Abstract
BACKGROUND: Patients with cardio-kidney-metabolic (CKM) disease are at high risk of long-term mortality. While several metabolic and inflammatory biomarkers have been investigated, their relative prognostic value in real-world CKM populations remains uncertain. We aimed to develop and internally validate a Cox-based prognostic model for long-term all-cause mortality, with a particular focus on serum albumin and inflammatory markers. METHODS: We conducted a prospective cohort study including 480 patients across the CKM spectrum. Participants were followed for up to 11 years, with all-cause mortality as the primary outcome. Cox proportional hazards models were constructed using prespecified clinical variables, including age, sex, estimated glomerular filtration rate (eGFR), serum albumin, high-sensitivity C-reactive protein (hs-CRP), body mass index, hypertension, coronary artery disease, and diabetes mellitus defined by laboratory criteria (HbA1c ≥ 6.5% and/or fasting plasma glucose ≥ 126 g/dL). Model performance was assessed using Harrell's C-index with internal validation by bootstrap resampling. Proportional hazards assumptions were evaluated using Schoenfeld residuals, with sensitivity analyses addressing time-varying effects. RESULTS: Among 480 participants, 114 deaths (23.8%) occurred during a median 10-year follow-up. In multivariable Cox models, age and coronary artery disease were consistently associated with higher all-cause mortality across all model specifications. Higher serum albumin levels were independently protective in models including laboratory-defined diabetes (HR 0.52 per 1 g/dL, 95% CI 0.30-0.93) and clinical history of diabetes (HR 0.49, 95% CI 0.28-0.86). Laboratory-defined diabetes mellitus was associated with an increased mortality risk (HR 2.24, 95% CI 1.36-3.69), whereas clinically documented diabetes mellitus showed an inverse association (HR 0.17, 95% CI 0.08-0.35). Measures of renal function, inflammatory burden assessed by log-transformed hs-CRP, body mass index, hypertension, and sex were not independently associated with mortality. CONCLUSIONS: In this real-world cohort spanning the cardiorenal-metabolic spectrum, serum albumin emerged as a robust and independent protective prognostic marker for long-term all-cause mortality, even after adjustment for age, comorbidities, renal function, and inflammatory burden. Age and coronary artery disease consistently predicted mortality across all model specifications. These findings highlight the value of serum albumin as an integrative biomarker reflecting nutritional status, systemic inflammation, and disease severity, with potential implications for refined risk stratification in patients with complex cardiorenal-metabolic conditions. Routine assessment of serum albumin may support pragmatic prognostic stratification in cardiorenal-metabolic populations.