Résumé de la thèse

Background: Socioeconomic position is associated with physical activity cross-sectionally during adulthood. Is this association purely cross-sectional, or is there also a life course association between socioeconomic position during childhood and physical activity during adulthood?

Methods: Systematic review and original research. Structural equation modeling was used to compare three competing models in life course epidemiology.

Results: In our systematic review, in a subset of 15 studies with more rigorous methodology, the proportion of studies finding a life course association was high for leisure-time physical activity: 10/11 (90.9%). In our own original research, finding were: (1) in Canadian men, for leisure-time physical activity, the critical period model fit the data best for education and income; (2) in Canadian women, for leisure-time physical activity, the accumulation of risk model with additive effects fit the data best for income, (3) in Canadian women, for leisure-time physical activity, education did not fit any model; (3) in British men and women, for physical activity during leisure time, at work, and during transports, the accumulation of risk model with additive effects fit the data best for social class.

Conclusion: The bulk of the evidence pointed to a pattern of associations best represented by the accumulation of risk model with additive effects.