Successful translation of preventive programs from the RCT phase of research to large-scale public health initiatives requires continued evaluation throughout the dissemination process (18
). Such initiatives often also require cross-sector collaboration, including some combination of public, nonprofit, and private sector partners (20
). Our study builds on prior RCT studies (9
) that established the protective effect of the Planet Health obesity-prevention curriculum on DWCB in middle schools by evaluating its implementation as part of a large-scale dissemination program led by a state public health department in collaboration with a major private health insurer in the state. This study appears to be both the first large-scale dissemination evaluation of the effects of an obesity prevention program on DWCB in youth and the first partnership between public and private sectors implementing an evidence-based DWCB prevention program statewide in the United States.
Findings indicated that when Planet Health was widely disseminated throughout Massachusetts as part of the HC program, a greater number of student lesson-exposures to Planet Health lessons on reducing television viewing was associated with reduced odds of student DWCB at follow-up. Students in schools offering programs for staff on reducing television viewing had reduced odds of DWCB at follow-up compared with students in other schools. It is possible that the staff programs may have enhanced the education that students received on why and how to reduce their own television viewing. The Planet Health efficacy trial found decreased television viewing was the strongest mediator of intervention effects on obesity (11
). In our study, other factors protective against DWCB at follow-up were having a greater number of partners from outside the school involved with HC activities and having a high level of teamwork among staff working on HC.
Of concern is that we found positive associations for some Planet Health components, which is counter to study hypotheses; however, this is not surprising, given the challenges that schools face in implementing and sustaining complex, long-term health-promotion programs (16
). A larger number of teachers teaching Planet Health lessons and a larger number of student lesson-exposures to the physical activity and fitness topics were both associated with small but significantly increased odds of DWCB at follow-up. The reasons underlying these associations are unclear, but it is possible that in schools where many teachers were teaching Planet Health, it was difficult to make sure all received training. In a separate process evaluation of HC, school staff sometimes found it difficult to ensure that teachers received training in the Planet Health curriculum even though training was offered both in person and online (ML Greaney, CK Hardwick, JL Spadano-Gasbarro, S Mezgebu, CM Horan, S Schlotterbeck, et al, unpublished data). In addition, prior research has found that many school teachers, like adults in general, hold stigmatizing attitudes toward obesity, and physical education and health teachers in particular are themselves at elevated risk of eating disorder symptoms and behaviors compared with other teachers (29
). These findings suggest that school staff involved in implementing Planet Health and other obesity and DWCB prevention programs likely need training not only in how to implement the programs but also in how to recognize and address their own possible weight-biased attitudes and unhealthy beliefs and behaviors related to food and weight (29
A greater number of Planet Health lessons taught on reducing television viewing also had a positive association with DWCB, but only after simultaneously controlling for number of student lesson-exposures to the television lessons, which, in contrast, was protective. Reasons for this seemingly paradoxical finding are unclear. Because these 2 variables were highly correlated and may be confounded, data from our study did not allow full disentanglement of their effects. One alternative interpretation is that if the number of television lesson-exposures is held constant, then increasing the number of Planet Health television lessons exposes students to a more diverse set of messages from teachers. In this light, it is possible that increasing lesson-exposures is protective, but only when done in a way that focuses the message and exposes the most children to the same narrow set of effective messages.
This study has several limitations. Survey data from students were anonymous and therefore did not allow estimation of change in individuals. Students attending the schools at baseline were overlapping but not identical to those at follow-up because of normal fluctuation resulting from graduation and changing residences. We did not use a randomized-controlled design; however, because Planet Health has already been shown to be protective against obesity and DWCB in efficacy and effectiveness RCTs (9
), MDPH held that it would not be appropriate to withhold a proven intervention through randomization of schools. More frequent assessments to monitor incremental behavior changes were not possible. Information was not available on idiosyncratic implementation styles of individual teachers; for instance, some teachers who themselves hold unhealthy beliefs about food, weight, and weight control may have conveyed these beliefs to students while teaching Planet Health lessons (29
). We did not assess potentially important symptoms, such as preoccupation with thinness, body dissatisfaction, fasting, and some symptoms that especially involve males, such as use of anabolic steroids.
Findings from our study provide modest evidence that Planet Health can help prevent DWCB when widely disseminated to middle schools by a cross-sector (ie, public-private) partnership as part of a multicomponent obesity prevention program. Messages related to reducing television viewing for both students and staff seem to be especially important. In addition, findings underscore the need to ensure adequate training for dissemination with regard to both implementation and staff biases. Further study of Planet Health and HC is needed to identify why some activities appear to be protective when implemented under real-world conditions while others are not and to identify what level and types of teacher training are required to achieve protective effects. With the increase in the number of programs to improve weight-related behaviors and prevent obesity and DWCB currently in efficacy trials, researchers need to turn next to effectiveness trials, and finally to dissemination studies, likely through cross-sector partnerships. Through these next steps, we can ensure that all youth benefit from evidence-based preventive programs.