Averaging attributable fractions in the multifactorial situation: assumptions and interpretation.

In multifactorial situations, the concept of attributable fraction quantifying the population impact of exposure factors on the disease load needs special methodologic care to take the interrelationship between exposures into account. Recently, new epidemiologic parameters have been introduced to address this issue. These approaches are reconsidered here from a game-theoretical perspective. The derivation of game-theoretical properties of these parameters provides additional motivation for their use in epidemiologic practice and offers new insights into their interpretation in epidemiologic studies. The nontechnical description of the results is illustrated by examples from the Hordaland study on obstructive lung disease.