Address University of Edinburgh, Centre for Population Sciences, Teviot Place, Edinburgh, EH8 9AG Summary of research Key words: Lifespan, Longevity, Biomarkers, Prediction, Mortality, Association
1a: The determinants of longevity are of wide interest and have been studied for over 100 years. Human lifespan is influenced by both genetic and environmental factors. We propose to study longevity in UK Biobank to better understand genetic and biological markers of lifespan, not focussed on one health condition but on overall mortality. We shall use the unprecedented scale and rich data of Biobank to investigate the degree to which lifespan is inherited, using the latest genomic methods, we shall then search for genetic variants and biomarkers which influence lifespan and estimate how well it is possible to predict lifespan. 1b: The proposed research is clearly in the public interest: individuals, health services, and the insurance and pensions industry will benefit from a better understanding of the genetic and environmental influences on lifespan and
ageing. There is a clear relationship to health and illness – mortality being the ultimate end point. 1c: We propose to analyse longevity in a number of ways. (a) Assess the degree to which lifespan is genetic, using new methods designed for unrelated people. (b) Search across the genome for regions that are associated with longer or shorter survival. (c) Use the DNA sharing between individuals to try to predict lifespan in UK Biobank and compare to how this works in other populations available to us where individuals area all related and so share more DNA. (d) Assess the contribution of environmental factors and biomarkers such as albumin to lifespan. 1d: The research will focus on the ~9,000 individuals who are already deceased and recapture data towards the end of 2015 when complete genotype information is available, but will also use data for all participants. PROJECT EXTENSION - APPROVED BY UK BIOBANK 07.10.2015 Request received to extend application 8304 to disease in the context of lifespan, and thus avoid any ambiguity. "We wish to study the association with lifespan of diseases in offspring and parent (and the other traits we already have access to, e.g. education and height), through use of linear models and disease as a binary explanatory factor. In particular, we also wish to look at SNPs singly and collectively, previously known or discovered by us in UKB using conventional GWAS techniques, to associate with disease and their effect on longevity, akin to Mendelian randomisation. Should we discover robust disease associations with genetic variants, individually or collectively, we would report these in their own right, as well as their consequences on lifespan, using Cox models. For genetic variants that individually or collectively are associated with longevity we wish to examine which if any diseases associate with the variant, by measuring the association between SNP and each disease, using logistic regression. We would like to measure pleiotropy in the context of lifespan: the degree to which a variant causing one disease may increase or reduce the risk of other diseases and the resultant overall magnified or diluted effect on lifespan, using Cox models and logistic regression". "We are also interested in exploring whether some non-disease characteristics are related to lifespan, and have been using height and weight for this purpose".
Project extension July 2017: “We would use the hospitalisation data as part of our analysis of lifespan and longevity to help build variables which capture "frailty", for instance through number and type of hospital visits, but also to further enrich the pool of phenotypes which can be interrogated vis a vis their relationship to lifespan or parental lifespan, for example using genetic instruments. This may involve "PheWAS"-type approaches, whereby phenotypes are generated from the record linkage data using algorithms.” “We plan to derive measures of fluctuating asymmetry, which are potentially a measure of body systems integrity, which in turn may be related to lifespan and aging. In our own cohort we have measures of fluctuating asymmetry derived from the long bones in the whole body DEXA scans (e.g. tibia, femur, ulna, radius) and we'd aim to perform the same measures on at least a subset of the UKB scans. The limiting factor is how much time we can spend on generating the FA measures from the scans. Clearly we'd share back FA measures with UKB."
Principal Investigator Dr James Wilson