Cardiovascular-kidney-metabolic syndrome, abbreviated CKM, describes a pathophysiological state in which cardiovascular disease, chronic kidney disease, and metabolic disorders feed on one another through shared mechanisms such as insulin resistance, chronic inflammation, and hemodynamic overload. The AHA’s 2023 Presidential Advisory formalized a staging system running from stage 0, with no risk factors, to stage 4, marked by established clinical cardiovascular disease, subdivided into stage 4a without kidney failure and stage 4b with kidney failure. Stage 2 is the pivotal intermediate phase, characterized by metabolic risk factors such as diabetes, hypertension, hypertriglyceridemia, or metabolic syndrome, or by moderate-to-high-risk chronic kidney disease, in the absence of clinical cardiovascular disease. What happens between stage 2 and stage 4 is, in essence, the story of how metabolic dysfunction converts into organ damage, and clinicians currently lack simple tools to predict which patients will make that transition.

Plasma volume has long been suspected to play a role in this conversion, but measuring it has been the problem. The gold-standard techniques, hemodynamic monitoring and radioisotope tracer dilution, are invasive, technically demanding, and unsuitable for routine practice. Estimated plasma volume status, or ePVS, emerged over the past decade as a validated surrogate calculated directly from two values already present in any complete blood count: hematocrit and hemoglobin. The formula divides the fraction of whole blood occupied by plasma, expressed as 100 minus the hematocrit percentage, by the hemoglobin concentration in grams per deciliter, yielding a value in milliliters per gram of hemoglobin mass. In a patient with 40 percent hematocrit and 14 grams per deciliter of hemoglobin, the calculation gives roughly 4.29 milliliters per gram. The dimensional logic is elegant: because hematocrit expressed as a percentage represents milliliters of red blood cells per 100 milliliters of whole blood, the quotient reflects plasma volume relative to the body’s total hemoglobin content, a proxy that has proven accurate and reliable across cardiovascular and metabolic disease cohorts.

The research team, led by Longrong Huang, Xiaofei Luo, Lin Zhang, and Qiang Xu from Tianjin University of Traditional Chinese Medicine and affiliated hospitals, drew on the Tianjin Heart Failure with Integrated Treatment cohort, registered as ChiCTR2300077220. Participants were enrolled from 2006 to 2018 across 43 hospitals in Tianjin. After excluding patients with missing baseline data, those outside stage 2, individuals younger than 18, and anyone lacking ePVS measurements at admission or discharge, the final population comprised 36,505 hospitalized stage 2 patients, 45 percent of them men, with a median age of 61 years. Hypertension was the dominant condition, present in 70 percent, followed by chronic kidney disease in 33 percent, diabetes in 32 percent, and anemia in 25 percent. Over a median follow-up of 10 years, 15,495 patients developed stage 4 disease, defined as new-onset clinical cardiovascular disease including coronary artery disease, stroke, peripheral arterial disease, heart failure, deep vein thrombosis, or pulmonary embolism.

The statistical machinery behind the analysis was considerable. Five nested Cox proportional hazards models progressively adjusted for demographics, length of stay, medical history, laboratory parameters, and medications, with all variance inflation factors well below the multicollinearity threshold of 5, peaking at 1.75 for hypertension. Missing data, never exceeding 4.77 percent for any variable, were handled through multiple imputation using predictive mean matching and logistic regression across five datasets pooled by Rubin’s rules. Fine-Gray subdistribution hazard models treated all-cause mortality as a competing risk, confirming that the standard Cox results were not materially biased. Restricted cubic splines with four knots mapped the nonlinear dose-response relationship, while Shapley Additive exPlanations, or SHAP, analysis applied to random survival forests quantified the predictive contribution of each of 21 variables, from age and sex to potassium, creatinine, and both ePVS measurements.

The results revealed a striking asymmetry between the two time points. Patients who progressed to stage 4 had significantly higher ePVS at both admission and discharge than those who did not, with median admission values of 5.20 versus 5.09 milliliters per gram and discharge values of 5.83 versus 5.72. But in the fully adjusted quartile analysis, the risk signal for admission ePVS was confined to the upper strata: the second quartile showed no statistical evidence of elevated risk, with a hazard ratio of 1.04 and a confidence interval crossing unity, while the third and fourth quartiles carried 19 percent and 43 percent higher risks of stage 4 progression, respectively. For discharge ePVS, significance emerged earlier, starting from the second quartile, with a hazard ratio of 1.06. The fourth quartile of admission ePVS was associated with 88 percent higher risk of stage 4b, the variant with kidney failure, and 76 percent higher risk of heart failure, the two most volume-dependent endpoints. Per-unit increases in admission ePVS raised the risk of each endpoint by 5 to 10 percent, with heart failure showing the steepest per-unit hazard at 10 percent.

The restricted cubic spline analysis exposed a spoon-shaped nonlinear association between admission ePVS and stage 4 progression, with the risk curve remaining relatively flat at lower values and climbing steeply once ePVS exceeded a critical inflection point of approximately 5.14 milliliters per gram. Discharge ePVS inflection points clustered around 5.78, and while significant nonlinear associations appeared for stage 4 and coronary artery disease, the remaining endpoints showed approximately linear dose-response profiles. This divergence is likely attributable to in-hospital volume management: diuretics and SGLT-2 inhibitors such as dapagliflozin demonstrably modify plasma volume, meaning discharge ePVS reflects treatment-altered status rather than the patient’s intrinsic baseline pathophysiology. Admission ePVS, captured before any pharmacological intervention, better characterizes the underlying condition, which is why the authors argue it is more suitable for hierarchical risk stratification. SHAP analysis reinforced this hierarchy: admission ePVS outperformed discharge ePVS at every endpoint and even exceeded the predictive contribution of white blood cell count and fasting blood glucose, established markers of poor prognosis in CKM, and surpassed estimated glomerular filtration rate for stage 4, heart failure, and stroke.

Age emerged as a powerful modifier of the association. Significant age interactions appeared for stage 4, stage 4a, stage 4b, coronary artery disease, and heart failure, with patients older than 60 years showing consistently higher hazard ratios than younger patients, while no interaction was observed for stroke. Sex produced no significant interactions at any endpoint, though men showed higher quartile-level hazards for stage 4b, coronary artery disease, and heart failure. The authors propose a mechanistic explanation rooted in the physiology of aging: plasma volume naturally contracts in older patients as muscle mass declines, renal concentrating ability deteriorates, and cardiovascular function weakens. A higher plasma volume in this population may therefore impose a disproportionate hemodynamic burden on an already compromised heart and kidney, accelerating the progression to stage 4. This age-related risk gradient provides a biological rationale for stratified management of older high-risk populations, where ePVS may carry the greatest clinical value.

The pathophysiological logic connecting CKM stage 2 conditions to elevated plasma volume is coherent. In hypertension, increased sodium intake raises plasma osmotic pressure, and water is retained simultaneously to maintain isotonicity, so the more severe the hypertension, the greater the plasma volume. In diabetes, insulin resistance promotes sodium and water retention through effects on the kidney, directly expanding circulating volume. Elevated circulating glucose and triglycerides similarly increase plasma osmotic pressure and promote water retention, while circulating lipids damage the vascular endothelium, increase its permeability, and further expand plasma volume. In this framework, higher ePVS may simply indicate more severe underlying stage 2 disease, a marker of the metabolic and hemodynamic burden that ultimately drives organ damage. The effect sizes observed here are comparable to those in established single-organ disease cohorts: in patients with myocardial infarction complicated by heart failure, the highest ePVS quartile carried a 51 percent increased risk of one-year mortality, and a discharge ePVS above 5.5 milliliters per gram was associated with 88 percent higher all-cause mortality in cardiovascular patients. That similar magnitudes appear in a relatively early-stage CKM population reinforces the metric’s clinical relevance.

The authors are careful to frame the findings as hypothesis-generating rather than practice-changing. ePVS is a surrogate calculated from hemoglobin and hematocrit, carrying inherent measurement noise and imperfectly reflecting actual intravascular volume. The observational design cannot exclude reverse causality, meaning elevated plasma volume may be a consequence of progressive CKM pathophysiology rather than a causal driver, so the results demonstrate prognostic associations without implying that modifying ePVS would improve outcomes. Residual unmeasured confounding cannot be fully eliminated despite extensive adjustment, and the cohort was recruited from a single regional center in Tianjin without external validation, leaving generalizability across ethnic groups and healthcare systems uncertain. ePVS was measured at only two time points, introducing potential misclassification and regression dilution. Still, the study is the first to examine the link between estimated plasma volume status and CKM stage progression, and its scale, decade-long follow-up, and layered sensitivity analyses give the signal considerable weight. If external cohorts replicate the threshold effect around 5.14 milliliters per gram, clinicians may one day glance at a routine blood count and, with a single division, identify which stage 2 patients are quietly heading toward the syndrome’s most dangerous stage.

Subject of Research: Association between estimated plasma volume status and progression of cardiovascular-kidney-metabolic syndrome from stage 2 to stage 4

Article Title: Association of estimated plasma volume status with progression from stage 2 to stage 4 cardiovascular–kidney–metabolic syndrome: a cohort from the Tianjin region of China

Article References: Huang, L., Luo, X., Ou, X., Liu, Y., Zhang, L., & Xu, Q. (2026). Association of estimated plasma volume status with progression from stage 2 to stage 4 cardiovascular–kidney–metabolic syndrome: a cohort from the Tianjin region of China. Frontiers in Endocrinology, 17, Article 1959815. https://doi.org/10.3389/fendo.2026.1959815

Image Credits: AI Generated

DOI: 10.3389/fendo.2026.1959815

Keywords: cardiovascular-kidney-metabolic syndrome, estimated plasma volume status, CKM staging, risk stratification, cohort study, heart failure, chronic kidney disease, diabetes, noninvasive biomarker, Tianjin cohort, Association, estimated