It is not controversial that study design considerations and challenges must be addressed when investigating the linkage between single omic measurements and human phenotypes. It follows that such considerations are just as critical, if not more so, in the context of multi-omic studies. In this review, we discuss (1) epidemiologic principles of study design, including selection of biospecimen source(s) and the implications of the timing of sample collection, in the context of a multi-omic investigation, and (2) the strengths and limitations of various techniques of data integration across multi-omic data types that may arise in population-based studies utilizing metabolomic data.