Sara Johnson, Senior Vice President of Research and Product Development at Pro-Change Behavior Systems, Inc. recently presented important findings regarding multiple behavior change at the 31st annual meeting of the Society of Behavioral Medicine (SBM) meeting in Seattle, Washington.
As part of a symposium on multiple behavior change interventions, Dr. Johnson shared results of three NIH-funded Small Business Innovative Research grants that suggest that behavior change co-varies. In other words, changing one behavior increases the likelihood of changing another behavior.
In the effectiveness trial of a Transtheoretical Model (TTM)-based computer-tailored intervention for medication adherence (n=404), participants in the treatment group were significantly more likely to exercise regularly and change dietary fat intake 12 months post treatment.
In a randomized trial of a TTM-based multiple behavior weight management intervention, participants were up to five times as likely to change a second behavior if they had changed one behavior. This finding was not true for the comparison condition, suggesting that the co-variation of behavior change was due in part to having received tailored interventions for the behaviors.
Finally, a randomized trial of a multiple behavior change program for obesity prevention in eight high schools revealed that students assigned to the treatment group who took action on one behavior, such as regular exercise, were more likely to take action on another behavior (e.g., fruit and vegetable intake). The same was not true for those randomized to the comparison condition.
Pro-Change is currently examining the effectiveness of novel approaches to maximizing the impact of tailored multiple behavior change interventions to leverage this co-variation effect.
Dr. Johnson stated, “Leveraging the co-variation of behavior change will allow us to enhance the effectiveness of multiple behavior change interventions while minimizing response burden and the disseminability of tailored interventions.”
The citation and abstract of this presentation appears below.
Co-Variation: A Promising Approach to Multiple Behavior Change Interventions
Authors: Sara Johnson, PhD, James Prochaska, PhD, Karen Sherman, MA Pro-Change Behavior Systems, Inc.
ABSTRACT
Transtheoretical Model (TTM)-based computer-tailored interventions for multiple behavior change (MBC) are effective. However, fully tailored interventions (FTI) for multiple behaviors increase response burden and delivery costs, limiting feasibility and dissemination potential. Leveraging co-variation of behavior change (i.e., when taking action on one treated behavior increases the odds of taking action on a second) represents an innovative approach to MBC interventions. Data from three randomized trials will be presented to demonstrate the potential of interventions utilizing co-variation. Among 404 adults (50.4% male; 83% White) in a medication adherence study, random effects modeling indicated that those receiving a FTI for adherence and minimal messaging for related behaviors were significantly more likely to reach Action (A) or Maintenance (M) for exercise (43.5% vs. 24.7%, Odds ratio (OR)=2.39, z=2.37, p<.05) and dietary fat reduction ((24.7% vs.12.5%; OR=1.57, z=2.06, p<.05). Logistic regression performed on data from 1206 adults (50.8% male; 78% White) in a weight management intervention indicated that on each pair of the three treated behaviors (exercise, healthy diet, and emotional eating), the ORs for taking action on one behavior if a participant took action on another ranged from 2.52 to 5.18 in the treatment group vs. 1.24 to 2.63 in control. Among 1800 high school students (mean age=15.97, 49.2% male; 71.5% White) who received FTIs for exercise paired with stage-matched interventions for fruit and vegetable and reducing TV time, logistic regression indicates that the OR for likelihood of reaching A/M for a behavior if A was reached on another behavior ranged from 1.99-2.36 among the treatment group at the end of treatment (6 months) as compared to 0.5-1.0 in the control group. Across all studies, progress on one behavior led to progress on another among treatment but not control participants. Suggestions for a planned co-variation approach that has the potential to produce the greatest impact with the least demands on participants will be presented.