Table 1_Diagnostic agreement of nutritional screening tools in obesity-related knee osteoarthritis.docx

CONUT showed the highest agreement with GLIM (κ = 0.58), MUST the lowest (κ = 0.29). Overall diagnostic discordance rate was 37.9%. MUST had the highest missed diagnosis rate for GLIM-defined malnutrition (42.5%), CONUT the lowest (13.0%).

Concealed malnutrition (defined as GLIM-defined malnutrition with BMI ≥ 28 kg/m² but without significant recent involuntary weight loss) was identified in 121 patients, accounting for 82.9% of GLIM-defined malnourished patients. Low muscle mass, high CRP, and higher WOMAC score were independently associated with concealed malnutrition. Adding CRP and muscle mass improved the AUC from 0.73 to 0.86 (NRI = 0.31, IDI = 0.09); sensitivity analyses using a modified GLIM reference that excluded these components confirmed that the improvement was not solely driven by criterion overlap.</p>Conclusion<p>In this cohort, concealed malnutrition was highly prevalent among patients with obesity-related KOA, accounting for 82.9% of GLIM-defined malnutrition cases.

Traditional BMI-dependent screening tools may systematically underestimate nutritional risk in this population. A strategy integrating laboratory indices, inflammatory burden, and muscle mass appears to provide better diagnostic agreement and risk stratification in this setting; however, these findings are derived from a single-center retrospective study and require prospective external validation.</p>