The Association between Proteomic Aging Clocks and the Risk of Cancer in Midlife Individuals

Background: To measure the aging process before a cancer diagnosis, we developed the first cancer-specific proteomic aging clock (CaPAC) and examined its association with cancer risk in the Atherosclerosis Risk in Communities (ARIC) and Multi-Ethnic Study of Atherosclerosis (MESA) studies. Methods: Using the SomaScan assay, ARIC measured 4,712 proteins in plasma samples collected in 1990-92 from 3,347 participants who developed cancer over follow-up until 2015 and 7,487 who remained cancer-free, all aged 46-70. We constructed CaPAC0 using elastic net regression among two-thirds randomly selected cancer-free participants (N=4,991, training set) and calculated age acceleration for CaPAC0 (CaPAA0) as residuals of CaPAC0 on chronological age in all remaining ARIC participants. We used multivariable-adjusted Cox proportional hazards regression to calculate hazard ratios (HRs) for the risk of overall, obesity-related, smoking-related, and the most common cancers (prostate, lung, breast, colorectal) with CaPAA0 using a case-cohort design. We replicated the analysis in 3,893 MESA participants aged 46-70 at Exam 1 (456 incident cancer). Results: CaPAC0 was correlated with chronological age in ARIC and MESA (r=0.82 and 0.86, respectively). In both ARIC and MESA, CaPAA0 was significantly (p<0.05) associated with the risk of overall [HRs per 5-years=1.08 and 1.23, respectively], smoking-related [HRs=1.30 and 1.54, respectively], and lung cancers [HRs=1.54 and 1.94, respectively]. CaPAA0 was also significantly associated with colorectal cancer risk in ARIC [HR=1.31], but not in MESA. CaPAA0 was not associated with obesity-related, breast, or prostate cancers. Conclusion: CaPAA0 was associated with several types of cancer with the strongest association observed for lung cancer risk.