Nutritional risk in hospitalized heart failure patients - a proteomic approach to pathway enrichment analysis

Abstract Background Malnutrition is a common yet overlooked factor in heart failure (HF), associated with poor outcomes. The Geriatric Nutritional Risk Index (GNRI) assesses nutritional status in hospitalized patients, but its biological pathways in HF remain unclear. Purpose To investigate the association between GNRI categories and circulating proteins in hospitalized HF patients using pathway enrichment analysis. Methods Three-hundred-twenty-four patients hospitalized for HF underwent clinical and biomarker assessments. GNRI was calculated based on admission albumin concentrations and the ratio of actual and ideal body weight, categorizing patients into no risk (≥ 98), low risk (92–97.9), moderate risk (82–91.9), and major risk (82) groups. Proteomic profiling of 92 proteins was carried out, and Kruskal-Wallis tests were used to assess differences in protein levels between GNRI categories, with false discovery rate (FDR) correction applied for multiple testing (Benjamini-Hochberg method). A heatmap was generated to visualize the mean expression levels of significant proteins across GNRI categories, allowing for the identification of patterns in protein abundance. Thereafter, GNRI categories were dichotomized, grouping "major risk" separately from all other categories, to facilitate enrichment analysis by specifically identifying biological pathways associated with the highest nutritional risk. Proteins with FDR-adjusted p-values 0.05 were mapped to Entrez Gene IDs, and pathway enrichment analysis was conducted using ReactomePA to identify overrepresented biological pathways. The enrichment results included generatio (proportion of significant genes in the pathway), background ratio (proportion in the total dataset), fold enrichment (observed vs. expected frequency), and adjusted p-values. Results The mean age in the cohort was 75.0 (±11.5) years, and 70.3% were men. A total of 26 proteins differed significantly between categories. Heatmap analysis showed distinct clustering of protein expression patterns across GNRI categories (Figure 1). Enrichment analysis revealed key pathways related to tumor-necrosis factor (TNF) signaling, neutrophil degranulation, integrin interactions, scavenger receptor activity, and extracellular matrix organization. The most enriched pathway, namely TNFs, bind their physiological receptors (adjusted p=0.0018), including TNFRSF1B, TNFRSF1A, and TNFRSF14, suggesting a potential role of inflammatory regulation in nutritional risk among HF patients. Additional pathways involved insulin-like growth factor transport, oxidative stress regulation, and interleukin signaling (Figure 2). Conclusion Inflammatory, metabolic, and extracellular matrix-related pathways are associated with nutritional risk in HF. TNF signaling suggests malnutrition may contribute to systemic inflammation, offering insights into mechanisms connecting nutritional risk to HF outcomes and potential targeted interventions.Figure 1. Figure 2.