When the cause of tuberculosis was discovered more than 100 years ago, the relationship between the infection and disease seemed rather straightforward. Koch's postulates for the general principles of such relationships became very famous and are still valid (table 1). When there is only one cause of a disease, we are inclined to think in a linear manner, as in the case of tuberculosis. However, it became clear subsequently that most diseases have more than one cause—that is, most diseases are multifactorial with several risk factors. Today's criteria for a causal relationship between a disease and its causes are therefore much more complicated, as listed in table2.1

However, we are still inclined to think of …