Evidence From Structural Equation Models

Objectives: This study examines the association between early-life socio-economic status (SES) at age 18 and physical activity (PA) at age 65, elucidates mechanisms explaining this association, and explores gender differences in mediating pathways. Methods: Multigroup structural equation modeling is applied to the 1957 to 2004 data from the Wisconsin Longitudinal Study (WLS). Results: Early-life SES is positively associated with exercise in later life. This association is mediated by socioeconomic resources, health prob-lems, obesity, and depressive symptoms (women only) in 1993 and sports participation in 1957. All mediators explain over 95 % of the effect of early-life SES. Discussion: This study emphasizes the importance of complex multiple pathways linking early family SES to later-life PA. We identify chains of risks that need to be broken to improve PA among older adults. Our findings also suggest that interventions aimed at maintaining optimal physical functioning in old age should begin at least at midlife. at PENNSYLVANIA STATE UNIV on May 18, 2016jah.sagepub.comDownloaded fro