Evaluating behavior change factors over time for a simple vs. complex health behavior

Background: Researchers are working to identify dynamic factors involved in the shift from behavioral initiation to maintenance—factors which may depend on behavioral complexity. We test hypotheses regarding changes in factors involved in behavioral initiation and maintenance and their relationships to behavioral frequency over time, for a simple (taking a supplement) vs. complex (exercise) behavior. Methods: Data are secondary analyses from a larger RCT, in which young adult women, new to both behaviors, were randomly assigned to take daily calcium (N = 161) or to go for a daily, brisk walk (N = 171), for 4-weeks. Factors (intentions, self-efficacy, intrinsic motivation, self-identity, habit strength) were measured weekly. Multi-level modeling evaluated their change over time. Bivariate correlations and multiple regression determined the relationships between factors and the subsequent-week behavioral frequency (self-report and objective). Finding: Results were partly in-line with expectations, in that individuals’ intentions and self-efficacy predicted initial behavioral engagement for both behaviors, and habit strength increased for both behaviors, becoming a significant predictor of behavioral frequency in later weeks of the study in some analyses. However, results depended on whether the outcome was self-reported or objectively measured and whether analyses were bivariate or multivariate (regression). Discussion: The factors theorized to play a role in behavioral maintenance (intrinsic motivation, self-identity, and habit strength) started to develop, but only habit strength predicted behavioral frequency by study-end, for both behaviors. Differences in initiation and maintenance between behaviors of differing complexity may not be as stark as theorized, but longer follow-up times are required to evaluate maintenance factors.