Statistical methods in genetic and molecular epidemiology and their application in studies with metabolic phenotypes.

Epidemiology is the study of distributions and determinants of diseases and their associated risk factors. Studies are typically based on large samples of individuals from the general healthy population and/or samples of disease cases from the same or a similar population. Molecular epidemiology is a relatively new discipline in epidemiology dealing with patterns of molecular risk factors of complex diseases. The investigation of broad panels of molecular intermediate phenotypes is central as they may provide a direct access to the underlying disease causing mechanisms. Genetically determined molecular phenotypes are typically quantitative traits that are known to be risk factors for complex diseases and they are assumed to be modulated by common genetic variants. Metabolite concentrations are examples of such genetically determined molecular phenotypes which we call genetically determined metabotypes. In contrast, genetic epidemiology means different things to different people. We regard it as a discipline that focuses on the familial, and in particular genetic, determinants of complex disease and investigates the joint effects of genes and non-genetic determinants such as environmental factors, lifestyles and diets. Both in molecular and genetic epidemiology appropriate account has to be taken regarding the biology that underlies the action of genes and the known mechanisms of inheritance. Important aspects in epidemiology are appropriate study design, data collection, data quality control, statistical analysis, as well as interpretation of results. In this chapter we will discuss some of these issues with a particular focus on genetic association studies in the general population or in case-control studies