Last Update: March 15, 2011

How to deal with multiple causes

Most diseases have multiple causes. Sometimes potential causes are so close to each other that it is easy to mistake one cause for the other: we may think that alcohol consumption seems related to lung disease, but it may simply be that smokers drink more alcohol than nonsmokers. At other times, causes augment or diminish each other’s effects. For example, a mutation in a gene might lead to a higher frequency of some disease, an environmental factor (say, smoking) might also lead to a higher frequency of the disease, but when the two causes are present together, the resulting disease burden is much greater than the sum of the effects of either one – this means that some people only get diseased when they are exposed to both factors. The task of the epidemiologist is sometimes to separate the effects of mixed causes (confounding) and at other times to assess their joint effect (interaction).

Background publications

Morabia A (2004) Epidemiology: An epistemological perspective. In History of epidemiological methods and concepts , Morabia A (ed) pp 1-126. Birkhšuser: Basel , sections 3.4, 3.8, 3.12,

Comstock GW (2004) Cohort analysis: W.H. Frost's contributions to the epidemiology of tuberculosis and chronic disease. In History of Epidemiological Methods and Concepts , Morabia A (ed), Birkhšuser: Basel, pp 223-231. Older version in SPM.

Vandenbroucke JP (2004) The history of confounding. In History of Epidemiological Methods and Concepts , Morabia A (ed). Birkhšuser: Basel, pp 313-326. Older version in SPM.

Key papers