Fig. 1

CONSORT flow chart of the study. ¹Sample size counted in 2015. ²Five-hundred-eighty-three of the 851 participants primarily assigned to the intervention, control, or active groups renewed their consent in 2019, granting access to the data for an additional 5-year period. ³PA assessment may have been corrupted for the following reasons: medical conditions impairing normal physical behaviors during the assessment period (pregnancy, injury, etc.), acceleration signal summary not meeting inclusion criteria (4 days with a minimum of 10 h/d wearing time) despite several measurement attempts, etc. ⁴The number of analyzed data varies for each variable (see N values within brackets for each group). Reasons for missing data include: baseline assessment of physical activity not meeting quality standards, non-eligibility to physical fitness tests (e.g., participant becoming over 65 during the course of the study), impossible extraction of some specific parameters (e.g., vascular echo parameters due to the quality of images), participant not willing/able to perform one specific test on the date of the visit, among others