Modifiable psychosocial constructs associated with physical activity participation in people with Multiple Sclerosis: a meta-analysis

Abstract

Background: This review aims to synthesis current knowledge of the modifiable psychosocial constructs associated with physical activity (PA) participation in people with Multiple Sclerosis (pwMS). Methods: A systematic search was conducted on 8 databases using keywords associated with MS, PA and behavioural theory. Cohort and intervention studies were included if they 1) included an objective or subjective measure of PA, 2) measured at least one modifiable psychosocial construct and 3) reported bivariate correlations (or these could be extracted) between the PA and psychosocial construct measures. Meta-analyses of correlations was conducted using the Hedges-Olkin method. Where a meta-analysis was not possible, results were reported descriptively. Findings: Meta-analyses indicated a pooled correlation coefficient between 1) objective PA and self-efficacy (n=7) of r=0.30 (p <0.0001) indicating a moderate, positive association, 2) subjective PA and self-efficacy (n=7) of r=0.34 (p <0.0001) indicating a moderate, positive association, 3) subjective PA and goal-setting (n=5) of r=0.44 (p <0.0001) indicating a moderate to large, positive association and 4) subjective PA and outcome expectancies (n=4) (physical: r=0.13, p=0.11; social: 0.19, p<0.0001; self-evaluative: r=0.27, p<0.0001) indicating small-moderate positive associations. Non-meta-analysed constructs, such as measures of health beliefs, enjoyment, social support and perceived benefits and barriers were reported to be significantly correlated with PA. Discussion: Future PA interventions should continue to focus on the psychosocial constructs of self-efficacy, goal-setting, and outcome expectancies. However, there is a need to explore the associations between other constructs outside of those reported in this review.