Proteomic aging clock and cancer risk and mortality

Abstract

Proteomic aging clocks (PACs) have been proposed as a measure of biological aging. The deviation of PAC from chronological age is called PAC acceleration. The aims of this dissertation were to construct and validate PACs in a large population-based study -- the Atherosclerosis Risk in Communities (ARIC) study and test their associations with mortality, cancer risk and survival. Using proteomics data from ARIC, in manuscript 1, we constructed PACs at middle age and older age using healthy participants. Both PACs were strongly correlated with chronological age, and the age acceleration for both PACs was significantly associated with all-cause mortality. Among middle-aged participants, we compared our newly created PAC to three published PACs. The newly created PAC was strongly correlated with all three published PACs and the age acceleration for all these PACs was similarly associated with mortality. In Manuscript 2, using Visit 2 proteomics data, we constructed a cancer-specific PAC (CaPAC) and compared it to the published PAC created by Lehallier (Lehallier’s PAC, this PAC was most strongly correlated with chronological age among published PACs). CaPAC was strongly correlated with chronological age and Lehallier’s PAC. Although only the age acceleration for CaPAC but not Lehallier’s PAC was significantly associated with risk of overall cancer and colorectal cancer, the magnitude of these associations was similar for both PACs. Acceleration for both PACs was significantly and similarly associated with lung cancer risk. In Manuscript 3, using Visit 5 proteomics data, we constructed a cancer-specific PAC (CaPAC5) and compared it to Lehallier’s PAC in cancer survivors and cancer-free participants. Both PACs were similarly associated with all-cause mortality among cancer survivors of overall cancer. These associations were similar to those among cancer-free participants. Overall, the findings from this dissertation showed that PACs that were developed in different populations were similarly associated with all-cause mortality, cancer risk and survival, suggesting the robustness of PACs.