Rationale and design of a randomized controlled trial examining the effect of classroom-based physical activity on math achievement

Physical activity (PA) in schools has historically received little attention and, in many countries, economic concerns and a greater focus on standardized testing have reduced the time spent on physical activity in favor of academic classes [1‐4]. A regular school day often includes very little PA and children are required to sit quietly in the same position for long periods of time [5‐7]. However, this may not be the most effective way of enhancing children’s academic achievement. Breaking up the sitting periods with physical activity might be a better approach to promote learning [8, 9].

PA has several cognitive and physiological benefits [10, 11]. A meta-analysis showed a significant positive relationship between PA in general and cognition in children [12]. This is consistent with the “executive function hypothesis”, in which PA is believed to enhance executive functions that are crucial to the learning process, such as organization and integration of information, and modulation of behavior through selective attention (e.g. [13‐16]). Studies report clear acute cognitive benefits of PA, typically increased information processing speed or response accuracy [17‐19]. Early experiments on the effects of regular PA showed either no effect of PA on cognitive outcomes [20] or limited improvements [21, 22]. However, recent studies report positive effects of regular PA on the structures and functions of the brain [14, 23‐25]. Possible neurophysiological mechanisms of PA include changes in brain blood flow and arousal level, improved conduction of information due to increased concentration of neurotransmitters and receptors, as well as increased concentrations of growth factors important for the development of new brain cells [26]. In addition, regular PA may increase brain volume (e.g. the volume of the anterior hippocampus), leading to improved spatial and relational memory [14, 27, 28]. Lastly, a recent study demonstrated a positive relation between mathematic achievement and decreased gray matter thickness in the superior frontal cortex, superior temporal areas, and lateral occipital cortex, in more fit children [29]. This suggests that aerobic fitness is essential for the normal thinning of the cortical gray matter during brain maturation, thus providing us with a predictor of math performance that could in turn help educators to identify interventions that can enhance learning.

Despite the increasing evidence for an association between PA or fitness and cognitive performance [30‐35], few high-quality longitudinal studies have examined the effect of PA on academic performance (rather than cognitive control or other educational outcomes) [8, 14]. Furthermore, to the best of our knowledge, only three studies have assessed academic outcomes following the integration of PA into the classroom. A study conducted among US elementary school children reported that children with 90 min/week of moderate to vigorous physical activity in academic lessons over three years showed greater academic improvement than controls [36]. However, these results were considered secondary and were based on a small subsample (n = 203) of the study. Another small-scale randomized experiment from the US conducted among 3rd grade children showed that integrating 30 min of PA three days per week for three months gave mixed results for academic improvement compared with controls [37]. Finally, a recent study showed that physically active academic lessons significantly improved mathematics and spelling performance of elementary school children [38].

Even the studies investigating the link between PA and cognition have typically examined cognition and PA separately, possibly due to a commonly held view that the mind and body function separately. However, recent theories of embodied cognition have emerged in several disciplines to challenge this traditional dualistic approach [39‐46]. These theories consider the body to be a key factor in shaping our cognition, where ‘non-cognitive’ processes of acting and perceiving providing bodily experience are fundamental to and inseparable from the development of important cognitive functions. This emphasizes the importance of sensory–motor interactions with our environment during learning, resulting in more endurable and richer consolidation of knowledge [47]. This psychological and neuroscientific research has implications for education as it emphasizes the importance of suitable sensory and motor interactions during learning for the development of human cognition [47].

We designed a randomized controlled trial with the primary aim of investigating the effect on mathematical achievement of incorporating physical activity in math teaching for 7-year-old schoolchildren. Secondary aims of the study were to investigate effects on executive function, and the longitudinal associations between PA and creativity, math, executive function, body mass index and aerobic fitness. The hypothesis was that PA integrated into classroom teaching would be superior to traditional, physically inactive classroom teaching in facilitating learning, and would be associated with a greater improvement in academic skills.

The active math intervention consisted of math teaching that implemented PA in the classroom as a facilitating instrument. The intervention period lasted one school year, during which the students received on average 6 math lessons of 45 min per week with physically active teaching. PA in this math intervention was defined as any bodily movement produced by skeletal muscles that resulted in increased energy expenditure [49]. This is similar to the definition known as movement integration (MI) [50].

Each 45-minute lesson consisted of at least 15 min of PA spread over the lesson. Math lessons usually took place in the classroom, but were also held in other school areas such as corridors and surrounding outdoor areas. Sedentary activities were limited to periods of maximum 20 min, given that most students are unable to sustain attention on one thing for more than about 20 min at a time [51]. The rationale behind the active math was that if an abstract phenomenon is connected to a concrete bodily experience other than just seeing or hearing it, the comprehension and memory of the given phenomenon could be optimized (e.g. [52, 53]). In addition, we expected that breaking up the cognitive load imposed by a learning task would limit overload of working memory in the children and thus help maintain focus [54]. Studies have detected limited working memory capacity in children, and spreading the load across different working memory subsystems might prevent overload in one specific subsystem [55‐57]. The active math lessons followed the principles of “learning by doing” and the development of practical life skills that are crucial to children’s education [58]. This is related to experiential learning theory, i.e. learning through action and experience as opposed to through repetition [59, 60]. The active math approach was adopted for all math lessons for the entire school year. Appendix 1 provides examples from the active math intervention.

Children in the control schools received regular classroom instruction, also with an average of 6 math lessons of 45 min per week. The math teachers in the control schools were asked not to make any changes to their usual teaching methods before the study endpoint measurements. This was accepted with the agreement that at the end of the study, these teachers would receive the same training in physically active math teaching as the teachers in the intervention group.

The study was carried out over nine months from August 2012 to May/June 2013. Unfortunately during this period, all Danish schools were closed for a month due to a teacher lockout as a result of a union-employer dispute. Thus in the final period of the intervention, participants did not attend school for 25 days during April/May 2013. This period ended 4 weeks prior to the end of the intervention.

The test battery included measurements of math skills, cognitive function, creativity, aerobic fitness, anthropometrics and level of PA. Prior to testing, the research assistants (i.e. master’s and PhD students) were thoroughly trained in conduction of the test battery to ensure valid and uniform data collection. The training included several test trials with feedback and was performed at the University of Southern Denmark by a researcher experienced in the techniques. The research assistants were blinded to the randomization result for measurement of the outcomes and for data entry. All tests were performed at the participating schools either in the classrooms or a gym, except for the assessment of executive function (Flanker task), which was performed in smaller rooms with a maximum of two students present at the same time. Each outcome measure was assessed at baseline and follow-up using the methods described below. The tests applied at baseline and follow-up were identical.

Intervention compliance was assessed using different methods targeting teachers and parents as well as the children. This included video analysis of lessons, SMS-tracking, and accelerometry.

The literature lacks well-designed large-scaled randomized controlled trials on the effect of classroom-based physical activity on academic achievement and cognition. The present study will contribute to the field by generating detailed knowledge about the relationship between physical experiences that are combined with abstract math information and the acquisition of math skills. We expect the study results to provide schools and education policy-makers with an evidence base for implementing physical activity in daily school life, as a change from schoolchildren sitting still much of the day.

Apart from the large sample size and the comprehensive follow-up of children’s learning experiences and physical behavior over nine months, this study includes detailed measurement of the primary and secondary outcomes using well-validated methods. The math test applied in this study has been used for over 25 years to assess the math skills of Danish schoolchildren. The complex flanker task that assesses the three core executive functions will allow us to verify the few previous research findings concerning the association between a change in physical activity level and the ability to focus on a target stimulus and inhibit distractions [14, 74]. Furthermore, we expect the combination of SMS-tracking and accelerometry to be an excellent solution for measuring physical activity in large cohort studies. This makes it possible to compensate for some of the problems associated with accelerometry measurements, for example by using SMS-tracking to ask questions about specific activities, including water activities, and to allow more precise measurement of exercise intensity during cycling. When all this information is coupled to the class school timetable, then we can analyze the physical activity undertaken in specific time periods such as lunch breaks or during the math lessons. However, accelerometry does not provide any qualitative detail about the types or modes of activity being undertaken. For this reason accelerometry measurements were combined with video recordings of math lessons.

An important limitation of this study is the lack of qualitative data collection. This could have been used to further describe the various interventions and the complexity of the social contexts in which the active math was tested, (i.e. through case studies of a specific child/teacher or a specific context, observations of classroom activities, or qualitative interviews with children, teachers, and parents). Due to economic reasons, qualitative assessment was not performed since the main objective of the study was of quantitative character.

In conclusion, the results of this study will expand the current evidence on the relationship between physical activity and academic achievement in schoolchildren. We expect the results to stimulate the debate about whether the integration of physical activity into the classroom can enhance children’s cognitive skills and creativity, and will help educational practitioners to design learning environments that are optimal for cognitive development and academic achievement.