Unveiling the role of lipid metabolism in haemorrhagic disorders: genetic insights and therapeutic perspectives

Abstract

Background

See More

Coagulation defects, including purpura and other haemorrhagic conditions, are a critical area of medical research because of their significant health effects worldwide. Understanding the metabolic basis of these conditions may improve therapeutic strategies.

Methods

A two-sample Mendelian randomization (MR) approach was employed to evaluate the causal relationships between the levels of 1,400 metabolites and coagulation defects. Colocalization analysis confirmed significant shared genetic influences. Pathway and protein‒protein interaction (PPI) analyses identified rate-limiting enzymes and drug targets. The impacts of lifestyle factors on metabolite levels were also explored through MR.

Results

MR analysis revealed four metabolites whose abundance was significantly associated with coagulation defects: docosapentaenoate n3 DPA 22:5n3 (DPA) (OR: 1.594, 95% CI: 1.263–2.011, P < 0.001), 1-palmitoyl-2-stearoyl-gpc (PSPC) (16:0/18:0) (OR: 1.294, 95% CI: 1.134–1.477, P < 0.001), 1-stearoyl-2-docosahexaenoyl-gpc (SDPC) (18:0/22:6) (OR: 1.232, 95% CI: 1.101–1.380, P < 0.001) and hydroxypalmitoyl sphingomyelin (HPSM) (d18:1/16:0 (OH)) (OR: 0.803, 95% CI: 0.719–0.896, P < 0.001). Colocalization analysis provided robust evidence for shared genetic loci. Pathway analysis highlighted the importance of lipid metabolism, identifying key enzymes such as FADS1, FADS2 and TCP1. PPI analysis revealed an interaction between TCP1 and plasminogen, indicating potential therapeutic synergy. Further analysis revealed that lifestyle factors, including dried fruit and oily fish intake, were linked to the abundance of metabolites associated with coagulation risk.

Conclusions

This study identifies specific metabolites and metabolic pathways involved in coagulation defects, proposes novel therapeutic targets and highlights the roles of dietary and lifestyle interventions in the management of these conditions. These findings pave the way for personalized strategies to manage coagulation-related conditions.

Introduction

Coagulation defects, including purpura and other haemorrhagic conditions, remain significant clinical and public health concerns because of their high morbidity and mortality worldwide [1, 2]. These disorders often result from complex disruptions in coagulation pathways involving coagulation factors, platelets, and endothelial function [3, 4]. Therapeutic approaches for coagulation defects focus primarily on managing symptoms and preventing complications [5]. Nevertheless, the complex pathophysiology of these conditions makes their management particularly challenging for healthcare providers.

Recent advances in metabolomic analysis have revolutionized our understanding of disease mechanisms by associating metabolic alterations with pathological conditions and genetic predispositions [6, 7]. The relationships between metabolic processes and coagulation have attracted increasing attention in recent years. For example, phosphatidylserine plays a crucial role in blood clotting through its ability to form a prothrombinase complex, and the exposure of phosphatidylserine on activated platelet membranes is a pivotal step in the coagulation process [8, 9]. However, establishing causal relationships between the levels of specific metabolites and coagulation defects remains challenging because of the complex nature of metabolic networks, and new approaches are needed.

Mendelian randomization (MR) has emerged as a tool for inferring causal relationships in complex biological systems [10]. Recent successful applications of MR in cardiovascular and metabolic disease research have demonstrated its utility in identifying causal risk factors and potential therapeutic targets [11, 12]. By utilizing genetic variants as instrumental variables, MR analysis can distinguish markers of disease processes from causal metabolites, providing a novel strategy to identify new therapeutic targets and pathways.

Colocalization analysis facilitates the identification of genetic regions where variants simultaneously influence both metabolite levels and disease susceptibility, strengthening the evidence for causal relationships [13]. Furthermore, examining protein‒protein interactions between rate-limiting enzymes and known drug targets may reveal novel therapeutic strategies and drug repositioning opportunities [14]. The integration of colocalization analysis and protein‒protein interaction networks provides more comprehensive insights into the mechanisms underlying disease processes. Lifestyle factors have long been recognized as important modulators of disease risk, but their specific effects on metabolic pathways related to coagulation disorders remain poorly understood [15]. Understanding how lifestyle modifications influence relevant metabolites may provide critical insights for developing preventive strategies and personalized interventions.

This study aims to address these knowledge gaps related to the relationships between the levels of metabolites and coagulation defects via a multifaceted approach. By combining MR analysis, colocalization studies, pathway analysis, protein‒protein interaction (PPI) network analysis and lifestyle factor assessments, we endeavoured to (1) identify causal relationships between the levels of specific metabolites and coagulation defects through robust genetic approaches, (2) validate these relationships through colocalization analysis to confirm shared genetic architecture, (3) elucidate the biological pathways and rate-limiting enzymes involved in these associations, (4) investigate potential therapeutic targets through PPI network analysis, and (5) assess the influences of lifestyle factors on disease-associated metabolites. Thus, this study not only improves our understanding of the metabolic basis of coagulation defects but also has important implications for clinical practice.

Methods

Study design

Figure 1 outlines the study workflow. We first applied two-sample MR to assess causal relationships between the levels of 1,400 plasma metabolites and coagulation defects [16]. Through this approach, genetic variants were leveraged to infer causality, offering mechanistic insights into metabolite‒disease associations. To further validate these findings, the colocalization analysis was performed to identify loci where the same variant influenced both metabolite levels and coagulation risk, thereby validating shared genetic signals [13]. Pathway enrichment analysis was subsequently conducted to explore relevant biological pathways, with a focus on rate-limiting enzymes because of their central roles in metabolic control [17]. PPI network analysis was then employed to examine the interactions between these enzymes and existing drug targets, offering insights into possible therapeutic strategies [18]. Finally, this study focused on lifestyle factors by applying MR to analyse the relationships between lifestyle habits and metabolites associated with coagulation defects. By integrating findings from metabolomic, genetic and lifestyle data, this study offers a comprehensive framework for understanding the metabolic underpinnings of coagulation defects, highlighting actionable targets for therapeutic development and prevention strategies.

Fig. 1

This flowchart shows the comprehensive analytical workflow employed in the study

Data sources

We obtained summary statistics for 1,091 metabolites and 309 metabolite ratios from the GWAS Catalogue on the basis of plasma samples from the Canadian Longitudinal Study on Aging (CLSA) [19, 20]. Variants were selected using stringent criteria of P-value < 5e-8, R² < 0.001, and a linkage disequilibrium (LD) threshold of 10,000 kb. Metabolites were categorized into eight superpathways: lipid, amino acid, xenobiotic, nucleotide, cofactor and vitamin, carbohydrate, peptide, and energy pathways. From this comprehensive dataset, 690 metabolites were found to be significantly associated with 248 loci, whereas 143 metabolite ratios were linked to 69 loci. Among the 850 previously characterized metabolites, 81 novel metabolites yielded 85 new associations, underscoring the complex genetic regulation of metabolite levels [21, 22]. The data on coagulation defects, specifically purpura and other haemorrhagic conditions, were obtained from FinnGen R11 [23]. Additionally, lifestyle data were retrieved from the IEU open GWAS database [24]. Together, these datasets have facilitated multifaceted investigations among metabolome profiles, genetics and lifestyle factors in coagulation defects.

MR analysis

In R software, the TwoSampleMR package was employed to perform the MR analysis, assessing the causal relationships between the levels of metabolites and coagulation defects. The Wald ratio method was used to calculate causal estimates for metabolites with a single associated single nucleotide polymorphism (SNP), and the Inverse Variance Weighted (IVW) method was employed as the principal analytical approach for metabolites linked to multiple SNPs [25, 26]. To ensure robustness, conventional sensitivity analyses, such as MR-Egger or leave-one-out analysis, were applied for metabolites linked with more than four SNPs [27]. For metabolites unsuitable for conventional sensitivity analyses, we conducted a thorough search in the GWAS Catalogue to evaluate whether the instrumental SNP was associated with other traits that could act as confounders. The Steiger directional test was performed to identify SNPs that primarily influence the exposure (metabolites) rather than the outcome (coagulation defects), ruling out reverse causality [28]. Any SNPs failing this criterion were excluded. Finally, the Benjamini‒Hochberg false discovery rate (FDR) correction was applied to ensure the reliability of significant associations [29].

Colocalization analysis

To assess the genetic overlap between metabolites and coagulation defects, the colocalization analysis was performed using the coloc R package to examine whether the same genetic variation influenced two distinct phenotypes, thereby revealing potential causal relationships [30].

Metabolite pathway analysis and the search for rate-limiting enzymes

The superpathway and subpathway classifications for all the metabolites were identified on the basis of the original metabolite dataset. Each metabolite was categorized into a broader superpathway, such as the lipid, amino acid, carbohydrate, or cofactor superpathway, and further categorized into more specific subpathways detailing their specific biological or chemical roles [31]. For metabolites showing significant associations with coagulation defects, detailed investigations of their respective subpathways and rate-limiting enzymes were performed using information from publicly available databases, pathway-specific annotations and the biochemistry literature [32].

Drug targets and PPI network analysis

We initially utilized DrugBank to compile a list of drug targets associated with medications commonly used to treat coagulation defects [33]. Proteins targeted by approved drugs or investigational compounds were identified and categorized as potential therapeutic targets, and relevant drug information was carefully documented. We subsequently constructed a PPI network using the STRING database, focusing on interactions between the identified drug targets and the rate-limiting enzymes of metabolites associated with coagulation defects [18].

The impact of lifestyle on metabolites

We performed the MR analysis to evaluate the influence of lifestyle factors on metabolites associated with coagulation defects. The aim was to identify specific metabolites that might be regulated by targeted lifestyle interventions, offering potential strategies for disease management.

Statistical analysis

All the statistical analyses were implemented with the TwoSampleMR package in R software (4.4.2). The coloc package was used for the colocalization analysis. The results were corrected via FDR using the Benjamin‒Hochberg method. We regarded the metabolites with corrected FDR values < 0.05 as positive.

Results

MR analysis of metabolites and coagulation defects (purpura and other haemorrhagic conditions)

In our analysis of metabolites and coagulation defects, initial preliminary associations were observed between the levels of 48 metabolites and coagulation defects (purpura and other haemorrhagic conditions) at a threshold level of P < 0.05 (Supplementary Material 2). However, after applying the FDR correction, only four metabolites remained significantly associated with coagulation defects (FDR < 0.05). The levels of docosapentaenoate n3 DPA 22:5n3 (DPA) (OR: 1.594, 95% CI: 1.263–2.011, P < 0.001), 1-palmitoyl-2-stearoyl-gpc (PSPC) (16:0/18:0) (OR: 1.294, 95% CI: 1.134–1.477, P < 0.001) and 1-stearoyl-2-docosahexaenoyl-gpc (SDPC) (18:0/22:6) (OR: 1.232, 95% CI: 1.101–1.380, P < 0.001) displayed strong positive associations with coagulation defects, whereas the levels of hydroxypalmitoyl sphingomyelin (HPSM) (d18:1/16:0 (OH)) (OR: 0.803, 95% CI: 0.719–0.896, P < 0.001) exhibited a protective effect against coagulation disorders, with a negative association (Figs. 2 and 3). Conventional sensitivity analyses, which typically require a minimum of four SNPs, were not feasible in this study. Thus, a comprehensive search of the GWAS Catalogue was conducted to investigate the potential pleiotropic effects of the instrumental SNP. Our findings indicated no evidence of pleiotropy, which strengthened the robustness of causal inference. The Steiger test was performed for all significant associations. The results confirmed that the identified SNPs influence metabolite levels rather than being affected by coagulation defects, ruling out reverse causality (Supplementary Material 3). These findings highlight the critical role of specific metabolites in the pathophysiology of coagulation defects and suggest potential targets for further investigations.

Fig. 2

Volcano plot illustrating the causal relationships between metabolite levels and coagulation defects after FDR correction. The x-axis represents the estimated effect size (Beta), whereas the y-axis represents the -log10 (P value)

Fig. 3

Forest plot showing the causal relationships between FDR-corrected metabolite levels and coagulation defects. Odds ratios (ORs) and 95% confidence intervals (CIs) for significant metabolites are shown

Colocalization results

Colocalization analysis was used to estimate five posterior probabilities (PPs) on the basis of the following assumptions. PP0: No association with the genomic region existed for either trait. PP1: The first trait was associated with the genomic region, whereas the second trait was not associated with that region. PP2: The second trait was associated with the genomic region, whereas the first trait was not associated with that region. PP3: Both traits were associated with the genomic region, but they were influenced by different genetic variations. PP4: Both traits were associated with the genomic region and were influenced by a shared causal genetic variant [34]. The primary focus was on PP4, since a high PP4 value > 0.8 indicated strong evidence that the same genetic variant led to both traits [35]. SNPs for the analysis were selected on the basis of their significance, using the top SNP (lowest P value) for each metabolite and considering SNPs within a ± 100 kb region around the lead SNP [36]. For this study, the colocalization analysis revealed a strong colocalization signal for one metabolite, DPA, with a posterior probability of hypothesis 4 (PPH4) greater than 0.8 (PPH4 = 0.821) (Fig. 4, Supplementary Material 4), indicating high-confidence evidence of a shared genetic variant that influences both the metabolite and susceptibility to coagulation defects. For the remaining three metabolites, the colocalization probabilities fell below the threshold (PPH4 < 0.8) (Supplementary Material 4), indicating lower confidence in a shared genetic influence. DPA exhibited strong alignment at rs4246215 on chromosome 11, indicating a specific genetic locus that may drive the association of DPA levels with coagulation defects (Fig. 4). In summary, this analysis presents compelling evidence linking DPA to coagulation defects, contributing to a deeper understanding of the genetic and metabolic pathways involved in coagulation defects.

Fig. 4

Colocalization analysis of metabolites and coagulation defects. This figure focuses on DPA with a high colocalization probability (PPH4 = 0.821) observed at rs4246215 on chromosome 11. The x-axis represents chromosomal positions, whereas the y-axis displays the -log10 (P value) for the metabolite and GWAS signals. The r² values indicate the degree of linkage disequilibrium, highlighting regions of genetic overlap

Search for metabolic pathways and rate-limiting enzymes

On the basis of the findings from MR and colocalization analyses, we further investigated the biological pathways associated with these four significant metabolites. At the superpathway level, all four metabolites were classified within the lipid pathway, emphasizing their roles in lipid metabolism and coagulation defects. At the subpathway level, specific metabolic pathways were identified for each metabolite. DPA was mapped to the long-chain polyunsaturated fatty acid (LCPUFA) (n3 and n6) pathway. The rate-limiting enzymes for this pathway are FADS1 and FADS2, which are essential for the biosynthesis and regulation of polyunsaturated fatty acids. Both PSPC (16:0/18:0) and SDPC (18:0/22:6) were mapped to the phosphatidylcholine (PC) pathway. This pathway is regulated by the rate-limiting enzyme TCP1, which is pivotal for maintaining phospholipid homeostasis. HPSM (d18:1/16:0 (OH)) was mapped to the sphingomyelin metabolism pathway, with the rate-limiting enzymes SGMS1 and SGMS2 facilitating sphingomyelin synthesis and turnover. By identifying these rate-limiting enzymes—FADS1, FADS2, TCP1, SGMS1 and SGMS2—as key control points, this study highlights the crucial roles of lipid metabolism pathways and their regulatory enzymes in the context of coagulation defects (Table 1; Supplementary Material 5).

Table 1 Summary of metabolite pathways and rate-limiting enzymes

Drug targets and PPI network construction

To further elucidate therapeutic targets for coagulation defects, we analysed four drugs that are currently approved or being investigated in clinical trials. The targets of these drugs were identified through the DrugBank database, with a focus on potential interactions with the rate-limiting enzymes identified in the metabolite pathways (Supplementary Material 6). TCP1, a critical rate-limiting enzyme in the PC pathway, was found to interact with plasminogen (PLG). PLG is a known target of tranexamic acid and a widely used antifibrinolytic drug. This interaction highlights the pivotal role of metabolic pathways in therapeutic mechanisms, suggesting a potential synergistic interplay between lipid metabolism and the pharmacological effects of tranexamic acid. For the other identified enzymes, including FADS1, FADS2, SGMS1 and SGMS2, no direct interactions with the remaining drug targets were identified. These enzymes remain critical components of lipid metabolism, influencing coagulation processes, but further exploration is needed to establish the possibility of direct pharmacological targeting. The PPI network analysis provided visual confirmation of the linkage between TCP1 and PLG (Fig. 5). The absence of significant connections with other enzymes underscores the specificity of the TCP1‒PLG interaction within this context, offering an opportunity for exploring lipid pathway modulation as a complementary approach to existing coagulation therapies.

Fig. 5

PPI network analysis of existing drug targets and metabolite rate-limiting enzymes. The network highlights a prominent interaction between TCP1 and plasminogen (PLG)

The impact of lifestyle on three types of proteins

Seventeen lifestyle factors (Supplementary Material 1), including dietary habits, physical activity and smoking status, were analysed via MR to identify causal relationships with the metabolites associated with coagulation defects [37, 38]. Among the 17 healthy lifestyle factors analysed, several significant associations were identified. Specifically, positive causal relationships were observed between dried fruit intake and oily fish intake with SDPC (18:0/22:6) levels. Similarly, a positive causal relationship was found between abstinence from smoking and DPA levels. Additionally, both cereal intake and dried fruit intake showed positive causal associations with PSPC (16:0/18:0) levels (Fig. 6). These results emphasize the potential influences of dietary and lifestyle modifications on the levels of metabolites associated with coagulation defects, offering novel insights into therapeutic interventions.

Fig. 6

The impacts of lifestyle factors on the levels of metabolites associated with coagulation defects. Significant positive causal associations were observed: dried fruit intake and oily fish intake increased the levels of SDPC (18:0/22:6), whereas abstaining from smoking increased the levels of DPA. Additionally, both cereal intake and dried fruit intake were positively correlated with PSPC (16:0/18:0) levels

Discussion

This comprehensive study provides innovative perspectives into the metabolic basis of coagulation defects through an integrated analysis combining MR, colocalization analysis, pathway investigation, PPI network analysis and lifestyle factor assessments. Our findings highlight specific metabolic signatures associated with coagulation defects and identify potential therapeutic targets and lifestyle interventions that can modulate the levels of metabolites associated with disease risk.

Coagulation defects can be classified on the basis of complex pathogenesis [39]. Primary haemostasis-related coagulation defects are caused mainly by abnormal platelet and vascular endothelium functions, thereby disrupting the formation of platelet thrombi [40, 41]. Secondary haemostasis-related coagulation defects occur due to abnormalities in the coagulation cascade, which disrupt the formation of fibrin protein [42]. Furthermore, elevated levels of plasminogen activators or decreased levels of antifibrinolytic substances can lead to the excessive degradation of fibrin protein, causing abnormal bleeding [43].

Our MR analysis identified four metabolites whose levels were significantly associated with coagulation defects after FDR correction. The levels of DPA were significantly associated with increased coagulation risk (OR: 1.594), highlighting the potential role of DPA in the pathophysiology of coagulation defects. Notably, the causal relationship was corroborated by the colocalization analysis, with substantial genetic overlap (PPH4 = 0.821) at rs4246215 on chromosome 11. This high PP value suggests that genetic variants influencing DPA levels may directly affect coagulation risk, underscoring its critical role in the disease process [44, 45].

Regarding biological functions, DPA possesses a marked ability to disrupt the cyclooxygenase pathway and accelerate the lipoxygenase pathway, thereby inhibiting platelet aggregation [46, 47]. These findings indicate the potential association between DPA and platelet function-related primary haemostasis defects. DPA, EPA, and DHA are omega-3 fatty acids with similar structures and shared biosynthetic pathways. In a randomized controlled trial, the high contents of DPA and EPA in seal oil were associated with decreased levels of the platelet activation marker p-selectin [48]. The meta-analysis of clinical studies indicated that high-dose EPA and DHA were associated with an increased risk of bleeding events, which may further support the potential role of DPA in bleeding [49, 50]. While omega-3 fatty acids are well-known for their beneficial effects on cardiovascular diseases, our results indicate that DPA may exert distinct and potentially adverse effects in the context of coagulation and further clinical trials are needed to investigate the association between the use of DPA supplements and the coagulation risk.

The identification of the levels of two phosphatidylcholine species, PSPC (16:0/18:0) (OR: 1.294) and SDPC (18:0/22:6) (OR: 1.232), as risk factors, provides new insights into the roles of membrane phospholipids in coagulation defects. Regarding their biological functions, phosphatidylcholine, the major external phospholipid of the cell membrane, plays an anticoagulant role by inhibiting the membrane binding of clotting proteins due to the steric hindrance of the choline head group [51]. This finding implicates the potential role of phosphatidylcholine in coagulation defects associated with secondary haemostasis dysfunction. Although specific clinical studies directly linking these two phosphatidylcholines to coagulation defects are scarce, their roles in modulating membrane integrity and exhibiting anticoagulant ability are consistent with the coagulation risks observed in our study.

The negative association between HPSM (d18:1/16:0(OH)) levels and coagulation defects (OR: 0.803) is notable, suggesting that elevated levels of this sphingomyelin may confer protective effects against disease. Regarding its biological functions, some studies have reported that sphingomyelin inhibits the activity of coagulation complexes and tissue factors, thereby guiding the inhibition of blood coagulation [52, 53]. This evidence suggests that sphingomyelin is potentially involved in secondary haemostasis-related coagulation defects. However, direct clinical studies linking HPSM (d18:1/16:0(OH)) levels to coagulation defects remain limited. The protective association observed here shows that the disruption of HPSM (d18:1/16:0(OH)) metabolism may lead to a high risk of coagulation defects, generating new insights into the coagulation function of sphingomyelin. Future functional studies and clinical trials are essential to validate the precise mechanisms by which the identified HPSM (d18:1/16:0(OH)) influences the coagulation processes.

Pathway analysis revealed that all metabolites significantly associated with coagulation defects are involved in the lipid superpathway, emphasizing the central role of lipid metabolism in coagulation defects. The identification of specific rate-limiting enzymes, FADS1, FADS2, TCP1, SGMS1, and SGMS2, provides concrete targets for therapeutic development, suggesting potential mechanisms involved in metabolic alterations and coagulation function.

PPI network analysis further revealed a significant interaction between TCP1, a rate-limiting enzyme in the PC pathway, and PLG, the target of tranexamic acid. This finding suggests that the therapeutic efficacy of tranexamic acid may be partially mediated through its effects on phosphatidylcholine metabolism, opening new avenues for drug development and optimization. To explore whether this interaction is unique to tranexamic acid or extends to other antifibrinolytic agents, we screened additional clinically approved or investigational antifibrinolytics, including aminocaproic acid, aprotinin, and epsilon-aminocaproic acid (EACA). However, no direct PPI links were identified between the rate-limiting enzymes and the targets of these drugs. These preliminary findings suggest that tranexamic acid is a unique intersection point between antifibrinolytic therapy and lipid-regulated coagulation, possibly due to interactions with plasminogen and membrane phospholipid components, providing potential opportunities for future combination therapies.

For other rate-limiting enzymes that lack direct interactions with approved drug targets, further exploration through computational modelling and experimental screening is needed. Structure-based drug discovery approaches, including virtual docking, homology modelling, and molecular dynamics, can be employed to identify candidate compounds capable of modulating enzymatic activity or binding allosteric sites. In parallel, in vitro high-throughput screening of compound libraries may reveal novel inhibitors or activators with therapeutic potential. FADS1/2 have well-characterized structures and are known to be involved in lipid homeostasis, making them promising candidates for structure-guided drug design [54]. Future studies will combine the structure-based computational approaches with biochemical and cellular validation to guide medicinal chemistry optimization and facilitate the discovery of novel therapeutics.

Our analysis of lifestyle factors revealed several significant associations with the levels of disease-related metabolites, suggesting potential preventive strategies. The positive correlations of dried fruit and oily fish consumption with SDPC (18:0/22:6) levels provide evidence-based dietary recommendations for patients at risk of coagulation defects. Similarly, the relationship between smoking cessation and DPA levels underscores the importance of lifestyle modifications in managing disease risk. These findings emphasize the potential to modulate bleeding risk through the complex interactions between metabolites and lifestyle factors. Future prospective cohort studies are needed to validate whether lifestyle-induced alterations in metabolite levels can effectively reduce disease incidence. Moreover, integrating metabolomic data with dietary and lifestyle information into comprehensive databases to develop risk scoring systems for coagulation disorders, may significantly enhances the ability of clinical decision-making. In addition, developing rapid assay kits to detect specific metabolites and their incorporation into routine diagnostic tools could facilitate early detection and support precision public health interventions.

Our findings are consistent with growing evidence that lipid metabolism plays fundamental roles in regulating haemostasis. Previous studies have shown that the anticoagulation abilities of phosphatidylcholine and omega-3 lipids in the context of coagulation, which are consistent with our findings. Sphingomyelin has been shown to be linked to the impaired function of coagulation complexes and tissue factors, indicating its role as a bleeding risk factor. However, the identification of HPSM (d18:1/16:0(OH)) as a protective factor in coagulation defects introduces a novel dimension to the role of sphingomyelin, suggesting that sphingomyelin subtype-specific effects must be considered in clinical contexts. These insights refine and expand upon existing models of lipid‒haemostatic interactions by identifying specific metabolic drivers with causal inference support.

While our study provides robust evidence for metabolic influences on coagulation defects, several limitations should be acknowledged. Firstly, the absence of specific subtypes and detailed pathogenesis on coagulation defects obtained from FinnGen R11 limits our ability to offer more detailed classifications or mechanistic information within the current scope of data. To address this issue, further clinical studies should focus on the classified subtype of coagulation defects to assess the universality of our findings. Secondly, owing to the limited information of summary-level GWAS data, analyses stratified by sex or age were not feasible in this study. Future work using individual datasets with stratified information is essential for determining whether metabolite‒coagulation associations differ across demographic subpopulations, helping refine risk prediction models and optimize targeted interventions for high-risk groups. Thirdly, the individuals in the CLSA and FinnGen R11 datasets used here are of European ancestry, so the results may not be directly applicable to other races or the wider population. Therefore, future studies of multiethnic populations should verify these findings to evaluate the universality of our results. Lastly, despite stringent criteria and rigorous quality measures, the residual confounding in MR analysis cannot be entirely ruled out. Thus, future studies using larger and more diverse datasets are needed to reduce the potential impact of bias and confounders.

Conclusions

This MR study provides compelling evidence that specific metabolites and metabolic pathways play significant roles in the risk of coagulation defects, spotlighting the need for novel therapeutic targets and underlining the importance of dietary and lifestyle interventions. These findings lay the foundation for personalized approaches in the management of coagulation-related conditions.