Distinct trajectories of leisure time physical activity and predictors of trajectory class membership: a 22 year cohort study.

Peter T Katzmarzyk,Lise Gauvin,Cora L Craig,Tracie A Barnett

doi:10.1186/1479-5868-5-57

Abstract

BackgroundProspective studies linking social factors to long term patterns of physical activity are lacking. In this 22 year longitudinal study, we seek to identify long term patterns of involvement in leisure time physical activity (LTPA), and explore socioeconomic and demographic predictors of distinct LTPA trajectories.MethodsAmong 2102 individuals aged 18–60 years in 1981 who participated in the 1981 Canada Fitness Survey/1988 Campbell's Survey of Well-Being, 1186 (56.4%) completed questionnaires for the 2002/04 follow-up. Complete data on LTPA at all 3 surveys were available for 884 participants. Latent class growth analysis was used to identify major classes of LTPA trajectories; predictors of class membership were identified using polytomous logistic regression.ResultsFour latent classes were identified: inactive, increasers, active, and decreasers (53%, 26%, 12%, and 9% of participants, respectively). Women, older participants, those with lower household income, and with lower educational attainment, were significantly less likely to follow active (Vs. inactive) trajectories of LTPA. Disadvantaged groups with respect to education and income were also significantly more likely to follow decreasing (Vs. active) trajectories.ConclusionThere is a need for continued efforts to increase overall population levels of LTPA, particularly among socially disadvantaged groups with respect to income and education, who are most likely to experience unfavorable trajectories of LTPA.

Highlights

Prospective studies linking social factors to long term patterns of physical activity are lacking
There were no differences with respect to baseline physical activity level or overweight status, but those retained for analysis were significantly more likely to view their general level of health favourably (Table 1)
Latent class analysis To identify the optimal number of classes, we initially modeled two groups; we systematically included one additional group, and compared the goodness of fit statistics of each incremental pair

Summary

Introduction

Prospective studies linking social factors to long term patterns of physical activity are lacking. In this 22 year longitudinal study, we seek to identify long term patterns of involvement in leisure time physical activity (LTPA), and explore socioeconomic and demographic predictors of distinct LTPA trajectories. We sought to identify distinct long term patterns of leisure time physical activity (LTPA), in a large, diverse cohort of adults who provided 3 measures of LTPA over a 22 year period. We examined socioeconomic and demographic predictors of following specific trajectories of LTPA involvement. These objectives were undertaken using data from the Physical Activity Longitudinal Study (PALS).[8]

Methods

Results

Discussion

Conclusion