Title: Exploring the reciprocal relationship between stress and physical activity
Thesis Supervisor: Dr. Eli Puterman
Committee Members: Drs. Bruno Zumbo & Mark Beauchamp
External Examiner: Dr. Catherine Sabiston
University Examiners: Drs. Nancy Sin & Amery Wu
Chair: Dr. Jason Snyder
Abstract: Continued participation in physical activity improves both mental and physical health, as well as decreases the risk of premature mortality. However, the majority of Canadians are not meeting the recommended levels of physical activity. One factor that is associated with continued physical activity behaviours is psychological stress, with those who are more stressed, generally, being less physically active. However, previous research has demonstrated that being physically active lowers perceptions of stress. As there is a potentially bidirectional relationship between psychological stress and physical activity, a thorough investigation of these associations is warranted. In this dissertation, I explored these relationships in three notable ways. In the first study, I assessed the prospective associations between an exercise intervention and psychological stress over the course of a six-week intervention in previously inactive individuals. In the second study, I examined the associations between physical activity behaviours and psychological stress during a two-month period of natural physical activity behaviours. In the third study, I assessed whether unique profiles of physical activity and psychological stress over a ten-year period exist, as well as whether potential predictors from four domains (i.e., psychosocial, behavioural, physical health, demographic) might predict membership of these profiles. In the first study there were no associations between psychological stress and physical activity at either the within- or between-person level. In the second study, there was an association between physical activity and stress, but only at the between-person level (e.g., those who are more physically active are, on average, less stressed). In the third study, I identified six unique profiles of physical activity and psychological stress. Four of the profiles matched, to varying extents, the predominantly negative associations that have been described in the literature (e.g., high physical activity and low stress), one matched a positive association (i.e., high physical activity and high stress) and one matched a negligible association (i.e., medium physical activity and medium stress). Further, I used a machine learning algorithm to identify the most salient predictors of profile membership which were: self-rated health, role limitations due to emotional problems, depressive symptoms, and sex.