Discriminative Performance of Anthropometric and Body Composition Measures for Identifying CONUT-Defined Nutritional Risk in Hospitalized Older Adults: A Cross-Sectional Study

Background/Objective: The Controlling Nutritional Status (CONUT) score is a laboratory-based tool for identifying nutritional risk. This study examined associations of anthropometric and bioimpedance-derived body composition measures with CONUT and their ability to discriminate CONUT-defined nutritional risk in hospitalized older adults. Methods: This cross-sectional study included 164 hospitalized adults aged >60 years. Anthropometric measurements, derived indices, and bioimpedance-derived body composition parameters were assessed, and CONUT was calculated from routine laboratory data. Results: CONUT-defined nutritional risk was present in 43.3% of participants. Participants with nutritional risk had lower body weight, waist, mid-upper arm and calf circumferences, fat mass, body mass index (BMI), and body adiposity index (BAI) (all p < 0.05). Correlations with the CONUT score were weak; the largest absolute correlations were observed for fat mass in kilograms and fat mass percentage (both ρ = −0.25). In the fully adjusted model, mid-upper arm circumference (OR = 0.900; 95% CI 0.825–0.982), fat mass in kilograms (OR = 0.955; 95% CI 0.921–0.991), and fat mass percentage (OR = 0.928; 95% CI 0.887–0.972) showed nominally significant associations with lower odds of CONUT-defined nutritional risk. After FDR adjustment, only fat mass percentage remained statistically significant (q = 0.024). Discriminatory performance was modest; the highest areas under the curve (AUCs) were 0.655 for fat mass percentage, 0.647 for fat mass in kilograms, and 0.638 for mid-upper arm circumference. Conclusions: Anthropometric and body composition measures showed limited ability to discriminate CONUT-defined nutritional risk and should not be considered substitutes for CONUT. Among the evaluated measures, fat mass percentage showed the most robust association with CONUT-defined nutritional risk. More complex derived indices did not provide additional discriminatory value.