Exploring Factors Associated with Dietary Improvement Among Participants Who Completed a Randomized Controlled Trial for Physical Activity: Evidence of a Transfer Effect.

Background: This research aimed at understanding how changes in one behavior affect other behaviors, known as "transfer effect," could extend knowledge of shared constructs among compound health risk behaviors, and improve methods for promoting simultaneous behavior changes. The current study explored whether participants who completed a randomized controlled trial for physical activity (PA) improved their diet with no intervention in diet or nutrition.

Methods: US adults (N = 283) were randomly assigned to 12 weeks of either exercise videogames, standard exercise, or an attention control. Secondary analyses were conducted to examine a possible transfer effect of the intervention on diet at end of intervention (EOT) and 6-month follow-up. Assessments of potential PA constructs (i.e., exercise enjoyment, self-efficacy) and demographics (e.g., age, gender) were conducted. PA, specifically moderate-to-vigorous levels of physical activity (MVPA), was measured using a self-report measure. Diet was measured using the Rate Your Plate dietary assessment.

Results: Findings support that a randomization effect was associated with a higher likelihood of increasing MVPA (β = 30.00, 95% CI = 4.46, 64.46) and dietary improvement at EOT (β = 1.48, SE = .83, P = .01) and follow-up (β = 1.74, SE = .52, P = .02). At EOT, changes in diet were associated with higher PA enjoyment (β = 0.41, SE = 0.15, P = .01). This intervention effect on diet was moderated by gender, with women showing more dietary improvement than men (β =  - .78, SE = 13, P = .03). At 6 months, dietary improvement was associated with greater self-efficacy (β = .04, SE = .01, P = .01).

Conclusion: This study shows evidence of a transfer effect among two synergistic behaviors and expands understanding of factors that predict the occurrence of this type of behavior change.