Study design
This was an open-label, parallel group, school-based 6-month cRCT with individual middle schools as the unit of allocation and individual students as the unit of analysis. The design conformed with the Consolidated Standards of Reporting Trials (CONSORT) guidelines. Before recruitment, we registered the study with UMIN-CTR and uploaded the study protocol [28].
Study setting
Voluntary middle schools in Kumamoto and Miyazaki Prefecture located in the south area in Japan.
Participants
School was the unit for cluster randomisation. Participants were all eligible adolescents of these middle schools in school years 1 and 2 (age 12–14 years). Therefore, participants having both healthy and unhealthy attitudes and behaviour were mixed. The participants were registered between October 2018 and December 2019. Parents/guardians in the SPRAT group were asked to join the programme to help their children.
JW and MW approached the Municipal Board of Education and the Principal Meetings in each area and got approval. Subsequently, JW identified volunteer middle schools and asked permission to enroll participants by telephone or a visit. Trial content was explained and efforts to obtain consent for schools’ cooperation were made. Several telephone meetings as well as face-to-face meetings with responsible persons and persons in charge of the Municipal Board of Education and the Meetings of Principals of Schools in each area were conducted. Further, we conducted several information sessions for parents/guardians in each school, for instance.
A minimum of 40 students was set as a requirement. Criteria for student participation are shown the study protocol and in Supplementary file 1. Written informed consent was obtained from all participants, including both students and their parents/guardians. As for participant recruitment and participant assignment, see the study protocol [28].
Randomisation, allocation concealment and blinding
Randomisation was conducted using a permuted-block technique using a randomisation list. Schools had been allocated manually by random numbers obtained using a personal computer (PC). KY generated the allocation sequences and JW enrolled schools and assigned participants to interventions. Due to the nature of the treatment, participants could not be blinded to the type of dietary and lifestyle education. However, data management team members, except for the project coordinator and research assistants, were blinded to the group allocations. To minimise the risk of bias, we developed strict protocols for follow-up assessment procedures (in Japanese) and trained research assistants to adhere to these protocols.
Interventions
The follow-up period was 6 months from randomisation (baseline). A 6-month intervention using SPRAT was conducted. The control was a usual school programme. This study was conducted according to the guidelines of the Declaration of Helsinki. The trial leader, a registered dietitian, provided training for the registered dietitians and learning support assistants based on the ‘registered dietitian training programme for SPRAT at the study management centres in Miyazaki and Kumamoto (around a 5-hour period). A summary of the intervention is as follows and details of the intervention are shown elsewhere [28].
SPRAT programme group (intervention group)
SPRAT is a dietary and lifestyle education programme of 6 monthly sessions involving parents/guardians to reduce SPS and to improve students’ health behaviour and dietary and lifestyle habits including physical activity and increase enjoyment of school life. Students are encouraged to increase intake of staple foods (rice, bread, etc.), main dishes (soy, fish, eggs and meat) and vegetables, particularly at breakfast and are assisted in developing skills in food selection by advice and education.
The FFQW82 questionnaire (at baseline and 6 months) [22], student’s SPS questionnaire (SPQ), and student’s Lifestyle Questionnaire (LQ) (at baseline and 2, 4, and 6 months) were administered. Participation self-check sheets (PPS) were provided to parents/guardians in the intervention group. Monthly sessions were conducted using a booklet ‘Smart Life’ which was delivered by registered dietitians and directed toward schoolteachers, students, and parents/guardians. Parent manuals (PPM) (every month) and a newsletter (4 times) were provided to parents/guardians in the SPRAT group. Thus, SPRAT also aimed to improve students’ health behavior and lifestyle habits as summarized in Table 4. Parents/guardians in the SPRAT group were required to participate regarding several issues according to the programme strategies. For instance, five homework assignments for students as well as parent manuals were distributed and practices to improve students’ lifestyle were conducted with their parents/guardians during the study period.
Usual programme (control group)
Students in the control group participated in the schools’ usual programme in health education classes. That programme was composed of existing health curriculum about diet and/or exercise that was routinely taught at each participating school. The registered dietitian, learning support assistants, or the teacher assessed dietary intake using the FFQW82. Teachers in junior high schools distributed and collected questionnaires for parents/guardians as shown in Fig. 1.
Study hypothesis
The hypothesis was that students who participated in the SPRAT group would have a greater decrease in their mean SPS score and more appropriate dietary intake than students in the control group after 6 months.
Outcome measures and background variables
Primary outcome
The primary outcome was the SPS score at 6 months from baseline measured by responses to the SPQ, which consists of questions on nine symptoms such as fatigue, headache, lassitude, irritation, lack of concentration, lack of motivation, poor ability to wake up in the morning, upset stomach and stiff shoulder. Using a Likert scale (5 levels: ‘0 = never’ to ‘4 = always’), we calculated the SPS score (0–36 points) as the sum of the category values for the nine items. The validity and reliability of the SPS score were shown elsewhere [21]. The primary endpoint was the difference in the change from baseline at 6 months (±1 month) between the two groups.
Secondary outcomes
Secondary outcomes were the SPS score at 2 and 4 months from baseline, dietary and lifestyle factors (sleeping habits, eating habits, enjoyable school life, and physical activity), and dietary intake (11 energy intakes by food group and 9 nutrients) assessed by the FFQW82 [22] and body mass index (BMI) at 6 months from baseline. A school nurse measured BMI. The dietary intake standard [29] states that the target BMI range (≥18 years old) can be applied to investigate the intake and output in the energy balance in adults but not children. For junior high school students, a growth curve indicating the level of obesity based on the School Health Law is used. Therefore, in this program, SPRAT used this growth curve to perform an exercise to plot the current height and weight for the purpose of students recognizing their own physique and maintaining an appropriate physique.
On the other hand, the potential significance of this study is to reduce mental and physical health problems in the study participants 12 to 13 years old- in the transitional period from adolescence to adult, and to track this o risk to the next generation [30]. This also helps prevent health problems in the future.
For example, regarding the health risks to the next generation, according to the National Health and Nutrition Examination Survey [31], the percentages of obese people in their 20s was 26.8% for males and 5.7% for females and the percentages of the underweight individuals were 9.1% for males and 21.7% for females. In this study, we thought that it was meaningful to get BMI change as a reference value for changes in the body. Therefore, the BMI endpoint was used.
Sample size
The required sample size was determined based on information needed to detect a difference in the primary outcome with a significance level of 5% and a power of 80%, under the assumptions of 40 students per cluster (same sample size for each cluster), an effect size (for severe SPS) of 0.3, and an intraclass correlation coefficient of 0.02. The effect size was estimated from our experience in a former study [21]. In total, participation by 28 schools (14 schools in each arm) was needed.
Statistical analysis
We conducted the statistical analyses following the New Guidelines for Statistical Reporting as previously described [32].
We used descriptive statistics to assess the balance between the trial arms at baseline. Continuous variables were summarized as mean, standard error (SE), and intraclass correlation coefficient (ICC) estimated using nested analysis of variance (ANOVA) considering a cRCT study design. The difference between the two groups was examined based on an intention-to-treat (ITT) principle with the full analysis set (FAS). For the primary outcome, we defined the FAS as full data for baseline and endpoint (at 6 months) SPS scores. Change from baseline (CFB) at 6 months was the primary endpoint. Dietary and lifestyle factors were categorised as dichotomous variables. A general linear random-effects mixed model employing the restricted maximum likelihood method was used to analyse continuous variables. For dichotomous secondary outcomes, random effects logistic models (subject-specific models) were used, and associations were shown as odds ratios (ORs) and 95% confidence intervals (CI). Primary outcome measure was used to examine the effects of the intervention by a crude model (model 1), a model adjusted for baseline values (model 2), and a multivariate-adjusted model (adjusted for baseline, sex, age, and school type [private or public].) (model 3).
As for the secondary outcomes such as dietary intake, on the other hand, the treatment effect was estimated as a baseline*treatment interaction in those models that include a baseline*treatment interaction term. We expected the interaction term baseline*treatment to be statistically significant. Since the study participants were both underweight and overweight, to accept the validity of the interaction between the treatment and dietary intake it is reasonable to consider baseline intake. Participants were encouraged to consume either lesser or greater amounts in consideration of estimates based on baseline intake.
For missing values, the analyses by the imputation of missing data were performed using the last observation carried forward method (LOCF) and a multiple imputation method (MI) using chained equations under the assumption of missing at random [33].
All tests for significance were conducted using a two-sided approach with a 5% significance level. P values were shown only for the primary analyses and 95% CI was shown for the secondary analyses. All statistical analyses were performed using SAS V.9.4 for Windows (SAS Institute).
Data management and monitoring
Personal information was coded and anonymised. All data and documents related to this study were managed as planned [28]. The ethics committee will receive a report at the conclusion of the study and after the final results.
Protocol amendments
No amendments to the protocol or of the follow-up of adverse events.
Ethics and dissemination
The All tests for significance were conducted to the Declaration of Helsinki and the ethical guidelines for medical research on humans. This study was approved by the Ethical Review Board of the Medical Ethical Committee of Minami Kyushu University in 2017 (Number 137).
Patient and public involvement
There was no patient or public involvement.