The endothelium is not merely a passive lining of blood vessels; it is a dynamic, metabolically active organ that governs vascular homeostasis. In diabetes mellitus, the endothelial monolayer undergoes profound functional and structural changes that precede and accelerate cardiovascular disease (CVD). Endothelial dysfunction — defined as a shift toward a pro-inflammatory, pro-coagulant, and vasoconstrictive state — is now recognized as both an early marker and a causal driver of atherosclerosis, microvascular complications, and adverse clinical outcomes. Identifying reliable biomarkers of this dysfunction allows clinicians to assess risk, track disease progression, and tailor therapeutic strategies. This review examines the most clinically relevant biomarkers of endothelial dysfunction in diabetes and CVD, their underlying mechanisms, and the emerging tools that promise to bring these measurements into routine care.

The Endothelial Barrier: Structure and Function

Endothelial cells line the entire vascular tree, forming a semi-permeable barrier that regulates the exchange of nutrients, gases, and immune cells between blood and tissues. Under healthy conditions, the endothelium maintains an anti-thrombotic and anti-adhesive surface. It produces nitric oxide (NO) through endothelial nitric oxide synthase (eNOS), a vasodilator that also inhibits platelet aggregation and leukocyte adhesion. Prostacyclin, tissue-type plasminogen activator, and heparan sulfate proteoglycans further contribute to a non-thrombotic environment. The endothelial glycocalyx — a gel-like layer of proteoglycans and glycoproteins on the luminal surface — serves as a mechanosensor and a barrier to protein leakage.

In diabetes, this elegant system unravels. Hyperglycemia, insulin resistance, dyslipidemia, and oxidative stress converge to damage the endothelium. The glycocalyx is shed, eNOS activity declines, and endothelial cells acquire a pro-adhesive, pro-coagulant phenotype. This transition is not merely a marker of existing disease but a direct contributor to atherogenesis, plaque progression, and microvascular complications such as nephropathy, retinopathy, and peripheral neuropathy. Understanding the molecular underpinnings of this shift is essential for interpreting the biomarkers discussed below.

Pathophysiologic Drivers of Endothelial Dysfunction in Diabetes and CVD

Endothelial dysfunction in diabetes arises from multiple interrelated pathways. Chronic hyperglycemia is the initiating insult. Excess glucose drives flux through the polyol pathway, consuming NADPH and depleting glutathione, thereby increasing intracellular oxidative stress. Mitochondrial superoxide overproduction activates protein kinase C (PKC), which in turn increases vascular permeability, promotes expression of adhesion molecules, and impairs eNOS activity. Advanced glycation end-products (AGEs) accumulate on long-lived proteins and activate their receptor (RAGE), triggering pro-inflammatory signaling that further compromises endothelial function.

Insulin resistance amplifies this damage. Under normal conditions, insulin stimulates eNOS via the PI3K-Akt pathway, promoting NO production. In the insulin-resistant state, this signaling branch is blunted, while the MAPK pathway remains active, leading to excessive secretion of the vasoconstrictor endothelin-1 (ET-1). The resulting imbalance — low NO and high ET-1 — favors vasoconstriction, smooth muscle proliferation, and thrombosis. Dyslipidemia compounds the problem: elevated oxidized low-density lipoprotein (oxLDL) binds to lectin-like oxLDL receptor-1 (LOX-1) on endothelial cells, triggering reactive oxygen species generation and upregulation of adhesion molecules. These processes create a self-reinforcing cycle of inflammation, oxidative stress, and endothelial injury that drives both macrovascular and microvascular complications.

Shear stress also plays a critical role. Disturbed flow at arterial bifurcations reduces eNOS expression and promotes a pro-atherogenic endothelial phenotype. In diabetes, the normal shear-responsive mechanisms are blunted, further predisposing to plaque formation. Understanding these pathways is essential because each offers a target for biomarker measurement and therapeutic intervention.

Key Biomarkers of Endothelial Dysfunction

Biomarkers of endothelial dysfunction fall into several categories: molecules secreted by endothelial cells, markers of endothelial damage or regeneration, and indicators of the local vascular microenvironment. The biomarkers discussed below represent the most extensively studied and clinically relevant in diabetes and CVD.

Asymmetric Dimethylarginine (ADMA)

ADMA is an endogenous inhibitor of eNOS that competes with L-arginine for binding to the enzyme's active site. Elevated ADMA levels reduce NO production, triggering vasoconstriction, platelet aggregation, and increased expression of adhesion molecules. In patients with type 2 diabetes, ADMA independently predicts cardiovascular events and progression of diabetic nephropathy. A large meta-analysis of over 2,000 individuals found that higher ADMA concentrations confer a 30-40% increased risk of fatal and non-fatal cardiovascular events, an effect that remained significant after adjustment for traditional risk factors.

ADMA is cleared primarily by the enzyme dimethylarginine dimethylaminohydrolase (DDAH), which is impaired under conditions of oxidative stress. This creates a feedback loop: oxidative stress reduces DDAH activity, raising ADMA levels, which in turn further impairs NO production and amplifies vascular injury. DDAH polymorphisms have been linked to cardiovascular risk in diabetic populations, making this pathway a potential therapeutic target. Clinically, ADMA is measured by high-performance liquid chromatography or ELISA, and reference ranges are well established, though the assay has not yet been widely adopted in routine care.

Endothelin-1 (ET-1)

ET-1 is a 21-amino-acid peptide that is the most potent endogenous vasoconstrictor known. It is produced by endothelial cells in response to hyperglycemia, insulin, angiotensin II, and shear stress. ET-1 acts on ETA and ETB receptors on vascular smooth muscle cells and endothelial cells, mediating vasoconstriction, proliferation, and fibrosis. Plasma ET-1 levels are elevated in patients with type 1 and type 2 diabetes, and correlate with hypertension, coronary artery disease, and diabetic nephropathy.

ET-1 plays a particularly important role in microvascular complications. In the retina, ET-1 contributes to pericyte loss and capillary occlusion; in the kidney, it promotes mesangial expansion and albuminuria. Urinary ET-1 levels have been proposed as a marker of early diabetic nephropathy, potentially preceding albuminuria. Selective ET receptor antagonists have shown renal protective effects in experimental models, but clinical trials in cardiovascular disease have been limited by fluid retention and other side effects. Nonetheless, ET-1 remains a valuable biomarker for assessing the vasoconstrictor component of endothelial dysfunction.

Vascular Endothelial Growth Factor (VEGF)

VEGF is a key regulator of angiogenesis. In diabetes, VEGF signaling is dysregulated: excessive VEGF drives pathological neovascularization in the retina and kidney, while insufficient VEGF impairs wound healing and collateral vessel formation. Serum VEGF levels are elevated in patients with proliferative diabetic retinopathy and are associated with albuminuria. VEGF also acts as a pro-inflammatory mediator, increasing vascular permeability and upregulating adhesion molecules.

The clinical impact of VEGF is most evident in the success of anti-VEGF therapies for diabetic macular edema and retinal neovascularization. Intravitreal injections of ranibizumab, aflibercept, and bevacizumab have become standard of care, dramatically reducing vision loss in patients with diabetic eye disease. Systemic VEGF inhibition, however, has not shown the same promise in cardiovascular disease and may even increase thrombotic risk in some populations. As a biomarker, VEGF is routinely measured in ophthalmology to guide treatment decisions, and systemic levels are increasingly studied as predictors of microvascular progression.

Soluble Adhesion Molecules (sICAM-1, sVCAM-1, sE-Selectin)

Endothelial activation upregulates cell surface adhesion molecules that mediate leukocyte rolling, adhesion, and transmigration. These molecules are cleaved and released into the circulation as soluble forms. Soluble intercellular adhesion molecule-1 (sICAM-1) and soluble vascular cell adhesion molecule-1 (sVCAM-1) are elevated in diabetes and predict incident cardiovascular events independently of traditional risk factors. In the Framingham Heart Study, individuals with the highest sICAM-1 levels had a 50% increased risk of coronary heart disease over 10 years compared with those with the lowest levels.

Soluble E-selectin is more specific to endothelial cells and reflects early activation. E-selectin mediates the initial rolling of leukocytes along the vessel wall, and its soluble form appears in the circulation within hours of endothelial activation. Elevated sE-selectin levels have been associated with the development of type 2 diabetes, suggesting that endothelial activation may precede the clinical diagnosis of diabetes by years. Measurement of these adhesion molecules by ELISA is straightforward and reproducible, making them attractive candidates for multi-biomarker panels. However, reference ranges vary by age, sex, and ethnicity, and standardization across laboratories is still evolving.

Oxidized Low-Density Lipoprotein (oxLDL)

oxLDL is a modified lipoprotein that plays a central role in the initiation and progression of atherosclerosis. Once in the subendothelial space, oxLDL is taken up by macrophages via scavenger receptors, leading to foam cell formation and fatty streak development. oxLDL also directly activates endothelial cells by binding to LOX-1, inducing reactive oxygen species generation and upregulation of adhesion molecules. Circulating oxLDL levels are elevated in patients with diabetes and correlate with endothelial dysfunction measured by flow-mediated dilation (FMD). A prospective study of over 1,400 individuals found that baseline oxLDL levels independently predicted the risk of myocardial infarction and stroke over 5 years.

The pro-atherogenic effects of oxLDL are amplified in diabetes because hyperglycemia and insulin resistance increase LDL oxidation and reduce its clearance. Antibodies against oxLDL are also elevated in diabetes and have been associated with cardiovascular events. oxLDL is available as a clinical assay through major commercial laboratories, but its use is limited by variability in assay methodologies and a lack of consensus on reference ranges. Immunoassays that measure circulating oxLDL bound to autoantibodies may offer greater specificity.

Circulating Endothelial Cells (CECs) and Endothelial Microparticles (EMPs)

CECs are mature endothelial cells that have detached from the vessel wall, typically in response to severe injury. Elevated CEC counts are found in acute coronary syndromes, vasculitis, and diabetes with microvascular complications. CEC enumeration provides a direct measure of endothelial damage but is technically challenging due to low cell numbers in peripheral blood. Flow cytometry with specific endothelial markers is the preferred method.

EMPs are small membrane vesicles (0.1–1 µm) shed from activated or apoptotic endothelial cells. They carry surface markers such as CD31, CD144, CD146, and CD62E, and their numbers increase in conditions of endothelial stress. EMPs are not merely debris; they can transfer bioactive molecules — including microRNAs, lipids, and cytokines — to target cells, propagating pro-coagulant and pro-inflammatory signals. In diabetes, elevated levels of CD31+/CD42b- EMPs have been associated with coronary artery disease and diabetic nephropathy. Flow cytometry is the primary detection method, but standardization of isolation, storage, and gating strategies is still needed before these markers can enter routine clinical use.

Nitric Oxide Metabolites (NOx) and eNOS Activity

Direct measurement of NO is difficult due to its short half-life (seconds). Instead, stable metabolites — nitrite and nitrate, collectively termed NOx — are measured as an index of NO production. Reduced NOx levels have been reported in patients with diabetes and coronary artery disease, reflecting impaired eNOS activity. However, NOx concentrations are influenced by dietary nitrate intake, renal function, and medications such as statins and ACE inhibitors, which can confound interpretation.

More sensitive assays now allow measurement of NO production ex vivo in cultured endothelial cells, but these are research tools. The most widely used functional measure of NO bioavailability is flow-mediated dilation (FMD) of the brachial artery. FMD correlates strongly with coronary endothelial function and predicts cardiovascular events. Combined measurement of FMD and circulating NOx provides a more complete picture of NO pathway integrity than either alone. In clinical trials targeting endothelial function, a composite endpoint that includes FMD, NOx, and a marker of oxidative stress is often employed.

Predictive Value and Clinical Utility in Diabetes and CVD

Biomarkers of endothelial dysfunction serve three main purposes: risk stratification before disease onset, monitoring of disease progression, and surrogate endpoints in clinical trials. Each biomarker offers a distinct window into the endothelial state, and their predictive value often exceeds that of traditional risk factors alone.

In type 2 diabetes, elevated ADMA, ET-1, and sICAM-1 independently predict cardiovascular mortality and progression to end-stage renal disease. The combination of high oxLDL and low NOx provides a particularly robust risk profile for coronary events. A recent prospective study of over 800 patients with type 2 diabetes found that those with both elevated sVCAM-1 and high-sensitivity C-reactive protein had a 3.5-fold increased risk of major adverse cardiovascular events over 5 years, even after adjustment for age, sex, and HbA1c.

In type 1 diabetes, elevated VEGF and urinary ADMA predict the development of proliferative retinopathy within 5 years, offering a window for preventive treatment. CEC and EMP counts are elevated in patients with early diabetic nephropathy and may detect renal injury before albuminuria develops. These markers could help identify which patients with microalbuminuria are at highest risk of progression to macroalbuminuria and kidney failure.

Clinically, several biomarkers have entered practice. ADMA is measured in specialized lipid clinics for patients with suspected endothelial dysfunction despite normal LDL levels. VEGF is routinely assessed in ophthalmology to guide anti-VEGF therapy. oxLDL is available through major lab networks, though its use is not yet guideline-recommended. The greatest barrier to wider adoption is the lack of standardized reference ranges. Large-scale, multi-ethnic cohort studies are needed to establish age-, sex-, and ethnicity-specific cutoffs for each biomarker. Multi-biomarker risk scores that incorporate traditional risk factors along with biomarkers such as ADMA, sICAM-1, and FMD could significantly improve risk discrimination in primary prevention.

Assessment of Endothelial Function: Methods and Practical Considerations

Beyond circulating biomarkers, endothelial function can be assessed directly with non-invasive vascular tests. Flow-mediated dilation (FMD) of the brachial artery using high-resolution ultrasound is the gold standard. FMD reflects NO-dependent vasodilation in response to reactive hyperemia. In diabetes, FMD is impaired and correlates with coronary endothelial dysfunction. A reduction of 1% in FMD has been associated with a 10-15% increase in the risk of cardiovascular events. FMD is operator-dependent and requires rigorous training and standardization.

Peripheral arterial tonometry (PAT) is an alternative method that measures digital pulse amplitude. PAT is easier to perform and less operator-dependent, but it reflects a combination of NO-dependent and independent vasodilation and has a weaker correlation with coronary function. Venous occlusion plethysmography is another research technique that measures forearm blood flow responses to intra-arterial infusions of vasoactive agents, but it is invasive and impractical for routine care.

Each method has trade-offs. FMD provides the most specific measure of NO bioavailability but requires specialized equipment and training. PAT is more accessible but less specific. Combining a functional test (such as FMD) with a biomarker panel (such as ADMA and sICAM-1) offers the most comprehensive assessment. For clinical trial purposes, a composite score that incorporates FMD, ADMA, and a measure of oxidative stress (such as urinary 8-isoprostane) is increasingly common. The development of point-of-care devices — such as microfluidic chips that can measure ADMA or EMPs from a fingerstick — could make endothelial function assessment a routine part of diabetes management.

Therapeutic Approaches for Restoring Endothelial Function

Restoring endothelial function is an important therapeutic target in diabetes and CVD. Lifestyle interventions — including aerobic exercise, a Mediterranean-style diet rich in antioxidants and omega-3 fatty acids, and smoking cessation — consistently improve FMD and reduce circulating adhesion molecules. A meta-analysis of randomized trials found that supervised exercise training increased FMD by an average of 2.5% in patients with type 2 diabetes, a magnitude comparable to that of many pharmacologic interventions.

Among pharmacologic agents, statins reduce oxLDL levels and improve NO bioavailability. ACE inhibitors lower ET-1 and reduce NADPH oxidase activity, thereby decreasing oxidative stress. Metformin has been shown to decrease ADMA levels in type 2 diabetes, likely through its effects on insulin sensitivity and DDAH activity. More recent drug classes have demonstrated direct endothelial benefits. SGLT2 inhibitors such as empagliflozin reduce EMPs and improve FMD in patients with type 2 diabetes, independent of their glucose-lowering effects. GLP-1 receptor agonists such as liraglutide and semaglutide decrease sICAM-1 and sVCAM-1 levels and improve endothelial function in both diabetic and non-diabetic populations.

Supplemental L-arginine has been investigated as a way to boost NO production, but results have been inconsistent. The likely reason is that ADMA competes with L-arginine for eNOS binding, so L-arginine supplementation is only effective when ADMA levels are low. Strategies that enhance DDAH activity — such as the investigational drug DDAH-1 activator — have shown promise in preclinical models and may offer a more targeted approach. Antioxidant supplementation with vitamins C and E has failed in large trials, likely due to pro-oxidant effects at high doses and poor cellular delivery. Nanoparticle-based delivery of antioxidant enzymes such as superoxide dismutase and catalase is currently under investigation. Gene therapy to upregulate eNOS expression has shown success in animal models but faces significant hurdles in clinical translation.

Emerging Directions and Open Questions

The discovery of novel endothelial biomarkers continues to accelerate. Proteomic and metabolomic analyses have identified several candidates including endoglin, thrombomodulin, angiopoietin-2, and circulating syndecan-1, a marker of glycocalyx degradation. MicroRNAs — small non-coding RNAs packaged into EMPs or circulating in protein-bound complexes — offer a dynamic readout of endothelial gene expression. miR-126, miR-21, and miR-146a are dysregulated in diabetes and have been linked to vascular complications. miR-126, in particular, is enriched in endothelial cells and promotes eNOS expression; reduced circulating levels predict cardiovascular events in diabetic patients.

The glycocalyx itself is emerging as an important biomarker. Shedding of glycocalyx components such as syndecan-1 and hyaluronan can be detected in the circulation and may reflect early endothelial injury before traditional markers become abnormal. Measuring glycocalyx thickness directly using microvascular imaging techniques is feasible but not yet widely available. Combining glycocalyx markers with classic biomarkers could enhance early detection of endothelial dysfunction.

Artificial intelligence and machine learning are beginning to integrate multi-omics data — including biomarkers, clinical variables, and imaging — into predictive algorithms for cardiovascular risk in diabetic populations. A recent algorithm that combined ADMA, sICAM-1, and retinal photography outperformed the Framingham Risk Score in identifying which patients with type 2 diabetes would develop nephropathy within 5 years. As these tools mature, personalized endothelial risk profiling will become clinically feasible. However, the black-box nature of some AI models raises questions about interpretability and generalizability, and rigorous external validation is needed before they can guide clinical decisions.

Standardization of assays and reference ranges remains the most pressing challenge. The European Society of Cardiology and the American Diabetes Association have both called for harmonized protocols for biomarker measurement and for large-scale cohort studies to define clinical decision thresholds. Until these efforts succeed, endothelial biomarkers will remain primarily research tools. With coordinated efforts across endocrinology, cardiology, and vascular biology, these markers have the potential to transition from the laboratory to the bedside, enabling earlier, more precise intervention for the millions of patients affected by diabetes and CVD.

Conclusion

Endothelial dysfunction is a central pathophysiologic feature of diabetes and cardiovascular disease, and its biomarkers provide a window into vascular health that extends beyond traditional risk factors. ADMA, ET-1, VEGF, adhesion molecules, oxLDL, CECs, EMPs, and NOx each capture distinct aspects of endothelial injury, activation, and repair. Their measurement — particularly when combined with functional assessment — improves risk stratification, guides therapy selection, and serves as a surrogate endpoint in clinical trials. As our understanding of the molecular pathways deepens and measurement technologies advance, integrating these biomarkers into routine diabetes care promises to reduce the enormous burden of cardiovascular morbidity and mortality that remains the leading cause of death in this population.

For further reading: a comprehensive meta-analysis on ADMA and cardiovascular risk in diabetes (PubMed); the American Heart Association scientific statement on endothelial function and biomarkers (AHA); clinical guidelines for endothelial dysfunction assessment from the European Society of Cardiology (ESC); a review of SGLT2 inhibitors and endothelial function (PubMed); and an overview of microRNAs as endothelial biomarkers (PMC).