Study design, study population and inclusion criteria
ScIM is a multicenter, school-based, three-arm cluster randomized controlled trial (RCT) recruiting ninth graders from lower secondary schools in Norway. The intervention was implemented during the school year 2017-18 and conducted by four collaborating study partners (Norwegian School of Sport Sciences, Western Norway University of Applied Sciences, University of Agder, and University of Stavanger). A random sample of lower secondary schools located in municipalities in the geographical area near the four study partners were included. In the selection of schools, population density was considered; consequently, a different number of schools were included from each of the four geographical areas.
Private schools, designated special schools and schools with less than 25 adolescents in ninth grade were excluded from the study sample. Schools that already worked systematically with PA as an integrated part of the school day were also excluded. The recruitment of schools started in January 2017. First, County Governors were informed about the project by information posted on their web page. We then contacted school owners in the different municipalities to get permission to contact the principals of the schools that should be invited to participate. If the school owner approved, the letter of invitation was sent to the school’s principal. This was followed by a phone call from the research team to give the principal and school management further information about the project and answering any questions they might have. A visit to the school was then scheduled to present the project in its entirety. Schools that wanted to participate in the study returned a signed consent form. Information meetings was held with all teachers at the intervention schools, and teachers involved in implementation of the intervention received further information regarding the intervention components.
A total of 103 schools were invited to participate. Of these, 30 schools returned the signed consent form and were included. A neutral third party randomly assigned the schools (clusters) to one of the two intervention arms or the control arm. The randomization was stratified by study center to assure that there were schools in all three study arms at all four study locations. Ten schools were randomized to the intervention arm called “Physical Active Learning” (PAL), ten schools to the intervention arm called “Don't worry, Be happy” (DWBH), and ten schools were randomized to the control arm. One of the included schools in the control arm withdrew from the study after the randomization procedure but prior to baseline testing, consequently, nine schools were left in the control arm. Blinding of participants and schools were not possible due to the contents of the intervention.
The flow diagram is presented in Figure 1. All adolescents in ninth grade (14-year-olds) within the 29 participating schools were eligible to participate (n=2,733). Adolescents and their parents received detailed written information about ScIM, and they were invited to a school meeting outlining the study. A total of 2,084 adolescents returned the consent form signed by a parent/guardian, yielding a participation rate of 76 %. Three months into the intervention, one of the schools in the DWBH-intervention arm withdrew from the study due to practical reasons.
(Insert figure 1 here)
The ScIM intervention
All schools participating in ScIM had two or three scheduled mandatory PE-lessons per week as part of their curriculum. Schools in the two intervention arms added one hour of PA and one hour of PE to the class schedule per week. This was done by redistributing five percent of time from other subjects in the curriculum to PA (corresponds to 60 minutes of PA per week), while the last 60 minutes were added to the classes’ schedule each week. The schools received financial resources from The Norwegian Directorate for Education and Training to account for the increased expenses for the schools. The amount received by each lower secondary school was based on the number of students attending the school.
To optimize adherence, the intervention components were established as part of the mandatory school curriculum for all adolescents attending the schools in both intervention arms. The two intervention arms and the control arm are described below.
The PAL-intervention arm included three additional components of PA per week:
- PE (60 minutes): One weekly additional lesson of PE. The activities taught in this lesson were in line with the curriculum, and the PE teacher planned and taught the class.
- Physical active learning (30 minutes): During this session play-based activities were integrated in regular subjects (i.e. maths, English, Norwegian). The aim was to increase the students' PA level while improving their academic performance. The PAL component should preferably be carried out on days without PE-lessons. The classroom teacher of the current subject planned and taught the session.
- PA session (30 minutes): This session included a variety of activities preferably with at least moderate intensity, with an emphasis of activities that were enjoyable, giving the adolescents a feeling of pleasure and well-being. The PA sessions often took place outdoors and without the adolescent changing clothes. A classroom teacher or a PE teacher planned and led the activity.
The DWBH-intervention arm included two extra PA components each week:
- Be happy-class (60 minutes): In this session the adolescents performed activities in self-organized groups of at least three students, developed according to the individual’s activity preferences (groups were made of individuals across regular classes). The chosen activities could be traditional sports, lifestyle sports, dancing, out-door recreation, drama etc. and was performed inside or outside the school. PE-teachers (often several teachers based on the size of the class) were present to support the students if necessary.
- Don’t worry class (60 minutes): This was a PE lesson, and was conducted in the adolescents' regular classes, however, it was organized and led by the students. The students practiced their Be Happy-activities, or they introduced their class peers to their Be Happy-activity. A PE-teacher was present to support the students if necessary.
During the first week, the students were introduced to the key features of the intervention. The students were told to organize Be Happy groups based on activity preferences. When the groups were established, every group had to hand in a document that showed:
- three aims to govern the group (e.g. we do street dance and will develop each other, be friends and have fun with dance),
- one major aim for the next six months (e.g. we will develop enough knowledge about yoga to be able to teach yoga in our PE class),
- a management structure (e.g. leadership will switch according to a plan),
- a strategy for impending conflicts (e.g. we must dare to address what may be difficult. If we do not fix it, we will ask the teacher for help), and
- routines for registration of attendance (e.g. we will take pictures and send to the teacher).
A teacher had to approve the document before the group could start their activities.
The schools in the control arm continued current practice and were asked not to make changes to increase PA or PE during the intervention period. In order to prevent changes, the intervention was offered to the schools in the control arm when the intervention period was finished.
The ScIM-intervention is based on a socioecological framework, that understands the complex interplay between the many personal and environmental influences on behavior . In short, the socioecological approach recognizes proximal individual and social factors and several distal determinants for behavior change such as individual factors (e.g. self-esteem, attitudes), social relationships (e.g. family, friends), the physical environment (e.g. schools, walkability), and policies (e.g. curriculum, schedules) as different levels of impact. Change at all levels is necessary to achieve lasting positive change in health behavior. In terms of individual and social factors, the PAL-intervention builds on Harter's competence motivation theory , Bandura's social-cognitive theory  and Ryan & Deci's self-determination theory . The theoretical rationale is thought to function as a mediating structure between intervention strategies and outcomes.
The primary focus of the DWBH-intervention arm was to promote friendship through PA and vice versa. The model was anchored to an integrative relational developmental systems (RDS) approach to human development , theories on Positive Youth Development (PYD)  and the concept of Positive Movement Experiences (PME) .
According to RDS-theories on human development, different subsystems of variables in humans are related to each other at different organizational levels (e.g. microbiology, cognitive function, family, friends and culture). Thus, when intervening the everyday life of adolescents these subsystems should be seen as interpenetrating dynamic dimensions and a PA intervention should be seen as mental and social as well as physical. In addition, diversity among adolescents involved represents a significant issue meaning that effects of a PA intervention will rely on the adolescents’ relational experience of meaning and relevance in the new movement context (intervention) and not on the activity dose per se. An intervention hits as many developmental trajectories as there are participants and every participant will experience the interaction going on in the intervention context differently. Therefore, such interventions cannot rely on a "one size fits all" approach.
According to the RDS perspective school-based interventions should promote individual-context relations that are mutually beneficial for everyone involved. According to Agans and colleagues  it is possible for all individuals to have PME if the characteristics of the individual and the characteristics of the context are aligned in such a way as to produce mutually beneficial relations among all students and teachers involved. From the perspective of PYD no one but the adolescents are capable to build these mutually beneficial person-context relations. According to theories on PYD all youth do have strengths, and these strengths need to be identified and acknowledged to promote healthy development. This view of adolescents regards them as resources to be developed, not problems to be managed .
All participants were tested in an identical set of outcome measures at baseline, and approximately 12 months after the baseline measures, when the participants were at the end of ninth grade and during the last phase of the intervention. Physical activity was also assessed midpoint during the intervention. Data were collected at the participating schools in a gymnasium or in the classroom. All research personnel were trained by members of the research team. Figure 2 provides an overview of the study time line and the data collection intervals.
(Insert figure 2)
Physical activity level (primary outcome)
PA was assessed by triaxial accelerometry (ActiGraph GT3X+, LLC, Pensacola, Florida, USA). The adolescents were instructed to wear the accelerometer on the right hip for seven consecutive days at all times, except during water activities or while sleeping. All raw accelerometer files were processed and analyzed using specifically developed and commercially available software (KineSoft version 3.3.20, Saskatchewan, Canada). A wear-time of ≥ 480 min/day was applied as a criterion for a valid day. Periods of ≥ 20 min of zero counts were defined as non-wear time . The number of ‘valid days’ vary depending on the analyses performed. Outcomes for PA are total PA level averaged per day (mean counts per minute [CPM]), time spent sedentary (min/day), and time spent in different PA intensities (min/day). Sedentary time was defined as all activity < 100 CPM, this cut-point has shown to provide a realistic estimate of the time adolescents spend doing sedentary activities . MVPA was defined as all activity > 2000 CPM, this cut‐point was developed for the European Youth Heart Study and is equivalent to a walking speed of adolescents of >4 km/h . All analyses were based on accumulation of data over 10 second epochs.
Cardiorespiratory fitness was measured with the Andersen-test, which is an intermittent running field test . The Andersen-test was administered as per standard procedures indoors on a wooden or rubber floor, however, due to different sizes of available indoor facilities we standardized the length to 16 meters (original protocol 20 meters). All adolescents were tested in groups of 6-12 individuals. They ran from one line to another line 16 meter apart in an intermittent to-and-from movement for a total of 10 minutes, with 15-seconds work periods and 15-seconds breaks standing still. We recorded the distance covered during the test in meter as a proxy for cardiorespiratory fitness.
Muscle strength (i.e. endurance, isometric and explosive strength) was measured using reliable and validated selected tests from the Eurofit test battery : 1) Upper limb strength – handgrip strength was measured using a hand dynamometer (Baseline® Hydraulic Hand Dynamometer, Elmsford, NY, USA); 2) Explosive strength in the lower body was measured using a standing broad jump test; and 3) Abdominal muscle endurance was tested using a sit-up test (number of correctly performed sit-ups within 30 seconds). The result from the muscle strength tests were recorded both individually and as a composite z-score for analyses.
Body mass (weight; 0.1 kg) was measured using an electronic scale (Seca 899, SECA GmbH, Hamburg, Germany) with individuals wearing light clothing. Stature (height; 0.1 cm) was measured using a portable stadiometer (Seca 213 GmbH, Hamburg, Germany). During the measurement the individual was facing forward, with shoes removed. In the analyses, 0.6 kg (light clothing) or 1.5 kg (more heavy clothing) was subtracted from each person’s weight to account for the clothing. We calculated body mass index (kg·m−2) as weight (kg) divided by the height squared (m2). Waist circumference was measured with an ergonomic circumference measuring tape (Seca 201 GmbH, Hamburg, Germany). The measure was taken at the midway between the lower rib and iliac crest with the adolescent’s abdomen relaxed at the end of a gentle expiration. The individual stood with arms hanging slightly away from the body. Two measurements were collected from each individual. If the difference between measures was a greater than one cm, a third measurement was obtained; and the mean of two closest measurements was used for analyses.
Socioeconomic status and birth country
The highest education level of the participant's parents was used as a proxy for socioeconomic status (SES). To obtain information regarding the parents’ education level, our database was linked to registry data collected by Statistics Norway. Four SES groups were computed: low (primary school, lower secondary school, vocational high school), middle (secondary school/high school), middle high (undergraduate degree) and high (graduate degree). We also used registry data collected by Statistics Norway to receive information regarding the birth country of participants' and their parents.
Mental health (SDQ, Kidscreen, Harter, Hopkins)
Overall psychosocial problems and strengths was assessed with the Strength and Difficulties Questionnaire (SDQ). This is a screening instrument consisting of 25 items equally divided across five scales measuring emotional symptoms, conduct problems, hyperactivity-inattention, peer problems, and prosocial behavior .
Adolescent’s health-related quality of life (QoL) was self-reported using the Kidscreen-27 questionnaire . The instrument consists of 27 items covering the following five QoL dimensions: 1) physical well-being (5 items); 2) psychological well-being (7 items); 3) parents/guardians relations & autonomy (7 items); 4) social support & peers (4 items); and 5) school environment (5 items).
We used Harter's Self-perception Profile for Adolescents (SPPA) to assess the adolescent's domain specific self-evaluation of competence or adequacy . The instrument consists of seven subscales covering the following domains: 1) scholastic competence, 2) social competence, 3) athletic competence, 4) physical appearance, 5) job competence, 6) close friendship, and 7) romantic appeal.
Symptoms of anxiety and depression were assessed by the short version of Hopkins Symptom Checklist (HSCL-10) . HSCL-10 consists of 10 symptoms or problems that people sometimes have in which adolescents answered how much the symptoms have bothered or distressed them during the last week.
Academic performance in reading and numeracy were measured using specific standardized Norwegian National tests designed and administrated by The Norwegian Directorate for Education and Training . The national test in reading measured the extent to which the individuals' reading skills are consistent with the descriptions of reading as basic skills in the curriculum, while the national test in numeracy measures the extent to which individuals' competency complies with the descriptions of numeracy as the basic skill of the curriculum. The National tests are designed for the students to show whether they have the basic skills in reading or numeracy that is necessary to achieve the competence goals in the subjects, and it is not a test in the Norwegian or mathematic subject per se. The National tests are extensively validity and reliability tested by The Norwegian Directorate for Education and Training.
The Classroom Climate Scale was used to assess learning environment . This instrument consists of 22 questions related to intrinsic and extrinsic motivation, anticipation and teacher-student relations. In addition, three questions were asked related to the learning environment at school in general, in the classroom and during recess. The answers to the various questions were summarized individually.
Three instruments were used to assess the adolescent's experience of PE-lessons: 1) change in satisfaction of basic psychological needs in the subject was assessed by Basic Psychological Needs in Exercise Scale (BPNES) ; 2) emotional responses to the subject was assessed by the Basic Emotions Trait Test (BETT) ; and 3) eagerness to join the subject was assessed by Eagerness for Physical Activity Scale (EPAS) .
In the initial phase of the implementation, teachers at the intervention schools completed a short questionnaire to assess the school’s readiness to change. Towards the end of the intervention period, four random schools from each intervention arm were invited to participate in a qualitative study where the aim was to investigate the following questions: a) what factors influenced the implementation of the intervention, and, b) how did the students and teachers receive the intervention? A semi-structured interview guide was developed to answer these questions, and interviews were conducted with principals, teachers and students. The interviews with principals and teachers were performed as individual interviews, as was considered a suitable method for investigating how the principals and teachers perceived, led and facilitated the intervention. Among the adolescents, on the other hand, focus groups were used as this can help reduce the power imbalance between an adult interviewer and young informants . In the latter interviews, the aim was to obtain common reflections and descriptions of how the various intervention components were received and integrated into the adolescents’ school life and the teachers' practice. All participates provided written consent to participate in the qualitative study.
During the intervention period, teachers at the intervention schools documented the extent to which the intervention was implemented (dose) as intended through a weekly online form. The report described activities performed throughout the school day, the intensity of the activities (one a 1 to 3 scale) and the number of minutes of the PA sessions. Researchers from the project group visited each intervention school twice per semester to assess the quality of the implementation (fidelity). To improve the implementation of the intervention, teachers from the intervention schools were invited to the coordinating test center to share their experience with the implementation of the intervention. Two webinars were also organized for the teachers during the intervention period, and participation was voluntarily.
Statistical considerations/Statistical methods
Power calculations/Sample size and power
The ScIM study was designed to detect a difference in total PA level of 7 % (49 cpm) between the participants in the intervention arms and the control arm. We assumed a standard deviation (SD) of 150 cpm, a power of 90 %, a significance of 0.05, leading to 492 individuals in each group. To allow for 20 % loss to follow-up we needed 590 individuals in each group. Further, we needed a minimum of ten clusters per study arm, consequently we aimed to recruit clusters and individuals until we had at least ten clusters and 590 individuals per study arm.
Plan for analysis
Data will be described with suitable measures of central tendency and variability, along with confidence intervals. Mixed models with school as random effect, adjusted for class, individual and baseline values, will be used to investigate the difference in the different variables when comparing the students in the two intervention groups with the students in the control group. By building up a mixed model, all participants who have data at baseline and everyone who has data at the post- test will be included, and thus all data will be utilized. A two-sided p-value < 0.05 will be considered statistically significant.
The analysis from the qualitative study is based on a thematic content analysis . The analysis work starts with the sound recordings being listened to repeatedly while immediate reflections are written down. Then all interviews will be transcribed and coded descriptively. The content will be discussed and organized by the researchers in collaboration, and a thematic framework will be identified.
The project was reviewed by the Regional Committee for Medical and Health Research Ethics (REK) in Norway, who according to the Act on medical and health research (the Health Research Act 2008) concluded that the study did not require full review by REK. The study was approved by the Norwegian Centre for Research Data. The study was conducted in accordance with the principles set forth by the Helsinki declaration. Written informed consent from the participants and their parents or caretakers was obtained prior to the data collection. Reporting are anonymous, and individual participants cannot be identified in any published materials. The design, conduct, and reporting of this trial adhere to the CONSORT statement .