Background: EPIC-Italy cohort study recruited subjects who voluntarily accepted to participate in the project. From the self-selected bases of the population sample, some bias could derive in the data interpretation when risk estimation for cancer disease related to life-style factors is the principal concern. Knowledge of the bias related to self-selected sampling is important for better directing the interpretation of the EPIC-Italy study results.
Methods: We investigated the characteristics of volunteer subjects recruited in the EPIC-Italy cohorts and compared them with those of the randomly selected subjects recruited in the Multipurpose ISTAT Surveys realized in the same period (1993-1998) in which the EPIC-cohorts were recruited.
Results: We found some differences, and in particular a different attitude towards cigarettes smoking and wine consumption, between the EPIC cohort and the Multipurpose ISTAT Surveys, as well as among geographical areas within the EPIC cohort.
Conclusions: The uneven distribution of some characteristics suggests that the self-selected subjects were characterized by an overall lower consumption of wine and cigarette smoking even when the educational level was considered. This could suggest a generally more healthy life-style among subjects recruited on a volountary bases.