Introduction: Constances is a general-population epidemiological cohort with analytical and descriptive objectives. Objective: To estimate prevalence rates that can be extrapolated to individuals aged 30 to 69 in 2013 who were eligible for inclusion in the Constances cohort, taking into account non-participation and combining data from two years of invitations. Methods: Based on randomly selected invitees in 2012 and 2013 aged 30 to 69 (n=471,152), participation is analyzed based on sociodemographic data, data from Sniiram (National Inter-Regime Health Insurance Information System), and data from Cnav (National Old-Age Insurance Fund). Several prevalence rates are estimated by gender using reweighting: without accounting for non-participation, and then accounting for non-participation using sociodemographic data, Sniiram data, and Cnav data. Results: the participation rate is 7.3%. Many variables from Sniiram and Cnav are associated with participation in Constances. A comparison of prevalence rates shows that accounting for non-participation has a small to significant impact, the direction of which may differ depending on whether the variables were measured at the health examination center or via self-administered questionnaire. Conclusion: These preliminary results show that data from Sniiram and Cnav are important for studying selection effects in Constances and that it is generally necessary to account for them when estimating prevalences. The combination of invitation years should be performed with caution.