WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
When we collect data such as blood pressure in research studies, there is always some variability in the measurements due to error. This can be caused by things like day-to-day fluctuations in the values or by imprecision in the equipment used to take the measurements. It is possible to get an idea of how large the measurement error is and what effect it has on our conclusions by re-measuring some of the participants in a repeatability sub-study and calculating correction factors for the associations seen between error-prone measures and outcomes such as mortality. Researchers have to re-measure enough people to estimate the correction factors precisely enough, but not so many that it becomes too expensive or time consuming. We have written a paper explaining to researchers how they can calculate how many people to re-measure in their repeatability sub-studies. We ve illustrated this using the International Project on Cardiovascular Disease in Russia as an example, and have provided estimates of several correction factors from the UK Biobank study to help researchers plan their own studies, in our paper Reflection on modern methods: calculating a sample size for a repeatability sub-study to correct for measurement error in a single continuous exposure published in the International Journal of Epidemiology
The impact of smoking, alcohol and adiposity on health outcomes in the UK Biobank
The proposal is to investigate the impact of three classic risk factors (smoking, alcohol and adiposity, each of which can affect many different diseases) on a wide range of health outcomes in UK Biobank (UKB). Of particular interest will be vascular outcomes and cancer along with cognitive performance and mood. This study will contribute to a greater understanding of how these classic risk factors combine to affect population health.
The initial focus of this work will be to analyse baseline data comparing risk of major chronic disease according to smoking behaviour, alcohol consumption and adiposity. When re-measurement data become available we will use these to improve the precision of our analyses. When prospective data become available we will extend our analyses to include incidence and mortality. Outcome data of interest will include death, cancer registry data and hospital episode data. When bio-marker data become available we will investigate possible mechanisms underlying the associations of these three risk factors with prevalent and incident outcomes. When genetic data become available we will use these to investigate the interaction between genetic and environmental factors and to further investigate possible mechanisms.
The value of the proposed analyses lies in the statistical power that is available from using the entire UKB cohort. This will enable us to provide the most detailed analyses of these important causes of morbidity and mortality to date