The results of this study confirmed that troops randomly assigned to the intervention condition provided more leader promotion and opportunities for PA and healthful eating than control condition troops. Evidence of preferential effects from the intervention on obesity and its behavioral influences was limited solely to objectively measured PA levels of girls while attending troop meetings. Intervention girls were less sedentary and accumulated greater levels of MVPA than control girls. This is an important result, as it demonstrates the feasibility of implementing enjoyable opportunities for MVPA in what otherwise could be rather inactive troop meetings. With regard to public health, more physically active troop meetings could help attending girls to reduce sedentary behavior, and perhaps over time be more likely to achieve health benefits associated with meeting PA guidelines.
A major finding of this effectiveness study was that the intervention components were implemented with good fidelity in real-world settings, and resulted in troop leader health promotion behaviors and environmental opportunities for PA and healthful eating in the troop meetings. However, our intervention resulted in no measured impact on the behavioral influences of obesity for individual girls or parents in settings beyond the troop meeting environment. Intervention troop meetings offered ample physically active content and FV snacks, while control condition meetings offered very little PA or healthful eating opportunities. It appears that troop leaders delivered the curriculum, promoted PA, FM, and FV, discouraged SSB and eating with TV, and instituted troop policies in accordance with their training for the study. However, we saw no evidence for the hypothesized changes to BMI z-scores, habitual PA, FM, FV consumption, eating with TV, or SSB consumption. Our study's statistical power appears to be adequate for the primary outcome, but low for PA and FV. However, in all cases, there were not any identified trends in the data to suggest potential non-significant effects in these variables.
The lack of detectable effects beyond the troop environment may mean that the program itself was not efficacious outside the meetings, that the troop leaders were ineffective, or that girls failed to attend to the core messages and active learning opportunities, among numerous possibilities. It is also quite possible that our intervention had some effects that were not measured in our study. For example, improved skills, attitudes, behavioral intentions, empowerment, self efficacy, proxy efficacy, and increased knowledge of health promotion behaviors are possible outcomes that were not assessed in this study. According to a mediating variable framework, interventions impact mediating variables, which then act on behaviors and other outcomes . Tests of mediation were beyond the scope of the paper, but further work using mediation analysis could help to answer the question as to why the intervention did not achieve measured outcomes beyond troop meetings. The intervention may have failed to impact potential mediators of behavior change, or perhaps the impacted mediators failed to result in detectable behavior change. Future studies of this sort should assess additional mediating variables such as skills, attitudes, intentions, and especially self efficacy for behaviors that impact obesity.
More than a decade ago, Cullen and co-authors conducted a nutrition education intervention in Girl Scout troops of similar age to the present study's sample . Cullen found significant increases in FV consumption among the intervention troop scouts, and suggested that troop positive norms and social support could be created by consistently serving FV at troop meetings, which may lead to increased consumption of FV in scouts. Although we did not assess troop meeting environmental norms, per se, it appears that our intervention policies created a troop positive snack norm of having FV that nearly all the girls ate each time. Also, the healthful eating promotion efforts of the troop leaders could constitute social support, but we did not assess whether other girls were supportive of FV consumption. We were unable to detect positive outcomes on habitual consumption levels in our intervention troop girls, but better measures combined with a greater focus on FV, may have shown more favorable results. Similar to our approach, Baranowski and colleagues also used a customized badge incentive with Boy Scouts, and were able to increase FV consumption in their intervention [9, 10].
On the PA side, intervention troop leaders provided girls with opportunities to be physically active at troop meetings, and accelerometer-based data show that the girls took advantage of those opportunities by being less sedentary and getting more MVPA than control girls. However, our data suggest that this bi-monthly troop meeting PA had no impact on girls' self-reported habitual PA levels. Similarly, Ievers-Landis and colleagues implemented an intervention to increase weight bearing PA (and calcium intake) in Girl Scouts, with a goal of primary prevention of osteoporosis . The results of their study showed no significant differences in PA among two intervention groups and a control group. Future studies of such interventions may benefit by better measures, such as use of accelerometer-based measures of habitual PA.
The present study held a number of limitations, including a relatively small sample studied over only a five-month period. Our study's size and scope were limited based on constraints in funding. Despite the size, we believe the high participation rate, combined with extensive multiple observations, all bode well for the external validity of the findings within our geographic region. We relied on several self-report measures for both parents and children, and some measures may have limited reliability and validity. Similar studies in the future could be strongly improved with better measures of behavioral influences on obesity. In contrast to these limitations, our direct observation of troop environments and rigorous objective measurement of PA in troop meetings and child weight status provide major strengths for this study.