Genetic evidence against a causal relationship between myocardial infarction and urological malignancies

Introduction

The relationship between myocardial infarction (MI) and urinary system cancers has gained increasing attention in recent years, particularly due to the cardiovascular risks associated with cancer treatments rather than the malignancies themselves. Emerging evidence suggests that therapeutic interventions for prostate cancer (PCa), bladder cancer (BCa), and malignant neoplasm of kidney (MRN) may substantially increase cardiovascular event risks, creating complex bidirectional associations that extend beyond shared risk factors. In PCa management, androgen deprivation therapy (ADT) has been identified as a major contributor to cardiovascular morbidity, with contemporary meta-analyses demonstrating increased risks of MI and stroke among patients receiving ADT.1,2 Notably, an Italian real-world analysis revealed that gonadotropin-releasing hormone (GnRH) agonists confer higher cardiovascular risks compared to GnRH antagonists, highlighting treatment-specific cardiovascular profiles.3 This finding was corroborated by a large multinational study using five databases, which confirmed that GnRH agonists are associated with increased cardiovascular disease risk while GnRH antagonists exhibit fewer atherosclerotic effects.4 For BCa, radical cystectomy (RC) presents significant perioperative cardiovascular challenges, with studies reporting 90-day mortality rates of 5.2%, where MI represents a leading cause of acute heart failure.5 Even minimally invasive approaches carry substantial risks, as evidenced by reports of acute MI following robot-assisted radical cystectomy.6 While direct evidence for kidney malignancies remains limited, these cancers are increasingly recognized in the context of multimorbidity studies examining cardiovascular-renal interactions.7 These treatment-related cardiovascular complications underscore the urgent need to distinguish between causal relationships and iatrogenic associations in understanding MI-urinary system cancer links.

The inconsistent results from observational studies can be attributed to several methodological limitations, including confounding by shared risk factors, reverse causation, and selection bias. Traditional epidemiological approaches struggle to disentangle the causal relationships between cardiovascular events and cancer risk due to the presence of common risk factors such as smoking, obesity, diabetes, and chronic inflammation.8 Furthermore, the competing risk of death from cardiovascular causes may mask true cancer associations, while surveillance bias following MI diagnosis could artificially inflate cancer detection rates.9 These limitations necessitate alternative analytical approaches that can overcome the inherent biases of conventional observational studies and provide more reliable evidence for causal inference.

Mendelian randomization (MR) has emerged as a powerful tool for causal inference in epidemiological research, leveraging genetic variants as instrumental variables to mimic randomized controlled trials.10 This approach exploits the random assortment of genetic variants at conception to overcome confounding and reverse causation inherent in observational studies. Recent applications of MR in cardio-oncology have provided valuable insights into causal relationships between cardiovascular risk factors and cancer outcomes. Notable studies have examined the causal effects of lipid levels on various cancer risks,11 blood pressure on cancer development,12 and inflammatory markers on malignancy outcomes.13 These investigations have demonstrated the utility of genetic epidemiological approaches in elucidating the complex interplay between cardiovascular health and cancer biology.

Despite growing interest in cardio-oncology, the causal relationship between acute cardiovascular events, specifically MI, and urinary system cancers remains unexplored using genetic approaches. Urinary system malignancies, specifically PCa, BCa, and MRN, represent significant health burdens with distinct epidemiological patterns and potential interactions with cardiovascular systems.14 PCa, the most common non-cutaneous malignancy in men, shares several risk factors with cardiovascular disease, including age, metabolic syndrome, and chronic inflammation.15 BCa, strongly associated with smoking and environmental exposures, may share inflammatory pathways with cardiovascular disease.16 Kidney malignancies, while less common, are increasingly recognized for their association with cardiovascular comorbidities, particularly hypertension and chronic kidney disease, which may create complex bidirectional relationships with cardiac events.17 The absence of causal evidence for MI-cancer relationships in these specific urinary system malignancies represents a critical knowledge gap that limits our understanding of cardio-oncological interactions and clinical management strategies.

To address this knowledge gap, we conducted a comprehensive bidirectional two-sample MR study to investigate the causal relationships between MI and three major urinary system cancers: PCa, BCa, and MRN. We hypothesized that genetic predisposition to MI would not demonstrate significant causal effects on urinary system cancer risk, based on the mixed results from observational studies and the distinct biological pathways underlying acute cardiovascular events versus cancer development. By employing multiple MR methodologies and utilizing both discovery and validation datasets, our study aims to provide robust genetic evidence regarding the causal nature of MI-urinary system cancer associations and inform clinical practice in cardio-oncology management.

Methods and Materials

Study design

We conducted a bidirectional two-sample MR analysis following STROBE-MR guidelines (Supplementary Material 1) to investigate potential causal relationships between MI and urinary system cancers, specifically PCa, BCa, and MRN. Our analytical framework adhered to three fundamental MR assumptions: relevance (genome-wide significance P < 5 × 10−8), independence (absence of confounding factors), and exclusion restriction (no horizontal pleiotropy). Genetic variants underwent linkage disequilibrium clumping to ensure independence (r² < 0.001, distance > 10,000 kb). The analytical strategy employed inverse variance weighted (IVW) as the primary method, complemented by sensitivity analyses including MR-Egger, weighted median, weighted mode, and MR-PRESSO approaches. Bidirectional causality was assessed through reverse MR analysis, with all computations performed using R software (version 4.2.1) utilizing TwoSampleMR (version 0.6.15) and MR-PRESSO (version 1.0) packages. The comprehensive study design and three core MR assumptions are illustrated in Figure 1. A key strength of the MR approach in addressing potential reverse causation lies in the temporal relationship between genetic variants and disease outcomes. Since genetic variants are randomly allocated at conception and remain fixed throughout life, they serve as proxies for lifelong exposure that precede disease development. This genetic randomization effectively addresses concerns about reverse causation, particularly the possibility that undiagnosed urological malignancies might influence myocardial infarction risk through subclinical metabolic or inflammatory pathways. Unlike observational studies where the temporal sequence of exposure and outcome may be uncertain, genetic instruments provide a natural experiment that establishes the direction of causation from genetically determined MI liability to cancer risk, rather than the reverse.

FIGURE 1. Conceptual framework and assumptions of bidirectional two-sample Mendelian randomization analysis examining causal relationships between myocardial infarction and urinary system cancers. Abbreviations: MR, Mendelian randomization; MI, Myocardial infarct; IV, Instrumental variable; PCa, Prostate cancer; BCa, Bladder cancer; MRN, Malignant neoplasm of kidney

The directed acyclic graph (DAG) illustrates the three fundamental assumptions of Mendelian randomization: (1) genetic instruments are significantly associated with the exposure (myocardial infarction); (2) genetic instruments are independent of confounding factors (indicated by crossed-out dashed lines); and (3) genetic instruments affect outcomes (prostate cancer, bladder cancer, and malignant neoplasm of kidney) only through the exposure pathway (exclusion restriction). The framework shows instrumental variable selection criteria, including genome-wide significance (P < 5 × 10−8) and independence requirements (r² < 0.001; kb > 10,000). Visual representations include DNA sequences as genetic instruments, anatomical illustrations of the heart (representing MI) and urinary system (representing cancer outcomes).

Data sources

All MR analyses utilized publicly available genome-wide association study (GWAS) summary statistics accessed through the IEU Open GWAS database. We selected MI phenotype (ID: ebi-a-GCST011364) as the exposure data,18 while outcome datasets comprised three urinary system cancers with discovery and validation cohorts for enhanced reliability. PCa discovery data originated from UK Biobank, with PRACTICAL consortium providing validation samples; BCa discovery cohort was sourced from MRC-IEU, with UK Biobank (ID: ieu-b-4874) serving as validation; MRN utilized FinnGen_R1219 for discovery and the MRC-IEU consortium for validation analyses. To minimize population stratification bias and maintain methodological consistency, we restricted analyses exclusively to individuals of European ancestry across both exposure and outcome datasets, thereby enhancing the reliability of causal inference. Detailed information regarding data sources is provided in Supplementary Table 1 and Table 1.

Instrumental variable selection

Single nucleotide polymorphisms (SNPs) served as genetic instrumental variables in our MR framework through a rigorous multi-stage selection protocol. We initially extracted genome-wide significant variants (P < 5 × 10−8) from exposure GWAS data, maintaining this threshold consistently across forward and reverse MR analyses; subsequently applied linkage disequilibrium filtering (r² < 0.001, clumping distance = 10,000 kb) to eliminate correlated variants that could introduce analytical bias; harmonized effect estimates and allele frequencies between exposure and outcome datasets to ensure analytical consistency. Instrumental variable strength was evaluated using F-statistics calculated as F = R²(n – k − 1)/k(1 − R²), with variants showing F < 10 excluded to maintain adequate statistical power.20 We systematically screened the GWAS catalog (https://www.ebi.ac.uk/gwas/) to identify and remove SNPs significantly associated (P < 1 × 10−5) with established confounders, including chronic kidney disease,21 lipid profiles,22 smoking behavior,23 and hypertension,24 effectively minimizing confounding risks. Comprehensive characteristics of the selected genetic instruments are detailed in Supplementary Table 2. Importantly, validation confirmed that all selected instrumental variables lacked significant associations with potential confounders or outcome variables through alternative pathways (Supplementary Table 3).

Statistical analysis

The IVW method constituted our primary analytical approach, assuming instrument validity while providing optimal statistical power under the no horizontal pleiotropy assumption through weighted averaging of SNP-specific causal estimates.25 Cochran’s Q test systematically assessed heterogeneity among exposure-associated SNPs, with fixed-effects models applied when heterogeneity was absent and random-effects models employed when detected.26 Multiple complementary MR approaches enhanced result robustness: MR-Egger regression utilized intercept terms to detect directional pleiotropy while maintaining validity under instrument invalidity (p < 0.05 indicating pleiotropy presence);27 weighted median maintained consistency when up to 50% of statistical weight was derived from valid instruments;28 weighted mode estimation identified patterns in weighted SNP-specific causal estimates, particularly valuable under horizontal pleiotropy conditions.29

Secondary analyses comprehensively evaluated heterogeneity, pleiotropy, and sensitivity through MR-Egger intercept analysis, where near-zero intercepts suggested low pleiotropy probability, combined with MR-PRESSO global testing for systematic outlier detection and management.30,31 Following identification of outlying variants, we implemented iterative optimization procedures, repeating MR analyses after outlier removal to ensure robust causal estimates.32 Extensive sensitivity testing included leave-one-out (LOO) assessment, which systematically excluded individual SNPs to evaluate their influence on overall effect estimates, alongside examination of result consistency across different MR methods to ensure stable and reliable causal inference under potential violations of instrumental variable assumptions.33

Bidirectional causal relationships were assessed through reverse MR analyses employing identical standardized protocols, with the three urinary system cancers serving as exposures and MI as the outcome variable. Steiger testing verified directional validity by confirming stronger correlations between genetic instruments and their respective exposure variables compared to outcome variables,34 thereby establishing appropriate causal direction. All statistical analyses were conducted using R software (version 4.2.1) with the TwoSampleMR (version 0.6.15) and MR-PRESSO (version 1.0) packages, providing comprehensive bidirectional validation of causal relationships between MI and urinary system cancers.

Results

Instrumental variable characteristics

Following the systematic selection process, we identified genetic variants as instrumental variables for MI across different analyses. For the forward MR analyses examining MI as exposure, 19 SNPs were selected for both PCa and MRN analyses, while 6 SNPs were used for the BCa discovery cohort and 21 SNPs for BCa validation cohort. The variation in SNP numbers reflected the availability of harmonized variants across different outcome datasets. All selected instrumental variables demonstrated robust statistical power, with F-statistics exceeding the conventional threshold of 10, ensuring adequate instrument strength for causal inference. For reverse MR analyses, we obtained 28 SNPs from the PCa discovery cohort, 108 SNPs from the validation cohort, and 4 SNPs from the MRN discovery cohort as instrumental variables. We applied the Benjamini-Hochberg false discovery rate (FDR) correction method to account for multiple testing across all exposure-outcome combinations. With this correction, the adjusted significance threshold was set at α = 0.025 (0.05/2 for bidirectional testing). Under this corrected threshold, both p-values (0.082 and 0.054) remained non-significant, supporting our overall null findings. The heterogeneity statistics (I² = 0% and Q-statistic p > 0.05 for most analyses) indicate minimal heterogeneity among the genetic instruments, which is within acceptable ranges for MR studies and suggests that our genetic instruments act through similar biological pathways. Low heterogeneity supports the reliability of our null findings and indicates that the genetic variants are not acting through diverse, contradictory mechanisms that would compromise the validity of our causal estimates. Where modest heterogeneity was observed (I² = 15%–25% in some sensitivity analyses), the consistency of effect directions across methods supports the robustness of our conclusions. Comprehensive characteristics of these genetic instruments, including individual SNP effects, allele frequencies, and statistical parameters, are detailed in Supplementary Table 2.

Forward MR analyses: effect of MI on urinary system cancer risk

MR analysis revealed no significant causal effect of genetically predicted MI on the risk of any examined urinary system cancer. For PCa, the primary IVW analysis showed null associations in both discovery (OR = 1.000, 95% CI: 0.995–1.004, p = 0.885) and validation (OR = 0.964, 95% CI: 0.904–1.027, p = 0.257) cohorts. Consistent conclusions were obtained through additional statistical methods, with MR-Egger yielding ORs of 1.000 (95% CI: 0.989–1.010, p = 0.963) and 0.998 (95% CI: 0.860–1.158, p = 0.980) for discovery and validation analyses, respectively. Similarly, weighted median and weighted mode approaches produced comparable results, with all confidence intervals encompassing the null value and p-values exceeding 0.05.

BCa analysis demonstrated remarkably consistent null findings across all methodological approaches. In the discovery cohort, IVW analysis yielded an OR of 1.000 (95% CI: 0.999–1.001, p = 0.795), with MR-Egger (OR = 1.000, 95% CI: 0.998–1.002, p = 0.787), weighted median (OR = 1.000, 95% CI: 0.999–1.001, p = 0.701), and weighted mode (OR = 1.000, 95% CI: 0.999–1.001, p = 0.706) methods producing virtually identical estimates. The validation cohort analysis further corroborated these findings, with all four methods yielding ORs of 1.000 and narrow confidence intervals tightly centered around the null value (all p > 0.05).

For MRN, forward MR analysis in the discovery cohort showed no significant causal association, with IVW producing an OR of 0.989 (95% CI: 0.866–1.130, p = 0.874). Alternative methods yielded consistent results: MR-Egger OR = 1.060 (95% CI: 0.773–1.454, p = 0.721), weighted median OR = 1.044 (95% CI: 0.867–1.256, p = 0.650), and weighted mode OR = 1.056 (95% CI: 0.863–1.293, p = 0.601). Notably, the validation cohort analysis revealed marginally significant results with the IVW method (OR = 0.999, 95% CI: 0.998–1.000, p = 0.082) and weighted median approach (OR = 0.999, 95% CI: 0.998–1.000, p = 0.054), though these remained non-significant after multiple testing considerations and were not consistently reproduced across all analytical methods. The comprehensive forest plot summarizing these null associations for forward MR analyses is presented in Figure 2 and Supplementary Table 4.

FIGURE 2. Forest plot of forward mendelian randomization analysis evaluating the causal effect of myocardial infarction on urinary system cancer risk using multiple statistical methods. Abbreviations: SNP, Single nucleotide polymorphism; 95% CI, 95% confidence interval; IVW, Inverse variance weighted; OR, Odds ratio; MI, Myocardial infarct; IV, Instrumental variable; PCa, Prostate cancer; BCa, Bladder cancer; MRN, Malignant neoplasm of kidney

Results are presented as odds ratios (OR) with 95% confidence intervals (CI) for discovery and validation cohorts across three cancer types. All analyses demonstrate null associations with confidence intervals encompassing unity (OR = 1.0). Prostate cancer analyses utilized 19 SNPs in both cohorts, bladder cancer employed 6 SNPs (discovery) and 21 SNPs (validation), while malignant neoplasm of kidney used 19 SNPs (discovery) and 6 SNPs (validation). The forest plot shows consistent null findings across inverse variance weighted (IVW), MR-Egger, weighted median, and weighted mode methods, with all p-values exceeding 0.05. Point estimates cluster tightly around the null value, indicating no causal effect of MI on urinary system cancer risk.

Reverse MR analyses: effect of urinary system cancers on MI risk

Reverse MR analysis examined whether genetic predisposition to urinary system cancers causally influences MI risk. For PCa, both discovery and validation cohorts demonstrated no significant causal effects on MI risk. The discovery cohort analysis using 28 SNPs yielded an IVW OR of 1.057 (95% CI: 0.559–1.997, p = 0.865), with consistent null findings across alternative methods: MR-Egger OR = 0.250 (95% CI: 0.062–1.011, p = 0.063), weighted median OR = 1.202 (95% CI: 0.481–3.003, p = 0.694), and weighted mode OR = 1.200 (95% CI: 0.235–6.120, p = 0.828). The validation cohort, utilizing 108 SNPs, corroborated these findings with IVW OR = 1.002 (95% CI: 0.975–1.031, p = 0.868) and consistent results from other methodological approaches (all p > 0.05).

MRN showed no causal effect on MI risk in the discovery cohort analysis. Using 4 SNPs as instrumental variables, the IVW method produced an OR of 0.983 (95% CI: 0.899–1.074, p = 0.700), with alternative approaches yielding comparable estimates: MR-Egger OR = 1.063 (95% CI: 0.898–1.259, p = 0.551), weighted median OR = 0.979 (95% CI: 0.883–1.087, p = 0.693), and weighted mode OR = 1.031 (95% CI: 0.908–1.171, p = 0.669). No validation cohort analysis was available for the reverse direction of MRN to MI. Figure 3 visually summarizes these reverse MR findings through forest plot representation.

FIGURE 3. Forest plot of reverse mendelian randomization analysis assessing the causal effect of urinary system cancers on myocardial infarction risk. Abbreviations: SNP, Single nucleotide polymorphism; 95% CI, 95% confidence interval; IVW, Inverse variance weighted; OR, Odds ratio; MI, Myocardial infarct; IV, Instrumental variable; PCa, Prostate cancer; BCa, Bladder cancer; MRN, Malignant neoplasm of kidney

Reverse causality analysis employed varying numbers of genetic instruments: 28 SNPs for prostate cancer discovery, 108 SNPs for validation, and 4 SNPs for malignant neoplasm of kidney discovery cohort. All analyses consistently demonstrate null associations with odds ratios clustering around unity and confidence intervals spanning the null value. The IVW, MR-Egger, weighted median, and weighted mode methods yield concordant results with non-significant p-values (all > 0.05), providing robust evidence against causal effects of genetic predisposition to urinary system cancers on MI risk. The wider confidence intervals for some analyses reflect the varying statistical power associated with different numbers of instrumental variables.

Sensitivity analyses

Comprehensive sensitivity analyses confirmed the robustness of our findings and absence of methodological violations. MR-Egger intercept analyses revealed no evidence of directional pleiotropy across most comparisons, with intercept values close to zero and non-significant p-values (Supplementary Table 5). Notable exceptions included marginally significant intercepts for MI-BCa discovery analysis, though this did not materially affect the overall null conclusions. MR-PRESSO global tests detected minimal evidence of horizontal pleiotropy, with most analyses showing non-significant global test p-values, indicating absence of systematic outlying variants.

Heterogeneity assessment using Cochran’s Q statistic demonstrated acceptable levels of between-SNP heterogeneity for most analyses (Supplementary Table 6). Significant heterogeneity was observed in MI-PCa validation analysis (Q = 37.78, p = 0.004) and borderline significance in PCa validation-MI reverse analysis (Q = 130.91, p = 0.058). However, the use of random-effects IVW models in these instances ensured appropriate handling of heterogeneity without compromising causal estimates.

Steiger testing confirmed correct causal directionality for all analyses, with genetic instruments demonstrating stronger associations with their respective exposure variables compared to outcome variables (all Steiger test p-values < 0.001 and direction = TRUE; Supplementary Table 7). LOO sensitivity analysis showed that no single SNP substantially influenced the overall causal estimates, confirming the stability of our findings across all exposure-outcome pairs. Detailed visualization of LOO analyses, scatter plots illustrating the relationship between genetic associations with exposures and outcomes, and comprehensive forest plots for all analytical methods are provided in Figures 4–6. The consistency of results across multiple MR methods, combined with successful sensitivity analyses, provides robust evidence against causal relationships between MI and urinary system cancers in either direction.

FIGURE 4. Leave-One-Out (LOO) analysis visualizing genetic associations between cardiovascular disease and urological cancers. (A) Myocardial infarction → Prostate cancer (Discovery cohort). (B) Myocardial infarction → Prostate cancer (Validation cohort). (C) Myocardial infarction → Bladder cancer (Discovery cohort). (D) Myocardial infarction → Bladder cancer (Validation cohort). (E) Myocardial infarction → Malignant neoplasm of kidney (Discovery cohort). (F) Myocardial infarction → Malignant neoplasm of kidney (Validation cohort). (G) Prostate cancer (Discovery cohort) → Myocardial infarction. (H) Prostate cancer (Validation cohort) → Myocardial infarction. (I) Malignant neoplasm of kidney (Discovery cohort) → Myocardial infarction

FIGURE 5. Mendelian randomization analysis of causal relationships between myocardial infarction and urological cancers. (A) Myocardial infarction → Prostate cancer (Discovery cohort). (B) Myocardial infarction → Prostate cancer (Validation cohort). (C) Myocardial infarction → Bladder cancer (Discovery cohort). (D) Myocardial infarction → Bladder cancer (Validation cohort). (E) Myocardial infarction → Malignant neoplasm of kidney (Discovery cohort). (F) Myocardial infarction → Malignant neoplasm of kidney (Validation cohort). (G) Prostate cancer (Discovery cohort) → Myocardial infarction. (H) Prostate cancer (Validation cohort) → Myocardial infarction. (I) Malignant neoplasm of kidney (Discovery cohort) → Myocardial infarction

FIGURE 6. Forest plot analysis of genetic associations between myocardial infarction and urological cancers. (A) Myocardial infarction → Prostate cancer (Discovery cohort). (B) Myocardial infarction → Prostate cancer (Validation cohort). (C) Myocardial infarction → Bladder cancer (Discovery cohort). (D) Myocardial infarction → Bladder cancer (Validation cohort). (E) Myocardial infarction → Malignant neoplasm of kidney (Discovery cohort). (F) Myocardial infarction → Malignant neoplasm of kidney (Validation cohort). (G) Prostate cancer (Discovery cohort) → Myocardial infarction. (H) Prostate cancer (Validation cohort) → Myocardial infarction. (I) Malignant neoplasm of kidney (Discovery cohort) → Myocardial infarction

Figure 4 shows Leave-One-Out (LOO) analyses visualizing genetic associations between myocardial infarction and urological cancers (prostate cancer, bladder cancer, and kidney malignancy) across discovery and validation cohorts. The reverse analyses (Figure 4G–I) contain notably fewer SNPs because, under strict screening thresholds, these datasets lacked sufficient SNPs meeting the inclusion criteria for analysis.

Figure 5 presents scatter plots of Mendelian Randomization analyses examining potential causal relationships between myocardial infarction and urological cancers (prostate, bladder, and kidney). All visualization results across Figure 5A–I demonstrate no significant slopes in the regression lines and widely scattered SNP effects, providing evidence that there is no genetic causal association between these conditions despite their epidemiological correlation.

Figure 6 presents comprehensive forest plots displaying SNP-specific and overall genetic associations between myocardial infarction and urological cancers (prostate, bladder, and kidney) across discovery and validation cohorts. The visualization results across all panels (Figure 6A–I) demonstrate that the confidence intervals for the overall effects (shown as red horizontal lines at the bottom of each panel) consistently cross zero, providing clear evidence that there is no significant genetic association between myocardial infarction and these urological cancers.

Discussion

This comprehensive bidirectional MR analysis provides robust evidence against causal relationships between MI and three major urinary system cancers—PCa, BCa, and MRN. Our findings consistently demonstrate null associations across multiple analytical approaches, discovery and validation cohorts, and both forward and reverse causal directions. The striking consistency of results, with odds ratios clustering tightly around unity and confidence intervals spanning the null value, suggests that despite shared epidemiological risk factors between cardiovascular disease and cancer, direct causal pathways linking MI to urogenital malignancies are unlikely to exist. This represents the first systematic evaluation of bidirectional causality between MI and urinary system cancers using large-scale genetic data, addressing a longstanding question in cardio-oncology research and helping to clarify inconsistent observations from previous epidemiological studies.

The absence of causal associations in our MR analysis contrasts with several observational studies that have reported both positive and negative correlations between cardiovascular disease and various cancers.35–37 While some longitudinal cohort studies have suggested increased cancer incidence following acute coronary events,38,39 others have documented protective effects or null associations depending on cancer type,40 follow-up duration, and population characteristics. These discrepancies likely reflect fundamental limitations inherent to observational research designs, including residual confounding from unmeasured variables, reverse causation bias where subclinical malignancy influences cardiovascular risk, and temporal confounding where shared risk factors operate over different time scales. Our MR approach circumvents these methodological challenges by utilizing genetic variants as instrumental variables, which are randomly allocated at conception and remain fixed throughout life, thereby providing unconfounded estimates of causal effects. The consistency between our null findings and recent meta-analyses questioning the strength of MI-cancer associations41 suggests that previously reported correlations may represent spurious relationships rather than genuine causal mechanisms.

The methodological strengths of our investigation substantially enhance confidence in the null findings. The bidirectional design enables evaluation of causality in both directions, addressing whether MI predisposes to urinary cancers or whether genetic liability to these malignancies influences MI risk. Our analytical framework incorporated multiple complementary MR methods, each operating under different assumptions about instrumental variable validity, with the remarkable concordance of IVW, MR-Egger, weighted median, and weighted mode results providing robust evidence against horizontal pleiotropy. The use of discovery and validation cohorts derived from independent populations further strengthens causal inference by demonstrating reproducibility across different genetic architectures and population structures. Additionally, our comprehensive sensitivity analyses, including Steiger directionality testing, leave-one-out analysis, and systematic evaluation of heterogeneity and pleiotropy, confirmed the stability of causal estimates and absence of methodological violations. The utilization of large-scale GWAS data encompassing hundreds of thousands of individuals provides adequate statistical power to detect clinically meaningful effect sizes, making it unlikely that true causal associations were missed due to insufficient sample sizes.

While our findings demonstrate no direct causal relationships, the well-established co-occurrence of cardiovascular disease and cancer in clinical practice warrants mechanistic consideration. Both conditions share numerous traditional risk factors, including advanced age, tobacco exposure, obesity, diabetes, and chronic inflammation,42,43 which may create apparent associations in observational studies without implying direct causation. The absence of genetic evidence for causal pathways suggests that these diseases may represent parallel consequences of common upstream determinants rather than sequential pathological processes. Furthermore, the complex interplay between cardiovascular medications, particularly antiplatelet agents and statins, and cancer development has been extensively studied with mixed results,44–46 indicating that pharmacological rather than pathophysiological mechanisms might explain some observed associations. The metabolic and inflammatory consequences of acute MI,47 while substantial, may not be sufficient to directly initiate or promote urogenital carcinogenesis over clinically relevant timeframes.48 This interpretation aligns with the understanding that cancer development typically requires prolonged exposure to carcinogenic stimuli,49 whereas the acute inflammatory response following MI, though intense, is relatively brief and may not provide sustained oncogenic pressure.47,50

Several limitations merit consideration when interpreting our results. The restriction to individuals of European ancestry, while necessary to minimize population stratification bias, potentially limits generalizability to other ethnic populations where different genetic architectures or environmental interactions might exist. The statistical power for detecting weak causal effects, particularly for MRN where instrumental variables were limited, may have been insufficient despite our large sample sizes. This is especially relevant for the marginally significant results observed in the MRN validation analysis, which, while not reaching conventional significance thresholds, raise the possibility of very weak causal effects, which warrant cautious interpretation. Additionally, while MR analysis effectively controls for measured and unmeasured confounding through genetic randomization, the potential for genetic pleiotropy or linkage disequilibrium with unmeasured variants could introduce bias, though our extensive pleiotropy testing provides reassurance against systematic violations. The temporal aspects of causation also deserve consideration, as our analysis captures lifelong genetic predisposition rather than the acute effects of MI, potentially missing short-term causal relationships that operate immediately following cardiac events. The temporal aspects of causation deserve particular consideration in interpreting our null findings. While MR analysis effectively controls for lifelong genetic predisposition, it may not capture acute post-MI effects that could theoretically influence cancer risk within shorter time windows (e.g., 1–2 years post-MI). The acute inflammatory response, immune suppression, and metabolic perturbations following MI could potentially create transient windows of increased cancer susceptibility that would not be detected through our genetic approach, which reflects average effects over entire lifespans. Additionally, competing mortality risk in the immediate post-MI period might mask cancer development that would otherwise become clinically apparent, creating a complex interplay between acute cardiovascular events and cancer detection that genetic instruments cannot fully address.

From a cardio-oncology guideline perspective,51 our findings suggest that current surveillance recommendations should focus on established risk factors rather than implementing MI-specific cancer screening protocols. Clinicians managing patients with both cardiovascular disease and urological malignancies can approach these conditions as potentially independent disease processes, with treatment decisions guided by individual risk stratification rather than assumed causal relationships. This evidence supports the development of parallel care pathways that address cardiovascular and oncological risks through evidence-based approaches tailored to each condition’s specific risk profile, rather than unified protocols assuming causal interdependence.

The clinical implications of our findings are substantial for both cardiovascular and oncological practice. Our results do not support intensified cancer surveillance protocols based solely on MI history, as the absence of causal relationships suggests that increased screening would not yield proportional benefits compared to standard risk-stratified approaches. However, this should not diminish the importance of addressing shared modifiable risk factors, as optimal management of obesity, smoking cessation, diabetes control, and blood pressure regulation benefits both cardiovascular and cancer outcomes through independent pathways. For clinicians managing patients with concurrent MI and cancer diagnoses, our findings suggest that these conditions should be viewed as potentially independent disease processes requiring parallel rather than sequential therapeutic attention. Patient counseling can be informed by these results, as individuals with MI history need not harbor excessive concerns about subsequent urinary cancer development beyond their baseline population risk. From a research perspective, our findings redirect attention toward investigating shared environmental determinants and identifying novel therapeutic targets that might simultaneously benefit both cardiovascular and cancer outcomes, rather than focusing on direct mechanistic pathways between these conditions.

Conclusions

This bidirectional MI study demonstrates no causal relationships between MI and urinary system cancers, including PCa, BCa, and MRN. The consistent null findings across multiple analytical methods, population cohorts, and causal directions provide robust evidence that apparent epidemiological associations likely reflect shared risk factors rather than direct biological causation. These results have important clinical implications, suggesting that MI-specific cancer screening protocols are not warranted, while emphasizing the continued importance of comprehensive risk factor modification for both cardiovascular and oncological health. Future research should focus on multi-ethnic populations and investigation of shared therapeutic targets that benefit both disease domains through independent mechanisms.

Acknowledgement

None.

Funding Statement

This work was supported by the 2021 Wuhan Medical Research Program (Grant No. WX21D12).

Author Contributions

Wei Zhang conceived and designed the study, performed the data analysis, and wrote the manuscript. Xixi Peng supervised the study and provided critical revision. All authors reviewed and approved the final version of the manuscript.

Availability of Data and Materials

All data generated or analyzed in this study are included in this published article and its supplementary information files and are available from the corresponding author upon request.

Ethics Approval

Not applicable.

Informed Consent

Not applicable.

Conflicts of Interest

The authors declare no conflicts of interest.

Supplementary Materials

The supplementary material is available online at https://www.techscience.com/doi/10.32604/cju.2026.072565/s1.

References

1. Hu JR, Duncan MS, Morgans AK et al. Cardiovascular effects of androgen deprivation therapy in prostate cancer: contemporary meta-analyses. Arterioscler Thromb Vasc Biol 2020;40(3):e55–e64. doi:10.1161/atvbaha.119.313046. [Google Scholar] [PubMed] [CrossRef]

2. Crawford DE, Albala D, Garnick MB et al. Optimizing outcomes in men with prostate cancer: the cardiovascular event lowering (CaELo) pathways. Can J Urol 2024;31(211820–11825. [Google Scholar] [PubMed]

3. Perrone V, Degli Esposti L, Giacomini E, Veronesi C, Blini V, Oderda M. Cardiovascular risk profile in prostate cancer patients treated with GnRH agonists vs. antagonists: an italian real-world analysis. Ther Clin Risk Manag 2020;16:393–401. doi:10.2147/tcrm.S249208. [Google Scholar] [PubMed] [CrossRef]

4. George G, Garmo H, Scailteux LM et al. Risk of cardiovascular disease following gonadotropin-releasing hormone agonists vs. antagonists in prostate cancer: real-world evidence from five databases. Int J Cancer 2021;148(9):2203–2211. doi:10.1002/ijc.33397. [Google Scholar] [PubMed] [CrossRef]

5. Kotov SV, Khachatryan AL, Kotova DP et al. Analysis of the results of eras protocol in real-life clinical practice after radical cystectomy (the first prospective multicenter study in Russia). Urologiia 2019;6:60–66. [Google Scholar]

6. Guliev BG, Komyakov BK, Bolokotov RR, Al-Attar TK. Robot-assisted radical cystectomy with neobladder formation. Urologiia 2020;5:54–60. [Google Scholar]

7. Carli V, Wasserman D, Hadlaczky G et al. A protocol for a multicentre, parallel-group, pragmatic randomised controlled trial to evaluate the nevermind system in preventing and treating depression in patients with severe somatic condi tions. BMC Psychiatry 2020;20(1):93. doi:10.1186/s12888-020-02494-3. [Google Scholar] [PubMed] [CrossRef]

8. Koene RJ, Prizment AE, Blaes A, Konety SH. Shared risk factors in cardiovascular disease and cancer. Circulation 2016;133(11):1104–1114. doi:10.1161/circulationaha.115.020406. [Google Scholar] [PubMed] [CrossRef]

9. Malmborg M, Christiansen CB, Schmiegelow MD, Torp-Pedersen C, Gislason G, Schou M. Incidence of new onset cancer in patients with a myocardial infarction—a nationwide cohort study. BMC Cardiovasc Disord 2018;18(1):198. doi:10.1186/s12872-018-0932-z. [Google Scholar] [PubMed] [CrossRef]

10. Su Q, Li J, Lu Y et al. Spleen volume in relation to ulcerative colitis and crohn’s disease: a mendelian randomization study. Sci Rep 2025;15(1):6588. doi:10.1038/s41598-025-90104-1. [Google Scholar] [PubMed] [CrossRef]

11. Nowak C, Ärnlöv J. A mendelian randomization study of the effects of blood lipids on breast cancer risk. Nat Commun 2018;9(1):3957. doi:10.1038/s41467-018-06467-9. [Google Scholar] [PubMed] [CrossRef]

12. Carreras-Torres R, Johansson M, Haycock PC et al. Role of obesity in smoking behaviour: mendelian randomisation study in Uk biobank. BMJ 2018;361:k1767. doi:10.1136/bmj.k1767. [Google Scholar] [PubMed] [CrossRef]

13. Georgakis MK, Gill D, Rannikmäe K et al. Genetically determined levels of circulating cytokines and risk of stroke. Circulation 2019;139(2):256–268. doi:10.1161/circulationaha.118.035905. [Google Scholar] [PubMed] [CrossRef]

14. Siegel RL, Miller KD, Wagle NS, Jemal A. Cancer statistics, 2023. CA Cancer J Clin 2023;73(1):17–48. doi:10.3322/caac.21763. [Google Scholar] [PubMed] [CrossRef]

15. Leong DP, Guha A, Morgans AK, Niazi T, Pinthus JH. Cardiovascular risk in prostate cancer: jacc: cardiooncology state-of-the-art review. JACC CardioOncol 2024;6(6):835–846. doi:10.1016/j.jaccao.2024.09.012. [Google Scholar] [PubMed] [CrossRef]

16. Leal J, Luengo-Fernandez R, Sullivan R, Witjes JA. Economic burden of bladder cancer across the European union. Eur Urol 2016;69(3):438–447. doi:10.1016/j.eururo.2015.10.024. [Google Scholar] [PubMed] [CrossRef]

17. Schuett K, Marx N, Lehrke M. The cardio-kidney patient: epidemiology, clinical characteristics and therapy. Circ Res 2023;132(8):902–914. doi:10.1161/circresaha.122.321748. [Google Scholar] [PubMed] [CrossRef]

18. Hartiala JA, Han Y, Jia Q et al. Genome-wide analysis identifies novel susceptibility loci for myocardial infarction. Eur Heart J 2021;42(9):919–933. doi:10.1093/eurheartj/ehaa1040. [Google Scholar] [PubMed] [CrossRef]

19. Kurki MI, Karjalainen J, Palta P et al. Finngen provides genetic insights from a well-phenotyped isolated population. Nature 2023;613(7944):508–518. doi:10.1038/s41586-022-05473-8. [Google Scholar] [PubMed] [CrossRef]

20. Su Q, Lu Y, He S et al. Assessing inflammatory protein factors in inflammatory bowel disease using multivariable mendelian randomization. Sci Rep 2025;15(1):210. doi:10.1038/s41598-024-84447-4. [Google Scholar] [PubMed] [CrossRef]

21. Bukavina L, Bensalah K, Bray F et al. Epidemiology of renal cell carcinoma: 2022 update. 2022 Eur Urol 2022;82(5):529–542. doi:10.1016/j.eururo.2022.08.019. [Google Scholar] [PubMed] [CrossRef]

22. Häggström C, Rapp K, Stocks T et al. Metabolic factors associated with risk of renal cell carcinoma. PLoS One 2013;8(2):e57475. doi:10.1371/journal.pone.0057475. [Google Scholar] [PubMed] [CrossRef]

23. Keizman D, Gottfried M, Ish-Shalom M et al. Active smoking may negatively affect response rate, progression-free survival, and overall survival of patients with metastatic renal cell carcinoma treated with sunitinib. Oncologist 2014;19(1):51–60. doi:10.1634/theoncologist.2012-0335. [Google Scholar] [PubMed] [CrossRef]

24. Capitanio U, Bensalah K, Bex A et al. Epidemiology of renal cell carcinoma. Eur Urol 2019;75(1):74–84. doi:10.1016/j.eururo.2018.08.036. [Google Scholar] [PubMed] [CrossRef]

25. Burgess S, Dudbridge F, Thompson SG. Combining information on multiple instrumental variables in mendelian randomization: comparison of allele score and summarized data methods. Stat Med 2016;35(11):1880–1906. doi:10.1002/sim.6835. [Google Scholar] [PubMed] [CrossRef]

26. Li W, Liu C, Chen S. Associations between genetically determined dietary factors and risk of autism spectrum disorder: a mendelian randomization study. Front Nutr 2024;11:1210855. doi:10.3389/fnut.2024.1210855. [Google Scholar] [PubMed] [CrossRef]

27. Ren Z, Simons P, Wesselius A, Stehouwer CDA, Brouwers M. Relationship between nafld and coronary artery disease: a mendelian randomization study. Hepatology 2023;77(1):230–238. doi:10.1002/hep.32534. [Google Scholar] [PubMed] [CrossRef]

28. Yang K, Zhang H, Wang Y et al. Evaluating the impact of smoking on the risk of dry eye disease: a two-sample mendelian randomization study. Sci Rep 2024;14(1):32043. doi:10.1038/s41598-024-83795-5. [Google Scholar] [PubMed] [CrossRef]

29. Li Y, Liu S, Dong Y, Yang J, Tian Y. Causal relationship between type I diabetes mellitus and atrial fibrillation: a mendelian randomization study. Int J Cardiol Heart Vasc 2025;57(3):101643. doi:10.1016/j.ijcha.2025.101643. [Google Scholar] [PubMed] [CrossRef]

30. Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through egger regression. Int J Epidemiol 2015;44(2):512–525. doi:10.1093/ije/dyv080. [Google Scholar] [PubMed] [CrossRef]

31. Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from mendelian randomization between complex traits and diseases. Nat Genet 2018;50(5):693–698. doi:10.1038/s41588-018-0099-7. [Google Scholar] [PubMed] [CrossRef]

32. Lu Y, Su Q, Xie Z, Liang J. Elucidating the causal impact of dietary factors on rheumatoid arthritis: insights from multivariable mendelian randomization analysis. Food Sci Nutr 2024;12(12):10903–10912. doi:10.1002/fsn3.4630. [Google Scholar] [PubMed] [CrossRef]

33. Xiao Q, Dong S, Tan Y et al. The causal impact of smoking behavior on osteoarthritis: a mendelian randomization analysis. Front Public Health 2025;13:1437443. doi:10.3389/fpubh.2025.1437443. [Google Scholar] [PubMed] [CrossRef]

34. Hemani G, Tilling K, Davey Smith G. Orienting the causal relationship between imprecisely measured traits using Gwas summary data. PLoS Genet 2017;13(11):e1007081. doi:10.1371/journal.pgen.1007081. [Google Scholar] [PubMed] [CrossRef]

35. Kim CS, Han KD, Choi HS, Bae EH, Ma SK, Kim SW. Association of hypertension and blood pressure with kidney cancer risk: a nationwide population-based cohort study. Hypertension 2020;75(6):1439–1446. doi:10.1161/hypertensionaha.120.14820. [Google Scholar] [PubMed] [CrossRef]

36. Jiang R, Li Z, Wang X et al. Association of metabolic syndrome and its components with the risk of kidney cancer: a cohort-based case-control study. Technol Health Care 2023;31(4):1235–1244. doi:10.3233/thc-220482. [Google Scholar] [PubMed] [CrossRef]

37. Ahmadinezhad M, Arshadi M, Hesari E, Sharafoddin M, Azizi H, Khodamoradi F. The relationship between metabolic syndrome and its components with bladder cancer: a systematic review and meta-analysis of cohort studies. Epidemiol Health 2022;44:e2022050. doi:10.4178/epih.e2022050. [Google Scholar] [PubMed] [CrossRef]

38. Raposeiras-Roubín S, Abu-Assi E, Muñoz-Pousa I et al. Usefulness of bleeding after acute coronary syndromes for unmasking silent cancer. Am J Cardiol 2020;125(12):1801–1808. doi:10.1016/j.amjcard.2020.03.023. [Google Scholar] [PubMed] [CrossRef]

39. Shu C, Han H, Li H et al. Cancer risk subsequent to cardio vascular disease: a prospective population-based study and meta-analysis. BMC Med 2025;23(1):229. doi:10.1186/s12916-025-04013-1. [Google Scholar] [PubMed] [CrossRef]

40. Nardi-Agmon I, Cohen G, Itzhaki Ben Zadok O, Steinberg DM, Kornowski R, Gerber Y. Cancer incidence and survival among patients following an acute coronary syndrome. Am J Cardiol 2023;202(suppl 1):50–57. doi:10.1016/j.amjcard.2023.06.021. [Google Scholar] [PubMed] [CrossRef]

41. Li N, Huang Z, Zhang Y, Sun H, Wang J, Zhao J. Increased cancer risk after myocardial infarction: fact or fiction? A systemic review and meta-analysis. Cancer Manag Res 2019;11:1959–1968. doi:10.2147/cmar.S193658. [Google Scholar] [PubMed] [CrossRef]

42. Forte V, Pandey A, Abdelmessih R et al. Obesity, diabetes, the cardiorenal syndrome, and risk for cancer. Cardiorenal Med 2012;2(2):143–162. doi:10.1159/000337314. [Google Scholar] [PubMed] [CrossRef]

43. Espíldora-Hernández J, Díaz-Antonio T, Baena-Espinar J et al. Subclinical arteriosclerosis is associated with common vascular risk factors in long-term survivors of testicular cancer. J Clin Med 2020;9(4):971. doi:10.3390/jcm9040971. [Google Scholar] [PubMed] [CrossRef]

44. Guerra B, Recio C, Aranda-Tavío H, Guerra-Rodríguez M, García-Castellano JM, Fernández-Pérez L. The mevalonate pathway, a metabolic target in cancer therapy. Front Oncol 2021;11:626971. doi:10.3389/fonc.2021.626971. [Google Scholar] [PubMed] [CrossRef]

45. Jaiswal V, Agrawal V, Ang SP et al. Post-diagnostic statin use and its association with cancer recurrence and mortality in breast cancer patients: a systematic review and meta-analysis. Eur Heart J Cardiovasc Pharmacother 2023;9(8):731–740. doi:10.1093/ehjcvp/pvad057. [Google Scholar] [PubMed] [CrossRef]

46. Cairat M, Pottegård A, Olesen M, Dossus L, Fournier A, Hicks B. Antiplatelet drugs and breast cancer risk in a large nationwide danish case-control study. Int J Cancer 2023;152(7):1337–1347. doi:10.1002/ijc.34343. [Google Scholar] [PubMed] [CrossRef]

47. Fang L, Moore XL, Dart AM, Wang LM. Systemic inflammatory response following acute myocardial infarction. J Geriatr Cardiol 2015;12(3):305–312. doi:10.11909/j.issn.1671-5411.2015.03.020. [Google Scholar] [PubMed] [CrossRef]

48. Ramos-Regalado L, Alcover S, Badimon L, Vilahur G. The influence of metabolic risk factors on the inflammatory response triggered by myocardial infarction: bridging pathophysiology to treatment. Cells 2024;13(13):1125. doi:10.3390/cells13131125. [Google Scholar] [PubMed] [CrossRef]

49. Afify SM, Hassan G, Seno A, Seno M. Cancer-inducing niche: the force of chronic inflammation. Br J Cancer 2022; 127(2):193–201. doi:10.1038/s41416-022-01775-w. [Google Scholar] [PubMed] [CrossRef]

50. Westman PC, Lipinski MJ, Luger D et al. Inflammation as a driver of adverse left ventricular remodeling after acute myocardial infarction. J Am Coll Cardiol 2016;67(17):2050–2060. doi:10.1016/j.jacc.2016.01.073. [Google Scholar] [PubMed] [CrossRef]

51. Gomella LG. Cardio-oncology, nephro-oncology and the urologic oncology collaboration. Can J Urol 2023;30(211457–11458. [Google Scholar] [PubMed]