Tiara Ratz

• Background: Digital physical activity interventions can be effective, but how their components influence complex health behavior change processes has rarely been investigated in older adults. Objective: Applying principles from social cognitive theory, missing value treatment and person-centered analyses, this thesis aims to tackle three research gaps in the form of barriers to physical activity behavior change in digital interventions targeted at older adults. Methods: Study 1 covers social-cognitive mechanisms in the effect of tailored, theory-based digital interventions on movement in the physical activity stage of change. In study 2, lifestyle profiles consisting of six self-reported, health-related behaviors were researched using latent profile analysis. Adjusted risk ratios were calculated to identify dropout-vulnerable risk profiles. In Study 3, latent class growth analysis was used to determine trajectory subgroups regarding physical activity and sedentary behavior. Results: In study 1, the hypothesized positive effects on stage of change were partly mediated by social-cognitive predictor changes. There were heterogenous intervention mechanisms. Four latent health-related lifestyle profiles were identified in study 2. Membership of the “socially inactive lifestyle” profile was associated with an elevated risk of dropping out. Study 3 identified two latent physical activity and sedentary behavior change trajectories, respectively. Significant positive trajectories were only observed in the highly sedentary. Discussion and Conclusion: This thesis lays out a theoretical and methodological basis of how areas of heterogeneity in the physical activity behavior change process of older adults participating in digital interventions can be analyzed. Characterizing distinct subgroups and their needs can advance tailoring of intervention components and behavior change strategies, and ultimately improve acceptance, retention, and long-term intervention effectiveness.