Evidence-Based Overview of Accelerometer-Measured Physical Activity during School Recess: An Updated Systematic Review

Interest in analyzing physically active behaviors during school recesses has grown in recent years as the school environment has consolidated (recess, physical education classes, lunch-time, before and after school) as a crucial space to bring these levels towards those recommended through intervention programs and improvements in the school environment. Unfortunately, in most of these studies, children do not achieve the 60 min a day of moderate to vigorous physical activity (MVPA) recommended by the World Health Organization. The aim of this systematic review is to analyze the cross-sectional, longitudinal, and intervention studies objectively measured with accelerometers that have emerged in recent years to determine the amount of MVPA of children at recess. This systematic review followed the PRISMA guidelines. The extraction process for the studies included in this systematic review yielded a total of 43 articles. The studies were classified according to the methodological nature of the research: cross-sectional (n = 34), longitudinal (n = 3) and quasi-experimental (n = 6). The results of the studies confirm that during the recess period younger children are physically more active than older ones and that in general, boys are more physically active than girls. In addition, the data show that the school contributes to more than 40% of the total MVPA. The intervention programs led to an increase in MVPA of up to 5%. Providing schools with equipment and facilities shows that intervention programs are beneficial for raising children’s levels of physical activity.


Introduction
The low levels of physical activity in the younger population are considered one of the main factors that cause health problems such as obesity or overweight, these levels being below the daily recommendations for moderate to vigorous physical activity [1]. In the United States, only about 42% of children between 6 and 11 years old meet the daily recommendations of 60 min a day, so it is necessary to study strategies to promote physical activity and health among children and adolescents [2]. Recent studies show that the environment affects the physical activity of children, since those who have large spaces, adequate equipment or a marking system in the schoolyard register higher levels of physical activity [3]. Children spend a very important proportional part of their daily work in the school environment, making school a great opportunity to influence levels of physical activity. However, each country is influenced by a series of important factors such as school policies, the programs that take place or the time allocated to recess [4]. Physical education classes, like recesses, are also periods that help students to reach the daily recommendations [5]. All these factors make physical activity levels vary from one country to another. In Japan, the proportion of moderate to vigorous physical activity performed by children is lower, around 18%, compared to European countries such as England or France where studies reported levels between 23.7% and 32.9% [3].

Search Strategy
This systematic review followed the PRISMA guidelines. The process of searching and the systematic review of the existing literature began in February 2020 and lasted approximately six months. The rigorous identification processes of the studies associated with the objective made it possible to group the most outstanding findings in a structured and organized way.
This systematic review of the scientific literature took as its starting point two previous ones carried out by Ridgers et al., in 2012 [4] and Parrish et al., in 2013 [11], corresponding to cross-sectional and intervention studies respectively. The reason that encouraged the realization of this work was the lack of updating in the large amount of scientific literature generated on the realization of physical activity during school recess. That is why for this documentary search, a screening of all the research articles between May 2011 and July 2020 was carried out for cross-sectional and longitudinal studies, and from April 2013 to July 2020 for quasi-experimental studies.
The Web of Science (WOS) database was mainly used for this literature review, although nine electronic literature databases were also involved in the process (Scopus, MedLine, SciELO, KCI, RSCI, Eric, PubMed, Dialnet and Google Scholar). The search strategies in the different databases were carried out by establishing descriptors (Table 1) in four large groups: population (child, children, youth, young, adolescent, student, students), school context (school, primary, elementary, middle school, high school, secondary school), recess time (break, breaks, recess, playtime, free play, free time), physical activity (physical activity). Only articles published in scientific journals were selected, excluding abstracts, books and conferences. To guarantee the rigor of the research, a search was carried out with boolean markers using the following keywords [("child" OR "children" OR "youth" OR "young" OR "adolescent" OR "student" OR "students") AND ("School" OR "primary" OR "elementary" OR "middle school" OR "high school" OR "secondary school") AND ("break" OR "breaks" OR "recess" OR "playtime" OR "free play" OR "free time") AND ("physical activity" OR "physical activities")]. In this way, it was possible to guarantee the non-exclusion of any results related to the research topic. It is important to note that regardless of the nature of the studies (cross-sectional, longitudinal or quasi-experimental) the keywords were the same for all search engines.

Inclusion/Exclusion Criteria
The inclusion criteria ( Table 2) that were used to define the set of papers in this systematic review were the following: (1) Scientific journal publications in English; (2) Research published from the latest systematic reviews carried out; (3) Participants aged between 6 and 18 years; (4) Framed in the context of school recess; (5) Accelerometry as an instrument for measuring physical activity; (6) Contribution of values of physical activity of moderate to vigorous intensity (MVPA); (7) Exclusion of those intervention studies that do not include randomized controlled trials (RCT).

Data Extraction and Risk of Bias
The potentially eligible studies were initially screened by two reviewers (with previous experience in systematic reviews on physical activity in children and adolescents) independently by reading titles and abstracts, following the stipulated criteria prepared in advance, included in the search protocol [14]. In those studies with unclear abstracts or titles, a peer review was performed and the articles were agreed, resolving discrepancies through discussion and consensus [15]. In a second phase, the two reviewers independently read the full text of the studies preselected in the previous phase, creating the final list of potentially eligible studies, going to a third external investigator, when no consensus was reached regarding acceptability [16]. Finally, the full texts of the screened articles were carefully examined and analyzed.

Results
This systematic review yielded a total of 1018 search results, of which 998 were within the time period delimited by the inclusion/exclusion criteria. The extraction process for the studies included in this systematic review yielded a total of 43 articles. The studies were classified according to the methodological nature of the research: cross-sectional (n = 34), longitudinal (n = 3) and quasi-experimental (n = 6). A summary of the PRISMA flowchart of each stage of the search can be seen in Figure 1. Table 3 shows the main characteristics of the 34 cross-sectional studies found. A global analysis allows identifying a large set of variables related to this research topic (SED, PA, LPA, MPA, VHPA, MVPA, obesity, overweight, parental education level, ambulatory and non-ambulatory activity, ethnic groups), all of them conditioning the amount of physical activity that the subjects perform during recess. In general, younger children have higher levels of MVPA. It is also worth highlighting the evident difference that exists between sexes, since the levels of MVPA in boys are higher than those of girls. In some studies carried out by different authors in relation to the MVPA, girls performed 30.5%, while boys 47.0% [17]. Others studies found that the AFMV for boys was 17.9% and for girls 15.5% [4]. On the other hand, values of 30.1% versus 19.7%, respectively, were also found [18].
There were also significant effects in relation to programs and courses. Children's MVPA at recess was higher than in physical education classes. In the lower grades, a lower sedentary behavior was demonstrated, accompanied by an increase in the MVPA of the first grade children during recess. The MVPA performed at recess is up to 39% higher than that performed in physical education classes [2] and the same happens in another study where PE only contributed to the 6.4% of the overall MVPA of children, while recess time contributed 18.7% and the results indicate that more than half of the children are reaching the daily recommendations and that recess and PE contribute to 43% of the total activity, so they are important moments for conducting MVPA [19].
This systematic review yielded a total of 1018 search results, of which 998 were within the time period delimited by the inclusion/exclusion criteria. The extraction process for the studies included in this systematic review yielded a total of 43 articles. The studies were classified according to the methodological nature of the research: cross-sectional (n = 34), longitudinal (n = 3) and quasi-experimental (n = 6). A summary of the PRISMA flowchart of each stage of the search can be seen in Figure 1.    43.2% compared to 25.0%, differencing between boys and girls respectively. The MVPA increased in climbing activities and in those that were under supervision or with the use of equipment. If girls didn't have to compete for play spaces it would be a great opportunity for them to be more active, but boys tend to occupy them. This study demonstrated that items in PAQ can be specified to predict MVPA minutes in youth groups during out-of-school periods. Recess is the second most active segment of the day and the results obtained from accelerometry and PAQ were similar, with a total of 20.4 ± 23.8 in the percentage of MVPA performed.

Viciana et al. (2016);
Chile [24] To compare the levels of MVPA objectively measured in PE and recess as well as to examine the influence of sex and weight.  Recess has an average duration of 19.7 ± 6.5 min.
There are no significant differences for MVPA between overweight and non-overweight students, although there are notable differences in relation to MPA and VPA, and also between boys (49.5%) and girls (34.8%).

Baquet et al. (2014);
France [33] The objective of this study was to objectively measure the schoolchildren PA at recess and relate it to socioeconomic levels.  The students in the group with long rhythms performed higher MVPA than those with short rhythms, 9.8 ± 7.9 versus 3.9 ± 4.0. PA recommendations were also poorer in the short rhythm group, 30.7% vs. 34.1% of long rhythms.
The results suggest taking advantage of school time to promote PA.  To examine the objectively measured differences between PA and SED in adolescents.  Table 4 shows the main characteristics of the 3 longitudinal studies found. More specifically, a study on a marking program in the schoolyard, another on physical activity in the different seasons of the year and finally another that relates physical activity to recess and lunchtime. In general, over time, benefits are achieved in terms of increased physical activity. The marking system in the patio or the improvement of the patio areas favors the practice of physical activity, with significant changes just six months after the start [45]. Regarding the seasons of the year, the data reveal that the physical activity carried out decreases in the summer months (19.5%) and spring (24.4%), with the highest levels being in winter (27.7%) [46]. The analysis of the data also allows us to observe that girls are more active in dance activities, a low court or a climbing structure, while boys are more active in an obstacle course [47]. As can be seen, the duration of recess time varies from one study to another depending on the country where they were carried out, being 30 min in a single period in some studies [46] and 60 min distributed in four periods in other studies [47]. Evidence-based solutions for schoolyard design can help in an effort to provide good opportunities for children to be more physically active during recess. Courtyards with large asphalt areas offer good opportunities for children. It is suggested from the study that the schoolyards have multifunctional areas such as courts, marking areas or dance areas. Ball courts or marking zones generated AF points for both sexes. Girls are most active in dance activities, a low court, or a climbing frame, while boys are on an obstacle course. In the first school it was 9.2 (6.  Table 5 shows the main characteristics of the six intervention studies found. Two of the studies are based on the "Ready for recess" program, two on the "Playworks" program, one on a marking system in play areas and the last one on changes in the playing environment. Thanks to these programs, five of the six studies corroborate the importance of designing recess programs to increase the practice of physical activity, in which increases in physical activity carried out by students of up to 19.4% are observed. However, one of the studies, despite indicating that these interventions are positive, does not show a significant increase in the physical activity of the sample in the accelerometry devices [48]. According to "Ready for Recess" program, boys in FP + EQ increased their MVPA by 14.1% while girls remain slightly more sedentary after the intervention [49]. In another study, a significant increase in MVPA was also observed in the experimental group compared to the control group [7]. Furthermore, teachers in schools that made use of the "Playworks" program reported an increase in levels of physical activity, but accelerometry measures and student surveys did not show significant impacts [48]. Nevertheless, future studies should redirect their analysis in such a way that they can explore the underlying mechanisms in the intervention proposals that are carried out in schools [7]. Age (years), weight (kg), height (cm), body mass index (kg m −2 ), body mass index (%) 1300 to 1500 m 2 of recreation between asphalt and green areas Brands in the yard worth €15,000 Two 15 min recesses Design of the playground by zones with three specific areas of play. The red one (sports), the blue one (multiple activities) and the yellow one (fast zone) At the beginning, the PA of the CG is slightly higher than that of the EG. After the intervention, the PA of the EG (27 ± 9.9) increased slightly while that of the CG (29.5 ± 11.1) hardly varied.
Painting markings in the play area has positive effects on students' PA levels at recess The aim of this article is to investigate the effects of Playworks on girls and boys and physical activity separately using data collected using accelerometers and structured recess observations. n = 1573 29 primary schools 823 girls 750 boys

MVPA (%) Gender
The measurement was made during at least 10 min of recess. The recess was divided into different observation areas.
The person in charge of carrying out Playworks in each school tries to involve students in physical activity by providing equipment, such as balls and cones, introducing organized games within different areas or designated zones on the playground.
Girls who participate in the Playworks program are more active than those who do not participate, going from 27.5% to 33.3%. A significant impact was found on the types of activities that girls participated in during recess.
Playworks had a significant impact on girls' PA measurements, however it was not as significant in boys.

Discussion
The aim of this systematic review was to investigate the determining factors of active behavior during school recess, analyzing the levels of physical activity performed by children in this time frame and their impact on the global values of daily physical activity of the school-age population. Although the results of these studies show heterogeneity and not all of them provide significant evidence on the potential of the school as an institution that promotes physical activity, we can observe that most of the investigations agree that recess is a key moment to promote the levels of physical activity towards the recommended 60 min of daily MVPA. Other variables of physical activity that are reflected on in this research topic are light physical activity (LPA), moderate physical activity (MPA), vigorous physical activity (VPA) and sedentary behaviors (SED), expressed in minutes or in percentage, always in relation to the 60 min recommended by the World Health Organization and that guarantee objective measures that allow knowing real data about what is happening in schoolyards.
The realization of this update of the existing literature has had a triple approach, collecting cross-sectional, longitudinal and quasi-experimental studies, which implies an important strength of this systematic review, focused specifically on physical activity during school recess. The sample of students from the different schools has presented a great variety from some studies to others, ranging between 25 and 2024 subjects. Of the studies analyzed, 14 of them have been carried out in the United States, with the vast majority of the rest coming mainly from countries of the European Union.
From a global point of view we can appreciate the wide range of variables that are collected in the scientific literature and how each of them influences in relation to the levels of physical activity measured objectively with accelerometry, most of them being individual variables that directly affect the active behavior of the subject, such as age and gender. First, boys engage in more physical activity than girls, probably due to their nature, since boys tend to spend most of their time doing physical activities and competing with each other, while girls spend their time mostly in social activities [17,18]. In addition, it is recorded in numerous studies that boys tend to occupy the vast majority of play spaces and that the violence of their activities relegates girls to being located in the corners of the patio [4]. Second, we find that age is also a limiting factor. Younger children are generally significantly more active than older children. This is because as many of them grow up they lose interest and motivation in the practice of physical activity and that day-to-day activities considerably reduce the time they have for it [41]. That is why the need arises to design intervention programs that encourage physical activity both within the school environment and outside.
Although the objective of this study is to examine the amount of physical activity that children perform in relation to MVPA, we can find other variables that have been included by other studies presented here and that may be related with levels of physical activity at recess (such as the socioeconomic status of students or disorders such as obesity and autism).Children from families with a higher socioeconomic level perform more MVPA than those who live in families with a low socioeconomic level [30]. Other students who find their physical condition limited by the different disorders that are collected in the studies (whether they refer to eating disorders or others such as autism or the loss of some sense) also present lower levels of MVPA than those students exempt from them. The need arises to devise programs to adapt the centers to the different needs of their students and to create habits in students that allow them to develop physical activities both inside and outside of school [29].
The duration of recess is also a variable that we can find in all studies. In some schools, the recess time is divided into the morning recess and the lunch time recess, being typical of the schools that have a split day. Other studies are carried out in schools with up to four recess periods, which positively favors the performance of physical activity by students, since more than 40% of the children's MVPA is carried out at school. The time for recess for the sample of articles is between 10 and 60 min.
Regarding the intervention studies, it should be noted that all of them correspond to studies carried out under the RCT premise so that their results are not biased by manipulation of the groups and the results are obtained by a randomized controlled trial. The six quasi-experimental studies analyzed in this review sought to increase the physical activity of students through intervention programs or improvements in the school environment (i.e., "Ready for recess" program, "Playworks" program, "marking system in play areas"). It is clear that teachers and school policies can have a significant impact on the physical activity carried out by children during school recesses. Most of the strategies reported benefits for students regardless of gender. The work of teachers is necessary to carry out these programs to prevent boys from dominating the entire play space and to compensate for the difference that exists with girls in increasing levels of physical activity. Some strategies such as the implementation of markings in the playground, the remodeling of the play areas or the implementation of play equipment are configured as potential strategies to increase the levels of physical activity in the school environment, but there is a need to continue working on it. The findings currently found are on the right track.

Conclusions
School recess is a space to potentially improve levels of daily physical activity in the lives of children. Gender and age are determinants, as at all levels girls are significantly less active than boys and prone to develop sedentary behaviors. As they grow, levels of physical activity also decreased, the smallest people performing the greatest MVPA. There are other variables that act positively or negatively related to physical activity levels present in children. Socioeconomic status is a variable to be highlighted because children with higher SES spend more time in MVPA. There is a need to advocate for interventions at school to promote moderate to vigorous physical activity (i.e., "Ready for recess" program, "Playworks" program, "marking system in play areas"). Environmental changes are as strong as the staff implements. Supervision is crucial because if it is not interactive, it can be detrimental to the PA, especially girls. Teachers should pay special attention to schedules to have more control of the variables of physical activity.
In conclusion, the school context is configured as a key space to meet the daily recommendations for physical activity. Studies show that by improving the conditions of the courtyard, providing it with equipment equipped according to the needs of students, improve play areas and designing programs that promote physical activity are to achieve positive improvements in levels of MVPA. For future research, it would be very useful for the scientific community to continue studying the relationship between the variables of physical activity and school recess, specifically in quasi-experimental studies, in which the capacity of these intervention programs to increase the levels of physical activity up to those recommended is analyzed, since most of the articles are cross-sectional in nature but there are very few longitudinal and intervention ones, in which this need for RCTs should be qualified as a criterion for the research design.

Conflicts of Interest:
The authors declare no conflict of interest.