PROTECT YOUR DNA WITH QUANTUM TECHNOLOGY
Orgo-Life the new way to the future Advertising by AdpathwayA systematic review and meta-analysis examining more than 1.8 million individuals across nine countries has found that while individual observational investigations consistently link declining kidney function with a higher occurrence of venous blood clots, pooled statistical estimates remain statistically non-significant and highly uncertain. Published in Frontiers in Molecular Medicine, the comprehensive investigation evaluated fourteen observational studies to clarify the relationship between chronic kidney disease and both arterial and venous thrombotic complications. The research team, led by investigators from Harvard Medical School and collaborating institutions, sought to address long-standing ambiguities surrounding the precise magnitude of clotting risks in renal patients. Chronic kidney disease is widely recognized in biomedical literature as a prothrombotic state driven by endothelial dysfunction, chronic inflammation, platelet abnormalities, and dysregulation of coagulation cascades. However, the newly synthesized data reveal that current epidemiologic evidence is severely constrained by methodological discrepancies, substantial between-study heterogeneity, and an absence of comparative incident data for arterial events, leaving the exact quantification of thrombotic risk unresolved.
To investigate the epidemiological connection, the researchers conducted a prospectively registered systematic review following PRISMA guidelines across four major biomedical databases, searching records indexed up to August 2026. From an initial pool of 9,110 identified publications, systematic multi-stage screening yielded fourteen eligible observational studies spanning cohorts from North America, Europe, and Asia. These fourteen studies encompassed a total sample of 1,836,579 individuals, with females comprising 51.6 percent of the aggregate population and cohort mean or median ages ranging from 49.1 to 75 years. Ten of the included investigations were longitudinal cohort studies, two were case-control studies, and two were cross-sectional evaluations. While all fourteen studies informed the qualitative narrative synthesis, seven provided sufficient raw event data to be included in quantitative random-effects meta-analyses of venous outcomes. The investigators restricted quantitative pooling strictly to venous endpoints—specifically composite venous thromboembolism, deep vein thrombosis, and pulmonary embolism—because reporting across published arterial studies was entirely inadequate to permit valid mathematical combination.
In the narrative synthesis of individual studies that featured comparative control groups, a consistent biological gradient emerged showing that lower kidney function corresponded with an elevated risk of venous thrombosis. Across mild-to-moderate chronic kidney disease, multiple cohorts documented moderate but statistically robust increases in adjusted hazard ratios, ranging generally from 1.54 to 1.71 when compared against individuals with normal renal parameters. For instance, an analysis of five pooled community cohorts by Mahmoodi and colleagues demonstrated an adjusted hazard ratio of 1.54 for incident venous thromboembolism, while a Canadian population-based cohort reported an adjusted hazard ratio of 1.61 among participants exhibiting severe albuminuria with preserved glomerular filtration. In much more advanced disease stages, the observed relative hazards climbed dramatically. A nationwide cohort study from Taiwan conducted by Lu and Liao identified an adjusted hazard ratio of 13.92 for deep vein thrombosis among patients with dialysis-dependent end-stage kidney disease compared to matched non-dialysis controls, highlighting a pronounced vulnerability among individuals undergoing chronic renal replacement therapy.
Despite the consistent directional pattern observed across individual adjusted cohorts, the quantitative meta-analysis of pooled raw event counts failed to establish statistically significant summary risk ratios. Pooling six cohort studies encompassing 954,129 participants for composite venous thromboembolism yielded an unadjusted risk ratio of 2.06, but with an extraordinarily broad 95 percent confidence interval spanning from 0.41 to 10.23 and extreme statistical heterogeneity reflected by an I-squared value of 99.2 percent. The pooled analysis for deep vein thrombosis, derived from three cohort studies involving 17,562 individuals, produced a point estimate of 9.52 with an even wider 95 percent confidence interval extending from 0.43 to 210.24, alongside an I-squared heterogeneity metric of 95.0 percent. Similarly, pooling two studies with 8,904 participants for pulmonary embolism yielded a summary risk ratio of 3.48 with a 95 percent confidence interval ranging from 0.14 to 88.84. Because the calculated confidence intervals for all three venous conditions crossed the null value of 1.0, the pooled summary figures could not definitively confirm an elevated risk.
The study authors emphasized that the statistical non-significance and extreme instability of the pooled estimates do not demonstrate an absence of biological association, but rather reflect severe methodological friction arising from unadjusted data aggregation. When synthesizing crude event counts, substantial confounding factors such as advanced patient age, diabetes, cardiovascular comorbidities, and varying exposure categorizations heavily distorted raw calculations. In two large community investigations, the crude risk ratios calculated directly from raw event totals were paradoxically below 1.0—yielding 0.33 in an Atherosclerosis Risk in Communities analysis and 0.28 in an analysis combining two community cohorts—even though both primary studies reported statistically significant positive associations after multivariable adjustment, with adjusted hazard and relative risk ratios of 1.60 and 1.71. Conversely, other cohorts displayed unadjusted risk ratios exceeding 6.0 that sharply contracted once adjusted for confounding baselines, demonstrating that unadjusted meta-analytic pooling introduces severe distortions that obscure true clinical relationships.
To evaluate the stability of their numerical conclusions, the investigators performed a series of rigorous sensitivity analyses, including Baujat diagnostic plots, outlier verifications, and leave-one-out influence models. One primary community cohort was confirmed as a statistical outlier within the composite venous thromboembolism synthesis; however, excluding this outlier did not resolve the high heterogeneity, yielding a recalibrated risk ratio of 3.08 with an unresolved confidence interval of 0.61 to 15.55 and an I-squared value of 98.3 percent. Sequential removal of individual investigations confirmed that no single publication drove the directional or statistical trajectory of the findings, with leave-one-out risk ratios fluctuating between 1.60 and 3.08 for composite venous events and between 5.38 and 17.96 for deep vein thrombosis. Furthermore, removing a study centered exclusively on recurrent rather than first-time venous events produced negligible impact on between-study variation, confirming that differing baseline definitions and underlying population profiles constituted the deeper source of variance.
Applying the standardized Grading of Recommendations Assessment, Development and Evaluation framework, the authors classified the certainty of evidence for all evaluated clinical outcomes as uniformly low. Because every included paper relied on observational methodologies, baseline certainty commenced at a lower threshold and was systematically downgraded due to very serious inconsistency and profound statistical imprecision. For composite venous thromboembolism and deep vein thrombosis, severe inconsistency was underscored by I-squared values exceeding 95 percent, while confidence intervals encompassing both protective effects and multi-fold risk increases warranted severe downgrades for imprecision. For pulmonary embolism, although statistical heterogeneity registered at zero percent across the two eligible studies, certainty was constrained to low because the small sample forced the Hartung-Knapp adjustment to utilize a t-distribution with only one degree of freedom. Funnel plot evaluations exhibited visible asymmetry for both composite thromboembolism and deep vein thrombosis, indicating that publication bias and small-study distortion could not be ruled out.
The investigation also highlighted an acute evidence gap concerning arterial thrombotic events, which encompass debilitating conditions such as myocardial infarction, ischemic stroke, and peripheral arterial occlusion. Although chronic kidney disease has long been mechanistically implicated in accelerated arterial atherothrombosis, only three retrieved studies addressed arterial conditions, and none offered a direct epidemiological comparison of incident arterial events between individuals with and without kidney impairment. One evaluated cohort focused exclusively on the prognosis of incident atrial fibrillation within a kidney disease population, reporting that new-onset atrial fibrillation doubled subsequent stroke risk, but lacked a non-renal comparator group. Another study documented high baseline frequencies of coronary and peripheral arterial disease among hemodialysis patients without reference controls, while a third evaluated prior arterial thrombosis merely as an extrinsic risk factor for venous complications. Consequently, quantitative meta-analysis of arterial thrombosis was impossible, leaving the attributable arterial risk entirely unquantified.
These nuanced findings carry direct implications for nephrology and hematology practice, cautioning clinicians against using current pooled risk estimates to formulate generalized thromboprophylaxis policies. While the biological plausibility of hypercoagulability in kidney disease remains robust—supported by systemic platelet hyperactivity, impaired fibrinolytic capacity, and uremic endothelial inflammation—the absence of precise summary metrics means universal anticoagulant thresholds cannot be justified by available meta-analytic data. The clinical dilemma is further intensified because advanced kidney disease simultaneously confers elevated bleeding liabilities alongside clotting predispositions, requiring meticulous individualization of anticoagulant therapies that account for altered renal drug clearance. The authors underscored that high-risk presentations, such as dialysis-dependent end-stage disease where catheter usage and vascular calcification multiply clotting frequency, demand vigilant surveillance, but therapeutic decisions must remain tailored rather than standardized until clearer quantitative evidence becomes available.
Ultimately, the researchers concluded that bridging the gap between observed individual cohort associations and imprecise pooled summaries requires a major overhaul in epidemiological research standards for cardio-renal medicine. Future investigations must move away from disparate exposure categories and instead adopt uniform diagnostic benchmarks aligned with standardized international guidelines based on both estimated glomerular filtration rates and urinary albumin-to-creatinine ratios. The authors called for prospective, well-powered longitudinal studies that specifically incorporate appropriate non-renal comparator cohorts, track objectively confirmed arterial and venous clinical endpoints using time-to-event survival frameworks, and report fully adjusted effect measures alongside granular covariate breakdowns. Furthermore, conducting collaborative individual-participant data meta-analyses represents the most viable pathway to eliminate the confounding distortions that currently plague aggregate-level syntheses, providing the clinical community with the reliable, high-certainty evidence required to guide antithrombotic management in chronic kidney disease.
Subject of Research: Systematic review and meta-analysis of arterial and venous thrombotic event risks in chronic kidney disease
Article Title: The kidney–thrombosis connection in chronic kidney disease through a meta‐analysis of arterial and venous thrombotic events
Article References: Sharma, A., Vamaraju, D. S., Sharma, B., Arora, A., Yoo, J. Y., Sharma, K., & Ajay, A. K. (2026). The kidney–thrombosis connection in chronic kidney disease through a meta‐analysis of arterial and venous thrombotic events. Frontiers in Molecular Medicine, 6, 1978556. https://doi.org/10.3389/fmmed.2026.1978556
Image Credits: AI Generated
DOI: 10.3389/fmmed.2026.1978556
Keywords: chronic kidney disease, venous thromboembolism, deep vein thrombosis, pulmonary embolism, arterial thrombosis, meta-analysis, systematic review, end-stage kidney disease, hemodialysis, glomerular filtration rate, albuminuria, GRADE assessment, Hartung-Knapp adjustment, observational studies, thromboprophylaxis, relative risk, heterogeneity


2 days ago
15




















English (US) ·
French (CA) ·