The aim of this paper was to assess the integration of ecological principles and theoretical constructs involved in a successful school-based obesity prevention program in Mexico City. We sought to unpack the intervention program along intervention domains in order to develop a map of the successful intervention program. If we compare our ecological map with recommendations in the literature ,,, we can ascertain that this intervention program was a genuine ecological effort within a single setting since it delivered a diversity of intervention strategies involving multiple targets (POL, ORG, INT, IND), across both intervention domains. The ecological mapping showed that several different targets (PE teachers, school teachers, food vendors, parents, children and the school environment) were engaged, and that these efforts are consistent with those recommended to promote behavior change and to prevent obesity in children ,,,-.
Our findings indicate that the project was only implemented within the school (ORG) setting; this finding is not surprising given that this was a school-based intervention program. It appears that the intensity of effort invested in the implementation of intervention activities across the two domains, varied. Within the Nutrition domain, more intervention strategies were delivered overall, relative to the PA domain. This finding is not unexpected given that when the project was implemented, the nutrition and PA contexts in the schools were different ,. The nutrition environment was disadvantaged in relation to the PA environment; for instance, there were no nutritional guidelines to regulate eating practices at school, whereas PE class, even though only offered once a week, was already in place as a regular, mandated practice . Thus, there may have been more “room for improvement” for INSP staff to influence nutrition in schools. We found that more than half of the strategies within the Nutrition domain were focused on organizational change, while the rest were mainly aimed at influencing policy to impact schools. Given significant improvements in food intake in children and food availability at schools , it would appear that the use of a diversity of strategies at different ecological levels was effective.
Overall, we observed that in relation to the intervention strategies implemented within the Nutrition domain, strategies implemented within the PA domain were less in number and in diversity, with the majority of strategies concentrated on targeting the Public Education Secretariat (POL) as a way to improve PE and PA at schools. This finding is not surprising given that the improvement of PE and PA opportunities typically requires an injection of new resources (e.g., new equipment, additional PE teachers) , whereas changes to the food environment do not necessarily require more resources per se, but instead, require an improvement of practices (e.g., replacing energy-dense foods with fresh fruits and vegetables ). Furthermore, within the Mexican context, curriculum change is not within the mandate of schools, but rather belongs to the state. Thus, the main way in which PE can be impacted in a sustainable manner at schools is through policy change at the state level (Public Education Secretariat). Given that during the 2007–2008 year period, the proportion of students taking more steps (according to pedometer counts) increased in the intervention schools as compared to comparison schools (p = .06) , it would seem that strategies to engage policy decision-makers to influence school PA opportunities may be particularly warranted in Mexico.
Overall, our findings are consistent with existing evidence that policy intervention strategies can impact different ecological levels of influence . This may be especially the case for hierarchical institutions such as school systems, where decisions are made at more than one level.
Finally, another notable finding from the current study is that only one intervention activity was implemented to engage parents in the support of healthy lifestyles for their children. Given growing evidence that family support is essential to the success of school-based health promotion activities ,, the lack of family oriented intervention activities would appear to be a shortcoming of the INSP program. Based on the existing literature -, it can be speculated that the inclusion of additional intervention activities to engage families in obesity prevention efforts might have resulted in a larger impact in child PA and nutrition behavior change. Future research should investigate the feasibility of engaging families to support school-based health promotion efforts in Mexico.
An analysis of SCT behavioral constructs showed no statistical differences in their use within and between intervention domains. However, some tendencies and variations were observed. The theoretical construct most often used by intervention staff to influence the different ecological levels was reciprocal determinism (RD), also referred to as the environmental construct ,,,,. Given that the aim of the intervention program was to improve the environment, the frequent use of this construct is not surprising. Intervention activities flowing from the RD construct intended to influence existing attitudes about the school food and PA environment, related policies, practices, and services. For instance, HP workers supported the improvement of the food environment by influencing the reduction of the sale of sweets during recess and by facilitating conditions for water consumption by organizing potable water deliveries to schools. HP workers improved the PA environment and modeled schoolyard use by organizing structured games during recess. The use of the RD construct to influence the food and PA environment has been shown to be effective for creating a supportive milieu for healthy behaviors in children .
Patterns of use of other SCT constructs appeared to vary slightly, although not significantly, across and within domains. Within the Nutrition domain, BC was used as frequently as RD. In the Nutrition domain, the BC construct was used in workshops to influence food and nutrition knowledge and to model skills that children need to carry out healthy eating behaviors. In addition to providing verbal prompts to intervention targets, HP workers used BC-oriented print materials to engage food vendors, key authorities and children as a way to correct misconceptions about healthy eating and to provide new options for improving food related practices at school. The frequent use of RD within the Nutrition domain was likely related to program aims to influence food availability at schools. The high use of these two constructs in the Nutrition domain is congruent with the INSP impact results that show significant improvements in the food environment and enhanced healthy food intake by children in intervention schools . These results are consistent with other studies that have effectively used these constructs to improve food intake practices in children -.
Within the PA domain, the use of SCT constructs appeared to be balanced. This may be because PE sessions were already occurring in schools on a regular basis and were being delivered by professionally trained experts (i.e., PE specialists). Thus, the implementation of activities within the PA domain may have required less guidance by INSP staff as compared to activities within the Nutrition domain, which required that food vendors be provided with guidance and information to learn new practices about how to create a healthy food environment. This finding is comparable to other research showing less use of theoretical constructs in the implementation of PA in interventions for children relative to other domains and other populations .
Overall, SCT behavioral constructs were used less frequently in the PA domain relative to the Nutrition domain, although this difference was not significant. This finding is consistent with our ecological mapping showing that the majority of PA intervention strategies were targeted to the Public Education Secretariat (POL) and aimed to modify the political environment rather than to change the behaviors of individuals (e.g., students and school staff). It may be that constructs from theories other than SCT (e.g., relevant to the policy process) are used to influence policy.
Our findings provide new insight into a promising combination of strategies and theoretical constructs that can be used to implement an ecologically founded school-based obesity prevention program. A limitation of this study was our inability to gauge the magnitude of effort devoted to various intervention activities. The IAP does not allow for the assessment of the “dose” of each intervention. Thus, the development and implementation of educational materials to enhance healthy eating might have been captured as one intervention activity (and coded as a single strategy), when in fact these same materials might have been used daily by teachers. In addition, the checklist used to assess SCT construct use only captured four of the SCT constructs; a more exhaustive list may have yielded a different theoretical picture.