3.1 Correction of the control sample

The allocated weightings varied from 0.37 for men aged between 18 and 24 years to 3.38 for women aged 65 years and above.

3.2 Sociodemographic description of the two samples

Eighty-nine percent of the course-takers were male (vs 54% in the control population). Table 1 shows the sociodemographic characteristics of the two samples, by gender.

The course-takers were younger (course-takers mean age 39 ±14 for women and 37 ± 13 for men, 46 years ± 16 for both genders in controls mean age), more frequently single and more often with an occupation (88% compared to 69% in the men, 78% compared to 61% for the women). When they were in employment, they were more often self-employed than the controls. In the case of men only, they had a lower educational level. The female course-takers were marked out by their executive or self-employed occupational status. In fact, women course-takers’ educational level was not only higher than the male course-takers, but also than the male controls.