Principal Investigator: Dr Paul O'Reilly
Institution: King's College LondonTags: 23203, morbidity, mortality, prediction, PRS, Stratification
1a: The main aim is to investigate the importance of genetic data, over and above detailed epidemiological data, in multivariable prognostic models that predict risk of: (i) major morbidity (defined as: cardiovascular disease, stroke, cancer, or diabetes), (ii) all-cause mortality. Our primary questions are:
(1) How accurately can the risk of major morbidity and mortality be estimated using multivariable prognostic models based on detailed epidemiological risk factor data?
(2) Can such prognostic models be significantly improved by including genotype data, in the form of polygenic risk scores?
(3) What is the genetic predictive ability in individuals with otherwise healthy profiles?
1b: Being able to accurately predict mortality and morbidity outcomes is at the heart of improving the prevention, diagnosis and treatment of a wide range of serious and life-threatening illnesses, the stated goal of the UK Biobank.
1c: Statistical models will be built to predict morbidity and mortality, using a multitude of relevant epidemiological and genetic risk factors. These models are standard in statistical epidemiology, namely, logistic regression and survival analysis, with variable selection and cross-validation. The aim is to produce models that can accurately predict morbidity and mortality risk in independent members of the UK population, for whom the same risk factors can be measured. The contribution to the prediction from genetics takes the form of a single genome-wide summary of the relevant risk alleles, aggregated into what is known as a ‘polygenic risk score’.
1d: Full cohort