Moderate-to-vigorous physically active academic lessons and academic engagement in children with and without a social disadvantage: a within subject experimental design

As a part of the larger F&V intervention, 86 elementary children (23 SDC) from second or third grade classes of four elementary schools from the Northern Netherlands were recruited to participate in this study. All children participated in the F&V lessons. Any missing demographic information eliminated a child from all comparisons, resulting in a final sample of 81 children (mean age: 8.2 ± 0.65 years; 41 girls; 40 boys). Children were classified as SDC or non-SDC based on parental education, children of whom the person(s) responsible for the daily care completed less than three years of secondary school were classified as SDC. In the context of the Dutch educational policy schools are required to obtain this data from every child. Based on this data, schools receive extra funds for SDC in their student population [20]. The SDC (n = 20) and non-SDC (n = 61) classification data were obtained from the personal school files of the children. Table 1 shows the demographic information of the children, it can be found that there were significantly more SDC in grade 2 in comparison with grade 3 (p < 0.05). A subgroup of 67 children (mean age: 8.2 ± 0.59 years; 33 girls; 34 boys; 15 SDC) wore heart rate monitors during the F&V intervention. Written informed consent was obtained from the school principals of the participating schools. The parents/legal guardians were informed before the start of the intervention and were given the option to withdraw their permission for their child to participate at any time. Furthermore, the parents/legal guardians gave written informed consent for wearing heart rate monitors during the physical active academic intervention. All procedures were approved by the ethical committee of the Center for Human Movement Sciences of the University Medical Center Groningen/University of Groningen.

The F&V intervention consisted of physical active mathematics and language lessons that were taught in the classroom. The main (long-term) goal of F&V was to improve academic skills and physical fitness of SDC and non-SDC. Children participated in the F&V program for 22 weeks, three times a week. During each lesson the children stood behind or beside their school desk. In each lesson 10–15 minutes were spent on mathematics and 10–15 minutes on language. The main focus was on reinforced concepts learned in earlier classes. The physical exercises were relatively easy to perform and aimed at exercising at moderate to vigorous intensity level. During the lessons all children performed basic exercises and specific exercises simultaneously. The specific exercises were performed when the children solved an academic task. For example, the word ‘dog’ must be spelled by jumping in place for every mentioned letter or the children had to jump 6 times to solve the multiplication ‘2x3’. Similar academic tasks with different words or sums were exercised during one lesson. The basic exercises were performed during the remaining part of the lesson (for example when the children were thinking about a sum). Basic exercises were for instance, marching, jogging or hopping in place. A more complete description of the intervention and its implementation has been published elsewhere [21].

In this study a within-subject design was used to determine whether physically active F&V lessons affected the time-on-task in the regular classroom lessons that immediately followed the F&V lessons. To do this, the time-on-task during a regular classroom lesson that followed a F&V lesson (post-intervention) was compared to the time-on-task during a regular classroom lesson that followed a regular inactive lesson (post-control). On every participating school post-intervention and post-control observations took place at the start, midway and the end of the 22-week F&V intervention period (see Table 2). In order to improve the reliability of the task-observations, at the start and midway observation two post-intervention and two post-control observations were performed on every school in one week. At the end observation, one post-intervention and one post-control observation was performed on every school in one week. On each school, the post-intervention and post-control lessons were observed at different schooldays at about the same time of day (always before lunch time). A post-intervention observation was followed by a post-control observation in the same week. All observations started when the regular classroom lesson, in which the children returned to their academic content, began. To assess the intensity of the F&V lessons, part of the children wore heart rate monitors during each physically active F&V lesson that was followed by observations.

The time-on-task of the children was observed through time sampling, by using a modified version of the observation method described by Grieco et al. [9]. Every child was observed for the duration of five seconds, before the observer moved on to the next child. To support the observer an audio file was played through a headphone with beeps with an interval of five seconds. When every child was observed once, the observers started with the first child again and repeated this sequence for the remaining part of the lesson. We aimed at observing each child 15 times a lesson (50 minutes of observations in an average class of 20 children), this criteria was not always met due to lesson interruptions (for example recess or lunch break). Eventually each child was observed 10–16 times a lesson. It was observed if a child showed on-task or off-task behavior. On-task behavior was defined as any behavior in which a child was attentive to the academic instruction or actively engaged in the appropriate task, as assigned by the teacher. All behaviors that were not on-task were noted as off-task behavior. Off-task behavior included fidget and listless behavior, for example wiggling on the chair, needless moving without paying attention, placing the head on the desk, talking to or looking at other children. During the start observation the inter-rater reliability of the observation method was determined. Five lessons were observed by two observers, the inter-rater reliability appeared to be of substantial agreement [22,23], κ = 0.73.

Prior to the F&V intervention a maximal endurance test (20 m Shuttle Run test) was performed during a physical education class. This test is an item of the Eurofit and consisted of running back and forward over 20 meter at an increasing speed [24]. Heart rate data were collected using team heart rate monitors [25]. The heart rate monitors averaged and stored heart rate every second throughout the test time. The maximum heart rate (HRmax) was determined by visual inspection of the heart rate curve that was obtained from the test.

Heart rate data were also collected during the F&V lessons. Exercising with heart rates between 60% and 90% of HRmax was considered as MVPA. Exercising above 90% of HRmax was considered as highly intensive exercise [26,27]. From the heart rate data, the lesson time spent in MVPA was calculated for each child for every observed F&V lesson.

The percentages of time-on-task were calculated for each child by dividing the number of observations in which the child performed on-task behavior, by the total number of observations during that lesson. Next, the two post-intervention and two post-control observations that were performed at the start were merged into one post-intervention and one post-control observation by calculating the mean time-on-task. The same merging took place for the midway observation. The time in certain heart rate zones was calculated (0-60%, 60-90% and 90-100% of HRmax) per child per lesson. By dividing the lesson time spent in 60-90% of HRmax with the total lesson time, the percentage of MVPA per child per F&V lesson was calculated. Pearson’s correlations were calculated between time-on-task, percentage of MVPA and participant demographics in order to detect important covariates. Demographics that significantly correlated with time-on-task or percentage of MVPA were used as covariates in further analyses.

Differences in time-on-task between SDC and non-SDC during post-control and post-intervention lessons (aim 1) were analyzed using independent samples t-tests (SPSS, version 20.0). Differences between post-intervention and post-control observations (aim 2) were analyzed using multilevel modeling (MLwiN 2.29). Time was nested within children and each individual had six observations. Multilevel models were calculated for the time-on-task of the children. The first model contained only the important covariates. Next, condition (post-intervention and post-control; model 2), the interaction between condition and SDC (model 3) and the interaction between condition and observation moment (model 4) were entered as possible predictors. Variables with a nonsignificant contribution to the model were removed from further analysis. Effect sizes were calculated as (mean post-intervention – mean post-control)/(pooled SD) [28,29].

The relationship between percentage of MVPA during F&V lessons and time-on-task was analyzed by calculation of partial correlations controlling for the effect of important covariates (SPSS, version 20.0). Statistical significance for all analyzes was set at 0.05.