Biomarker Signatures Predicting 10-Year All-Cause and Disease-Specific Mortality

Abstract

Novel wide array blood biomarkers of multisystem dysregulation, compared to conventional clinical and blood biomarkers, are potentially able to provide more accurate prognostic information of long-term mortality risks. We identified biomarker signatures of all-cause and disease-specific mortality from a comprehensive range of analytes related to six major physiological functions: cytokine, chemokine, and growth factors; glucose metabolism regulators and adipokines; adhesion molecules; acute phase response; pathogen-specific antibodies; and bone remodeling. A total of 144 elderly were prospectively followed up on mortality for median 136 months. Plasma levels of 93 biomarkers measured by enzyme-linked immunosorbent assay, Luminex, FlowCytomix, DNA quantification, and recombinant DNA technology at baseline were compared among deceased and surviving elderly and in a referent group of 72 healthy young adults. The elderly, and especially deceased elderly, exhibited differential profiles of the composite index of each physiological function from young adults. In Cox regression, we identified and validated in an independent cohort of 357 elderly the specific and common biomarkers predicting all-cause, cardiovascular disease-related, neoplasm-related, and respiratory disease-related mortalities after controlling age, sex, and major comorbidities. These biomarkers had 74.3% correct classification for deceased elderly from surviving elderly. We reported for the first time, stem cell growth factor-β and gastric inhibitory polypeptide as specific biomarkers of mortality risk.

Keywords: 10-year mortality; Biomarker signature; Pathway analysis; Physiological function.