Genetic epidemiology of markers of genomic ageing
Background Ageing is associated with changes in physical functioning, generally leading to a progressive decline in health and development of age-related diseases. Age-related changes also affect our genome and markers of genomic ageing, such as telomere length and chromosomal loss, have been linked to cancer. The genetic architecture of these markers is not well understood and studies investigating associations with common age-related cardiometabolic conditions have been limited in their design, analytical methods, power and genetic instruments used. Only a few studies have investigated prospective changes in these markers with age. Objectives To study the epidemiology of two heritable markers of genomic ageing, leukocyte telomere length (LTL) and mosaic loss of chromosomal Y (mLOY) and test their causal relevance for cardiometabolic and other age-related disorders. Methods Large-scale, genome-wide meta-analysis, two-sample Mendelian Randomization (MR), and prospective observational case-cohort analysis methods were used to (1) identify novel genetic determinants of LTL, (2) investigate causal associations of genetic differences in LTL with disease, (3) review the evidence on and assess the feasibility of studying longitudinal changes of LTL, and (4) assess observational associations between LTL and mLOY and future risk of type 2 diabetes (T2D) in a large, international case-cohort study. Results Genome-wide meta-analyses including 78,592 individuals identified 49 regions associated with LTL at FDR<0.05 including 17 (6 novel) at p-value<5X10-8. A total of 32 candidate genes were prioritised with strong suggestive evidence for their roles in telomere homeostasis, DNA repair and nucleotide metabolism. Targeted and phenome-wide MR analyses suggested causal associations of shorter LTL with an increased risk of cardiovascular conditions, and decreased risks of multiple cancer types and diseases of excessive growth. LTL shortening was observed even in young and healthy individuals, and baseline LTL was strongly associated with the rate of shortening, questioning the usefulness of LTL shortening rate as an outcome in genetic association studies. No evidence was found for strong associations of mLOY or LTL either measured or genetically predicted with the risk of T2D. Conclusion Our findings substantially expand current knowledge on genes and mechanisms regulating LTL, as well as refine our understanding of the impact of genetic differences in LTL on human health and disease, while providing no strong evidence for prospective observational or causal associations between markers of genomic ageing and T2D risk.