power of both was greater than 80%, indicating high reliability. Cochran’s Q test of IVW and MR-Egger showed no heterogeneity (all P > 0.05). The results of pleiotropy test were also negative (all P > 0.05), and no outliers were found (Table 1). The visualization results also verify the above results (Supplementary Figure S1). However, UC, which was also a type of IBD, had nothing to do with ED (OR: 1.112, 95% CI: 0.897-1.377, P = 0.333), and the variance explained was only 0.222%. Detailed information was listed in Supplementary Table S3. The causal effect of ADs on ED. Sensitivity analyses of MR analysis. MR, Mendelian randomization; IVW, inverse-variance weighted; ED, erectile dysfunction; IBD, inflammatory bowel disease; CD, Crohn’s disease; CeD, Celiac disease; TT, total testosterone; BT, bioavailable testosterone. aHeterogeneity test after removing outliers. bTo achieve a significance threshold of 0.05, the number of distributions was set to 2000. cTo achieve a significance threshold of 0.05, the number of distributions was set to 3000.
3.3 The causal effect of ADs on abnormal spermatozoa
Only asthma was positively associated with abnormal spermatozoa (OR: 1.270, 95% CI: 1.045-1.545, P = 0.017), and this result was confirmed by the results of MR-Egger and weighted median (Figure 3). The results of both the heterogeneity test and the pleiotropy test were negative (all P > 0.05), and no outliers were detected, accompanied by a statistical power greater than 80% (Table 1). The visualization results were consistent with the above results (Supplementary Figure S2). No causal relationship was found between other ADs and abnormal spermatozoa. Detailed information was listed in Supplementary Table S4. The causal effect of ADs on abnormal spermatozoa.
3.4 The causal effect of ADs on sex hormones
Based on the results of IVW, we found no causal association between ADs and SHBG (all P > 0.05; Figure 4). Focusing on the next indicator, we found that asthma and eczema were negatively associated with TT (asthma, β: -0.045, 95% CI: -0.063 - -0.027, P < 0.001; eczema, β: -0.268, 95% CI: -0.403 - -0.134, P < 0.001). For asthma, the results of sensitivity analysis suggested the existence of heterogeneity (IVW, P < 0.001; MR-Egger, P < 0.001) and horizontal pleiotropy (MR-PRESSO, P < 0.001; while MR-Egger, P > 0.05). After excluding 6 outliers, the results of re-MR analysis remained unchanged (β: -0.042, 95% CI: -0.057 - -0.027, P < 0.001), but heterogeneity still existed (IVW, P = 0.002; MR-Egger, P = 0.003). For eczema, trends in the results of sensitivity analysis and re-MR analysis were consistent with asthma (Table 1 and Figure 4). The causal effect of ADs on sex hormones.
The MR analysis also suggested that CeD and eczema were negatively associated with BT (CeD, β: -0.007, 95% CI: -0.014 - -0.000, P = 0.036; eczema, β: -0.108, 95% CI: -0.191 - -0.024, P = 0.012; Figure 4). The statistical results were as follows in both diseases: (1) Heterogeneity and pleiotropy were present in sensitivity analyses. (2) Heterogeneity still existed after excluding outliers. For CeD, no causal relationship was found in re-MR analysis (β: -0.006, 95% CI: -0.017 - 0.005, P = 0.305). For eczema, the results of re-MR analysis remained unchanged (β: -0.103, 95% CI: -0.172 - -0.034, P = 0.003; Table 1 and Figure 4). Except for CeD, the statistical power for all exposures and outcomes mentioned above was greater than 80%. The visualization results were in line with the results of the above analysis (Supplementary Figure S3). Detailed information was listed in Supplementary Table S5. It was worth noting that MR analysis and re-MR analysis results between Ced and BT should be interpreted as statistically inconclusive due to being underpowered at the current sample horizon.
4. Discussion
Epidemiological studies have found that ADs adversely affect male reproductive health. This study used GWAS genetic data to explore the causal relationship of ADs to male reproductive health, including ED, abnormal spermatozoa, and sex hormones. Our two-sample MR analysis provides genetic evidence suggesting a potential causal influence of IBD and CD on the risk of ED. Furthermore, genetic liability to asthma was found to be consistently linked with a higher risk of abnormal spermatozoa and a potential reduction in TT levels. Similarly, eczema exhibited a potential negative causal alignment with both TT and BT. Beyond these findings, no robust evidence was found to support genetic causal associations between other ADs and male reproductive health.
Given that most ADs are more common in women, the impact of ADs on male reproductive health is often overlooked, and there is also a little research data in this area. In previous studies, evidence suggests that ADs may affect male reproductive health in several ways.9–12 (1) Systemic inflammation or local inflammation caused by the disease itself spreads to the male reproductive system. Long-term chronic inflammation directly damages the corpus cavernosum of the penis and the spermatogenic cells in the testicles. Secondary pathological changes such as increased ROS production also aggravate tissue damage and dysfunction. (2) Multiple drugs used to control the progression of ADs may have a negative impact on male reproductive health. These drugs include cyclophosphamide, methotrexate, corticosteroids, anti-tumor necrosis factor drugs, and among others, but drug toxicity appears to be disease heterogeneous. (3) The mental and psychological pressure caused by long-term illness is not conducive to the normalization of male reproductive function. (4) The factors mentioned above also cause hormone disorders, especially sex hormones. Normal levels of testosterone are crucial to the normalization of erectile function and spermatogenesis. (5) There are special factors in some ADs: high levels of blood glucose in diabetes exacerbate the degree of ED and spermatogenesis impairment,17,18anti-sperm antibodies in systemic lupus erythematosus prevent the formation of fertilized eggs and normal embryonic development. 12
IBD includes two subtypes, CD and UC, and the active phase may be accompanied by symptoms such as diarrhea, abdominal pain, fatigue, and depression. A recent meta-analysis showed that IBD was significantly positively associated with ED, and that surgery, disease activity, and depression significantly increased the risk of ED in patients with IBD. 19 The above view was also confirmed by cross-sectional studies from the United States and China.20,21 Chronic inflammation, secondary hypogonadism, psychological stress, drugs, and unhealthy lifestyle were all related to ED caused by IBD. 22 Our results suggested a positive causal relationship of IBD and CD on ED. However, no causal relationship between UC and ED had been found. The explained variance was only 0.2%, and the subsequent calculated statistical power was only 5.5%, which may limit this study’s exploration of the relationship between UC and ED.
Asthma is a common respiratory disease characterized by inflammation of the airways. In a cross-sectional study of more than 6,000 men, men with asthma had lower sperm concentration and total sperm count than men without asthma. 23 Animal experiments showed that asthma-mediated inflammation and hypoxic environment activated the HIF-1 signaling pathway, thereby initiating the cell apoptosis process, resulting in a decrease in sperm number and activity and an increase in abnormal sperm.24,25 In addition, hypoxia caused by asthma could also lead to reduced testosterone levels and aggravate spermatogenesis disorders. 24 In this study, the result of MR analysis suggested a positive causal relationship between asthma and abnormal spermatozoa, which was consistent with previous research.
In the circulation, TT is predominantly bound to SHBG and albumin, with the remaining unbound portion and albumin-bound fraction constituting BT. 26 Testosterone is essential for testicular somatic cell maturation, germ cell meiosis, spermatogenesis, the normalization of the structure and function of the corpus cavernosum.27,28 In a large cross-sectional study, testosterone levels were inversely associated with the incidence of asthma in men. 29 Testosterone was found to appear to play a protective role in asthma in animal study with mice. 30 In this study, we also found that asthma had a negative causal relationship with TT, which was consistent with the results of asthma and abnormal spermatozoa.
So far, there have been few clinical studies on the relationship between eczema and sex hormones. Limited evidence suggested a higher incidence of eczema in adult women, and androgens seemed to have anti-inflammatory effects. 31 Meanwhile, our results implied eczema was negatively related to TT and BT, but the current data were not sufficient to support a causal relationship between eczema and other outcomes (ED and abnormal spermatozoa).
MR analysis is one of the greatest strengths of this study. Compared with observational studies, MR analysis can effectively reduce the impact of confounders and reverse causation, thereby enabling high-efficiency causal inference. Besides, the exposure and outcome data in this study were all from the large-sample GWAS database, which enhanced the ability to judge causal relationships. Furthermore, the rigorous selection of IVs and subsequent sensitivity analyses jointly enhanced the reliability of the results. Inevitably, some limitations also existed in this study. Firstly, the statistics in the MR analysis came only from the European population. Whether our results could be generalized to the entire population required further investigation. Secondly, the lack of individual details in GWAS data limited our exploration of overlap of between exposure and outcome populations. Because all selected genetic instruments possessed substantial statistical power, the potential biases and exaggerated Type 1 error rates introduced by any latent sample overlap are expected to be critically low, posing no threat to the structural validity of our causal inferences. Finally, even though we had excluded outliers using the MR-PRESSO method, heterogeneity still existed in the MR analysis between ADs and sex hormones. This may limit the reliability of the results in this part. Given that this study is essentially an exploratory screening research, the results should be interpreted with caution. To obtain more reliable and accurate results, further MR analysis of larger sample sizes or high-quality randomized controlled trials is still needed.
5. Conclusion
In conclusion, the genetic evidence from this MR analysis suggested that IBD and CD were positively associated with the risk of ED. Additionally, genetic liability to asthma was associated with abnormal spermatozoa and decreased total testosterone levels. These findings offer valuable etiological insights into the biological links between systemic autoimmunity and male reproductive health. Future research should focus on exploring the pathological mechanisms between the aforementioned ADs and male reproductive health, and on investigating potential therapies to slowdown disease progression.
Supplemental material
Supplemental material - Association between autoimmune diseases and male reproductive health: A mendelian randomization study
Supplemental material for Association between autoimmune diseases and male reproductive health: A mendelian randomization study by Taotao Sun, Yipiao Liu, Penghui Yuan, Zhankui Jia, Jinjian Yang in Science Progress
Supplemental material
Supplemental material - Association between autoimmune diseases and male reproductive health: A mendelian randomization study
Supplemental material for Association between autoimmune diseases and male reproductive health: A mendelian randomization study by Taotao Sun, Yipiao Liu, Penghui Yuan, Zhankui Jia, Jinjian Yang in Science Progress
Footnotes
Acknowledgements
We would like to thank all participants and investigators of the included GWAS studies. Parts of this work were previously accepted as a conference abstract for the 27th World Meeting on Sexual Medicine and the 22nd Urological Association of Asia Congress, but not formally presented due to scheduling conflicts.
Author contributions
TS and JY conceived and designed the experiments. TS, YL and PY contributed to data acquisition and analysis. TS and YL drafted the work. PY, ZJ and JY contributed to the revision of the manuscript. All authors read and approved the final manuscript.
Funding
This work was supported by the National Natural Science Foundation of China (No. 82201775), the China Postdoctoral Science Foundation (No.2024M763006), the Henan Provincial Natural Science Foundation Youth Project (No.252300420540), and the Medical Science and Technology Research-related joint construction project of Henan Province (No. LHGJ20220343).
Declaration of conflicting interests
The authors declare no competing interests.
Data Availability Statement
All data used in this study are included in the article or uploaded as supplementary materials. There are no additional, unpublished data available from this study. Software codes and data are available upon request.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
