Inflammatory–lipid residual risk phenotypes delineate the HFpEF probability spectrum independent of cardiometabolic mediation

BACKGROUND: Heart failure with preserved ejection fraction (HFpEF) is increasingly recognized as a systemic syndrome characterized by chronic low-grade inflammation, metabolic dysregulation, and microvascular dysfunction. Although residual inflammatory risk and residual cholesterol risk have each been associated with adverse cardiovascular outcomes, their joint relationship with HFpEF probability spectrum, and the extent to which this relationship is mediated by conventional cardiometabolic pathways, remain unclear. METHODS: In this cross-sectional study, participants were classified into four inflammatory-lipid residual risk phenotypes according to low-density lipoprotein cholesterol (LDL-C) and high-sensitivity C-reactive protein (hs-CRP): no residual risk, residual inflammatory risk (RIR), residual cholesterol risk (RCR), and combined residual inflammatory and cholesterol risk (RCIR). HFpEF probability was assessed using the Heart Failure Association Pre-test assessment, Echocardiography and natriuretic peptide, Functional testing, Final etiological work-up (HFA-PEFF) score and categorized as low (≤ 1), intermediate (2-4), or high (≥ 5). Associations between residual risk phenotypes and the HFpEF probability spectrum were examined using ordinal logistic regression and marginal predicted probabilities analyses. Exploratory mediation analyses were performed to evaluate the contribution of conventional cardiometabolic components. RESULTS: Residual risk phenotypes were distributed heterogeneously across HFA-PEFF strata. The proportion of RCIR increased progressively with higher HFpEF probability, accounting for 26.4% in the low stratum, 34.9% in the intermediate stratum, and 43.5% in the high stratum, whereas the proportion of RCR decreased from 68.4% to 60.8% and 46.7%, respectively (P < 0.001). In the fully adjusted ordinal logistic model, RCIR (OR 0.59, 95% CI 0.41-0.84), RIR (OR 0.53, 95% CI 0.32-0.90), and particularly RCR (OR 0.38, 95% CI 0.27-0.54) were associated with lower odds of remaining in lower HFA-PEFF categories, consistent with a shift toward higher HFpEF probability strata. Marginal predicted probability analyses across Models 1-3 showed a consistent redistribution of HFA-PEFF probability across phenotypes, and the overall ordering pattern remained stable after full adjustment. Exploratory mediation analyses showed no meaningful mediation by BMI or systolic blood pressure, while glucose demonstrated only a modest borderline signal. Overall, the association between RCIR and higher HFpEF probability appeared not to be materially mediated by the selected conventional cardiometabolic variables. CONCLUSIONS: Inflammatory-lipid residual risk phenotypes were differentially associated with the HFpEF probability spectrum. RCIR showed progressive enrichment with increasing HFpEF probability, whereas RCR exhibited an inverse distributional pattern but remained strongly associated with higher-probability strata in multivariable models. These associations were largely independent of conventional cardiometabolic mediators, supporting a potential role for inflammatory-lipid residual risk phenotyping in refining HFpEF risk stratification.