Myocardial infarction (MI) is a critical cardiovascular event with multifaceted etiology, involving several genetic and environmental factors. It is essential to understand the function of plasma metabolites in the development of MI and unravel its complex pathogenesis.
METHODS
This study employed a bidirectional Mendelian randomization (MR) approach to investigate the causal relationships between plasma metabolites and MI risk. We used genetic instruments as proxies for plasma metabolites and MI and conducted MR analyses in both directions to assess the impact of metabolites on MI risk and vice versa. In addition, the large-scale genome-wide association studies datasets was used to identify genetic variants associated with plasma metabolite (1400 metabolites) and MI (20,917 individuals with MI and 440,906 individuals without MI) susceptibility. Inverse variance weighted was the primary method for estimating causal effects. MR estimates are expressed as beta coefficients or odds ratio (OR) with 95% CI.
RESULTS
We identified 14 plasma metabolites associated with the occurrence of MI (P < 0.05), among which 8 plasma metabolites [propionylglycine levels (OR = 0.922, 95% CI: 0.881–0.965, P < 0.001), gamma-glutamylglycine levels (OR = 0.903, 95% CI: 0.861–0.948, P < 0.001), hexadecanedioate (C16-DC) levels (OR = 0.941, 95% CI: 0.911–0.973, P < 0.001), pentose acid levels (OR = 0.923, 95% CI: 0.877–0.972, P = 0.002), X-24546 levels (OR = 0.936, 95% CI: 0.902–0.971, P < 0.001), glycine levels (OR = 0.936, 95% CI: 0.909–0.964, P < 0.001), glycine to serine ratio (OR = 0.930, 95% CI: 0.888–0.974, P = 0.002), and mannose to trans-4-hydroxyproline ratio (OR = 0.912, 95% CI: 0.869–0.958, P < 0.001)] were correlated with a decreased risk of MI, whereas the remaining 6 plasma metabolites [1-palmitoyl-2-arachidonoyl-GPE (16:0/20:4) levels (OR = 1.051, 95% CI: 1.018–1.084, P = 0.002), behenoyl dihydrosphingomyelin (d18:0/22:0) levels (OR = 1.076, 95% CI: 1.027–1.128, P = 0.002), 1-stearoyl-2-docosahexaenoyl-GPE (18:0/22:6) levels (OR = 1.067, 95% CI: 1.027–1.109, P = 0.001), alpha-ketobutyrate levels (OR = 1.108, 95% CI: 1.041–1.180, P = 0.001), 5-acetylamino-6-formylamino-3-methyluracil levels (OR = 1.047, 95% CI: 1.019–1.076, P < 0.001), and N-acetylputrescine to (N (1) + N (8))-acetylspermidine ratio (OR = 1.045, 95% CI: 1.018–1.073, P < 0.001)] were associated with an increased risk of MI. Furthermore, we also observed that the mentioned relationships were unaffected by horizontal pleiotropy (P > 0.05). On the contrary, MI did not lead to significant alterations in the levels of the aforementioned 14 plasma metabolites (P > 0.05 for each comparison).
CONCLUSIONS
Our bidirectional MR study identified 14 plasma metabolites associated with the occurrence of MI, among which 13 plasma metabolites have not been reported previously. These findings provide valuable insights for the early diagnosis of MI and potential therapeutic targets.
Myocardial infarction (MI),[1–5] commonly known as a heart attack, is a pivotal cardiovascular event, accompanied by substantial global morbidity and mortality.[6–8] A previous study involving large sample sizes has reported a global prevalence of 3.8% of MI for individuals aged under 60 years and 9.5% of MI in individuals aged over 60 years.[9] Thus, timely identification and potential therapeutic interventions are imperative to ameliorating the deleterious consequences associated with MI. There has been a recent growing interest in utilizing plasma metabolites as potential biomarkers for the early detection of MI and as feasible targets for innovative therapeutic interventions.[10,11] Plasma metabolites, comprising small metabolic molecules, are known to play crucial intermediary functions in different physiological pathways. Their concentrations provide valuable insights into metabolic status and potential pathological conditions, including the MI, making them excellent candidates to explore biomarkers. Currently, the literature on the relationship between plasma metabolites and MI is scarce. Further studies on the exact connection between plasma metabolites and MI would contribute to the discovery of early marker for MI and potential therapeutic targets.
Bidirectional Mendelian randomization (MR)[12–15] studies represent a robust methodological approach, with numerous advantages not commonly present in traditional research methodologies. These include mitigating the impact of confounding factors on conclusions and exploring reverse causation, thereby providing a more reliable foundation for causal inferences. This study employed a bidirectional MR approach to investigate the relationship between plasma metabolites and MI, offering new insights into the early diagnosis and potential treatment of MI.
METHODS
Study Design
We used a two-sample MR approach, initially using the large-scale genome-wide association studies (GWAS) to investigate the relationship between exposure (plasma metabolites) and outcome (MI). Subsequently, a reverse MR analysis was conducted to explore the relationship between exposure (MI) and outcome (plasma metabolites). The study design was based on three key assumptions of MR[16,17]: (1) genetic variants should display a robust correlation with the exposure; (2) genetic variants should not share any associations with confounding factors; and (3) the impact of genetic variations on the outcome should be exclusively mediated through the exposure, precluding involvement with alternative pathways. This study was conducted in accordance with the guidelines[17] outlined in the MR study. Figure 1 briefly describes the process of this bidirectional MR study.
We consistently adhered to rigorous ethical standards. Publicly available genetic tools and previously published GWAS data were used. In each study, all participants provided informed consent, and ethical approval was obtained from the Respective Institutional Review Boards.[18,19] Thus, ethical approval was considered unnecessary, as we did not use any individual-level data.
Data Sources
The overview of data sources used is provided in Table 1. Plasma metabolites were derived from Chen, et al.[18] study, involving the cohort of approximately 8000 individuals of European descent. The summary metrics for GWAS focusing on plasma metabolites are available in GWAS catalog (https://www.ebi.ac.uk/gwas/). Accession numbers of these GWAS data ranged from GCST90199621 to GCST90201020. The MI data for MR were obtained from the GWAS analysis conducted by Sakaue, et al.[19] (https://gwas.mrcieu.ac.uk/, dataset ID: ebi-a-GCST90018877). This analysis encompassed 461,823 individuals of European descent, comprising 20,917 individuals with MI and 440,906 individuals without MI. A total of 24,172,914 single nucleotide polymorphisms (SNPs) were identified to be associated with MI through genetic analysis.
Table
1.
The overview of data sources used in this bidirectional Mendelian randomization study.
Variables
PubMed ID
Journal
Study accession
Year of publication
Sample size
Plasma metabolites
36635386
Nature Genetics
GCST90199621 to GCST90201020
2023
Approximately 8000 individuals of European descent
Myocardial infarction
34594039
Nature Genetics
ebi-a-GCST90018877
2021
461,823 individuals of European descent, comprising 20,917 individuals with myocardial infarction and 440,906 individuals without myocardial infarction
We adhered to the three essential assumptions outlined earlier to select IVs using publicly accessible GWAS databases. We implemented a thorough screening process to address the concerns related to linkage disequilibrium and explore the causal relationship between plasma metabolites (exposure) and MI (outcome). This involved using a clump window of r2 = 0.001 and kb = 10,000 to alleviate associated issues. Furthermore, we filtered IVs using a significance threshold of P < 1 × e-5 for more comprehensive coverage of plasma metabolites. In addition, we examined the selected IVs for associations with potential confounding factors such as age and body mass index. This screening was performed using the online resource PhenoScanner[20] (http://www.phenoscanner.medschl.cam.ac.uk/), with a threshold of r2 = 0.8. While investigating the causal relationship between MI (exposure) and plasma metabolites (outcome), we did not investigate the causal connection between MI and the entire spectrum of plasma metabolites (1400 metabolites). Instead, we exclusively focused on the causal link between MI and plasma metabolites previously identified to have associations with MI in the initial step. Throughout this process, a significance threshold of P < 5 × e-8 was applied, setting it apart from the initial threshold of P < 1 × e-5. Other parameters remained consistent with the first step. F-statistics were computed to estimate the potential impact of sample and overlap and address concerns related to weak instrument bias resulting from the relaxation of significance thresholds.[21–24] The F-statistics were calculated as mentioned below:
EAF stands for effect allele frequency for exposure, N represents the sample size for exposure, SE stands for the standard error of β, and β denotes the estimated effect for exposure. IVs with F-statistics < 10 were excluded as weak instruments and were excluded.
Power Calculation
We used an online tool (https://sb452.shinyapps.io/power/) to calculate the power. The power for each plasma metabolite associated with MI was found to be 100%, indicating a substantial sample size to investigate the relationship between plasma metabolites and MI.
Statistical Analysis
We examined the causal relationships between plasma metabolites and MI, as well as the reverse causal relationship between MI and plasma metabolites. The inverse variance weighted (IVW) method under a random-effects model was the primary method for estimating causal effects.[25] The IVW method with multiplicative random effects was used to provide a succinct estimate, considering potential heterogeneity among Wald ratio estimates from SNPs.[26] Consequently, we used random-effects IVW analysis in the presence of heterogeneity (P < 0.05), whereas we used a fixed-effects IVW analysis in the absence of heterogeneity. In addition to the IVW method, four additional MR methods were used, including MR-Egger, weighted median, simple mode, and weighted mode, to assess causal relationships. The IVW method operates under the assumption that all SNPs included in the analysis were effective IVs.[27] The weighted median method requires that at least 50% of genetic variations are effective, meeting three basic assumptions, and is suitable when the majority of IVs do not exhibit horizontal pleiotropy.[28] Conversely, the MR-Egger regression method assumes more than 50% of genetic variations to be invalid.[29] MR estimates are expressed as beta coefficients or odds ratio (OR) with 95% CI. The results of IVW, MR-Egger, weighted median, simple mode, and weighted mode were considered consistent only if the direction of estimation, whether positive or negative, aligned. The association between metabolites and the occurrence of MI was not considered in case of inconsistent directions.
We conducted a variety of sensitivity analyses, including Cochran’s Q test and examination of the MR-Egger intercept, to explore potential heterogeneity and pleiotropy, ensuring the robustness of our findings. Heterogeneity was assessed using both IVW and MR-Egger regression, with Cochran’s Q statistic serving as the evaluation metric. Significant heterogeneity was absent if the P-values for Cochran’s Q tests in both IVW and MR-Egger exceeded 0.05. The absence of pleiotropy in our MR results was confirmed by examining the intercept in the MR-Egger regression, where a P-value more than 0.05 suggested an absence of pleiotropy. In addition, the MR-PRESSO analysis was conducted to eliminate significant outliers. Analytical procedures for this MR study involved the use of RStudio statistical software (version 4.2.2) and the TwoSampleMR package (version 0.5.6). Result is considered significant when the P-values are below 0.05.
RESULTS
Causal Relationship between Plasma Metabolites and MI
The selection of IVs is provided (supplemental material, Excel). For a P-value less than 0.05, whether from IVW, MR-Egger, weighted median, simple mode, or weighted mode, 198 unique plasma metabolites were initially identified to have a significant association with MI, as illustrated in Figure 2. Further analysis revealed exclusively considered results selected by IVW (P < 0.05) while simultaneously excluding IVs with horizontal pleiotropy. Ultimately, 14 plasma metabolites were identified, including propionylglycine levels, gamma-glutamylglycine levels, hexadecanedioate (C16-DC) levels, 1-palmitoyl-2-arachidonoyl-GPE (16:0/20:4) levels, behenoyl dihydrosphingomyelin (d18:0/22:0) levels, 1-stearoyl-2-docosahexaenoyl-GPE (18:0/22:6) levels, pentose acid levels, alpha-ketobutyrate levels, X-24546 levels, 5-acetylamino-6-formylamino-3-methyluracil levels, glycine levels, N-acetylputrescine to (N (1) + N (8))-acetylspermidine ratio, glycine to serine ratio, and mannose to trans-4-hydroxyproline ratio. These plasma metabolites displayed a significant association with the occurrence of MI.
Figure
2.
The association between plasma metabolites and myocardial infarction was first investigated.
A total of 198 unique plasma metabolites were preliminarily identified as significantly associated with myocardial infarction.
Among them, 8 plasma metabolites [propionylglycine levels (OR = 0.922, 95% CI: 0.881–0.965, P < 0.001), gamma-glutamylglycine levels (OR = 0.903, 95% CI: 0.861–0.948, P < 0.001), hexadecanedioate (C16-DC) levels (OR = 0.941, 95% CI: 0.911–0.973, P < 0.001), pentose acid levels (OR = 0.923, 95% CI: 0.877–0.972, P = 0.002), X-24546 levels (OR = 0.936, 95% CI: 0.902–0.971, P < 0.001), glycine levels (OR = 0.936, 95% CI: 0.909–0.964, P < 0.001), glycine to serine ratio (OR = 0.930, 95% CI: 0.888–0.974, P = 0.002), and mannose to trans-4-hydroxyproline ratio (OR = 0.912, 95% CI: 0.869–0.958, P < 0.001)] were correlated with a decreased risk of MI. Conversely, the remaining 6 plasma metabolites [1-palmitoyl-2-arachidonoyl-GPE (16:0/20:4) levels (OR = 1.051, 95% CI: 1.018–1.084, P = 0.002), behenoyl dihydrosphingomyelin (d18:0/22:0) levels (OR = 1.076, 95% CI: 1.027–1.128, P = 0.002), 1-stearoyl-2-docosahexaenoyl-GPE (18:0/22:6) levels (OR = 1.067, 95% CI: 1.027–1.109, P = 0.001), alpha-ketobutyrate levels (OR = 1.108, 95% CI: 1.041–1.180, P = 0.001), 5-acetylamino-6-formylamino-3-methyluracil levels (OR = 1.047, 95% CI: 1.019–1.076, P < 0.001), and N-acetylputrescine to (N (1) + N (8))-acetylspermidine ratio (OR = 1.045, 95% CI: 1.018–1.073, P < 0.001)] were associated with an increased risk of MI.
The causal relationship between plasma metabolites and MI is depicted in Figure 3. Furthermore, among the identified 14 plasma metabolites associated with MI, only 1 plasma metabolite (glycine levels) has been previously reported in an existing study,[30] whereas the remaining 13 plasma metabolites have not been documented in the literature.
Figure
3.
Causal relationship between plasma metabolites (exposures) and myocardial infarction (outcome).
A total of 14 plasma metabolites were associated with myocardial infarction, with 8 plasma metabolites correlated with a decreased risk of myocardial infarction, whereas the remaining 6 plasma metabolites were associated with an increased risk of myocardial infarction. GCST90199705: propionylglycine levels; GCST90199741: gamma-glutamylglycine levels; GCST90199807: hexadecanedioate (C16-DC) levels; GCST90200054: 1-palmitoyl-2-arachidonoyl-GPE (16:0/20:4) levels; GCST90200059: behenoyl dihydrosphingomyelin (d18:0/22:0) levels; GCST90200065: 1-stearoyl-2-docosahexaenoyl-GPE (18:0/22:6) levels; GCST90200259: pentose acid levels; GCST90200438: alpha-ketobutyrate levels; GCST90200630: X-24546 levels; GCST90200680: 5-acetylamino-6-formylamino-3-methyluracil levels; GCST90200707: glycine levels; GCST90200726: N-acetylputrescine to (N (1) + N (8))-acetylspermidine ratio; GCST90200758: glycine to serine ratio; GCST90200782: mannose to trans-4-hydroxyproline ratio; OR: odds ratio; SNPs: single nucleotide polymorphisms.
Heterogeneity Testing of Plasma Metabolites (Exposure) and MI (Outcome)
Table 2 describes the assessment of heterogeneity of plasma metabolites (exposure) and MI (outcome). As depicted in Table 2, irrespective of using MR-Egger or IVW, no significant heterogeneity was observed in the relationship between propionylglycine levels, hexadecanedioate (C16-DC) levels, 1-palmitoyl-2-arachidonoyl-GPE (16:0/20:4) levels, 1-stearoyl-2-docosahexaenoyl-GPE (18:0/22:6) levels, pentose acid levels, alpha-ketobutyrate levels, X-24546 levels, 5-acetylamino-6-formylamino-3-methyluracil levels, N-acetylputrescine to (N (1) + N (8))-acetylspermidine ratio, and mannose to trans-4-hydroxyproline ratio with MI (all P > 0.05). However, for gamma-glutamylglycine levels, behenoyl dihydrosphingomyelin (d18:0/22:0) levels, and glycine to serine ratio, both MR-Egger and IVW detected significant heterogeneity with MI (both P < 0.05).
Table
2.
Heterogeneity testing of plasma metabolites (exposure) and myocardial infarction (outcome).
Pleiotropy Testing of Plasma Metabolites (Exposure) and MI (Outcome)
Table 3 delineates the pleiotropy testing of plasma metabolites (exposure) and MI (outcome). As indicated in Table 3, each Egger intercept value was very small, and the corresponding P-values were greater than 0.05, suggesting that the relationships between the identified 14 plasma metabolites and MI remained unaffected by horizontal pleiotropy.
Table
3.
Pleiotropy testing of plasma metabolites (exposure) and myocardial infarction (outcome).
Causal Relationship between MI (Exposure) and Plasma Metabolites (Outcome)
We next assessed whether MI substantially altered identified 14 plasma metabolites. We used MI as the exposure and those identified 14 plasma metabolites as the outcome, further conducting a reverse MR analysis. The results are illustrated in Figure 4. As depicted in Figure 4, no significant relationship existed between MI and those identified 14 plasma metabolites, indicating no significant impact of MI on these plasma metabolites.
Figure
4.
Causal relationship between myocardial infarction (exposure) and plasma metabolites (outcome).
No significant association was discovered between myocardial infarction and specified 14 plasma metabolites. GCST90199705: propionylglycine levels; GCST90199741: gamma-glutamylglycine levels; GCST90199807: hexadecanedioate (C16-DC) levels; GCST90200054: 1-palmitoyl-2-arachidonoyl-GPE (16:0/20:4) levels; GCST90200059: behenoyl dihydrosphingomyelin (d18:0/22:0) levels; GCST90200065: 1-stearoyl-2-docosahexaenoyl-GPE (18:0/22:6) levels; GCST90200259: pentose acid levels; GCST90200438: alpha-ketobutyrate levels; GCST90200630: X-24546 levels; GCST90200680: 5-acetylamino-6-formylamino-3-methyluracil levels; GCST90200707: glycine levels; GCST90200726: N-acetylputrescine to (N (1) + N (8))-acetylspermidine ratio; GCST90200758: glycine to serine ratio; GCST90200782: mannose to trans-4-hydroxyproline ratio; OR: odds ratio; SNPs: single nucleotide polymorphisms.
Heterogeneity testing results for MI (exposure) and plasma metabolites (outcome) are presented in Table 4. Apart from significant differences observed in the analysis of 1-palmitoyl-2-arachidonoyl-GPE (16:0/20:4) levels, behenoyl dihydrosphingomyelin (d18:0/22:0) levels, and 5-acetylamino-6-formylamino-3-methyluracil levels, MI exhibited no significant heterogeneity with the other 11 plasma metabolites.
Table
4.
Heterogeneity testing of myocardial infarction (exposure) and plasma metabolites (outcome).
Table 5 describes the pleiotropy testing results for MI (exposure) and plasma metabolites (outcome). The relationship between MI and behenoyl dihydrosphingomyelin (d18:0/22:0) levels and alpha-ketobutyrate levels was affected by horizontal pleiotropy (P < 0.05). However, because MI exerted a significant impact on these 2 plasma metabolites, this pleiotropic effect is unlikely to significantly affect the conclusion. Moreover, we did not find horizontal pleiotropy affecting the relationship between MI and the other 12 plasma metabolites.
Table
5.
Pleiotropy testing of myocardial infarction (exposure) and plasma metabolites (outcome).
In the present bidirectional MR study, we unraveled the causal relationship between plasma metabolites and MI to uncover the potential diagnostic markers and therapeutic targets for MI. We identified 14 plasma metabolites associated with MI, of which 8 plasma metabolites were linked to a decreased risk and 6 plasma metabolites were linked to an increased risk, underscoring the complicated nature of metabolic pathways influencing MI. The robustness of our findings was strengthened by the application of bidirectional MR, enabling a thorough exploration of causality. Observed associations were insignificantly impacted by horizontal pleiotropy, supporting the reliability of our conclusions. We demonstrated that MI did not substantially alter the levels of the identified 14 plasma metabolites. Furthermore, the novelty of our study lies in the identification of 13 plasma metabolites not previously reported in the literature in association with MI. This widens the spectrum of potential biomarkers and therapeutic targets, providing researchers and clinicians with a more comprehensive understanding of the metabolic landscape linked to MI.
Aa, et al.[31] investigated the alterations in plasma metabolites between individuals with and without MI. They demonstrated that individuals with elevated plasma levels of deoxyuridine, homoserine, or methionine displayed an increased risk of MI. In contrast to this study, our study did not identify a significant association between plasma metabolites (deoxyuridine, homoserine, or methionine) and the occurrence of MI. Instead, we identified a distinct set of metabolites, including propionylglycine levels, gamma-glutamylglycine levels, hexadecanedioate (C16-DC) levels, 1-palmitoyl-2-arachidonoyl-GPE (16:0/20:4) levels, behenoyl dihydrosphingomyelin (d18:0/22:0) levels, 1-stearoyl-2-docosahexaenoyl-GPE (18:0/22:6) levels, pentose acid levels, alpha-ketobutyrate levels, X-24546 levels, 5-acetylamino-6-formylamino-3-methyluracil levels, glycine levels, N-acetylputrescine to (N (1) + N (8))-acetylspermidine ratio, glycine to serine ratio, and mannose to trans-4-hydroxyproline ratio, which was significantly associated with the occurrence of MI. The study performed by Aa, et al.[31] consisted of 230 participants and comprised 146 suspected MI participants and 84 control individuals. In contrast, our study included 20,917 individuals with MI and 440,906 individuals without MI. The substantial difference in the sample size could be a significant factor contributing to inconsistent conclusions. Furthermore, in a study involving 4109 participants suspected of stable angina pectoris, Ding, et al.[30] explored the relationship between plasma glycine and acute myocardial infarction (AMI). After a median follow-up of 7.4 years, they revealed a more favorable baseline lipid profile associated with plasma glycine. After adjusting for potential confounding factors, they reported a negative correlation between plasma glycine and the risk of AMI, suggesting that elevated plasma glycine concentrations contributed to a reduced risk of AMI. Consistent with the findings of Ding, et al.[30], we similarly identified an association between glycine levels (OR = 0.936, 95% CI: 0.909–0.964, P < 0.001) and a decreased risk of MI.
The specific mechanisms linking the identified plasma metabolites to MI remain limited because of a shortage of previous research elucidating their functions in this cardiovascular condition. The lack of comprehensive studies hampers the understanding of the underlying mechanisms. Nevertheless, we speculated that the mechanisms potentially involved in the association between the identified plasma metabolites and MI could be attributed to the following factors: (1) anti-inflammatory properties: metabolites such as glycine have known anti-inflammatory properties. Hasegawa, et al.[32] reported that the presence of glycine significantly reduced the activation of nuclear factor-kappa B (NF-κB) and the degradation of inhibitor κBα (IκBα) in human coronary arterial endothelial cells subjected to stimulation by tumor necrosis factor (TNF)-α; and (2) oxidative stress response: 8 plasma metabolites identified were associated with a reduced risk of MI and could potentially be linked to oxidative stress response. One plausible hypothesis is that these metabolites, correlated with a decreased risk of MI, could alleviate oxidative stress within the body, thereby exerting a protective effect against cardiovascular diseases, including MI. However, considering the current lack of studies elucidating potential mechanisms, the hypothesis associated with oxidative stress response still requires further investigation and substantiation through additional research.
We elucidated crucial clinical insights, particularly in the context of early diagnosis, with significant implications for MI research. The identified plasma metabolites associated with MI risk could be potentially used for transformative advancements in early diagnostic strategies. The metabolites identified could serve as robust biomarkers for early detection of MI. Early diagnosis is a pivotal factor in improving patient outcomes and implementing timely interventions to mitigate the severity of MI. The identification of specific plasma metabolites linked to MI will allow clinicians access to an efficient diagnostic tool. Other clinical implications include the development of targeted diagnostic assays incorporating these metabolites into routine screening protocols. Such innovations could empower healthcare professionals to identify individuals at elevated risk for MI, allowing for personalized and preemptive interventions. Timely detection using metabolic signatures could usher in a new era of preventive cardiology, where interventions are tailored to an individual’s metabolic profile. Furthermore, understanding the metabolic underpinnings of MI will contribute to the development of point-of-care diagnostic tools, providing rapid and accessible assessments. Thus, the findings of the study can revolutionize clinical practice by enabling early and precise diagnoses, ultimately causing more effective and tailored treatment strategies.
LIMITATIONS
Although our study provides valuable insights into the relationship between plasma metabolites and MI, it is essential to acknowledge certain limitations impacting the interpretation of our findings. Firstly, we used publicly available GWAS data for plasma metabolites. Although these datasets provide a broad range of metabolites, they could not capture the full spectrum of metabolomic variations, potentially causing incomplete exposure characterization. Secondly, the identified plasma metabolites were based on GWAS summary statistics, and we did not have access to individual-level metabolite measurements. This limits our ability to explore dose–response relationships and assess the impact of metabolite concentration on MI risk continuously. Last but not least, the study population included in this research comprised individuals of European descent. Further research is warranted to validate whether the study findings are applicable to other populations as well.
CONCLUSIONS
In conclusion, our bidirectional MR study identified 14 plasma metabolites associated with MI, providing potential markers for early diagnosis and therapeutic targets for MI. Among these plasma metabolites, 8 plasma metabolites were linked to a decreased risk of MI, whereas 6 plasma metabolites were associated with an increased risk of MI. Altogether, our findings contribute to the understanding of the metabolic landscape associated with MI, as the identification of 13 plasma metabolites have not been previously reported in the literature. Future research is warranted to investigate the clinical implications and mechanisms underlying these associations, and additional studies across diverse populations are essential to enhance the generalizability of our results. Overall, our study provides valuable insights into the complex interplay between plasma metabolites and MI, opening possibilities for further investigation and potential clinical applications.
ACKNOWLEDGMENTS
The study was supported by the Guangxi Natural Science Foundation (No.2020GXNSFDA238007), the Key Research and Development Program of Guangxi (No.2023AB22024), and the Chongzuo Science and Technology Bureau Planning Project (No.FA2018026). All authors had no conflicts of interest to disclose.
Anderson JL, Morrow DA. Acute myocardial infarction. N Engl J Med 2017; 376: 2053−2064. doi: 10.1056/NEJMra1606915
[3]
Chang C, Cai R, Wu Q, et al. Uncovering the genetic link between acute myocardial infarction and ulcerative colitis co-morbidity through a systems biology approach. Cardiovasc Innov Appl 2023; 8: e978. doi: 10.15212/CVIA.2023.0034
[4]
Li YX, Wang BN, Fan FF, et al. Thirty-day outcomes of in-hospital multi-vessel versus culprit-only revascularization strategy for ST-segment elevation myocardial infarction with multivessel coronary disease. J Geriatr Cardiol 2023; 20: 485−494. doi: 10.26599/1671-5411.2023.07.005
[5]
Wu Q, Li LF, Chen YD. Advances in Journal of Geriatric Cardiology over the course of a decade. J Geriatr Cardiol 2020; 17: 733−739. doi: 10.11909/j.issn.1671-5411.2020.12.001
[6]
Johansson S, Rosengren A, Young K, et al. Mortality and morbidity trends after the first year in survivors of acute myocardial infarction: a systematic review. BMC Cardio vasc Disord 2017; 17: 53. doi: 10.1186/s12872-017-0482-9
[7]
Agarwal M, Agrawal S, Garg L, et al. Effect of chronic obstructive pulmonary disease on in-hospital mortality and clinical outcomes after ST-segment elevation myocardial infarction. Am J Cardiol 2017; 119: 1555−1559. doi: 10.1016/j.amjcard.2017.02.024
[8]
Hayıroğlu Mİ, Keskin M, Uzun AO, et al. Predictors of in-hospital mortality in patients with ST-segment elevation myocardial infarction complicated with cardiogenic shock. Heart Lung Circ 2019; 28: 237−244. doi: 10.1016/j.hlc.2017.10.023
[9]
Salari N, Morddarvanjoghi F, Abdolmaleki A, et al. The global prevalence of myocardial infarction: systematic review and meta-analysis. BMC Cardiovasc Disord 2023; 23: 206. doi: 10.1186/s12872-023-03231-w
[10]
Floegel A, Kühn T, Sookthai D, et al. Serum metabolites and risk of myocardial infarction and ischemic stroke: a targeted metabolomic approach in two German prospective cohorts. Eur J Epidemiol 2018; 33: 55−66. doi: 10.1007/s10654-017-0333-0
[11]
Ward-Caviness CK, Xu T, Aspelund T, et al. Improvement of myocardial infarction risk prediction via inflammation-associated metabolite biomarkers. Heart 2017; 103: 1278−1285. doi: 10.1136/heartjnl-2016-310789
[12]
Yu Z, Coresh J, Qi G, et al. A bidirectional Mendelian randomization study supports causal effects of kidney function on blood pressure. Kidney Int 2020; 98: 708−716. doi: 10.1016/j.kint.2020.04.044
[13]
Choi KW, Chen CY, Stein MB, et al. Assessment of bidirectional relationships between physical activity and depression among adults: a 2-sample Mendelian randomization study. JAMA Psychiatry 2019; 76: 399−408. doi: 10.1001/jamapsychiatry.2018.4175
[14]
Park S, Lee S, Kim Y, et al. Atrial fibrillation and kidney function: a bidirectional Mendelian randomization study. Eur Heart J 2021; 42: 2816−2823. doi: 10.1093/eurheartj/ehab291
[15]
Guan B, Chen XQ, Liu Y, et al. Causal effects of circulating vitamin levels on the risk of heart failure: a Mendelian randomization study. J Geriatr Cardiol 2023; 20: 195−204. doi: 10.26599/1671-5411.2023.03.007
[16]
Burgess S, Davey Smith G, Davies NM, et al. Guidelines for performing Mendelian randomization investigations: update for summer 2023. Wellcome Open Res 2023; 4: 186. doi: 10.12688/wellcomeopenres.15555.3
[17]
Skrivankova VW, Richmond RC, Woolf BAR, et al. Strengthening the reporting of observational studies in epidemiology using Mendelian randomization: the STROBE-MR statement. JAMA 2021; 326: 1614−1621. doi: 10.1001/jama.2021.18236
[18]
Chen Y, Lu T, Pettersson-Kymmer U, et al. Genomic atlas of the plasma metabolome prioritizes metabolites implicated in human diseases. Nat Genet 2023; 55: 44−53. doi: 10.1038/s41588-022-01270-1
[19]
Sakaue S, Kanai M, Tanigawa Y, et al. A cross-population atlas of genetic associations for 220 human phenotypes. Nat Genet 2021; 53: 1415−1424. doi: 10.1038/s41588-021-00931-x
[20]
Staley JR, Blackshaw J, Kamat MA, et al. PhenoScanner: a database of human genotype-phenotype associations. Bi oinformatics 2016; 32: 3207−3209. doi: 10.1093/bioinformatics/btw373
[21]
Palmer TM, Lawlor DA, Harbord RM, et al. Using multiple genetic variants as instrumental variables for modifiable risk factors. Stat Methods Med Res 2012; 21: 223−242. doi: 10.1177/0962280210394459
[22]
Levin MG, Judy R, Gill D, et al. Genetics of height and risk of atrial fibrillation: a Mendelian randomization study. PLoS Med 2020; 17: e1003288. doi: 10.1371/journal.pmed.1003288
[23]
Gill D, Efstathiadou A, Cawood K, et al. Education protects against coronary heart disease and stroke independently of cognitive function: evidence from Mendelian randomization. Int J Epidemiol 2019; 48: 1468−1477. doi: 10.1093/ije/dyz200
[24]
Shi X, Yuan W, Cao Q, et al. Education plays a crucial role in the pathway from poverty to smoking: a Mendelian randomization study. Addiction 2023; 118: 128−139. doi: 10.1111/add.16019
[25]
Larsson SC, Burgess S, Michaëlsson K. Association of genetic variants related to serum calcium levels with coronary artery disease and myocardial infarction. JAMA 2017; 318: 371−380. doi: 10.1001/jama.2017.8981
[26]
Bowden J, Del Greco M F, Minelli C, et al. A framework for the investigation of pleiotropy in two-sample summary data Mendelian randomization. Stat Med 2017; 36: 1783−1802. doi: 10.1002/sim.7221
[27]
Davey Smith G, Hemani G. Mendelian randomization: genetic anchors for causal inference in epidemiological studies. Hum Mol Genet 2014; 23: R89−R98. doi: 10.1093/hmg/ddu328
[28]
Bowden J, Davey Smith G, Haycock PC, et al. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol 2016; 40: 304−314. doi: 10.1002/gepi.21965
[29]
Burgess S, Thompson SG. Interpreting findings from Mendelian randomization using the MR-Egger method. Eur J Epidemiol 2017; 32: 377−389. doi: 10.1007/s10654-017-0255-x
[30]
Ding Y, Svingen GF, Pedersen ER, et al. Plasma glycine and risk of acute myocardial infarction in patients with suspected stable angina pectoris. J Am Heart Assoc 2015; 5: e002621. doi: 10.1161/JAHA.115.002621
[31]
Aa N, Lu Y, Yu M, et al. Plasma metabolites alert patients with chest pain to occurrence of myocardial infarction. Front Cardiovasc Med 2021; 8: 652746. doi: 10.3389/fcvm.2021.652746
[32]
Hasegawa S, Ichiyama T, Sonaka I, et al. Cysteine, histidine and glycine exhibit anti-inflammatory effects in human coronary arterial endothelial cells. Clin Exp Immunol 2012; 167: 269−274. doi: 10.1111/j.1365-2249.2011.04519.x
Table
1.
The overview of data sources used in this bidirectional Mendelian randomization study.
Variables
PubMed ID
Journal
Study accession
Year of publication
Sample size
Plasma metabolites
36635386
Nature Genetics
GCST90199621 to GCST90201020
2023
Approximately 8000 individuals of European descent
Myocardial infarction
34594039
Nature Genetics
ebi-a-GCST90018877
2021
461,823 individuals of European descent, comprising 20,917 individuals with myocardial infarction and 440,906 individuals without myocardial infarction
DownLoad:
