GDF-15 is associated with sarcopenia and frailty in acutely admitted older medical patients.

Abstract

Growth differentiation factor-15 (GDF-15) has been associated with senescence, lower muscle strength, and physical performance in healthy older people. Still, it is not clear whether GDF-15 can be utilized as a biomarker of sarcopenia and frailty in the early stages of hospitalization. We investigated the association of plasma GDF-15 with sarcopenia and frailty in older, acutely admitted medical patients. The present study is based on secondary analyses of cross-sectional data from the Copenhagen PROTECT study, a prospective cohort study including 1071 patients ≥65years of age admitted to the acute medical ward at Copenhagen University Hospital, Bispebjerg, Denmark. Muscle strength was assessed using handgrip strength, and lean mass was assessed using direct segmental multifrequency bioelectrical impedance analyses and used to clarify the potential presence of sarcopenia defined according to guidelines from the European Working Group on Sarcopenia in Older People. Frailty was evaluated using the Clinical Frailty Scale. Plasma GDF-15 was measured using electrochemiluminescence assays from Meso Scale Discovery (MSD, Rockville, MD, USA). We included 1036 patients with completed blood samples (mean age 78.9±7.8years, 53% female). The median concentration of GDF-15 was 2669.3pg/mL. Systemic GDF-15 was significantly higher in patients with either sarcopenia (P<0.01) or frailty (P<0.001) compared with patients without the conditions. Optimum cut-off points of GDF-15 relating to sarcopenia and frailty were 1541 and 2166pg/mL, respectively. Systemic GDF-15 was higher in acutely admitted older medical patients with sarcopenia and frailty compared with patients without. The present study defined the optimum cut-off for GDF-15, related to the presence of sarcopenia and frailty, respectively. When elevated above the derived cutoffs, GDF-15 was strongly associated with frailty and sarcopenia in both crude and fully adjusted models.