Individual- and Environmental-Level Predictors of Recess Activity and Sedentary Behavior: Findings from the I-CAN! Study : Translational Journal of the American College of Sports Medicine

Secondary Logo

Journal Logo

Controlled Trial

Individual- and Environmental-Level Predictors of Recess Activity and Sedentary Behavior: Findings from the I-CAN! Study

Bartholomew, John B.1; Clutton, Jon1; Burford, Katie2; Aadland, Eivind3; Resaland, Geir Kare4; Jowers, Esbelle M.1; Errisuiz, Vanessa1,5

Author Information
Translational Journal of the ACSM: Fall 2022 - Volume 7 - Issue 4 - e000212
doi: 10.1249/TJX.0000000000000212
  • Free

Abstract

INTRODUCTION

Physical activity provides a host of lifelong health benefits that begin in childhood (1–3). As such, “The physical activity guidelines for Americans” (4) recommend that children and youth participate in ≥60 min of moderate-to-vigorous physical activity (MVPA) daily. However, 76% of children and youth fail to engage in this optimal dose of daily activity. Given the time children spend at school and the range of opportunities for intervention, schools have been recognized as one of the critical settings in which to intervene (5). Recess is a key opportunity to affect physical activity in children, especially because it does not interfere with academic class time (6,7). Unlike other school-based physical activity opportunities, recess increases physical activity through free play while simultaneously benefitting children’s cognitive, social, and emotional competencies (8,9). Free play allows children to enhance their creativity, experience freedom, build leadership, and learn problem-solving skills (9–11). Apart from lunch, recess is one of the few unstructured contexts in school.

Although there is clear benefit for children, at the time these data were collected (2012–2015), 60% of U.S. school districts had no recess policy, and only 22% of school districts in the U.S. required daily recess (12). Thus, physical activity during recess is highly variable, with studies reporting that children spend between 15% and 47% of time in MVPA (6,13–15). Understanding this variation in physical activity during recess may stimulate the establishment of national recess policies to support greater opportunities for physical activity. Social–ecological models are used by researchers to determine correlates and predictors of physical activity (16,17). These models are based on the work of Bronfenbrenner’s Ecological Systems Theory (18), which hypothesizes that there are multiple levels of influence from the individual to the micro (social), meso (environment), exo (policy, media, etc.), and macro (cultural) systems. For physical activity during recess, the emphasis has been on the individual, environmental, and policy levels of influence (6,15,19–23). As we better understand the influence of each of these levels on physical activity during recess, we can then tailor interventions to enhance physical activity.

Unfortunately, there is a lack of consensus regarding the correlates of physical activity during recess (15,19). Previous studies have consistently associated recess physical activity with gender (i.e., boys are more active than girls) and increased levels of unfixed play equipment (e.g., number of balls or jump ropes) (15,19,24,25). However, mixed results have been shown for age, socioeconomic status (SES), ethnicity, body mass index (BMI), and aerobic fitness (15,25–27). Similarly, associations with several environmental and school-level factors, including weather, fixed equipment, total outdoor space, and number of outdoor physical activity areas (i.e., sports fields, ball areas, green spaces), have been largely inconclusive (15,19,21–23). Intervention studies have supported the importance of environmental factors for increased MVPA, such as increasing playground markings and adding physical structures (e.g., soccer goals and basketball hoops), with the least active children showing the greatest improvement (20,28,29). However, several potential changes to the built environment, such as increasing the number of playground areas or providing sedentary features (i.e., benches), have been understudied and may serve to increase time spent sedentary rather than active. Among school-policy predictors, recess duration and timing (before or after lunch) were the most commonly studied factors, but results are, again, inconclusive (15,19). There are also a number of limitations to the existing research base that must be addressed and might lead to more firm conclusions.

The most common limitations within the existing literature are the sample of schools studied and the measure of physical activity. Most existing cross-sectional studies only assess one school and, thus, lack variation across multiple settings (15). This not only limits the generalizability of the resulting data but also leads to underpowered designs that prevent any meaningful subgroup comparisons and prevents appropriate analysis. For example, the bulk of the existing data do not consistently support hypothesized school-level factors (i.e., number of play areas, size of play area). These can only be assessed via a hierarchical model, which requires a larger number of schools. The second major concern is the overreliance on observational methods to assess MVPA. Although these studies enhance the depth of data collected, they lack the precision of accelerometer-based assessment and the ability to conduct subgroup analyses. Finally, there has been insufficient study of sedentary behavior in studies of recess. Thus, there is a need for research that uses objective measures of physical activity and sedentary behavior along with environmental- and individual-level models to estimate determinants of recess physical activity.

The primary purpose of this study is to elucidate how school-level variables (i.e., duration of recess, recess before or after lunch, the recess environment, playground spaces available, playground space per student, bad weather days) and individual-level variables (i.e., age, gender, SES, ethnicity, fitness) affect children’s physical activity and sedentary time during recess. The secondary purpose is to address existing measurement and sample limitations by analyzing accelerometer data from a large sample of students, drawn from 14 elementary schools, to allow for meaningful subgroup analyses.

METHODS

Study Sample

Data for the current study were obtained from the Texas Initiative for Children’s Activity and Nutrition (I-CAN!) study (30). This was a 3-yr (2012–2015) school-based physical activity intervention designed to introduce physically active, academic lessons to fourth-grade students in central Texas. Schools (n = 28) were recruited from four suburban school districts and randomly assigned to condition (n = 19 intervention and n = 9 control) using a random number generator; there were no statistically significant differences for participant characteristics between conditions (31). To prevent differences in recess activity that might have accrued between the intervention and the control schools while maximizing the sample, we limited the present analysis to intervention schools. Of these, five did not provide full data (e.g., lack of demographics, most recent fitness scores, or set time for recess) and were eliminated from the analysis. The final sample consisted of 14 schools and 1049 students. Of the 14 schools, 5 completed data collection in 2013, 6 in 2014, and 3 in 2015. This included every fourth-grade classroom within each school, and nearly 90% of children provided consent and were included in the study. All study protocols were approved by the University of Texas at Austin Institutional Review Board (2017-01-0069), with permission to collect data from the school district, school principal, and teachers before beginning the study and active parent consent obtained for all children.

Measures

Individual-Level Variables

Accelerometer-measured physical activity

Physical activity data were collected during school hours over the course of one school week (five consecutive days). Schools were randomly assigned to have their students’ data collected during the Fall or Spring semester of the academic year. Children’s physical activity was assessed with a triaxial accelerometer (model GT3X+; ActiGraph LLC, Pensacola, FL) (32). To ensure a complete collection of school-day physical activity, accelerometers were distributed at the beginning of the school day by research staff and taken off as students left class at the end of each school day. The accelerometers were worn in an elastic belt around the waist, positioned on the right hip (33). Data were collected in 5-s epochs to best capture children’s sporadic activity (34). Periods of greater than 90 min of zero counts were defined as non–wear time (35). Physical activity data were downloaded onto a computer and analyzed with ActiLife software (v6.13.3, ActiGraph LLC). In ActiLife, filters were applied to calculate the average minutes per day and percent time spent sedentary and in MVPA during recess as determined using Evenson cut points (i.e., MVPA ≥2296 counts per minute) (36,37). For students with fewer than 5 d of recess sessions (e.g., absent 1 d), weekly time sedentary and in MVPA for the existing days was extrapolated to 1 wk.

Demographic variables

Participant demographic information (i.e., sex, age, race/ethnicity, and eligibility for free/reduced lunch) was obtained through school records. Eligibility for free/reduced lunch was used as a proxy for low SES. Although the validity of free/reduced lunch as a proxy for SES has been contested (38), it has been commonly used as a measure for SES and is highly related to other community SES measures (39).

Cardiorespiratory fitness and BMI

Cardiorespiratory fitness (V̇O2max) and BMI were collected according to FITNESSGRAM® standards by school staff (32,40). Height was measured rounding down to the nearest 1/4 inch with students standing shoeless, facing forward. Body weight was measured rounding down to the lowest tenth of a pound using a calibrated scale without shoes or excessive clothing. BMI was calculated by dividing weight (kg) by height (m) squared as part of the FITNESSGRAM® protocol. The Progressive Aerobic Cardiovascular Endurance Run (20-m PACER) test (number of laps) was used as a proxy for aerobic fitness (V̇O2max). The 20-m PACER is a validated and reliable, progressively challenging aerobic test completed yearly by physical educators in Texas (41). These data were obtained from school records for all participating students and were collected in the same academic year as the collection of accelerometry data.

School-Level Environment Variables

Factors within the recess environment, benches, and area size

Recess environment characteristics for each school were collected in the spring of 2017, via direct observation after school hours. Because this occurred after the completion of the I-CAN! study, school personnel were consulted to confirm that no additions were made to the recess environment in the interim. Schoolyard permanent area improvements, or the number of environmental factors to the recess area, were coded according to the mapping strategies component of the validated System for Observing Play and Leisure Activity in Youth (42). This included area factors that supported physical activity (e.g., lines painted on courts, basketball hoops, and soccer goals) and those that supported sedentary behavior (benches or other seating areas). To deal with overlapping improvements, we counted the highest number of environmental factors that could be used at one time. We also noted the total number of water fountains. Estimates of playground spatial space (1000 ft2) were found using aerial pictures and the polygon measurement tool in Google™ Earth Pro software (Google LLC, Mountain View, CA). Average area was calculated from three estimations of playground spatial area.

Bad weather days

The number of bad weather days during collection of accelerometer data was determined using time-stamped archived weather data from Weather Underground (43). This Web site stores daily weather indicators (temperature, precipitation, etc.) over time, which were used to identify weather on the days of the original accelerometry assessments. Bad weather was considered as temperatures below 40°F, above 100°F, or any amount of precipitation during recess time.

Recess length and recess time of day

Recess length (min) was provided by school records, as was recess time of day, and categorized into two variables: 1) scheduled directly before or after lunch versus scheduled apart from lunch, and 2) scheduled in the morning (before 12 pm) versus the afternoon (after 12 pm). Because accelerometers provide time-stamped data, this was used to calculate the time students spent during the recess period. Although an accurate means to assess scheduled recess time, this prevented us from capturing situations where the school changed schedule. To minimize this effect, we did not collect data during weeks including a school assembly, which often causes broad shifts in schedule. This method also did not capture any instance where individual children were removed from recess for disciplinary or academic reasons.

Statistical analysis

We reported descriptive statistics as means and standard deviations (SD), percentages (of children), or numbers (of schools). Association analyses were performed using a linear mixed model including school as a random effect. The analyses were performed in four steps. First, we determined the crude associations for child characteristics by including each variable in a separate model. Second, we included all child characteristics in one model to mutually adjust these variables for each other. Third, we determined the crude associations for school characteristics by including each variable in a separate model, including adjustment for all child characteristics. Fourth, we included all school and child characteristics in one model to mutually adjust these variables for each other. Intraclass correlation coefficients (ICC), showing the clustering of MVPA within schools, were reported for the full child and school models. All associations were reported as unstandardized regression coefficients (B) and 95% confidence intervals (CI) along with their P values. All analyses were performed using SPSS v. 26 (IBM Corp, Armonk, NY). A two-sided P value ≤0.05 was considered statistically significant.

RESULTS

Individual and School Environment Characteristics

The final sample included 1049 children (48% female, 43% non-White) from 14 schools (27–128 children per school) who provided valid data on all variables of interest in the analysis. The children’s characteristics are shown in Table 1 (44), with a range of ethnicity and SES that mirrors the total population of Central Texas. In total, 89.6% of the children had four or five valid days of recess accelerometer monitoring and spent a mean (range) of 5.9 (0.1–17.3) min·d−1 and 23% (0%–69%) of recess time in MVPA and 10.3 (1.1–26.2) min·d−1 and 42% (5%–90%) of recess time sedentary. To place this in context, the mean MVPA across the full day for these participants was 30.05 min (SD = 9.99), and the mean for total sedentary time was 341.65 min (SD = 36.29). Thus, recess provided approximately 20% of all MVPA during the school day and 3% of total sedentary time.

TABLE 1 - Individual characteristics.
Total Boys Girls
n 1049 546 503
Age (years) 9.6 (0.5) 9.7 (0.5) 9.6 (0.5)
BMI (kg/m2) 18.3 (3.8) 18.3 (3.7) 18.3 (3.9)
Weight status (%)a
 Normal weight 73.2 72.7 73.8
 Overweight 16.4 17.0 15.7
 Obese 10.4 10.3 10.5
Ethnicity (%)
 White 56.7 60.1 53.1
 Not white 43.3 39.9 46.9
SES (%)
 No lunch support 76.3 76.2 76.3
 Eligible for free lunch 23.7 23.8 23.7
PACER (laps) 31.8 (16.1) 35.7 (17.6) 27.5 (13.1)
MVPA
 Time (min/day) 5.9 (3.6) 6.7 (3.9) 5.1 (3.0)
 Proportion of recess (%) 23 (12) 26 (13) 20 (10)
SED
 Time (min/day) 10.3 (4.5) 9.7 (4.6) 10.9 (4.3)
 Proportion of recess (%) 42 (16) 40 (17) 44 (16)
Numbers are presented as mean (standard deviation) unless otherwise noted. BMI, body mass index; MVPA, moderate-to-vigorous physical activity; SED, sedentary behavior; SES, socioeconomic status.
aClassified according to International Obesity Task Force criteria (44)

Recess environment characteristics are shown in Table 2. The scheduling of recess varied across the schools with recess lengths from 15 to 30 min·d−1. The ranges of area size, the number of area improvements, the number of sedentary upgrades, and the bad weather days were 25,037–188,660 ft2, 4–27 items, 0–18 items, and 0–3 d, respectively.

TABLE 2 - School Environment Characteristics.
Mean (SD) or n
Scheduling of recess
Recess vs lunch (n)
 With lunch (0) 7
Recess time of day (n)
 Before 12 pm (0) 9
After 12 pm (1) 5
 Recess length (min·d−1)
 15 2
 20 5
 25 1
 30 6
Physical environment
Area size (1000 ft2) 93.0 (39.4)
Area improvements (count) 17.3 (7.0)
Sedentary upgrades (count) 7.5 (5.7)
Water fountain (n)
 Yes 10
 No 4
Bad weather days (count) 1.2 (1.2)
For the schedule of recess and the water fountains, n indicates the number of schools that met the scheduling criteria listed. For the remaining indicators of the physical environment, data are presented as mean (SD).

Associations with MVPA and Sedentary Behavior

The absolute values of time spent in MVPA and sedentary (min·d−1) were significantly and directly associated with recess length (MVPA: B (95% CI) = 0.29 (0.11–0.46), P = 0.004 in a bivariate model, and 0.27 (0.08–0.46), P = 0.009 when controlling for child characteristics; sedentary: B (95% CI) = 0.34 (0.07–0.61), P = 0.017 in a bivariate model, and 0.34 (0.07–0.62), P = 0.019 when controlling for child characteristics). All further analyses were performed using percent MVPA and sedentary during recess as the outcome variable to take this association into account.

Of the child characteristics (Table 3), gender and aerobic fitness were directly associated with MVPA and sedentary behavior (P ≤ 0.002). Specifically, boys spent 5.9% more of their recess in MVPA and 4.2% less of their recess sedentary compared with girls. MVPA increased by 0.13% and sedentary behavior decreased by 0.10% for every lap completed of the PACER test. The ICC values for school in the full child level models were 0.39 for MVPA and 0.43 for sedentary. Of the school characteristics (Table 4), the number of environmental supports for activity was associated with MVPA (1.2% increased MVPA per item) and sedentary behavior (1.7% decreased sedentary behavior per item), whereas sedentary/seating areas showed the opposite pattern of effects with percent MVPA (2.0% decreased MVPA per item) and sedentary behavior (2.8% increased sedentary behavior per item) (Table 4). The ICC values for school in the full model were 0.30 for MVPA and 0.25 for sedentary behavior.

TABLE 3 - Associations for Individual Characteristics with Percent MVPA and SED.
Crude Model a Mutually Adjusted Model b
Coefficient 95% CI P ES Coefficient 95% CI P ES
MVPA
 Age (yr) 0.43 −0.77 to 1.64 0.48 0.04 −0.35 −1.48 to 0.78 0.54 −0.03
 Gender (ref = boy) −6.91 −8.05 to −5.77 <0.001 −0.58 −5.94 −7.11 to −4.77 <0.001 −0.50
 Ethnicity (ref = White) −0.47 −1.84 to 0.91 0.50 −0.04 0.16 −1.13 to 1.45 0.81 0.01
 SES (ref = low) 0.77 −1.03 to 2.57 0.40 0.06 0.63 −1.07 to 2.33 0.47 0.05
 BMI (kg·m−2) −0.08 −0.25 to 0.09 0.36 −0.05 0.12 −0.05 to 0.28 0.17 0.08
 PACER (laps) 0.17 0.13 to 0.21 <0.001 0.46 0.13 0.09 to 0.17 <0.001 0.35
SED
 Age (yr) −0.67 −2.29 to 0.95 0.42 −0.04 −0.12 −1.72 to 1.48 0.88 −0.01
 Gender (ref = boy) 4.84 3.24 to 6.44 <0.001 0.29 4.15 2.49 to 5.81 <0.001 0.25
 Ethnicity (ref = White) −0.93 −2.77 to 0.90 0.32 −0.06 −1.40 −3.24 to 0.43 0.13 −0.09
 SES (ref = low) −0.47 −2.88 to 1.94 0.70 −0.03 −0.67 −3.08 to 1.74 0.59 −0.04
 BMI (kg·m−2) 0.11 −0.11 to 0.34 0.32 0.05 −0.03 −0.26 to 0.21 0.82 −0.01
 PACER (laps) −0.13 −0.18 to −0.08 <0.001 −0.26 −0.10 −0.15 to −0.04 0.002 −0.19
a Crude model: associations for each variable separately and including the random intercept of school.
b Mutually adjusted model: associations mutually adjusted for all child characteristics and including the random intercept of school.
ref, reference; SED, sedentary behavior.

TABLE 4 - Associations for School Environment Characteristics with Percent MVPA and SED.
Crude Model a Mutually Adjusted Model b
Coefficient 95% CI P ES Coefficient 95% CI P ES
MVPA
Scheduling of recess
 Recess vs lunch (ref = apart from lunch) 2.93 −6.01 to 11.88 0.49 0.25 11.8 −4.58 to 28.25 0.12 1.00
 Recess time of day (ref = after 12 pm) −1.87 −11.32 to 7.57 0.67 −0.16 11.3 −2.88 to 25.55 0.10 0.96
Physical environment
 Area size (1000 ft2) −0.02 −0.13 to 0.09 0.65 −0.16 −0.01 −0.15 to 0.13 0.88 −0.06
 Area improvements (count) 0.33 −0.33 to 1.00 0.30 0.39 1.23 0.14 to 2.33 0.03 1.45
 Sedentary upgrades (count) −0.07 −0.87 to 0.72 0.84 −0.07 −2.04 −3.83 to -0.25 0.03 −1.98
 Water fountain (ref = no) −4.79 −14.42 to 4.84 0.30 −0.40 −1.20 −16.64 to 14.25 0.85 −0.10
 Bad weather days (count) −2.00 −5.29 to 1.29 0.21 −0.40 −2.84 −6.46 to 0.77 0.10 −0.56
SED
Scheduling of recess
 Recess vs lunch (ref = apart from lunch) −5.67 −19.21 to 7.87 0.38 −0.34 −18.41 −39.55 to 2.72 0.08 −1.12
 Recess time of day (ref = after 12 pm) 5.54 −8.64 to 19.72 0.41 0.34 −10.88 −29.19 to 7.42 0.19 −0.66
Physical environment
 Area size (1000 ft2) 0.05 −0.11 to 0.22 0.49 0.26 0.07 −0.10 to 0.25 0.33 0.36
 Area improvements (count) −0.52 −1.54 to 0.50 0.29 −0.44 −1.73 −3.14 to -0.32 0.03 −1.46
 Sedentary upgrades (count) 0.03 −1.20 to 1.26 0.96 0.02 2.80 0.50 to 5.11 0.03 1.96
 Water fountain (ref = no) 7.72 −6.98 to 22.41 0.28 0.47 −4.82 −24.70 to 15.05 0.56 −0.29
 Bad weather days (count) 3.45 −1.49 to 8.39 0.15 0.49 4.61 −0.05 to 9.27 0.05 0.66
a Crude model: associations for each variable separately with adjustment for all child characteristics and including the random intercept of school.
b Mutually adjusted model: associations mutually adjusted for all school characteristics with adjustment for all child characteristics, group (intervention vs control), and including the random intercept of school.
ref, reference; SED, sedentary behavior.

DISCUSSION

This study was designed to examine variation in the levels of physical activity and sedentary behavior during the recess period across multiple elementary schools, and to assess both individual- and school-level correlates of physical activity and sedentary behavior during recess. Results indicated that children, on average, spent approximately 25% of their time during recess in MVPA, which is similar to previous studies regardless of national setting (15,22,28,45–47). With regard to individual-level predictors, boys were more (d = 0.43) active in recess than girls. Fitness, as measured by laps completed during the PACER test, was positively associated with percentage of time spent in MVPA during recess. The pattern of effects for sedentary time largely mirrored the effects of MVPA. Fitter children and boys spent less time sedentary, and schools with more benches had children who spent more time sedentary during recess. Other studies have not found such a clear relationship between MVPA and sedentary time because their measurements are often confounded by the inclusion of time at lunch with recess (19). This study was able to limit the assessment of MVPA and sedentary behavior to recess alone, which we hope provides a more accurate reflection of the relationship between these variables. The difference between boys and girls is one of the most common findings in this literature (15). Although it has consistently been hypothesized that children who are more aerobically fit would be more active, regardless of gender, this has received surprisingly little research. Research instead has centered on the relationship between fitness and overall engagement in physical activity (48,49). At the time of the initial systematic review by Ridgers et al. (19), there was only one study that directly tested the association between fitness and physical activity during recess, and this study focused on students with learning challenges (50). Since then, we have found only one study examining this association (51). In that sample of 256 fifth-grade children, fitness was shown to be associated with physical activity during recess, but only for girls (51). As such, the present data provide the largest assessment to date of aerobic fitness and physical activity in elementary children. No other demographic variables (i.e., race, BMI, or SES) were associated with activity at recess.

There were two important school-level correlates of percent time spent in MVPA and sedentary during recess. Time spent in MVPA and sedentary was directly associated with the number of factors in the recess environment that supported movement (e.g., playground equipment and sport courts) and indirectly associated with the number of factors that support sedentary behavior (e.g., benches). Variation in recess MVPA between schools has been estimated to be as high as 40% (15), with counts of environmental factors generally being directly associated with physical activity (52). By contrast, the effect of interventions to enhance the recess infrastructure is more mixed, explaining about 5% of the variance in total MVPA (15,19,20,29). We would expect the cross-sectional and intervention outcomes to be more similar. The difference between these study designs may be explained by differences in the type of improvements implemented in the intervention research (29). For example, it appears that the increase in physical activity varies widely for boys and girls based on the types of improvements made (23). In addition, if these improvements include increased seating areas, then they are likely to undermine physical activity. Clearly, future research should be designed to compare forms of environmental support for physical activity at recess to guide the most effective improvement strategies. We did not find an effect for the size of the recess area and time spent in either MVPA or sedentary behavior. This may be because the 14 included schools were all from suburban areas and had large areas dedicated to recess. Future studies might purposely recruit schools with a range of recess areas to better test this question and to collect specific data on student density and movements within the recess space.

Our findings confirm one of the primary challenges with conceptualizing recess as an opportunity for physical activity. The percent of time spent in MVPA is greater in boys and more fit children than in girls and less fit children (i.e., children who are more likely to be active in recess and also more likely to be active outside of school) (16). This finding is comparable with findings for younger children, where boys and older and more active children benefited more from the preschool setting than their peers (53). As a result, an expansion of recess is not likely to achieve significant increases in MVPA for less fit children who are the target of public health efforts to increase activity. This is in contrast to adult-led activity interventions that occur in school, which have been shown to increase MVPA across all subgroups (31). This, of course, should not be considered an argument against the provision of recess. First, our data do indicate that more time in recess is associated with increased time spent in MVPA, if primarily for those students who tend to already be active. Second, although a modest amount of time is spent in MVPA across children, recess remains an important contributor to total day physical activity in children (54). Third, recess provides one of the only school-based opportunities for free play and unregulated social interaction (8). Given the large amount of time spent in school, this time is critical for maintenance of student focus, development of social skills, and creativity to the extent that recess allows for student directed, free play (9,55,56). Thus, there is strong evidence for the maintenance of daily recess that is structured around free play. The findings from this study merely suggest that an expansion of recess time is not likely to result in significant increases in MVPA for children who are less active. In addition, interventions to enhance the other benefits of recess may serve to reduce physical activity. For example, there have been efforts toward directed physical activity during recess in adolescents, which results in increased MVPA (57). This, however, is likely to undermine the social and emotional development that is the goal during free-play recess. Likewise, although our data indicate the number of benches to be associated with reduced physical activity, the addition of benches has been used to reduce social isolation at recess (58). Thus, although physical activity occurs for many children during recess, the broader goals of recess may be incompatible with any efforts to increase the health benefits of recess.

Limitations

The interpretation of these results should consider a number of limitations. We ascribed recess as a function of time-stamped, accelerometer data within the time assigned according to each school’s schedule. It is possible that school schedules shifted. To minimize this possibility, we did not collect data during weeks with a school assembly, which is a common reason for a shift in school schedule. It is more likely that we have included some children who were held back from recess due to disciplinary or academic reasons because we did not collect these data. Given the debate with regard to adjustment for type 1 errors (59,60), we did not adjust our findings for multiple comparisons. Thus, our findings should be interpreted with a possible increased type-1-error rate in mind. Although the number of participants (n = 1049) would be considered a strength, these children were drawn from only 14 schools. Because our goal was to assess the effect of school-level variables on MVPA during recess, this provides limited variance in these factors, and the data should be considered preliminary. For example, all of our schools were drawn from suburban districts and had at least minimally sufficient recess space and equipment. It may be that the effect of SES on activity would only be seen in settings with poor space and equipment. Other studies have collected data from more schools (n = 22), but fewer students (n = 408), which would provide a stronger test of these factors (61). We used PACER and BMI data collected by the school’s physical education teachers. This holds two limitations. Although teachers receive training in these methods, the accuracy of these data may vary across schools. In addition, although the data collection occurred in the same academic year as the accelerometer data, they were not tightly linked. Ideally, these data would be collected by research staff in conjunction with the outcome measures. Finally, these data were limited to MVPA and sedentary behavior during recess. This fails to consider MVPA in other times during and outside of the school day that might be more important for child outcomes.

Conclusions

Although there have been larger studies internationally (61), this is the largest U.S. study to date to assess physical activity during recess using objective measures (15). Results replicated the long-standing finding that boys are more active during recess than girls. More importantly, it found that the percent of time spent in MVPA during recess was directly associated with aerobic fitness, as measured by the 20-m PACER. Although long hypothesized, this is the first large-scale study to test this relationship. Moreover, these data supported existing research on the direct association between the infrastructure for recess and the percent of time in MVPA. We break new ground in demonstrating the negative effect of sedentary upgrades on percent of time in MVPA. Thus, although we strongly support the use of recess to support social and emotional development, we are less enthusiastic about the potential of recess to increase MVPA and reduce sedentary behavior in less fit children and question the provision of infrastructure changes that invite sedentary behaviors.

This study was funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (award no. 1R01HD070741). The authors report no conflict of interest. The results of this study do not constitute endorsement by the American College of Sports Medicine.

REFERENCES

1. Chaput J-P, Willumsen J, Bull F, et al. 2020 WHO guidelines on physical activity and sedentary behaviour for children and adolescents aged 5–17 years: summary of the evidence. Int J Behav Nutr Phys Act. 2020;17(1):141.
2. Janssen I, LeBlanc AG. Systematic review of the health benefits of physical activity and fitness in school-aged children and youth. Int J Behav Nutr Phys Act. 2010;7(1):40.
3. Poitras VJ, Gray CE, Borghese MM, et al. Systematic review of the relationships between objectively measured physical activity and health indicators in school-aged children and youth. Appl Physiol Nutr Metab. 2016;41(6 Suppl 3):S197–239.
4. Piercy KL, Troiano RP, Ballard RM, et al. The physical activity guidelines for Americans. JAMA. 2018;320(19):2020.
5. Pate RR, Davis MG, Robinson TN, et al. Promoting physical activity in children and youth: a leadership role for schools: a scientific statement from the American Heart Association Council on nutrition, physical activity, and metabolism (physical activity committee) in collaboration with the councils on cardiovascular disease in the young and cardiovascular nursing. Circulation. 2006;114(11):1214–24.
6. Erwin HE, Ickes M, Ahn S, Fedewa A. Impact of recess interventions on children’s physical activity—a meta-analysis. Am J Health Promot. 2014;28(3):159–67.
7. van Sluijs EMF, Ekelund U, Crochemore-Silva I, et al. Physical activity behaviours in adolescence: current evidence and opportunities for intervention. Lancet. 2021;398(10298):429–42.
8. Ramstetter CL, Murray R, Garner AS. The crucial role of recess in schools. J Sch Health. 2010;80(11):517–26.
9. Waite-Stupiansky S, Findlay M. The fourth R: recess and its link to learning. Educ Forum. 2002;66(1):16–25.
10. Jarrett O, Maxwell D. What research says about the need for recess. In: Clements RL, editor. Elementary School Recess: Selected Readings, Games, and Activities for Teachers and Parents. Lake Charles (LA): American Press; 2000. pp. 12–23.
11. Rogers CS, Sawyers JK. Play in the lives of children. Washington (DC): National Association for the Education of Young Children; 1988. 135 p.
12. National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Division of Adolescent and School Health. Results from the School Health Policies and Practices Study 2014. Washington (DC): Centers for Disease Control and Prevention; 2015. 180 p.
13. Ridgers ND, Stratton G, Fairclough SJ. Assessing physical activity during recess using accelerometry. Prev Med. 2005;41(1):102–7.
14. Ridgers ND, Saint-Maurice PF, Welk GJ, et al. Differences in physical activity during school recess. J Sch Health. 2011;81(9):545–51.
15. Pulido Sánchez S, Iglesias Gallego D. Evidence-based overview of accelerometer-measured physical activity during school recess: an updated systematic review. Int J Environ Res Public Health. 2021;18(2):578.
16. Sallis JF, Prochaska JJ, Taylor WC. A review of correlates of physical activity of children and adolescents. Med Sci Sports Exerc. 2000;32(5):963–75.
17. Sallis JF, Owen N. Ecological models of health behavior. In: Glanz A, Glanz K, Rimer BK, Lewis FM, editors. Health Behavior and Health Education: Theory, Research and Practice. 3rd ed. San Francisco (CA): Jossey-Bass; 2002. pp. 462–82.
18. Bronfenbrenner U. Ecological systems theory. In: Vasta R, editor. Six Theories of Child Development: Revised Formulations and Current Issues. London: Jessica Kingsley Publishers Ltd; 1992. pp. 187–249.
19. Ridgers ND, Salmon J, Parrish A-M, et al. Physical activity during school recess: a systematic review. Am J Prev Med. 2012;43(3):320–8.
20. Suga ACM, da Silva AAP, Brey JR, et al. Effects of interventions for promoting physical activity during recess in elementary schools: a systematic review. J Pediatr (Rio J). 2021;97(6):585–94.
21. Baquet G, Aucouturier J, Gamelin FX, Berthoin S. Longitudinal follow-up of physical activity during school recess: impact of playground markings. Front Public Health. 2018;6:283.
22. Ridgers ND, Salmon J, Timperio A. Seasonal changes in physical activity during school recess and lunchtime among Australian children. J Sports Sci. 2018;36(13):1508–14.
23. Andersen HB, Christiansen LB, Pawlowski CS, Schipperijn J. What we build makes a difference—mapping activating schoolyard features after renewal using GIS, GPS and accelerometers. Landsc Urban Plan. 2019;191:103617.
24. Parrish A-M, Okely AD, Stanley RM, Ridgers ND. The effect of school recess interventions on physical activity: a systematic review. Sports Med. 2013;43(4):287–99.
25. Ridgers ND, Timperio A, Crawford D, Salmon J. Five-year changes in school recess and lunchtime and the contribution to children’s daily physical activity. Br J Sports Med. 2012;46(10):741–6.
26. Gonzalez-Suarez CB, Grimmer-Somers K. Physical activity pattern of prepubescent Filipino school children during school days. J Sch Health. 2009;79(7):304–11.
27. Parrish A-M, Iverson D, Russell K, Yeatman H. Observing children’s playground activity levels at 13 Illawarra primary schools using CAST2. J Phys Act Health. 2009;6(Suppl 1):S89–96.
28. Ridgers ND, Stratton G, Fairclough SJ, Twisk JWR. Long-term effects of a playground markings and physical structures on children’s recess physical activity levels. Prev Med. 2007;44(5):393–7.
29. Parrish A-M, Chong KH, Moriarty AL, et al. Interventions to change school recess activity levels in children and adolescents: a systematic review and meta-analysis. Sports Med. 2020;50(12):2145–73.
30. Bartholomew JB, Jowers EM, Errisuriz VL, et al. A cluster randomized control trial to assess the impact of active learning on child activity, attention control, and academic outcomes: the Texas I-CAN trial. Contemp Clin Trials. 2017;61:81–6.
31. Bartholomew JB, Jowers EM, Roberts G, et al. Active learning increases children’s physical activity across demographic subgroups. Transl J Am Coll Sports Med. 2018;3(1):1–9.
32. Welk GJ, Schaben JA, Morrow JR. Reliability of accelerometry-based activity monitors: a generalizability study. Med Sci Sports Exerc. 2004;36(9):1637–45.
33. Trost SG, Mciver KL, Pate RR. Conducting accelerometer-based activity assessments in field-based research. Med Sci Sports Exerc. 2005;37(11 Suppl):S531–43.
34. Nettlefold L, Naylor PJ, Warburton DER, et al. The influence of epoch length on physical activity patterns varies by child’s activity level. Res Q Exerc Sport. 2016;87(1):110–23.
35. Choi L, Liu Z, Matthews CE, Buchowski MS. Validation of accelerometer wear and nonwear time classification algorithm. Med Sci Sports Exerc. 2011;43(2):357–64.
36. Evenson KR, Catellier DJ, Gill K, et al. Calibration of two objective measures of physical activity for children. J Sports Sci. 2008;26(14):1557–65.
37. Trost SG, Loprinzi PD, Moore R, Pfeiffer KA. Comparison of accelerometer cut points for predicting activity intensity in youth. Med Sci Sports Exerc. 2011;43(7):1360–8.
38. Harwell M, LeBeau B. Student eligibility for a free lunch as an SES measure in education research. Educ Res. 2010;39(2):120–31.
39. Nicholson A, Lowe MC, Parker J, et al. Systematic review and meta-analysis of enhanced recovery programmes in surgical patients. Br J Surg. 2014;101(3):172–88.
40. Plowman SA. Muscular strength, endurance, and flexibility assessments. In: Plowman SA, Meredith MD, editors. Fitnessgram/Activitygram Reference Guide. 4th ed. Dallas (TX): The Cooper Institute; 2013. pp. 8-1–55.
41. Morrow JR Jr., Martin SB, Jackson AW. Reliability and validity of the FITNESSGRAM: quality of teacher-collected health-related fitness surveillance data. Res Q Exerc Sport. 2010;81(3 Suppl):S24–30.
42. Mckenzie T. System for Observing Play and Leisure Activity in Youth (SOPLAY) [Internet]. San Diego (CA): San Diego State University; 2006 [cited 2022 Jun 9]. Available from: http://activelivingresearch.org/sites/default/files/SOPLAY_Protocols.pdf.
43. Historical Weather. Weather Underground [Internet]. Brookhaven (GA): The Weather Company; [cited 2022 Jun 9]. Available from: https://www.wunderground.com/history/.
44. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ. 2000;320(7244):1240–3.
45. Nettlefold L, McKay HA, Warburton DER, et al. The challenge of low physical activity during the school day: at recess, lunch and in physical education. Br J Sports Med. 2011;45(10):813–9.
46. Suzuki I, Okuda M, Tanaka M, et al. Variability in school children’s activity occurs in the recess and before-school periods. Pediatr Int. 2018;60(8):727–34.
47. Tercedor P, Segura-Jiménez V, Ávila García M, Huertas-Delgado FJ. Physical activity during school recess: a missed opportunity to be active?Health Educ J. 2019;78(8):988–99.
48. Sallis JF. Habitual physical activity and health-related physical fitness in fourth-grade children. Arch Pediatr Adolesc Med. 1993;147(8):890.
49. Ruiz JR, Rizzo NS, Hurtig-Wennlöf A, et al. Relations of total physical activity and intensity to fitness and fatness in children: the European Youth Heart Study. Am J Clin Nutr. 2006;84(2):299–303.
50. Faison-Hodge J, Porretta DL. Physical activity levels of students with mental retardation and students without disabilities. Adapt Phys Activ Q. 2004;21(2):139–52.
51. Chen W, Hammond-Bennett A, Hypnar A, Mason S. Health-related physical fitness and physical activity in elementary school students. BMC Public Health. 2018;18(1):195.
52. Delidou E, Matsouka O, Nikolaidis C. Influence of school playground size and equipment on the physical activity of students during recess. Eur Phy Educ Rev. 2016;22(2):215–24.
53. Nilsen AKO, Anderssen SA, Resaland GK, et al. Boys, older children, and highly active children benefit most from the preschool arena regarding moderate-to-vigorous physical activity: a cross-sectional study of Norwegian preschoolers. Prev Med Rep. 2019;14:100837.
54. Mota J, Silva P, Santos MP, et al. Physical activity and school recess time: differences between the sexes and the relationship between children’s playground physical activity and habitual physical activity. J Sports Sci. 2005;23(3):269–75.
55. Brez C, Sheets V. Classroom benefits of recess. Learning Environ Res. 2017;20(3):433–45.
56. Hyndman B, Mahony L. Developing creativity through outdoor physical activities: a qualitative exploration of contrasting school equipment provisions. J Adventure Educ Outdoor Learn. 2018;18(3):242–56.
57. Casado-Robles C, Mayorga-Vega D, Guijarro-Romero S, Viciana J. Sport education-based irregular teaching unit and students’ physical activity during school recess. J Educ Res. 2020;113(4):262–74.
58. Clarke KM. Benching playground loneliness: exploring the meanings of the playground buddy bench. Int Electron J Elem Educ. 2018;11(1):9–21.
59. Althouse AD. Adjust for multiple comparisons? It's not that simple. Ann Thorac Surg. 2016;101(5):1644–5.
60. Feise RJ. Do multiple outcome measures require P-value adjustment?BMC Med Res Methodol. 2002;2:8.
61. Martin K, Bremner A, Salmon J, et al. School and individual-level characteristics are associated with children’s moderate to vigorous-intensity physical activity during school recess. Aust N Z J Public Health. 2012;36(5):469–77.
Copyright © 2022 by the American College of Sports Medicine