Modelling longitudinal directional associations between self-regulation, physical activity, and habit: results from a cross-lagged panel model

Background: The directionality of associations between self-regulatory variables, behavior, and automaticity is seldomly tested. In this study, we aimed to examine a volitional, self-regulatory sequence of variables proposed in the Health Action Process Approach framework (intention → action plans → action control → behavior) and its relationship with the construct of automaticity of the physical activity habit. Methods: Longitudinal data was collected from high school students (N = 203, Mage = 15.39 (SD = 1.43), 52% women) at three measurement points. First, a confirmatory factor analysis measurement model was used to examine the study variables across measurement points. Next, a cross-lagged panel model was used to test directionality between variables. Results: After adequate fit of the measurement model was confirmed, a mechanism integrating self-regulation with behavior and automaticity was examined. The hypothesized directionality between variables was verified overall by cross-lagged analysis. However, for the intention-action plan association, the inverse relationship was found: plans were associated with subsequent intentions, but intentions did not predict plans. Moreover, automaticity was not associated with subsequent physical activity behavior. Conclusions: In general, our findings supported the hypothesized longitudinal direction of the associations, confirming that self-regulation may lead to behavior performance and automaticity. Unexpected findings and implications for intervention and future research are discussed.