The combination of GDF15, FGF21, sRAGE and NfL plasma levels can identify frailty in community-dwelling people across old age

Frailty is a complex medical condition characterized by decline in physiological functions and global health of older people, representing a strong risk factor for disability, hospitalization, and mortality. The identification of biomarkers reliably associated with frailty could provide more information on the actual health status and outcome of older subjects. To this aim, we investigated possible associations of four biomarkers related to age and age-related diseases, namely GDF15, FGF21, sRAGE, and NfL, with frailty, measured using frailty index (FI), in community-dwelling subjects of different age. The study was conducted on a cohort of 463 subjects (50–113 years) enrolled before the Covid-19 pandemic and categorized as frail and non-frail, based on a 45-item FI, according to the deficit accumulation model. Plasma levels of the four biomarkers were analysed by Ella Automated ELISA and investigated for their possible association with FI. A Random Forest Decision model was used to assess the biomarkers’ discrimination power with respect to FI. In our cohort, FI was associated with plasma levels of GDF15, NfL and FGF21, but not sRAGE. The first two were also associated with survival. The model based on those four biomarkers estimated frailty with 82 % accuracy. Moreover, frailty estimate obtained with this model led to a more refined prediction of survival on a 3-year follow-up. Our data suggest that GDF15, NfL, FGF21 and sRAGE plasma levels can be proposed as parameters to provide additional information about frailty status and survival with respect to FI in community-dwelling subjects.