The training focuses on promoting a healthy lifestyle via activities designed to reduce obesity, and to prepare the HPA to implement these educational interventions in schools.
The intervention approach combines constructs from education- and nutrition-based evidence.
Two courses are proposed:
Course #1: Methodology for the promotion of health in schools
Description of the course
Bases of health education and health behaviour
Theory and program planning: A study of determinants of health behaviour, factors influencing health behaviour, health behaviour theories and application of methodology are examined.
Health education curriculum and instruction: Research methods in health education; principles of evidence-based nutrition.
12 lectures of 1 h duration each, delivered over 15 weeks per academic year. Eight topics in nutrition are chosen for scientific evidence to improve consumption of some foods. Increased physical activity and healthy habits such as teeth-brushing and hand-washing are highlighted.
The objectives are:
Healthy lifestyle. Taste (knowledge of previously-unknown food items)
Candies and pastry vs. nuts
Healthy habits: timetable (home meals preferably) and physical activity
Course #2. Interdisciplinary implementation of health education in schools
This course provides a careful examination of strategies of design, implementation, and health program evaluations, including training and standardisation of each activity as well as the performance of these activities in schools.
In the activity training, the primary school teachers are included in the jury to assess the performance of the HPA in the primary schools.
4 lectures of 1 h duration each; in 12 h for training and standardisation and 12 activities (1 h/activity/classroom) over 15 weeks per academic year
The school-based lifestyle modification program is designed as an interdisciplinary health promotion program performed by HPA.
All strategies are focused on children aged between 7 and 8 years. The program targets the whole school community including parents, pupils, staff and teachers within the school environment. The schools for the program needed to be representative of the child population. We offered the program to all schools whether public (funded by the government and termed "charter" schools) or private. To maintain independence between intervention and control schools, we selected schools from Reus (a town with about 100,000 citizens) to represent the intervention group, with Cambrils, Salou and Vilaseca (3 towns on the outskirts of Reus with about 70,000 citizens in total) to represent the control group. There are very good communications between schools within each town and, hence, to avoid cross influence between control schools that were masked with respect to intervention, the towns themselves were the units for randomisation.
The coordinating Centre developed a randomization scheme in which the schools in Reus (designated as group A) and the schools in the three others towns of Cambrils, Salou and Vilaseca (designated as group B) had similar pupil populations.
Randomisation defined A group as the intervention and B as the control group.
In our area of Spain, each classroom contains approximately 25 children (at most 27 children and, occasionally, 20-22 children). When the numbers of children exceed the state-recommended level, a new classroom at the same education level is opened in the school. Thus, there could be 2 or more classrooms for any specific level; some schools may even have 3 classrooms per level with 25-27 children per classroom. All classes are co-educational
Participating intervention institutions consist of 24 schools involving 36 classrooms (97% public or state funded "charter" schools and 3% private) involving at least 700 pupils in Reus. Participating control institutions consist of 14 schools involving 39 classrooms (96% public or state funded "charter" and 4% private) involving at least 700 pupils in the 3 surrounding towns of Cambrils, Salou and Vilaseca. The ethnic origin of control and intervention pupils are: 83.2% of Western European descent (mainly from Catalunya) and 16.8% are non- European; 7.8% from Latin America, 5% of North-African Arab descent, 3% recent immigrants from Eastern Europe (mainly Rumanian) and from other countries from the Asian Far East (such as China).
For a school to be included in the study, at least 50% of the children in the classrooms needed to have volunteered to participate. In the intervention group, all children of the selected classroom are exposed to the intervention. For logistics reasons, the first scholastic group was enrolled in 2006 and followed-up for 3 years (2006-2009), and a second scholastic group was enrolled in 2007 and also followed-up for 3 years. Thus the final measures are performed in the year 2010. The data are collected on all the children, but only the data from individuals (and their parents) who provided informed consent are included in the final analyses.
Name, gender, date and place of birth are recorded at the start of the program, while weight, height, body mass index (BMI) and waist circumference variables (identified set of anthropometric measures) are recorded in each of the 3 years of the study. The measurements were performed in May of the years 2006, 2007, 2008, 2009, and 2010.
Body weight was measured to the nearest 0.05 kg using a standard beam balance (Tanita TBF-300 Body Composition Analyzer, Brooklyn NY, USA). The set of anthropometric measurements was performed three times each academic year. The mean values were used in all subsequent statistical analyses.
To assess intra-observer variability, the anthropometric measurements were repeated in 20 children (10 boys and 10 girls). To assess inter-observer variability, all 5 observers conducted the 5 anthropometric measurements in 10 children. The standardisation of observers was performed in each year of the study .
If the child is lost to follow-up (the child is relocated to a different school, or the parents move to a different area) we are able to contact the child/parents via the name and date of birth; data that are held in the records of the local education authority. If the child moves to a school that is participating in the study, he/she is identified and followed-up within the new school. If, instead, the children move to another city outside the study area, they are lost to the study and are not replaced. We anticipate a loss of about 20% of pupils which, by chance, would be similar in intervention and control schools. The assumption would be that all losses are, also, by chance and the statistical methods used are robust enough to accommodate for randomly-produced missing values.
Questionnaires regarding eating habits (Krece-Plus) developed by Serra-Majem et al  and physical activity as well as the level of parental education and their habits  are filled-in by the parents at the start and conclusion of the study.
We estimated that with a sample of 700 pupils per group, the study would have 83.5% power to detect a difference of 5 percentage points between the intervention and control schools (9%-14%), with respect to the primary outcome (prevalence of obesity), setting the bilateral level of statistical significance at 5%.
Descriptive data are presented as means ±SD (95%CI) or percentages. Generalised linear mixed models were used to analyse differences between the intervention and control pupils with respect to the primary outcome.
Measurements were conducted at baseline when the pupils are 2nd-3rd graders (6 or 7 year olds) and at the end of the study when they are 4th-5th graders (9 or 10 year olds).
Primary outcomes include obesity (BMI ≥95th percentile) and overweight (BMI ≥85th percentile) based on the 1988 BMI tables of Hernandez . We analysed obesity and overweight and a measure of thinness according to the Cole criteria [20, 21] as well as other measures of adiposity such as BMI z score and waist circumference. The numbers of subjects having a particular dietary habit are expressed as percentages of the total number of individuals being evaluated.