Study design
A randomized controlled trial was designed through the application of a random cluster sampling method. The research was carried out in metropolitan districts of Hanoi, Vietnam. Among the four metropolitan districts, two districts were randomly selected. Four schools were selected among 29 public secondary schools in two districts. In each school, five grade six classes were selected. The schools were assigned randomly to the intervention and control groups by independent statisticians from Hanoi Medical University. All the students and their parent(s) in the selected classes were invited to participate in the study. In total, 936 secondary school students aged 11–12 years were selected. At baseline (recruitment periods from January 05 to January 16, 2014), 821 students participated. After two years of follow-up (recruitment periods from January 14 to January 28, 2016), 733 students remained (Fig. 1) because some students transferred to different schools, some were absent on the day of blood collection, and some withdrew from the study. The parents and students were required to provide written consent to participate in the study.
Intervention program
The intervention program against OW/OB was carried out at two schools among four selected schools, with a chief focus on providing preventive solutions for OW/OB and encouraging healthy lifestyles. The intervention program was based on two sources: the recommendation and report of the School Health Guideline by the United States Centers for Disease Control (CDC), which includes 6 groups of intervention activities with the central value being nutritional education [28]. We applied these guidelines for our main activities along with the Vietnam National Institute of Nutrition Guidelines on Prevention and Intervention for Overweight and Obesity [29]. The intervention program also invited cooperation and comments from parents and teachers to make appropriate applications in caloric counts and physical activities that suit Vietnamese students to optimize intervention effectiveness. During the first month of the study, we collected body measurements and blood samples for testing and assessed lunch menus. Immediately afterward, all activities were carried out simultaneously to the end of the study. After conducting a baseline survey, intervention activities were carried out, including modifications of the students’ lunch menus. The modifications of students’ lunch menus at two intervention schools are shown in the following books from the Vietnam National Institute of Nutrition [30, 31].
Activity 1: School policies on healthy diet and physical activity
Based on the existing lunch menu offered at the intervention schools, the research team made adjustments for each daily lunch meal every school week (5 days/week). We assessed the actual lunch menus of schools and compared possible caloric intake with recommendations from the Vietnam National Institute of Nutrition to provide modifications to lunch menus by age [30, 31]. To provide appropriate and timely adjustments in diet and physical exercise, all the students had their height and weight measured and monitored monthly. Since 11- to 12-year-old students are experiencing puberty, their body weight and height can change very rapidly; therefore, there is a need for monthly height and weight measurements to provide appropriate dietary advice for each individual student, as well as for assessing intervention effectiveness. Moreover, frequent measurements of height and weight metrics will help students pay attention to and maintain their body weight. Some researchers have also shown that frequently measuring weight also helps prevent OW/OB [32].
Activity 2: Comprehensive physical activity program
A comprehensive physical activity program was developed based on recommendations from the United States National Association for Sport and Physical Education (A Statement of Guidelines for Children Aged 5–12. 2004, Reston, VA: National Association for Sport and Physical Education) and the Australian Government Department of Health [33], in which children and teenagers participate in a minimum of 60 min of physical exercise daily at home or at school, equivalent to 10,000–11,700 steps/day for teenagers aged 12–17 years old [34]. To help students and parents assess their own dietary intake and physical activity frequency, each student was supplied with one Dretec BS-150 WT electronic weight measuring scale with a BMI calculation function (Dretec, Saitama, Japan) and one OMRON pedometer HJ-109 with time, step-counting and distance-measuring functions. In classes and seminars, students were instructed on how to optimize physical exercise to reach their expected weight goals.
Activity 3: Nutritional education program
The design of the nutritional education program included 60 h (3 h/class/month × 10 months × 2 years) for all intervention group students. To improve the intervention education for the target subjects, experts customized a separate program for students with OW/OB and their parents. The training contents of each session are never repeated in the next session throughout the two years of the study. The goal was to enable all students to access and practice their knowledge, thereby improving their nutritional knowledge and habits.
Activity 4: Training for school staff members
While administering the nutritional education program, the research group provided training on physical exercise and nutrition to 3 groups of school staff members: head teachers, physical education teachers, and social coordinating staff. This program aimed to help school staff establish a good base of knowledge, thereby enabling them to directly participate in instructing students to change their understanding and behavior toward healthy eating and frequent physical exercise.
Activity 5: Social promotion program
Posters on healthy eating and physical exercise were displayed in all classes, school entrances, and functional rooms. Each student was also provided one copy of the booklet on healthy eating and physical exercise. Club activities by students were held once each month to deliver knowledge and practices on having a healthy lifestyle. A competition of knowledge on healthy eating was held in the second year of the intervention to further promote the exploration and practice of healthy dieting. These activities aimed to improve the promotion of good practices for healthy diets among students.
Activity 6: Family consultations
Family factors can contribute to the increase in OW/OB. The research team and nutritional experts consulted students’ families in teacher-parent meetings, group discussions, and consultations with parents of students with OW/OB every quarter. The research team provided phone consultations to each parent to monitor and assess the nutritional state of each student in detail, in addition to providing information and answering questions from parents. The phone numbers of 5 consultants were provided to parents of students and students with OW/OB to help consultants contact and answer questions from all students and parents. Each student and parent received a consultation at least once a month.
Data collection
Data from all categories were collected at baseline and at the 2-year follow-up.
Anthropometric data collection
All measurements were performed by two groups, each consisting of 2 doctors and 3 nutrition experts. An electronic Dretec BS-150 WT (Dretec, Saitama, Japan) scale with a standard error of ± 100 g was used for performing the measurements [35].
Body mass index (BMI)-for-age was calculated using the formula BMI = m/h2 (m: body mass in kilograms, and h: body height in meters), and BMI was categorized based on the z score table for boys and girls following the WHO guidelines for BMI in children aged between 5 and 19 years [36]. The BMI z scores were categorized into segments according to WHO criteria for the 5- to 19-year-old age group [36]. (Thinness: < -2 SD, Normal: -2 SD ≤ x ≤ 1 SD, Overweight: > + 1 SD, Obesity: > + 2 SD). The software WHO Anthro ver 3.2.2 (free of charge) was used to calculate the Z score [36].
Data on nutritional and physical exercise habits included day physically active total at least 60 m/day, spend time for video, game (hour/day), time eat green salad (time(s)/day), meals in an average day (time(s)/day), and main meals (time(s)/day) were collected via survey questionnaires. Students completed the surveys in their classes after the surveyors explained the questionnaire in detail.
Biochemical sample collection and analysis
Biochemical sample collection was performed at the four schools over eight days at baseline and at the 2-year follow-up. Blood samples were collected by pediatric nurses with support from Bach Mai Hospital (Vietnam). Before the day of blood collection, the students were instructed to not eat anything at least 10 h before blood samples were collected. Blood samples were stored in dry ice containers and transported that same day from the site of the study to Bach Mai Hospital (a distance of approximately 30 min by car). Blood samples were collected to investigate glucose (mmol/L), HbA1c (%), and insulin (μU/mL) for analysis by the Biochemistry Department of Bach Mai Hospital.
Statistical analysis
Data analysis was performed using SPSS Statistics desktop version 21.0 (IBM, Armonk, NY, USA). We used a Mann Whitney test with an α = 0.05 level of significance to test the associations of different risk factors.
To assess the primary outcome of incidence and the secondary outcomes of prevalence and remission, a generalized estimating equation (GEE) was used to model a binary outcome. For the secondary outcome, a GEE was also used to model a Poisson distribution for continuous and count variables (e.g., BMI, BMI z score, biochemical metrics, average hours of inactivity) in 2 intervention groups and control groups adjusted for age (in months) and sex to determine the odds ratios (ORs) and 95% CIs.
Reducing the incidence of OW/OB after 2 years was the primary objective of the intervention program. The incidence of OW was calculated as the percentage of normal weight individuals who became OW after the intervention relative to the total number of normal weight individuals at baseline. The incidence of OB was calculated as the percentage of OB cases after the intervention over the total number of OW cases at baseline.
The secondary objective was to reduce the prevalence of OW/OB after the intervention. The prevalence of OW/OB was calculated as the percentage of OW/OB cases relative to the total number of students who participated in this study at baseline and at the 2-year follow-up.
The OW remission proportion was calculated as the percentage of participants who were OW at baseline and achieved a normal weight at the 2-year follow-up over the total number of OW patients at baseline. Similarly, the remission rate of OB was calculated as the percentage of participants who were OB at baseline and who were OW or normal at the 2-year follow-up among the total number of OB patients at baseline.
The cutoffs for OW/OB were > + 1 SD for overweight and > + 2 SD for obesity [36]. Additional secondary analyses included BMI z scores, biochemical data, physical activity and daily eating habits.
