Study protocol: the Fueling Learning through Exercise (FLEX) study – a randomized controlled trial of the impact of school-based physical activity programs on children’s physical activity, cognitive function, and academic achievement

Aims

The primary aim of the FLEX Study is to evaluate the impact of two school-based PA programs, 100 Mile Club® and Just Move™, on children’s school-time MVPA and total daily MVPA, compared to a control group. In addition, the RCT will evaluate the effects of these innovative, school-based PA programs on children’s cognitive performance and academic achievement. The study will also evaluate the reach of the programs by examining factors that influence participation among different demographic, weight status, and fitness groups.

Schools

Identification of eligible schools began at the district level. The FLEX study is designed to examine health disparities among children associated with income and race/ethnicity. Therefore we targeted lower income and racially/ethnically diverse public school districts in Massachusetts. Initially, school districts were approached for participation if they had greater than 40 % of students qualifying for free or reduced price lunch and/or had greater than 40 % non-Caucasian students. In addition, districts were only approached if they had at least four elementary schools serving 3rd-5th graders.

Overall, twenty school districts were approached. Twelve of these districts declined to participate, citing changes in administration, ongoing participation in similar programs, and/or competing priorities. From the remaining eight districts, 24 schools agreed to participate and have at least one school enrolled. Six schools were included in Wave 1 and 18 schools in Wave 2. All districts serve a low-income and diverse population and are located in urban, suburban, and peri-urban areas. One exception to the initial inclusion criteria was made, such that one school from a district with 34 % low-income students was included.

Students

All third and fourth grade students at participating schools were eligible to enroll. Third and fourth grade students were recruited in year one and followed in year two into fourth and fifth grade, respectively.

School recruitment & randomization

Schools were initially recruited by email and phone, followed with in-person meetings. Initial contact was made at the district level to superintendents and district wellness/health/PE coordinators, with secondary outreach at the individual school level (principals/assistant principals). Where districts have research boards/IRBs, contact was simultaneously made to those entities, and the study protocol and recruitment materials were submitted for approval. Once enrolled, schools were randomized to receive one of two school-based PA programs – 100 Mile Club or Just Move – or to the control group. Schools were block randomized in groups of three, stratified by district, to ensure comparable numbers of schools in each arm. Once a school agreed to participate, the study statistician informed recruiting staff of the school’s group assignment. Throughout the school recruitment process, the statistician and recruitment staff worked independently. Figure 2 outlines the flow of district, school, and participant recruitment.

Student recruitment

Study staff conducted presentations in schools to explain the study and enrollment procedures, and to distribute recruitment materials to all 3rd and 4th grade students. Presentations were given assembly-style or in individual classrooms. Permission packets were sent home with each student and included: 1) a flyer with information about data collection procedures; 2) a plain language consent form for the parent/guardian; 3) a child assent form; and 4) a demographic form for the parent/guardian to complete. Permission packets were available in English, Spanish, Portuguese, Haitian Creole, Arabic, Vietnamese, and Mandarin, the primary languages spoken in targeted communities. Students were asked to return completed packets approximately 1 week after they were distributed. Enrollment was closed at the start of baseline data collection.

Overview

Briefly, child-level data collection takes place at each participating school during regular school hours at each of the three time points (baseline, midpoint, and post-intervention). Exceptions are demographic data, which are collected once at baseline, and fitness measurements, which are conducted at a separate time once per school year. Child-level standardized test scores for Math and English Language Arts (ELA) are collected from the Massachusetts Department of Elementary and Secondary Education (DESE) after testing is completed each spring. School-level data is collected once per year (physical activity environment and food environment surveys, along with attendance data). Trained research assistants (RAs) and study staff administer all instruments and assessments according to the written study protocol. At baseline, there were 174 participants enrolled from Wave 1 schools and 1008 from Wave 2 schools. Baseline measurements were conducted in the fall/early winter of Year 1 (Wave 1 – February-March, 2015; Wave 2 – September 2015-February 2016). Short-term measurements were conducted at the end of the first school year (Wave 1 – May-June 2015; Wave 2 – April-June 2016) and post-intervention measurements will be conducted approximately 18 months after baseline at the end of the second school year (Wave 1 – March-April 2016; Wave 2 – April-June 2017). At time of enrollment, participants are assigned a unique 5-digit code which is used in place of name or other identifiers to link participant to all their study data. The study project manager maintains an electronic, password-protected list linking participant with their unique ID.

Demographics

Demographic information was collected at baseline by paper-and-pencil questionnaire included with the recruitment packet and returned with informed consent documents. The 10-item questionnaire included questions on child’s date of birth, grade and age at time of enrollment, sex, race/ethnicity, maternal and paternal education levels, and free- or reduced-price lunch status. Parents/guardians were also asked to report whether or not their child has behavioral difficulties, such as learning, understanding or paying attention, or communicating [24] and whether their child was on an individualized education program (IEP). English language learner status will be obtained at the child-level from the Massachusetts DESE.

Anthropometrics

Height and weight are measured in light clothing with shoes removed using a portable stadiometer (Model 213, Seca Weighing and Measuring Systems, Hanover, MD) and portable digital scale (Model 803, Seca Weighing and Measuring Systems, Hanover, MD). Height and weight are measured in triplicate to the nearest 1/8 in. and 0.1 kg, respectively. Body mass index (BMI) will be calculated (kg/m²) and converted into z-score using the Centers for Disease Control and Prevention (CDC) age- and sex-specific growth charts [25]. BMI percentiles are classified accordingly as: <5th percentile as underweight, 5th-85th percentile as normal weight, 85th-95th percentile as overweight, and ≥95th percentile as obese.

Physical activity

Waist-worn tri-axial accelerometers are used to objectively measure participants’ physical activity (Actigraph accelerometers, models GT3X+ and wGT3X-BT, ActiGraph, LLC, Pensacola, FL) and have been validated and calibrated for use among children [26]. Study staff instruct each participant how to wear the accelerometer. Participants are asked to wear accelerometers during all waking hours, except when bathing or swimming, for 7 consecutive days. Accelerometers are initialized to store activity counts beginning at 00:00:01 on the first day the child will wear the device; data will be processed using a 15-s epoch. To support wear-time compliance, participants are also given a paper calendar-style tracking log on which they are instructed to write down the time they put the accelerometer on each morning and the time they remove it before bed each night. The log also asks participants to self-report any days they were ill during the wear week, as well as any sports or physical activities they played during the wear week and their favorite sports/physical activities. In addition, weather data are collected from the National Oceanic and Atmospheric Administration (NOAA) [27], and the high temperature (continuous) and precipitation (binary: yes/no) are recorded for each day accelerometers are worn. Data are collected from the weather station nearest to the participating school.

Accelerometer data will be categorized into minutes of sedentary, light, moderate and vigorous activity using thresholds developed specifically for children [28]. In-school and out-of-school time MVPA will be extracted. In-school hours will be calculated for each participant based on specific start and end times of the school day for each day the accelerometer was worn. Weekday out-of-school time will be calculated as the sum of before school time and after school time, accounting for school hours and average awake time.

Cognitive performance

Two cognitive assessments are administered one-on-one in a quiet area. A digit span test, with forward and backward subtests, is used to measure short-term memory, attention and concentration [29]. Digit span tests have been positively linked to PA [29, 30]. In the test, the administrator orally presents number lists of increasing length to the participant, and the participant is asked to repeat these digit spans back in the same order either forward or backward (for forward and backward subtests, respectively.)

A Stroop Color-Word test [31] is administered to assess inhibitory control and selective attention, aspects of cognitive function that have been positively linked with PA and fitness in previous studies [30, 32]. The test includes both a congruent task and an incongruent task, each lasting 45 s. Participants are first given the congruent task in which they are presented with a set of 100 color words printed in the same color ink as each word (ex. the word “red” is printed in red ink). They are asked to read these words aloud and complete as many as possible in 45 s. After completing this task, they are given the incongruent task, in which they are provided another set of 100 color-words printed in a different color ink (ex. the word “red” is printed in green ink). Participants are asked to identify and say aloud the color of the ink rather than the word, completing as many as possible in 45 s. The test is scored by calculating a ratio of items completed correctly to total number of items completed in 45 s.

Academic outcomes

Individual child-level standardized test scores will be collected from the Massachusetts DESE and used as an indicator of academic achievement. PA in children has been shown to be positively associated with scores on both Math and ELA [33, 34]. The Massachusetts Comprehensive Assessment System (MCAS) test has been used in Massachusetts annually for all students in grades 3–10. During the 2014–2015 academic year, the DESE began the process of transitioning to a different standardized test, the Partnership for Assessment of Readiness for College or Careers (PARCC). The MCAS/PARCC tests are administered annually in the spring and coincide with both mid- and post-intervention measurements. Baseline standardized test scores (year prior to enrollment) will be obtained retrospectively and for each year of participation for all participants. For both tests, raw scores are converted to scaled scores, which are linked to four levels that describe the performance levels of individual students, schools, and districts: Advanced, Proficient, Needs Improvement, and Warning. Additionally, grade-level standardized test scores will be collected across grades 3–5, and will be aggregated to analyze school-level impact of 100 Mile Club and Just Move.

School attendance will be collected and examined as an exploratory outcome at the individual and school level. Attendance will be assessed as the number of days present per academic school year and converted to a percentage.

Physical activity social support and self-efficacy

A PA social support questionnaire is administered to participants in small groups, with individuals completing their own questionnaire. The 10-item questionnaire, developed by Trost and colleagues [35], is modified from the Social Influences Scale developed and validated by Saunders et al. [36]. The modified questionnaire assesses the participant’s self-reported social support for engaging in physical activity by family, friends, classmates and teachers. Questions are answered on a three-point scale (“yes,” “no,” or “don’t know”). Example scale items include “My mother/father/caretaker thinks I should be physically active,” “The students in my class think I should be physically active,” and “My teacher has encouraged me to be physically active in the last two weeks.”

Participants also complete the “What Kind of Kid Are You?” 18-item questionnaire which is designed to assess how children evaluate themselves in various domains including athletic competence, behavioral conduct, and global self-worth. Questions are taken from the Harter Self-Perception Profile for Children [37]. Participants are asked which of two opposing statements best describes them and then whether it is “sort of true” or “really true” for them. This structured alternative format is designed to minimize socially desirable responses and ensure internal consistency and reliability [38].

Dietary intake

Dietary patterns are collected by child self-report using two instruments, both administered in small groups. One questionnaire focuses specifically on weekday breakfast consumption, and the other is a 7-day recall focused on diet quality. The 6-item breakfast questionnaire developed for this study asks about typical weekday breakfast consumption, including whether or not the participant usually eats breakfast on school days, source of breakfast (school, home, restaurant), and foods and beverages typically consumed for breakfast. The same questions are asked about “today” (the day of data collection). Breakfast consumption has been linked with stronger cognitive functioning in children [39] and data on breakfast consumption will be explored as a potential covariate or moderator of intervention effects on cognition.

Diet quality is assessed through a 39-question 7-day recall. Research assistants explain to participants how to complete the questionnaire. The FLEX Dietary Questionnaire is adapted from several published, validated instruments [40–42]. The recall is divided into 5 categories: beverages, fruits, vegetables, salty snacks, and sweet snacks. The structure is similar to the Block Food Frequency Questionnaire for Ages 8–17 [43] in which respondents indicate how many days in the last week they consumed an item (none, 1 day, 2 days, 3–4 days, 5–6 days, or everyday) and how much they had in one day (a little, some, or a lot). The Block Food Frequency Questionnaire for Ages 8–17 was used during Wave 1, but field administration showed the foods included to be less relevant to the consumption patterns of the study participants and the format (scantron form) to be challenging for the 8–10 year old participants to complete. Foods and beverages included in the FLEX Dietary Questionnaire, administered in Wave 2, reflect the foods and beverages frequently consumed by children in this age group as well as foods and beverages to limit (sugar-sweetened beverages, salty snacks, desserts) and those to promote (fruits, vegetables, water) [44]. The portion size for each item is matched to a standard serving size to represent “a little” (1/2 standard serving), “some” (1 standard serving) and “a lot” (1.5 times standard serving). A final question asks how many days the respondent has eaten out (including restaurants, fast-food, or take-out) in the last 7 days. Each questionnaire receives an overall score for diet quality and will be used as a covariate in analyses.

Fitness assessment

Cardiorespiratory fitness (CRF) is measured as a covariate using the FITNESSGRAM^®’s Progressive Aerobic Cardiorespiratory Endurance Run (PACER) test, which is a 20-meter maximal-effort shuttle-run. The 20-meter shuttle-run has been validated with young populations [45] and shown to be highly correlated with VO₂max in children [45, 46]. The test is administered once per school year (for 2 school/study years) in the study schools by trained research staff who follow a standard protocol based on the Cooper Institute’s published guidelines [47]. Aerobic capacity as measured by the PACER will be analyzed using standards described by FITNESSGRAM [47].