Relative validity of surrogate measures within the global leadership initiative on malnutrition (GLIM) phenotypic criteria for diagnosing adult malnutrition in resource-constrained settings

<h2>Abstract</h2><h3>Background and aim</h3> Implementing the Global Leadership Initiative on Malnutrition (GLIM) diagnostic criteria may improve adult malnutrition identification. Some GLIM parameters are challenging to obtain in resource-constrained settings. We assessed the relative validity of accessible surrogates in substituting accepted phenotypic diagnostic criteria within GLIM. <h3>Methods</h3> In a prospective diagnostic accuracy study, adult ambulatory patients with diverging diagnoses from five South African hospitals were consecutively sampled. The Malnutrition Universal Screening Tool (MUST score≥1) and GLIM diagnostic criteria determined malnutrition risk and diagnosis respectively. Surrogates investigated for appendicular skeletal muscle mass index (ASMMI), measured by multifrequency bio-electrical impedance, were mid-upper arm circumference (MUAC) and body mass index-adjusted calf circumference (CC). Pearson correlation coefficients, receiver operating characteristics curve analysis, Youden's index, sensitivity (Se), specificity (Sp), and Cohen's kappa were used. <h3>Results</h3> Among 480 patients screened (51% male; mean age: 47.03 years, SD 14.87), 73% (<i>n</i> = 350) were at malnutrition risk. Of these, 54% (<i>n</i> = 189) were malnourished (GLIM criteria). MUAC and CC moderately correlated with ASMMI (<i>r</i> = 0.645; <i>P</i><0.001 and <i>r</i> = 0.515; <i>P</i><0.001 respectively). Optimal cut-offs for surrogates, with corresponding diagnostic accuracy, were: MUAC for ASMMI: 24.9 cm (area under the curve (AUC) = 0.857, Se=85%, Sp=74%) and CC for ASMMI: 29.1 cm (AUC = 0.819, Se=73%, Sp=83%). Substituting MUAC and CC for ASMMI at study-specific cut-offs - within the GLIM phenotypic criteria - was relatively valid (MUAC: kappa=0.925, Se=100%, Sp=91%; CC: kappa=0.849, Se=100%, Sp=91%). <h3>Conclusion</h3> Within GLIM and using the reported cut-offs, MUAC reflects ASMMI. BMI-adjusted CC can replace low skeletal muscle mass. Valid, easy-to-use surrogates may enhance the feasibility of the GLIM diagnostic framework, particularly where resources are constrained.