AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
Predicted-Risk Sub-Classification of Cardiovascular-Kidney-Metabolic Syndrome Stage 2 in a Nationally Representative Cohort
By Prabhakar P
Background: The 2023 American Heart Association Cardiovascular-Kidney-Metabolic (CKM) syndrome framework places roughly half of US adults in Stage 2, a single category of widely varying prognosis. The American Heart Association Predicting Risk of Cardiovascular Disease Events (PREVENT) equations estimate 10-year risk and set the threshold defining Stage 3, yet this predicted risk is not applied within Stage 2, leaving its largest group undifferentiated by risk. Objective: To sub-classify CKM Stage 2 using the existing predicted-risk thresholds and characterize its prognosis. Methods: We studied adults aged 30 to 79 years in the National Health and Nutrition Examination Survey (2007-2018) with mortality linkage through 2019. CKM stage followed the 2023 framework; Stage 3 was assigned without the predicted-risk criterion, so all Stage 2 adults had predicted risk below 20% by construction. Base 10-year total-cardiovascular PREVENT risk defined sub-stages 2A (under 5%), 2B (5% to under 20%), and 2C (20% or higher, a Stage 3-equivalent risk). Survey-weighted analyses estimated all-cause mortality (Kaplan-Meier) and cardiovascular mortality (competing-risks cumulative incidence); crude Cox hazard ratios summarized prognostic separation, as predicted risk already integrates age, sex, and clinical factors. Reclassification versus single-category Stage 2 used integrated discrimination improvement (IDI) and inverse-probability-of-censoring-weighted net reclassification improvement (NRI). A stricter Stage 2 definition tested robustness. Results: Among 23,047 adults, 17,507 had CKM Stage 2; sub-stages 2A, 2B, and 2C comprised 64.6%, 31.6%, and 3.8%. Over a median of 6.8 years (860 deaths, 195 cardiovascular), 10-year all-cause mortality rose across 2A, 2B, and 2C (2.4%, 10.5%, 29.4%; hazard ratios 4.8 and 14.1 vs 2A), as did cardiovascular cumulative incidence (0.3%, 2.6%, 9.3%). Each 5-percentage-point higher predicted risk was associated with all-cause (hazard ratio 1.71; 95% CI, 1.63-1.79) and cardiovascular mortality (1.85; 1.71-2.01), with good within-stage discrimination (C-index 0.76 and 0.82). Sub-classification improved discrimination (IDI 0.071) and reclassification (continuous NRI +0.83; event +0.35, non-event +0.49). The small 2C subgroup had 10-year all-cause and cardiovascular mortality 12- and roughly 30-fold higher than 2A, approaching Stage 3 levels. Results were unchanged under stricter staging. Conclusions: Applying the framework's own predicted-risk thresholds sub-classifies the large, heterogeneous CKM Stage 2 population into prognostically distinct strata and identifies a subgroup whose 10-year mortality approaches Stage 3 yet receives no high-risk designation. Requiring no added computation, this refinement could sharpen risk assessment and prioritize prevention; prospective validation is warranted.