Our study shows that a nutrition program, delivered and taught by in-service teachers trained in nutrition, is effective in lowering the increase in the incidence of overweight among schoolchildren. Intervened children have a significantly lower increase in BMI and a lower proportion of children become overweight. This is particularly important as it provides further evidence to the WHO recommendations for schools to include both dietary and physical activity components in the curriculum taught by trained teachers [24
Our study has some weaknesses that should be mentioned. We did not explore differences between schools selected and not selected given human and materials constraints, although the schools were from the same geographical area and to the best of our knowledge, no data is available reporting sociodemographic and income significant differences between schools from this particular area. In addition, to the best of our knowledge there were no other “anti-obesity” programs going on in intervention and control schools.
It is possible that our sample size was not large enough to detect other significant differences than those reported. Furthermore, we failed to obtain identically equivalent groups after randomization, namely in parents education level and children’s height, mainly because we randomized by school and not by subjects, aiming to avoid cross contamination between intervention and control groups. Nevertheless, these differences were taken in account in all of the statistical models. Also, physical activity levels were obtained upon self-reported data making recall bias and overestimation possible. However, the questionnaire was validated for Portuguese adolescents [19
] and we have no reason to assume these biases would affect groups differently. Future studies would be improved by using accelerometry to better classify physical activity.
On the other hand, the present study has important strengths that also should be acknowledged. First, to the best of our knowledge this is the first study that included the program in the progression of teaching career. This probably allowed teachers to increase their motivation in the delivery of the intervention. It would be desirable that other similar programs could be recognized on the career progression of teachers in Portugal and other countries in order to engage and reward teachers for their efforts. Second, the long term teachers’ in-service training and the subsequent network developed between themselves, researchers and children. We know that in Portugal students of education courses do not have health promotion in their training curricula, neither consider its changes towards an increase in health content important [25
]. In addition, teachers are described as a group that is not able to devote much time and energy to provide obesity prevention interventions as dedicated interventionists [8
]. Being aware of this need and that long-term programs are more effective than those of short duration [8
] we promoted a six months duration training program expecting teachers to become also nutrition educators. We believe this period allowed teachers to recognize just how important healthy eating habits and physical activity are. Third, our approach was to standardize recommendations to teachers, allowing them enough flexibility to create interactive interventions and pedagogic instruments to be used with children. The teachers were assessed after their intervention for material and lessons developed, and it was found that the information incorporated in their lessons and/or the quality of teaching materials did not vary between them. This is in line with the development of “scholarship of teaching and learning” [15
] and is contrary to previous school-based interventions that have used tight controls to ensure uniform implementation but required frequent staff training and ongoing supports [26
]. Fourth, the intervention mapping developed to identify the real needs in knowledge of teachers and children helped to make the translation of objectives to change strategies. Finally, we estimated dietary intake using the 24 h dietary recall method, which is the most commonly used method in Europe [27
]. In addition, we assessed the effectiveness of the program using BMI values expressed as z-scores avoiding age and gender effects. It would be desirable that children from the present study could be followed to see whether the results still hold through life cycle, but no resources exist to further support this study.
We observe that most children are in the sedentary or low activity group. These results are in line with previous pediatric studies [28
We further observe a very high prevalence of overweight and obesity in our subjects, emphasizing the epidemic state of overweight/obesity in Portuguese children and adolescents [29
]. Actually, this is also a global problem, as confirmed by other studies where the prevalence of childhood overweight continues to increase [1
]. Therefore, during the last years several programs have been developed aiming to reduce childhood overweight and obesity incidence and prevalence [10
]. However, the results of these studies also indicate the resiliency of body weight because only few of them significantly reduced average body weight or adiposity in intervention groups of girls or boys [33
]. In this context, this study yielded encouraging results as it reflects the positive effect of an educational program on anthropometric status, little seen in Europe [35
]. Nevertheless, we should be cautious in the interpretation of our results, given the small sample size and the short follow-up in our study. Similar results as in the present study were found by James et al.
], Borys et al.
] and Panunzio [11
]. The intervention duration of the studies that reported effect on anthropometry varied from thirty six weeks [11
] to five years [12
], although the latter did not have a positive effect on anthropometry [12
]. Besides the effect on overweight, the intervention had no effect on the prevalence, incidence and remission of obesity, and further studies should find out probable reasons for inefficacy on obesity. Progression to or remission from the upper end of BMI distribution may be more likely to result from targeted and/or clinic-based programs than from primary prevention such this program. Moreover, the program was tailored to the specific study population and may not be as effective in a different group with different needs. The positive results may reflect in part aspects of the Portuguese educational system and motivation of teachers may be less in countries where the in-service training is not included in the progression of teaching career.