Application of a conceptual model to predict physical activity identity among Canadian adults

BACKGROUND: Evidence supports the benefits of physical activity (PA); however, many adults do not achieve recommended PA levels. Identity theories have been used to understand PA, but the antecedents of PA identity remain less clear. PURPOSE: This study presents an exploratory test of a conceptual model predicting PA identity strength using candidate antecedents identified in a recent narrative review. METHODS: A three-week prospective survey was administered to 570 Canadian adults. Time 1 recorded weekly self-reported minutes of moderate-to-vigorous physical activity (MVPA), self-regulation domains and five candidate antecedents-social relatedness, personal investment, perceived capability, PA alignment and priority. PA identity was measured 3 weeks later (Time 2). Using a structural equation modelling approach, we explored the proposed pattern of direct and indirect paths. RESULTS: = 795.92, df = 324, CFI = .95, TLI = .95, RMSEA = .05 (90% CI = .05-.06) and SRMR = .07. Significant direct paths to Time 2 identity were observed for Time 1 MVPA, reactive regulation, self-monitoring, relatedness, perceived capability, alignment and priority (β = .11-.34, ps ≤ .04). Indirect paths from Time 1 relatedness, personal investment and priority to Time 2 identity-operating via Time 1 MVPA and self-regulation strategies (especially reactive regulation and self-monitoring)-were also significant (β = .09-.18, ps ≤ .01). CONCLUSION: PA identity may have multivariate inputs. These findings provide preliminary evidence for the proposed conceptual model and experimental work is needed to determine whether modifying these inputs changes PA identity.