Dual-track nutritional assessment improves prognostic stratification in elderly patients with HFpEF: a 24-month longitudinal study
K Shi, H L Che, M Q Ye, X K QuAbstract
Background
Heart failure with preserved ejection fraction (HFpEF) is increasingly recognized as a multisystem "geriatric syndrome" characterized by multisystem impairment rather than isolated cardiac dysfunction. In older adults, nutritional status serves as a critical marker of biological aging, yet the "obesity paradox" often masks underlying protein-energy wasting. This study evaluated the synergistic prognostic value of a dual-track assessment combining clinical screening via Nutritional Risk Screening 2002 (NRS 2002) with biochemical indexation via the Prognostic Nutritional Index (PNI) to refine risk stratification.
Methods
This retrospective cohort study included 414 elderly patients (age≥60 years) diagnosed with HFpEF at a comprehensive tertiary hospital between July. 2018 and Dec. 2023. Nutritional risk was established using NRS 2002 (threshold = 3) and biochemical reserve was assessed via PNI (threshold = 41). All-cause mortality was systematically monitored at 3, 12, and 24 months post-discharge. Predictive performance was quantified through Harrell’s C-index, continuous Net Reclassification Improvement (NRI), and Integrated Discrimination Improvement (IDI).
Results
Baseline BMI did not differ between nutritional risk groups (23.5 vs. 23.4 kg/m², p = 0.571), highlighting the insufficiency of weight-based measures. Kaplan-Meier analysis revealed distinct survival trajectories: NRS 2002 ≥ 3 and PNI < 41 were both associated with consistently lower survival, with PNI showing a more rapid early decline (Figure 1A-B). A significant stepwise increase in mortality was observed as risk dimensions accumulated; dual-risk patients (NRS ≥ 3 and PNI < 41) reached a 24-month mortality of 48.5%, compared to 17.0% in low-risk patients (p < 0.001; Figure 1C). In multivariable Cox models, both NRS 2002 ≥ 3 (HR: 1.55, 95% CI: 1.07–2.23, p = 0.019) and PNI < 41 (HR: 1.79, 95% CI: 1.23–2.59, p = 0.002) remained independent predictors of 24-month mortality (Figure 2) . While PNI showed a stronger association with early mortality at 3 months (HR: 2.01, p = 0.047), the prognostic impact of NRS 2002 became most consistent over the long term. The integrated dual-track model significantly improved discrimination (C-index: 0.705 vs. 0.677, p < 0.001) and reclassification accuracy (Continuous NRI: 0.423, p < 0.001) over the clinical base model.
Conclusions
Nutritional vulnerability, assessed through integrated clinical and biochemical dimensions, is an independent driver of mortality in elderly HFpEF. This dual-track approach identifies a "masked" nutritional risk independent of BMI and provides superior prognostic insight, offering a robust framework for personalized management in geriatric cardiovascular care.Kaplan–Meier survival curvesForest Plot