Underlying pathophysiological mechanisms behind frailty are not fully understood. To evaluate the sex- and system-specific association of 35 blood-based biomarkers with frailty. Baseline data from the population-based Activity and Function of the Elderly study (≥65 years), collected between March 2009 and April 2010, was used. Frailty was defined through a frailty index (FI). Biomarkers associations with frailty were analysed sex-, and organ-/system-specific. Frailty models were built using backwards selection in Generalized Linear Models (GLM) for continuous and logistic regression (LR) for dichotomized FI (FI ≥0·2 frail), adjusting for age, education, smoking and alcohol intake, with further adjustment for medications when needed. Residual mean squared error (RMSE), area under the curve (AUC), sensitivity, specificity, positive and negative predictive value (PPV, NPV) were estimated. Among 1180 participants (57·9% men) GLMs showed a good fit of the data with gamma-glutamyl transferase, high-density lipoprotein-, low-density lipoprotein-cholesterol and growth differentiation factor 15 overall, and sex-specific transferrin, alanine transaminase, testosterone, vitamin D, lactate dehydrogenase, NT-proBNP in men (RMSE 0·064, specificity 0·96, NPV 0·86), and leucocytes, cystatin C, DHEA, fT3, hs-cTnT in women (RSME 0·074, specificity 0·94, NPV 0·87). LR models included less biomarkers with similar properties (AUC 0·83, specificity 0·80, 0·93 NPV in men; AUC 0·85, specificity 0·72, NPV 0·94 in women). Obtained models provide insight into sex-specific differences related to frailty. Surprisingly, inflammation does not play an important role when taking all other biomarkers into account. Obtained models offer a good framework for the identification of blood-based biomarkers to be used in frailty prediction models.
Boehm et al. (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: