Prospective changes in global DNA methylation and cancer incidence and mortality

Brian T. Joyce, Tao Gao, Yinan Zheng, Lei Liu, Wei Zhang, Qi Dai, Martha J. Shrubsole, Elizabeth A. Hibler, Massimo Cristofanilli, Hu Zhang, Hushan Yang, Pantel Vokonas, Laura Cantone, Joel Schwartz, Andrea Baccarelli, Lifang Hou

Research output: Contribution to journalArticlepeer-review

31 Scopus citations


Background:Methylation of repetitive elements Alu and LINE-1 in humans is considered a surrogate for global DNA methylation. Previous studies of blood-measured Alu/LINE-1 and cancer risk are inconsistent.Methods:We studied 1259 prospective methylation measurements from blood drawn 1-4 times from 583 participants from 1999 to 2012. We used Cox regression to evaluate time-dependent methylation as a biomarker for cancer risk and mortality, and linear regression to compare mean differences in methylation over time by cancer status and analyse associations between rate of methylation change and cancer.Results:Time-dependent LINE-1 methylation was associated with prostate cancer incidence (HR: 1.38, 95% CI: 1.01-1.88) and all-cancer mortality (HR: 1.41, 95% CI: 1.03-1.92). The first measurement of Alu methylation (HR: 1.39, 95% CI: 1.08-1.79) was associated with all-cancer mortality. Participants who ultimately developed cancer had lower mean LINE-1 methylation than cancer-free participants 10+ years pre-diagnosis (P<0.01). Rate of Alu methylation change was associated with all-cancer incidence (HR: 3.62, 95% CI: 1.09-12.10).Conclusions:Our results add longitudinal data on blood Alu and LINE-1 methylation and cancer, and potentially contribute to their use as early-detection biomarkers. Future larger studies are needed and should account for the interval between blood sample collection and cancer diagnosis.

Original languageEnglish
Pages (from-to)465-472
Number of pages8
JournalBritish Journal of Cancer
Issue number4
StatePublished - Aug 9 2016


Dive into the research topics of 'Prospective changes in global DNA methylation and cancer incidence and mortality'. Together they form a unique fingerprint.

Cite this