The causal relationship between serum metabolites and the development of infertility: A Mendelian randomization analysis.

In conclusion, this MR study provides robust evidence supporting a causal role of specific serum metabolites in the risk of both male and female infertility. By leveraging large-scale genetic datasets, we identified several metabolites (including bilirubin, 1,5-AG, and glutaroyl carnitine) that significantly influence infertility risk, highlighting the potential involvement of metabolic pathways related to oxidative stress, glucose metabolism, and mitochondrial dysfunction in reproductive health. These findings not only enhance our understanding of the etiological mechanisms underlying infertility but also underscore the value of integrating genomic and metabolomic data to elucidate biologically plausible risk factors. Further research is warranted to validate these associations, clarify the biological functions of unidentified metabolites, and explore the translational potential of these findings in clinical diagnostics and targeted therapeutic strategies.

Metabolites play a pivotal role in the fundamental biological processes of the human body, serving as intermediates or final products of metabolism. [ 1 ] Serum metabolites, which are detectable in an individual’s blood serum, refer to the metabolites present in the liquid component of blood subsequent to coagulation and the elimination of blood cells. [ 2 ] The quantification and examination of serum metabolites offer significant insights into an individual’s metabolic condition and general well-being. The field of serum metabolomics, which focuses on the analysis of metabolites found in blood serum, has garnered considerable interest in recent times due to its potential applications in disease diagnosis, prognosis, and treatment monitoring. [ 3 – 5 ] Through the examination of serum metabolite levels and patterns, researchers are able to identify metabolic irregularities linked to a range of diseases, encompassing cardiovascular disease, [ 6 ] diabetes, [ 7 ] cancer, [ 8 ] and neurological disorders. [ 9 ] In conclusion, serum metabolites serve as a valuable asset in comprehending an individual’s metabolic condition and well-being. The examination of serum metabolites holds the potential to enhance disease identification, treatment, and prevention approaches while also offering valuable insights into the influence of lifestyle factors on metabolism. Infertility is a complex and sensitive phenomenon encountered by numerous couples globally. It encompasses the failure to achieve pregnancy after a year or more of unprotected intercourse, as well as the incapacity to sustain a pregnancy until its completion. [ 10 ] The origins of infertility can be attributed to a range of factors, including reproductive system dysfunctions, [ 11 ] hormonal imbalances [ 12 ] and structural anomalies. [ 13 ] This deeply personal and psychologically impactful ordeal can profoundly affect individuals and couples on physical, emotional, and psychological planes. [ 14 ] Consequently, comprehending the etiology, available treatments, and support systems for infertility is of utmost importance in the academic realm. The investigation of the correlation between serum metabolites and infertility is a subject that continues to be explored academically. Numerous studies have examined the potential links between serum metabolites and infertility in both males and females. For instance, in females, changes in metabolite profiles have been observed in conditions such as polycystic ovary syndrome, [ 15 ] endometriosis, [ 16 ] and unexplained infertility. [ 17 ] Metabolomic profiling of serum has been employed to discern potential biomarkers associated with these conditions and gain a deeper understanding of fundamental metabolic dysregulations. Furthermore, investigations in men have also explored the involvement of serum metabolites in infertility. Certain studies propose that distinct metabolites may be linked to compromised sperm function, [ 18 ] diminished sperm count or motility, and other factors that may contribute to male infertility. Mendelian randomization (MR) is a statistical technique that employs genetic variants as instrumental variables (IVs) to estimate the causal impact of an exposure or risk factor on a specific outcome. [ 19 ] This methodology capitalizes on the fortuitous allocation of genetic variants during meiosis, which serves as a natural randomization mechanism akin to a randomized controlled trial. The fundamental tenet of MR posits that genetic variants are inherited in a random manner, thereby minimizing the influence of confounding factors or reverse causation. [ 20 ] By employing these genetic variants as surrogates for the exposure under investigation, MR can furnish substantiation for a causal association between the exposure and the outcome.

Conceptualization: Sicheng Ma, Wenbang Liu. Funding acquisition: Chenming Zhang, Zixue Sun. Methodology: Jing Hu, Wenlin Yu. Resources: Yizhe Gao, Yifei Wang. Writing – original draft: Yinuo Zhang, Sicheng Ma. Writing – review & editing: Tingyu Wang.

This study used a summary dataset from genome-wide association studies (GWAS) to conduct 2-sample MR analysis to evaluate the causal relationship between serum metabolites and infertility. In addition, sensitivity analysis was performed to test the reliability of the results. MR uses genetic variants as IVs to infer potential causal relationships between exposure and outcome, with its core lying in the premise that genotypes are randomly allocated at fertilization, which can effectively avoid confounding and reverse causation problems in traditional observational studies. This study utilized summary-level data from previously published GWAS that had already received ethical clearance from the respective institutional review board. Therefore, no additional ethical approval was necessary for this study. Genetic data for these metabolites were obtained from the metabolomics GWAS server. By performing genome-wide association scans, the researchers identified almost 2.1 million single nucleotide polymorphisms (SNPs) associated with 486 distinct metabolites in human genetic variants. The study included a large cohort of 7824 individuals of European ancestry, comprising 1768 participants aged 32 to 77 from the cooperative health research in the Augsburg region F4 study in Germany and 6056 participants aged 16 to 85 from the United Kingdom Twin Study. Remarkably, 107 out of the 486 investigated metabolites remain unknown, awaiting further understanding of their chemical characteristics. [ 21 ] Nonetheless, the study successfully authenticated and categorized 309 metabolites into 8 distinct metabolic categories, referencing the trusted Kyoto Encyclopedia of Genes and Genomes database. Infertility data were obtained from the Finnish Genome Project Consortium R9 release, [ 22 ] with 680 male infertility cases and 72,799 controls and 6481 female infertility cases and 68,969 controls with diagnostic criteria for infertility based on International Classification of Diseases (ICD)8, ICD9, and ICD10. Assumption of association: Only by ensuring a significant association between genetic variation and exposure can the accuracy of research results be ensured. Studying genetic variations with weak associations may lead to biased results. Assumption of independence: Genetic variation must be independent of confounding factors that affect “exposure and outcome.” Therefore, when choosing the IV, it must be ensured that the IV is only related to the exposure factor. Rigorous variable selection and control can improve the reliability and effectiveness of the results. Assumption of exclusivity: Genetic variation can only affect the outcome through exposure and not through other pathways. Only by excluding other possible pathways can accurate causal inference conclusions be drawn. To identify IVs for our study, we selected SNPs up to the genome-wide significance threshold ( P  < 5*10 −8 ). However, due to the limited number of qualifying SNPs, we expanded our criteria. Significant IVs were identified using a P  < 1*10 −5 threshold based on previously published studies. To ensure no linkage disequilibrium between IVs, the clumping process ( R 2  < 0.1, clumping distance = 500 kb) was used to assess linkage disequilibrium among the included SNPs for MR analysis. [ 23 ] We extracted information on SNPs related to serum metabolites from summary GWAS data of infertility, removing missing SNPs and setting the minor allele frequency at 0.01. Palindromic SNPs were excluded to prevent allele effects on the results. We used the f -value to detect bias in the causal relationship between serum metabolites and infertility due to weak IVs, deleting SNPs with an F -statistic < 10. This study used 3 different methods to explore the causal effect between serum metabolomics and infertility, including inverse variance weighted (IVW) analysis, the MR-Egger method, and weighted median estimator (WME) in the MR analysis. The IVW method does not consider the intercept term and only uses the weights of the inverse variances of each IV to fit the data. [ 24 ] By calculating the weighted average of IV effect estimates, the final results can be obtained. The MR-Egger method serves as an extension of the MR method, specifically designed to assess the assumptions of MR pertaining to balance and unbiasedness. [ 25 ] Within the framework of MR-Egger regression analysis, the estimation of a causal effect is achieved by regressing the genetic variation against the outcome variable. Diverging from conventional regression analysis, MR-Egger regression analysis mitigates the influence of confounding factors by incorporating the stochastic nature of genetic variation, thereby enhancing the dependability of causal inference. The WME method, a frequently employed weighted calculation approach, necessitates an effective IV surpassing 50%. [ 26 ] When organizing SNPs, it is advisable to arrange them based on their weights and subsequently derive the median as the outcome. This method consistently yields dependable causal estimates, thereby justifying its extensive utilization in practical scenarios. Moreover, it significantly enhances the precision and dependability of data analysis. Consequently, the WME method is strongly recommended for studies that necessitate causal estimation. To ensure the accuracy and stability of the test results, statistical software R 4.3.1 was used in this study. To verify the reliability of the IVW method and ensure the accuracy of the research results, heterogeneity tests are used. The heterogeneity between each IV is examined using the Cochran Q test, where larger differences indicate stronger heterogeneity. MR–Egger regression and MR-Pleiotropy RESidual Sum and Outlier (MR-PRESSO) methods were used to explore the gene pleiotropy between SNPs and outcomes. [ 27 ] When the intercept term of MR-Egger regression is close to 0, it indicates a lower possibility of gene pleiotropy. The core of the MR-PRESSO method is to calculate the IVW results after removing each SNP and calculate the sum of squared residuals between the effect of each SNP and the IVW results. Finally, by summing the squared residuals of all SNPs, a larger value indicates a more significant presence of pleiotropy. If pleiotropy was detected, the MR-PRESSO method was used to remove outlier SNPs and estimate the results after correction. These methods aim to reduce the impact of gene pleiotropy on the research results. In this study, 3 regression models, namely, MR-Egger, IVW, and WME, from the TwoSampleMR package in R 4.3.1 software were used to verify the causal relationship between serum metabolites and the risk of male infertility. [ 28 ] The significance level was set at α  = 0.05, and the final results are presented as odds ratios (OR) with 95% confidence intervals (CI), where P  < .05 indicated statistical significance.

After controlling for IV quality, a total of 486 metabolites and 10,877 SNPs were included in the MR study. The filtered IVs consisted of SNPs ranging from 3 to 497, with fructose having 3 SNPs and 2-methoxyacetaminophen sulfate having the highest number of genetically proxied SNPs at 497. Furthermore, all of the identified IVs showed a stronger association with exposure than with outcome ( P exposure  10, indicating that the IVs were strong and valid instruments. For more detailed information on the identified IVs, please refer to Table S1 , Supplemental Digital Content 1. A total of 369 SNPs associated with confounders were excluded; please refer to Table S2 , Supplemental Digital Content 2. As shown in Figure 1 , 19 serum metabolites were found to be statistically significant in the MR analysis with male infertility, with 1 belonging to the amino acid class, 1 to the carbohydrate class, 1 to the cofactors and vitamins class, 7 to the lipid class, 1 to the nucleotide class, 2 to the peptide class, and 6 to the unknown class. Among these metabolites, amino acid class alanine (OR: 6.384, 95% CI: 1.595–25.547, P  = .009), carbohydrate class erythritol (OR: 4.396, 95% CI: 1.241–15.579, P  = .022), cofactor and vitamin class bilirubin (OR: 1.791, 95% CI: 1.013–3.167, P  = .045), lipid class linoleic acid (OR: 7.457, 95% CI: 1.298–42.832, P  = .024), lauric acid (OR: 4.302, 95% CI: 1.254–14.756, P  = .020), 1-stearoylglycerol (OR: 3.355, 95% CI: 1.007–11.184, P  = .049), and hexanoic acid (OR: 7.662, 95% CI: 1.820–32.257, P  = .005), peptide class HWESASXX (OR: 3.020, 95% CI: 1.149–7.940, P  = .025), as well as unknown metabolite class X-08402 (OR: 4.298, 95% CI: 1.645–11.229, P  = .003) and X-11444 (OR: 2.763, 95% CI: 1.087–7.023, P  = .033) were found to be risk factors for male infertility. Lipid class 10-undecenoate (OR: 0.549, 95% CI: 0.311–0.967, P  = .038), 1-docosahexaenoylglycerophosphocholine (OR: 0.177, 95% CI: 0.036–0.885, P  = .035), octadecanedioate (OR: 0.268, 95% CI: 0.094–0.762, P  = .014), nucleotide class N2,N2-dimethylguanosine (OR: 0.313, 95% CI: 0.127–0.772, P  = .012), peptide class aspartylphenylalanine (OR: 0.229, 95% CI: 0.078–0.668, P  = .007), and unknown metabolite class X-10810 (OR: 0.327, 95% CI: 0.139–0.733, P  = .011), X-12798 (OR: 0.342, 95% CI: 0.179–0.651, P  = .001), X-13741 (OR: 0.439, 95% CI: 0.207–0.931, P  = .032) and X-14662 (OR: 0.655, 95% CI: 0.450–0.952, P  = .026) were found to be protective against male infertility. Please refer to Table S3 , Supplemental Digital Content 3 for detailed causal effects of the 19 serum metabolites on male infertility. Table S4 , Supplemental Digital Content 4 shows detailed causal effects of the 486 serum metabolites on male infertility. Forest plot for the causality of blood metabolites on male infertility derived from IVW analysis. CI = confidence interval, IVW = inverse variance weighted, OR = odds ratio. The 3 metabolites, X-10810 ( P IVW  = .010, P MR-Egger  = .006, P WME  = .045), bilirubin ( P IVW  = .045, P MR-Egger  = .035, P WME  = .039), and X-12798 ( P IVW  = .001, P MR-Egger  = .028, P WME  = .022), consistently exhibited strong cause-effect relationships across all MR methods. Figure 2 displays the genetically predicted impact magnitudes of these metabolites on male infertility. Three metabolites exhibited strong associations consistently across all MR methods. Each data point on the graph corresponds to an IV, with the line associated with each point representing the 95% CI. The horizontal axis represents the impact of SNPs on exposure, while the vertical axis represents the effect of SNPs on the outcome. The colored lines depict the outcomes of MR fitting. By analyzing the slope of the line, a positive value signifies a risk factor, whereas a negative value indicates a protective factor. CI = confidence interval, IV = instrumental variable, IVW = inverse variance weighted, MR = Mendelian randomization, OR = odds ratio, SNP = single nucleotide polymorphism, WME = weighted median estimator. As shown in Table 1 , the result of the Cochran Q test indicates that there is no heterogeneity. The MR-Egger intercept test and MR-PRESSO global test results indicated that there was no horizontal pleiotropy in the IVs of 19 serum metabolites associated with male infertility. The P values for both the MR-Egger intercept test and the global MR-PRESSO test were > .05. Sensitivity analysis results for MR analysis of serum metabolites on male infertility. MR = Mendelian randomization, MR-PRESSO = Mendelian Randomization Pleiotropy RESidual Sum and Outlier. After controlling for IV quality, a total of 486 metabolites and 10,877 SNPs were included in the MR study. The filtered IVs consisted of SNPs ranging from 3 to 497, with fructose having 3 SNPs and 2-methoxyacetaminophen sulfate having the highest number of genetically proxied SNPs at 497. Furthermore, all of the identified IVs showed a stronger association with exposure than with outcome ( P exposure   10, indicating that the IVs were strong and valid instruments. For more detailed information on the identified IVs, please refer to Table S5 , Supplemental Digital Content 5. As shown in Figure 3 , 29 serum metabolites were found to be statistically significant in the MR analysis with female infertility, with 2 belonging to the amino acid class, 2 to the carbohydrate class, 6 to the lipid class, 1 to the nucleotide class, 4 to the peptide class, 13 to the unknown class, and 1 to the xenobiotics class. Among these metabolites, amino acid class glutaroyl carnitine (OR: 1.461, 95% CI: 1.135–1.882, P  = .003), carbohydrate class glucose (OR: 2.213, 95% CI: 1.296–3.778, P  = .004), lipid class 1-linoleoylglycerol (OR: 1.270, 95% CI: 1.027–1.570, P  = .028), propionylcarnitine (OR: 1.669, 95% CI: 1.181–2.360, P  = .004), docosapentaenoate (OR: 1.478, 95% CI: 1.050–2.081, P  = .025), peptide class gamma-glutamylisoleucine (OR: 1.419, 95% CI: 1.002–2.011, P  = .049), unknown metabolite class X-11552 (OR: 1.523, 95% CI: 1.153–2.011, P  = .003), X-11787 (OR:1.838, 95% CI: 1.069–3.159, P:0.028), X-12092 (OR: 1.102, 95% CI: 1.029–1.179, P  = .005), X-12231 (OR: 1.341, 95% CI: 1.039–1.731, P  = .024), X-12844 (OR: 1.504, 95% CI: 1.025–2.206, P  = .037), X-12850 (OR: 1.254, 95% CI: 1.013–1.552, P  = .037), X-13619 (OR: 1.869, 95% CI: 1.043–3.348, P  = .036), and xenobiotics class 7-methylxanthine (OR: 1.294, 95% CI: 1.010–1.658, P  = .041) were found to be risk factors for female infertility. Amino acid class serotonin (5-HT) (OR: 0.675, 95% CI: 0.486–0.938, P  = .019), carbohydrate class 1,5-anhydroglucitol (1,5-AG) (OR: 0.677, 95% CI: 0.493–0.931, P  = .016), lipid class epiandrosterone sulfate (OR: 0.841, 95% CI: 0.752–0.941, P  = .002), 2-oleoylglycerophosphocholine (OR: 0.562, 95% CI: 0.337–0.936, P  = .027), 1-myristoylglycerophosphocholine (OR: 0.549, 95% CI: 0.357–0.844, P  = .006), nucleotide class 7-methylguanine (OR: 0.588, 95% CI: 0.399–0.867, P  = .007), peptide class glycylvaline (OR: 0.694, 95% CI: 0.533–0.903, P  = .006), DSGEGDFXAEGGGVR (OR: 0.738, 95% CI: 0.584–0.933, P  = .011), ADpSGEGDFXAEGGGVR (OR: 0.623, 95% CI: 0.449–0.864, P  = .005), and unknown metabolite class X-06307 (OR: 0.541, 95% CI: 0.307–0.954, P  = .034), X-08988 (OR: 0.643, 95% CI: 0.493–0.839, P  = .001), X-11529 (OR: 0.915, 95% CI: 0.850–0.985, P:0.018), X-11538 (OR: 0.813, 95% CI: 0.680–0.971, P  = .022), X-12063 (OR: 0.761, 95% CI: 0.660–0.878, P  = .000), and X-12851 (OR: 0.940, 95% CI: 0.896–0.998, P  = .042) were found to be protective against female infertility. Please refer to Table S6 , Supplemental Digital Content 6 for detailed causal effects of the 29 serum metabolites on female infertility. Table S7 , Supplemental Digital Content 7 shows detailed causal effects of the 486 serum metabolites on female infertility. Forest plot for the causality of blood metabolites on female infertility derived from IVW analysis. 5HT = 5-hydroxytryptamine, CI = confidence interval, IVW = inverse variance weighted, OR = odds ratio. 1,5-AG ( P IVW  = .016, P MR-Egger  = .001, P WME  = .003), X-08988 ( P IVW  = .001, P MR-Egger  = .003, P WME  = .021), X-11529 ( P IVW  = .017, P MR-Egger  = .012, P WME  = .003), X-12063 ( P IVW  = .000, P MR-Egger  = .011, P WME  = .002), and glutaroyl carnitine ( P IVW  = .003, P MR-Egger  = .001, P WME  = .015) exhibited strong cause–effect relationships consistently across all supplementary MR methods. Figure 4 displays the genetically predicted impact magnitudes of these metabolites on female infertility. Five metabolites exhibited strong associations consistently across all MR methods. Each data point on the graph corresponds to an IV, with the line associated with each point representing the 95% CI. The horizontal axis represents the impact of SNPs on exposure, while the vertical axis represents the effect of SNPs on the outcome. The colored lines depict the outcomes of MR fitting. By analyzing the slope of the line, a positive value signifies a risk factor, whereas a negative value indicates a protective factor. CI = confidence interval, IV = instrumental variable, IVW = inverse variance weighted, MR = Mendelian randomization, OR = odds ratio, SNP = single nucleotide polymorphism, WME = weighted median estimator. As shown in Table 2 , the result of the Cochran Q test indicates that there is no heterogeneity. The MR–Egger intercept test and MR-PRESSO global test results indicated that there was no horizontal pleiotropy in the IVs of 29 serum metabolite taxa associated with female infertility. The P values for the MR–Egger intercept test and global MR-PRESSO test were both > .05. Sensitivity analysis results for MR analysis of serum metabolites on female infertility. 5HT = 5-hydroxytryptamine, MR = Mendelian randomization, MR-PRESSO = Mendelian Randomization Pleiotropy RESidual Sum and Outlier.

In this study, we integrated 2 extensive GWAS datasets to examine the potential causal impacts of 486 blood metabolites on infertility through a robust MR design. A total of 19 serum metabolites were identified as being causally related to male infertility, with 3 of them showing a strong causal relationship. Furthermore, 29 serum metabolites were discovered to be causally associated with infertility, with 5 of them demonstrating a strong causal relationship. Our results demonstrated that individuals with genetically determined elevated levels of X-10810 and X-12798 exhibited a decreased likelihood of experiencing male infertility. Conversely, those with a genetic predisposition to elevated bilirubin levels displayed an augmented risk of male infertility. Individuals with genetically determined elevated levels of 1,5-AG, X-08988, X-11529 and X-12063 exhibited a decreased likelihood of experiencing female infertility. Conversely, those with a genetic predisposition to elevated glutaroyl carnitine levels displayed an augmented risk of female infertility. To the best of our knowledge, this study is the first to represent the initial integration of metabolomics and genomics, aiming to assess the causal impact of serum metabolites on infertility. Our investigation offers original perspectives that elucidate the involvement of genetic-environmental interactions in the etiology of human diseases. Male infertility is defined as the circumstance in which a couple engages in regular sexual intercourse without employing any contraceptive methods for a duration exceeding 1 year, yet the female partner is unable to conceive as a result of male-related factors. [ 29 ] Statistical data indicate that the worldwide prevalence of infertility ranges from 8 to 12%, with approximately half of these cases attributed to male factors. [ 30 ] Male fertility can be influenced by a range of factors, including age, [ 31 ] detrimental habits such as smoking [ 32 ] and excessive alcohol consumption, [ 33 ] obesity, [ 34 ] unhealthy lifestyle practices such as inadequate sleep [ 35 ] and lack of physical exercise, [ 36 ] exposure to various environmental chemicals, ionizing radiation, [ 37 ] specific heavy metals, [ 38 ] toxic gases, [ 39 ] prolonged exposure to high temperatures, [ 40 ] and diminished psychological well-being. [ 41 ] Previous studies have provided evidence of a plausible correlation between serum metabolites and male infertility. Serum metabolites have been found to exhibit changes in their concentrations or compositions that are associated with compromised sperm production, functionality, and overall male fertility. Studies have identified certain metabolites that potentially contribute to male infertility. Notably, dysregulations in lipid metabolism, characterized by heightened triglyceride and cholesterol levels, have been correlated with diminished sperm quality and fertility prospects. [ 42 ] Likewise, perturbations in amino acid metabolism, specifically arginine [ 43 ] and carnitine, [ 44 ] have been associated with compromised sperm motility and viability. The pursuit of novel biomarkers is a cornerstone of modern medicine, as exemplified in neurology where biomarkers like neurofilament light chain have revolutionized the monitoring of neuroaxonal damage in multiple sclerosis. [ 45 ] Our study identifies specific serum metabolites as potential causal biomarkers for infertility, underscoring the translational potential of biomarker research across disparate medical fields. Our study found that a genetic predisposition to elevated bilirubin levels displayed an augmented risk of male infertility. Bilirubin, a yellow pigment resulting from heme degradation, is predominantly metabolized and excreted by the liver through bile excretion. [ 46 ] Elevated concentrations of bilirubin in the bloodstream, referred to as hyperbilirubinemia, can serve as a diagnostic marker for liver dysfunction and other pathological conditions. [ 47 ] Current evidence indicates a potential association between bilirubin and male infertility, with suggestions that heightened bilirubin levels may detrimentally impact sperm quality and function. [ 48 ] Numerous studies have documented inverse associations between heightened bilirubin levels and various sperm parameters, [ 49 ] such as diminished motility and viability. [ 50 ] Therefore, our investigations definitively establish a causal connection and elucidate the intricate mechanisms underlying the potential influence of bilirubin on male infertility. This study provides new ideas for the diagnosis and treatment of male infertility. Furthermore, our investigation identified 2 serum metabolites that exhibit a potential protective role against male infertility. Regrettably, the precise nature of these metabolites remains ambiguous, necessitating further research to comprehensively elucidate their therapeutic implications for male infertility. Female infertility is a multifaceted reproductive health condition characterized by the incapacity of a woman to achieve and maintain a pregnancy to full term, even after engaging in regular, unprotected sexual intercourse for a minimum duration of 1 year. [ 51 ] This phenomenon is influenced by multiple factors of diverse origins, which collectively contribute to compromised fertility outcomes. Prominent causative factors encompass ovulatory disorders, [ 52 ] fallopian tube abnormalities, [ 53 ] uterine or cervical abnormalities, [ 54 ] endometriosis, [ 55 ] age-related factors, and other genetic [ 51 ] and lifestyle factors. [ 56 ] Notably, endocrine dysregulation is a cornerstone of many ovulatory disorders. In this context, the kisspeptin system, a key regulator of the hypothalamic-pituitary-gonadal axis, has emerged as a critical area of investigation. [ 57 , 58 ] A recent prospective case-control study demonstrated that serum kisspeptin levels were significantly elevated in subfertile women with polycystic ovary syndrome, a leading cause of anovulatory infertility, and were strongly correlated with increased luteinizing hormone levels. [ 59 ] This highlights the profound impact of hypothalamic-pituitary-gonadal axis dysregulation on fertility, suggesting that our identified metabolites might interact with or reflect similar underlying endocrine pathways. The investigation of the correlation between serum metabolites and female infertility has become a prominent focus in the field of reproductive medicine. [ 60 ] Dysregulated levels or imbalances of serum metabolites have been implicated in the pathophysiology of female infertility. A multitude of studies have examined the relationship between serum metabolites and various aspects of female reproductive health, such as ovarian function, [ 61 ] ovulation, [ 62 ] and embryo implantation. [ 63 ] Metabolomic analyses have revealed distinct categories of metabolites linked to female infertility, encompassing amino acids, [ 64 ] lipids [ 65 ] and oxidative stress-related molecules. [ 66 ] Perturbed amino acid metabolism, modified lipid composition, and heightened oxidative stress have all been implicated in the etiology of female infertility. Our study found that individuals with genetically determined elevated levels of 1,5-AG exhibited a decreased likelihood of experiencing female infertility. 1,5-AG serves as a distinctive glycemic marker, indicative of immediate glucose fluctuations and postprandial hyperglycemia. [ 67 ] Consequently, it has been suggested as a prospective biomarker for evaluating glucose metabolism and glycemic regulation in various individuals, including those afflicted with diabetes. [ 68 ] Our research findings indicate that 1,5-AG may serve as a protective factor against female infertility. However, the relationship between 1,5-AG and female infertility has not been extensively investigated or firmly established in the existing scientific literature. Nevertheless, these findings offer potential avenues for novel treatment strategies in addressing female infertility. Furthermore, our investigation identified 3 serum metabolites that exhibit a potential protective role against female infertility. Regrettably, the precise nature of these metabolites remains ambiguous, necessitating further research to comprehensively elucidate their therapeutic implications for female infertility. Our study found that a genetic predisposition to elevated glutaroyl carnitine levels displayed an augmented risk of female infertility. Glutaroyl carnitine serves as a biomarker for the identification of glutaryl-CoA dehydrogenase deficiency, [ 69 ] an infrequent hereditary disorder that disrupts the metabolic processes of lysine, hydroxylysine, and tryptophan. The insufficiency of glutaryl-CoA dehydrogenase results in the buildup of glutaric acid and its associated metabolites within diverse bodily tissues and fluids, thereby instigating potential neurological impairments [ 69 ] and motor dysfunctions. [ 70 ] The correlation between glutaroyl carnitine and female infertility has not been thoroughly investigated or firmly established within the academic literature. Glutaroyl carnitine is predominantly recognized as a biomarker for metabolic disorders, specifically glutaric aciduria and its associated ailments. In contrast, female infertility is a multifaceted condition characterized by various potential etiologies, encompassing hormonal imbalances, structural anomalies within the reproductive system, ovulatory dysfunction, and genetic predispositions. Although certain metabolic disorders may conceivably impact fertility, [ 71 ] the existing body of scientific research lacks adequate evidence to establish a direct association between glutaroyl carnitine and female infertility. Our study has certain limitations. First, the metabolites identified in our study as being associated with infertility need further validation through additional research. Additionally, there are 5 unknown metabolites that require further investigation for identification. Our findings represent the initial discovery phase, and future work must focus on the analytical and clinical validation of these metabolic biomarkers in diverse and larger cohorts. Second, this study was conducted based on a European population database, and further research is needed to study diverse ethnic samples to more accurately assess the genetic influences on metabolites. It is noteworthy that the metabolomics and MR methods employed in this study still face challenges in promotion within routine clinical practice, and their future clinical translation will depend on further simplification of detection technologies and cost reduction.