The Step Up pilot trial was designed to determine the acceptability and feasibility of a phone-based, cognitive-behavioral, MRF intervention targeted to depressed smokers. We also sought insight into how to best design and implement this type of treatment program, to guide a future randomized effectiveness trial and to inform future work in this area. Our results support the acceptability and feasibility of the Step Up program. Two-thirds of people who were eligible for the trial enrolled, two-thirds of enrollees completed six or more counseling calls, and a third completed all twelve calls. Participation may have been greater if the treatment window were not truncated at six months to accommodate the pilot timeline. Participants were also engaged during the calls, as evidenced by the average call duration (30 minutes). They rated their satisfaction with the program as high and each of the program components as moderately helpful or greater.
As a pilot trial, the study was not powered to detect clinically significant outcomes, so caution must be used in interpreting the results, but two important findings are worth noting. First, mood significantly improved (statistically and clinically) in the experimental group. This is not surprising since the depression module was largely adapted from a prior empirically-validated program , but positive effects were maintained in the absence of the medication management component included in the original treatment design, and the intervention's effect on mood did not appear to be negatively impacted by the inclusion of counseling for physical activity or smoking cessation. Second, several physical activity and smoking outcomes appeared to trend in the expected direction. More intervention participants attempted to stop smoking, and this group appeared to have a greater initial reduction in their daily smoking level. They also reported walking more days for exercise each week and more positive improvements in the amount of moderate physical activity they engaged in. At the same time, however, usual care participants reported a greater increase in vigorous physical activity. It is unclear why experimental participants would report greater moderate physical activity and controls report increased vigorous physical activity, but each of the study findings must be viewed with caution. Positive changes could reflect response biases rather than true behavioral changes, since they are based on self-report in a small sample. Moreover, given the small number of participants, one cannot conclude that similar effects would be observed in a larger sample. To truly evaluate the impact of the Step Up intervention on the behavioral outcomes of interest, a larger randomized trial is needed.
This study provides insight into several important design considerations for future research. First, while making a quit attempt is an important indicator of motivation and behavior change, smoking cessation is needed to ultimately reduce disease risk. We cannot draw conclusions about the relative impact of the Step Up intervention on cessation based on our pilot data, but based on our knowledge of the actual counseling call discussions, we are unsure whether the current intervention would have a meaningful impact on cessation in a larger trial. We intentionally targeted all smokers - regardless of their interest in quitting - to see if we could engage individuals who might not otherwise be reached by conventional cessation treatment programs. In this respect, we were successful, but the intervention may not be intensive enough to promote and support quitting among those with no interest in cessation. In the future, it is worth considering increasing the intensity of the cessation treatment (i.e., longer duration, pharmacotherapy), targeting smokers with some interest in quitting, or both. More intensive treatment and pharmacotherapy have both been shown to enhance quit rates , so these enhancements should increase the likelihood that participants will successfully quit smoking.
The findings from this study also provide evidence of the acceptability of targeting multiple risk behaviors concurrently. Physical activity is a natural component of behavioral activation, which is an effective form of depression intervention . Thus, integrating both behaviors into behavioral activation experiments was relatively straightforward and well received by participants. Moreover, participants were receptive to the integrated focus on smoking, which included weekly self-monitoring and self-efficacy building behavioral experiments such as practicing how to delay smoking in response to urges, changing their smoking environment, and making practice quit attempts. Furthermore, the current design allowed people to self-select to what extent they chose to attempt sequential or concurrent behavior change, since they ultimately chose which behavioral experiments to try. Thus, we cannot conclude that our concurrent method is truly preferable to sequential intervention, but this research provides a basis to recommend a similar intervention format to others.
We also found it better to focus on increasing self-directed physical activity than to require participation in a group-based, structured walking program. The walking group program was designed to provide social support, which is an important therapeutic component of CBT, but the increased social anxiety and avoidance that are often characteristic of depression created an insurmountable barrier to participation in the walking groups. According to participants, their guilt about not participating in the groups, in turn, adversely influenced their participation in the pre-pilot counseling program. Thus, despite the sound theoretical argument for group-based physical activity for people with depression, future interventions targeting this population may be better served by promoting personalized, self-paced activity.
Finally, it is worth noting that only 13% of those contacted were eligible for participation in this trial, even with our more lenient physical activity criterion which screened out highly active people but retained those who needed to improve or maintain their activity level. While this rate is not expected to generalize to other settings or recruitment methods, it does raise an important issue. Requiring criterion performance on multiple target behaviors can significantly reduce the number of potentially eligible MRF program candidates. Those eligible represent a high-risk, high-need population, but from a pragmatic standpoint, more people would be eligible and efforts to change may be more successful if people with three or more risk behaviors are not targeted. Requiring criterion performance on any two behaviors would make this type of program more generalizable, but doing so opens up important questions about how to measure the success of the program if not all participants are attempting to change the same behaviors. In this event, it would be important to measure success using an index of change across behaviors, such as that proposed by Prochaska and colleagues .