Terms such as 'molecular epidemiology' and 'genetic epidemiology' have been coined to depict the change from 'traditional epidemiology', concerned with disease determinants at the community or society level, over to 'modern epidemiology', which is concerned with determinants operating at the individual level or even below, i.e. at the organ, tissue, cell, or molecular level. In this commentary, we point out to the limitations of this development and suggest that more emphasis is placed on making the presumed causal disease models explicit, when investigating the relationship between putative determinants and disease. Understanding the disease processes at the micro-level is insufficient for understanding disease at the individual level; and disease patterns at the population level cannot be understood unless it is realized that individuals exist in a variety of contexts that cannot be reduced to individual attributes.