Abstract
Introduction/purpose:
In the United States, it is recommended that schools provide at least 20 minutes of daily recess, but the optimal amount for health benefits is unknown. We examined associations between amount of recess and health indicators using National Health and Nutrition Examination Survey data (NHANES; 2013–2016).
Methods:
For this cross-sectional analysis, parents/guardians of 6–11 year olds (n=738) reported recess provision which was classified as low (22.8%; approximately 10–15 min, 5 days per week), medium (54.9%; approximately 16–30 min, 5 days per week), or high (22.3%; approximately >30 min, 5 days per week). Outcomes measured included parent/guardian-reported and accelerometer-measured physical activity (PA), blood pressure, cholesterol, grip strength, bone mineral content, weight status, percent body fat, vitamin D level, and C-reactive protein level. Linear and logistic regression compared outcomes by level of recess provision accounting for the NHANES complex survey design.
Results:
The odds of meeting PA guidelines according to parent/guardian reports were 1.70 and 2.05 times higher in those with medium and high (respectively) versus low recess provision. Accelerometer-measured weekday activity was highest in those with high recess provision while weekend activity was highest in those with low recess provision (Cohen’s d = 0.40–0.45). There were no other significant associations.
Conclusion:
At least 30 minutes of daily recess is associated with two-fold greater odds of achieving recommended PA levels according to parent/guardian reports; accelerometer data suggest this is through increased weekday activity. This finding suggests current national recess recommendations are insufficient for PA promotion. More detailed data on the frequency and duration of recess are needed to quantify optimal provision more precisely.
Keywords: School, accelerometer, surveillance, metabolic traits, bone health
Introduction
The school setting allows an opportunity to positively influence child health because most children attend school for approximately seven hours, five days per week. Most of the United States (US) school day is spent sedentary (~64%) and indoors (~92%) while focusing on academic subjects (~95%) in a structured classroom setting (1). In contrast, recess is an unstructured break which affords time for physical activity (PA), socialization, creative and risky play, and outdoor experiences and serves as a respite from academic pressures and the expectations of adults.
Much research has focused on recess as an opportunity for PA (2). Positive associations between recess and children’s cardiometabolic or bone health may occur through this increased opportunity for PA or alternative pathways like stress reduction. Recess or unstructured PA has been related to lower blood pressure and improved high density lipoprotein cholesterol (3). While they have not been studied directly, recess physical activities may be particularly well suited to benefit bone development because they are vigorous in intensity, intermittent, weight bearing, or include unusual strain distributions, all of which have been shown to improve bone mineral content (4). Recess is primarily held outdoors so children may also receive other potential benefits (or detriments) from sun exposure, like increases in vitamin D levels (5), or exposure to natural elements like greenery that lead to reductions in blood pressure (6) or stress-related biomarkers like C-reactive protein. Despite support for the established and potential benefits of recess overall, only limited research is available on the association of the amount of recess provided (“recess provision”) with these or other outcomes (7–9).
Recess provision varies among schools because scheduling decisions are often made at the school level. Specifically, an analysis of US surveillance data from 2014 indicated most elementary schools provided regularly scheduled recess, but the amount ranged from 5 to 65 min∙day−1 (10). While little research has aimed to identify the optimal recess provision to accrue psychosocial, cognitive, or physical health benefits (7, 8), the Society of Health and Physical Educators of America (SHAPE), the Centers for Disease Control and Prevention (CDC), and the National Association for Sport and Physical Education (NASPE) recommend schools provide children with at least 20 minutes of daily recess (11, 12); it is unclear what research supported this specific recommendation. Findings from interventions and cross-sectional analyses indicated that providing more recess, often more than is currently recommended, was beneficial to children’s overall PA level or body mass index (BMI) (8, 9, 13).
A better understanding of the relationship between the amount of recess provided and outcomes in children will inform evidence-based recess recommendations for policy and practice. The purpose of the present study was to examine the association of recess provision with key health indicators related to PA and adiposity in 6- to 11-year-old children using data from the National Health and Nutrition Examination Survey (NHANES). Outcomes measured included parent/guardian-reported and accelerometer-measured PA, blood pressure, cholesterol, grip strength, bone mineral content, weight status, adiposity, vitamin D level, and C-reactive protein level.
Methods
Participants
This study involved secondary analyses of publicly available, deidentified data and therefore did not entail human subjects research that required ethical approval. This is a cross-sectional analysis of data from NHANES, a nationally representative sample of the non-institutionalized US population based on a complex, multi-stage, probability design. Participants complete an interview at their home and an in-person examination at a mobile examination center (MEC). Data are released in two-year cycles which can be combined with appropriate modification of sample weights. We combined data from the 2013–2014 and 2015–2016 two-year cycles, when possible, to increase sample size. These are the only cycles in which parents/guardians were asked about their child’s school recess.
Children 6–11 years of age whose parents/guardians reported that their children had recess were included in the present analysis. Children who had never attended school or who had only attended kindergarten were excluded. Additional exclusion criteria included having a physical limitation or health condition that affected their ability to walk, run, or play; receiving special education or early intervention services (i.e., services provided by the state or school system for children with special needs or disabilities); going to the emergency room for asthma in the past year; or being underweight according to BMI percentile (<5th percentile). If a child was not in school at the time of the assessment, the parent/guardian responded to questions according to when the child was last in school. Because BMI and PA are known to worsen during school breaks, we excluded children whose exams were conducted in the 6-month measurement period between May 1 and October 31, although results for this sub-group are provided in the supplementary material. More precise data collection dates are only available through the Research Data Center (closed due to the pandemic). Of the 2,703 children in this age group, 738 were included in the primary analysis (331 in 2013–2014 and 407 in 2015–2016). Flow charts describing exclusion are found in Supplemental Content 1–3 (figures).
Exposure
Parent/guardians of children who had recess provided the frequency (days·week−1) in response to the question “How often does (child) have recess?” Duration was reported in response to the question “On average, how long is the recess period?” as “less than 10 minutes,” “10–15 minutes,” “16–30 minutes,” or “more than 30 minutes.” We classified recess provision as low, medium, or high based on the duration and frequency variables (Table 1). Most children (~90%) fell into three clear categories: five days∙week−1 of 10–15 minutes (17.2±2.2%), 16–30 minutes (50.5±2.7%), or more than 30 minutes (20.3±2.9%) of recess which we termed low, medium, and high recess provision, respectively. Estimates for other specific combinations of recess duration and frequency had insufficient reliability (>30% relative standard error) to be individually reported. Thus, while actual recess provision in minutes cannot be computed due to the NHANES categorical response options, the remaining combinations of duration and frequency (~10% of children) were sorted into the three aforementioned categories based on the approximate amount of recess provision to which they would equate.
Table 1.
Percent of children, with standard error (SE) with each level of recess provision
| Recess Provision | Frequency and Duration | 2013–2016 | 2013–2014 | 2015–2016 | |||
|---|---|---|---|---|---|---|---|
| Mean | SE | Mean | SE | Mean | SE | ||
|
| |||||||
| Low | 1–2 day·week−1, any duration 3–4 days·week−1, up to 30 min·recess−1 5 days·week−1, <10–15 min·recess−1 |
22.8 | 2.5 | 24.7 | 3.7 | 20.1 | 3.3 |
| Medium | 3 days·week−1, > 30 min·recess−1 5 days·week−1, 16–30 min·recess−1 |
54.9 | 2.5 | 52.3 | 3.1 | 58.6 | 3.9 |
| High | 4–5 days·week−1, >30 min·recess−1 | 22.3 | 3.1 | 23.1 | 4.5 | 21.3 | 3.8 |
Frequency and duration of recess was reported by parents of 6- to 11-year-old children; estimates account for the complex sampling design of the National Health and Nutrition Examination Survey.
Outcomes
Parent/guardian-reported child PA, and MEC-assessed weight status, percent body fat, bone mineral content, blood pressure, and cholesterol outcomes were available across both NHANES cycles, from 2013–2016. Accelerometer-measured activity, grip strength, and vitamin D level were available for the 2013–2014 two-year cycle and C-reactive protein level was available for the 2015–2016 two-year cycle. Unless otherwise specified, all children 6 years or older were eligible for these measurements.
Whether or not the child met US PA guidelines, which state that children should participate in 60 min of moderate-to-vigorous PA each day, was determined from the parent/guardian’s response to the question “During the past week, on how many days did this child exercise, play a sport, or participate in PA for at least 60 minutes?” Stature and mass were measured, and weight status was classified as healthy weight (BMI ≥5th and <85th percentile) or overweight/obese (BMI ≥85th percentile) according to age- and sex-specific BMI percentiles. Percent body fat and bone mineral content were measured in children 8 years of age or older using dual-energy x-ray absorptiometry (DEXA; QDR 4500A, Hologic, Marlborough, MA, USA). For both measures, the outcome of interest was a percentile score based on age-, sex-, and race-specific reference curves (14). Bone mineral content is a more appropriate measure of bone health in children because it is less affected by bone size compared to bone mineral density.
Systolic and diastolic blood pressure of participants 8 years or older were measured using an auscultatory approach after a five-minute seated rest. Elevated systolic (or diastolic) blood pressure was classified using screening values based on the age- and sex-specific 90th percentile for children at the 5th percentile for height (15). High-density lipoprotein cholesterol and total cholesterol were measured enzymatically in serum and used to calculate non-high-density lipoprotein cholesterol. Low high-density lipoprotein cholesterol (<40 mg·dL−1) and elevated total cholesterol (≥170 mg·dL−1) or non-high-density cholesterol (≥120 mg·dL−1) were classified based on age-appropriate guidelines (16).
Children wore an ActiGraph GT3X+ (Pensacola, FL, USA) on their non-dominant wrist for one week. Days with at least 10 hours of valid wake minutes were included in this analysis (>70% of days). Instead of activity counts data that were released in prior NHANES cycles, the NHANES 2013–2014 accelerometer data were analyzed and released as monitor-independent movement summary (MIMS) units. MIMS are a non-proprietary summary metric that are meant to be device and sampling rate independent. MIMS are generated using a multi-step processing method that includes interpolation of raw acceleration data to 100 Hz, extrapolation to account for differences in device dynamic range, bandpass filtering, rectification and integration, and summation across the three axes of data collection (17). For the present study, the derived outcomes of interest were average daily MIMS overall and for weekdays and weekends separately. Average daily MIMS were compared to age- and sex-specific percentiles (17). Of note, MIMS is an indication of the total volume of activity, not specifically time spent in moderate-to-vigorous PA (17).
Dynamometer-measured relative grip strength was calculated as maximum grip strength from each hand combined and divided by body weight, which was compared to age- and sex-specific percentiles (18). Vitamin D (25-hydroxyvitamin D2 + D3) level was assessed using ultra-high-performance liquid chromatography-tandem mass spectrometry. Vitamin D insufficiency was classified as <50 nmol·L−1 (5). High-sensitivity C-reactive protein level was assessed using the SYNCHRON Systems (Beckman Coulter, Inc., Brea, CA, USA) which relies on turbidimetrics and changes in electromagnetic wavelength absorption to determine concentration. Elevated C-reactive protein was classified as >0.15 mg·dL−1 (19).
Statistical Analysis
Linear regression was used to compare continuous outcomes (blood pressure, cholesterol, vitamin D level, C-reactive protein level, percentiles for body fat, bone mineral content, grip strength, and MIMS) across recess provision categories (treated as a factor variable). Logistic regression was used to estimate odds ratios for meeting PA guidelines according to parent/guardian reports; being overweight/obese; or having elevated blood pressure, low high-density lipoprotein cholesterol, elevated total cholesterol, elevated non-high-density lipoprotein cholesterol, vitamin D deficiency, or elevated C-reactive protein using low recess provision as the referent group. Models were adjusted for race/ethnicity, age, sex, income-to-poverty ratio accounting for family size and geographic residence, sport participation, and weight status (when not the outcome) which are all available directly in the NHANES dataset. Due to insufficient sample size, models were not stratified by sex, but means by sex are presented.
All analyses accounted for the complex sampling design and incorporated the examination survey weights which were adjusted for analyses combining data across cycles. Analyses were conducted in RStudio (version 1.3.1056; R Foundation for Statistical Computing, Vienna, Austria), using the survey and emmeans packages (versions 4.0 and 1.5.5, respectively), which uses Taylor series linearization to estimate variance. Unadjusted means that accounted for survey design but not covariates were estimated using the “svyby” command, while means adjusted for the survey design and covariates were derived using “svypredmeans.” Significance was indicated when pairwise comparisons had a p-value of <0.05 (with Tukey adjustment) or the 95% confidence intervals surrounding the odds ratios did not include 1. Pairwise effect sizes are presented as Cohen’s d which can be interpreted as small (0.2), medium (0.5), or large (0.8).
Results
The average age of the children was 9 years of age with the largest portion of participants being white (35–45%; Table 2) and having medium recess provision (54.9%; Table 1). Participant characteristics overall and for each two-year cycle are reported in Table 2 for those measured between November 1 and April 30 (the focus of this analysis) and in Supplemental Content 4 (table) for participants measured between May 1 and October 31. Sample sizes varied for each outcome (parent/guardian-reported PA, n=738; weight status, n=738; DEXA, n=523; blood pressure, n=526; cholesterol, n=604; accelerometer-measured activity, n=340; grip strength, n=391; vitamin D, n=336; C-reactive protein, n=267).
Table 2.
Participant characteristics (mean or percent and standard error (SE))
| Characteristic | 2013–2016 | 2013–2014 | 2015–2016 | |||
|---|---|---|---|---|---|---|
|
| ||||||
| Mean | SE | Mean | SE | Mean | SE | |
|
| ||||||
| Age (years) | 8.98 | 0.07 | 8.92 | 0.11 | 9.06 | 0.09 |
| Income to poverty ratio | 2.33 | 0.16 | 2.37 | 0.21 | 2.28 | 0.26 |
|
| ||||||
| Percent | SE | Percent | SE | Percent | SE | |
|
| ||||||
| Sex | ||||||
| Male | 50.2 | 2.2 | 53.3 | 3.3 | 46.2 | 3.0 |
| Female | 49.8 | 2.2 | 46.7 | 3.3 | 53.8 | 3.0 |
| Race | ||||||
| Mexican-American | 24.0 | 3.6 | 19.6 | 2.8 | 29.8 | 7.4 |
| Other Hispanic | 9.6 | 1.3 | 6.8 | 1.8 | 13.3 | 2.1 |
| Non-Hispanic White | 40.8 | 4.2 | 45.0 | 4.3 | 35.3 | 8.0 |
| Non-Hispanic Black | 14.8 | 2.5 | 18.0 | 3.9 | 10.7 | 2.6 |
| Other race or multi-racial | 10.7 | 2.1 | 10.5 | 2.0 | 10.9 | 4.1 |
| Highest grade completed | ||||||
| 1st | 24.3 | 2.1 | 27.1 | 3.5 | 20.7 | 2.0 |
| 2nd | 21.0 | 1.6 | 18.2 | 1.3 | 24.6 | 3.0 |
| 3rd | 21.1 | 1.8 | 21.3 | 2.0 | 20.8 | 3.1 |
| 4th | 21.0 | 1.5 | 21.5 | 2.2 | 20.4 | 1.9 |
| 5th | 10.7 | 1.2 | 9.8 | 1.6 | 12.0 | 1.8 |
| 6th | 1.5 | 0.6 | 1.8 | 0.9 | 1.0 | 0.6 |
Participants are 6- to 11-year-old children in the National Health and Nutrition Examination Survey assessed between November 1 and April 30
Marginal means adjusted for covariates and pairwise effect sizes (Cohen’s d) are reported in Table 3 for continuous outcomes (unadjusted in Supplemental Content 5 (table)). Weekday MIMS percentile was significantly higher in those with high compared to medium recess provision (d=0.40). Weekend MIMS was significantly higher in those with low compared to medium recess provision (d=0.45). There was no significant association of recess provision with body fat, bone mineral content, blood pressure, cholesterol, grip strength, vitamin D level, or C-reactive protein level (d=0.01–0.23). The effect sizes for some outcomes, like overall MIMS, systolic blood pressure, and C-reactive protein were small but not statistically significant (e.g., d=0.22–0.30).
Table 3.
Continuous physical activity and cardiometabolic outcomes by level of recess provision (marginal means ± standard error (SE))
| Recess Provision | Effect Sizes | ||||||||
|---|---|---|---|---|---|---|---|---|---|
|
|
|||||||||
| Low | Medium | High | Low vs. Medium | Low vs. High | Medium vs. High | ||||
|
|
|||||||||
| Mean | SE | Mean | SE | Mean | SE | d | d | d | |
|
| |||||||||
| All Years (2013–2016) | |||||||||
|
| |||||||||
| Parent/guardian-reported physical activity (days) | 5.1 | 0.2 | 5.5 | 0.1 | 5.8 | 0.2 | −0.20 | −0.38* | −0.17 |
| Percent body fat (percentile) | 57.8 | 2.9 | 53.0 | 2.3 | 56.9 | 4.6 | 0.18 | 0.03 | −0.15 |
| Bone mineral content (percentile) | 65.7 | 4.2 | 64.0 | 3.2 | 65.3 | 2.8 | 0.06 | 0.01 | −0.05 |
| Systolic blood pressure (mmHg) | 102.2 | 1.3 | 101.6 | 0.7 | 100.3 | 1.1 | 0.07 | 0.23 | 0.16 |
| Diastolic blood pressure (mmHg) | 49.8 | 2.2 | 50.6 | 0.9 | 52.1 | 1.5 | −0.05 | −0.16 | −0.11 |
| Total cholesterol (mg·dL−1) | 157.2 | 2.8 | 157.7 | 3.3 | 158.0 | 1.9 | −0.02 | −0.03 | −0.01 |
| High-density lipoprotein cholesterol (mg·dL−1) | 54.3 | 1.0 | 54.6 | 1.4 | 56.1 | 1.1 | −0.03 | −0.15 | −0.12 |
| Non-high-density lipoprotein cholesterol (mg·dL−1) | 102.9 | 3.0 | 103.1 | 2.3 | 101.8 | 2.0 | −0.01 | 0.05 | 0.05 |
|
| |||||||||
| 2013–2014 Only | |||||||||
|
| |||||||||
| Overall MIMS (original unit) | 19168 | 492 | 18503 | 448 | 18889 | 341 | 0.23 | 0.10 | −0.13 |
| Overall MIMS (percentile) | 51.8 | 3.9 | 44.2 | 4.3 | 52.4 | 3.3 | 0.28 | −0.02 | −0.30 |
| Weekday MIMS (original unit) | 19464 | 445 | 18906 | 417 | 19537 | 410 | 0.19 | −0.02 | −0.21* |
| Weekday MIMS (percentile) | 53.3 | 3.8 | 47.7 | 3.8 | 58.6 | 3.9 | 0.20 | −0.19 | −0.40* |
| Weekend MIMS (original unit) | 18738 | 761 | 17451 | 610 | 17464 | 824 | 0.31 | 0.31 | −0.00 |
| Weekend MIMS (percentile) | 50.3 | 5.5 | 35.9 | 5.1 | 38.8 | 6.5 | 0.45* | 0.36 | −0.09 |
| Relative grip strength (original unit) | 0.91 | 0.02 | 0.91 | 0.01 | 0.89 | 0.02 | 0.02 | 0.13 | 0.11 |
| Relative grip strength (percentile) | 45.1 | 2.5 | 46.0 | 2.2 | 43.6 | 3.8 | −0.04 | 0.07 | 0.11 |
| Vitamin D (nmol·L−1) | 64.3 | 1.7 | 62.6 | 2.1 | 63.5 | 2.4 | 0.12 | 0.06 | −0.06 |
|
| |||||||||
| 2015–2016 Only | |||||||||
|
| |||||||||
| C-reactive protein (mg·L−1) | 1.49 | 0.47 | 1.80 | 0.43 | 1.24 | 0.33 | −0.12 | 0.10 | 0.22 |
Participants were assessed between November 1 and April 30. Estimates account for the complex sampling design of the National Health and Nutrition Examination Survey and are adjusted for sex, race/ethnicity, age, family-to-income ratio, weight status, and sport participation (percent body fat was not adjusted for weight status). Effect sizes are for pairwise differences (Cohen’s d). MIMS: monitor-independent movement summary.
Indicates significant pairwise difference, p<0.05.
Marginal means for binary outcomes and odds ratios with low recess provision as the referent group are presented in Table 4 (unadjusted means in Supplemental Content 5 (table)). The odds of meeting PA guidelines (7 days of PA) according to parent/guardian reports were 1.70 times higher (95% confidence interval: 1.04–2.80) in those with medium recess provision and 2.05 times higher (95% confidence interval: 1.22–3.44) in those with high recess provision when compared to low recess provision. This corresponds to an estimate that children were active approximately one additional day if they had high recess provision compared with low recess provision (unadjusted means of 5.8 vs. 4.9 days; Supplemental Content 5 (table)). There was no significant association between recess provision and weight status or any of the cardiometabolic outcomes.
Table 4.
Binary physical activity and cardiometabolic outcomes by level of recess provision (marginal percent ± standard error (SE))
| Recess Provision | Odds Ratio | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
|
|
||||||||||
| Low | Medium | High | Medium | High | ||||||
|
|
||||||||||
| Percent | SE | Percent | SE | Percent | SE | OR | 95% CI | OR | 95% CI | |
|
| ||||||||||
| All Years (2013–2016) | ||||||||||
|
| ||||||||||
| Meeting physical activity guidelines | 44.3 | 5.6 | 56.7 | 2.3 | 60.8 | 4.7 | 1.70 | 1.04, 2.80* | 2.05 | 1.22, 3.44* |
| Overweight or obese | 44.3 | 6.4 | 31.6 | 2.7 | 39.3 | 8.2 | 0.62 | 0.38, 1.01 | 0.83 | 0.36, 1.92 |
| Elevated systolic blood pressure | 27.2 | 6.7 | 19.0 | 2.7 | 14.2 | 4.7 | 0.59 | 0.25, 1.38 | 0.40 | 0.14, 1.18 |
| Elevated diastolic blood pressure | 5.2 | 2.9 | 2.4 | 1.2 | 5.3 | 2.5 | 0.44 | 0.08, 2.32 | 1.02 | 0.20, 5.23 |
| Elevated total cholesterol | 32.0 | 6.0 | 30.1 | 5.3 | 32.3 | 4.8 | 0.91 | 0.49, 1.68 | 1.01 | 0.43, 2.38 |
| Low high-density lipoprotein cholesterol | 12.3 | 3.2 | 12.0 | 2.2 | 8.6 | 3.4 | 0.96 | 0.49, 1.88 | 0.64 | 0.19, 2.20 |
| Elevated non-high-density lipoprotein cholesterol | 20.4 | 4.7 | 24.6 | 4.1 | 22.1 | 4.9 | 1.29 | 0.60, 2.79 | 1.11 | 0.45, 2.72 |
|
| ||||||||||
| 2013–2014 Only | ||||||||||
|
| ||||||||||
| Vitamin D deficient | 18.0 | 4.0 | 22.5 | 5.2 | 23.5 | 5.9 | 1.43 | 0.54, 3.79 | 1.54 | 0.59, 4.02 |
|
| ||||||||||
| 2015–2016 Only | ||||||||||
|
| ||||||||||
| Elevated C-reactive protein | 26.2 | 9.2 | 26.6 | 4.7 | 19.3 | 5.2 | 1.03 | 0.35, 3.04 | 0.59 | 0.13, 2.65 |
Participants were assessed between November 1 and April 30. Estimates account for the complex sampling design of the National Health and Nutrition Examination Survey and are adjusted for sex, race/ethnicity, age, family to income ratio, weight status (when not the outcome), and sport participation. Odds ratios (OR) are provided with low recess provision as the referent group.
Indicates significance, p<0.05.
Results for participants measured between May 1 and October 31 are in Supplemental Content 6–8 (tables). There were no significant associations in this group except that the odds of being overweight/obese were 38% lower in children with medium compared to low recess provision (odds ratio: 0.62, 95% confidence interval: 0.42–0.93). Estimates stratified by sex are in Supplemental Content 9 and 10 (tables).
Discussion
We examined the association of different levels of recess provision with PA, body composition, bone mineral content, and cardiometabolic outcomes in a nationally representative sample of 6- to 11-year-old US children. Providing more than 30 min of daily recess was associated with a two-fold-greater likelihood of meeting PA recommendations according to parent/guardian reports; accelerometer data indicate that this is likely due to increases in weekday, not weekend, activity. In contrast, there were no statistically significant associations of recess provision with cardiometabolic or anthropometric outcomes. Because this study focused on how levels of recess provision (not recess vs. no recess) were associated with outcomes, null associations are not interpreted as a lack of association between recess and these outcomes. Instead, null associations indicate no association of providing a certain level of recess (e.g., 30 min per day) with these outcomes.
Our findings support that recess is an important opportunity for PA (2, 20). These findings are similar to a recent analysis of nationally representative US data from the 2012 National Youth Fitness Survey (NYFS) which also indicated that the odds of meeting PA guidelines according to parent/guardian reports were two-fold greater in those with the highest levels of recess provision, who also had higher accelerometer-measured weekday PA levels (13). While we could not study this directly, our finding of higher overall PA on weekdays in those with high levels of recess provision may be due to either direct PA participation during recess or participation in additional PA after school on days when recess is provided, as reported in prior research (21). However, while prior analysis of NYFS data indicated children with high levels of recess provision had higher overall and weekend PA levels (13), this was not confirmed in the present study and there is no clear explanation for this inconsistency. Analysis of the participants assessed between May 1 and October 31, a timeframe including non-school months, also revealed no association of recess provision with parent/guardian-reported or accelerometer-assessed activity, suggesting that the benefits of recess on children’s PA levels may be confined to the school year. Put together, there appear to be acute benefits of recess provision on PA that do not translate to longer-term opportunities such as weekends or other times of the year when recess was not provided.
While parent/guardian-reported PA followed a more typical dose response in which more recess resulted in higher PA participation, the accelerometer-measured PA followed more of a U- or J-shaped curve, where children with low recess provision had similar levels of overall, week and weekend day PA as those with high levels of recess provision. This pattern is similar to prior analysis of NYFS where children with medium levels of recess provision have the lowest levels of PA (13). Of note, these two approaches are measuring two different aspects of PA participation. While parent/guardian reports captured whether children met PA guidelines of 60 min∙day−1 of moderate-to-vigorous PA, MIMS units capture total activity volume. There are currently no established methods for translating MIMS units to time spent in moderate-to-vigorous PA. Thus, it is possible that children with low levels of recess provision participate in more light PA compared to those with high recess provision, resulting in comparable total activity (MIMS). This would be captured by the accelerometer but not parent/guardian reports. Alternatively, children may compensate for having less opportunity for PA by participating in other types of PA or increasing PA intensity in ways that may not be considered by parents when reporting their child’s participation in moderate-to-vigorous PA. For example, accelerometer MIMS would capture even small compensatory behaviors like running instead of walking up the stairs or short bouts of activity, which are unlikely to be captured by parent/guardian reports. Further exploring the association of recess provision with specific PA outcomes, such as time spent in moderate-to-vigorous PA, may clarify our findings.
We found no association of recess provision with cardiometabolic or anthropometric outcomes. There may be no association of recess provision with these outcomes if PA is the pathway through which children may acquire these benefits because there are relatively weak correlations between PA and our primary outcomes (22, 23). However, while not statistically significant, the effect size for the difference in systolic blood pressure between low and high recess provision (adjusted mean difference of 1.9 mmHg) was small (d=0.23). This magnitude of difference in systolic blood pressure is in line with prior interventions (e.g., 1.25 mmHg; 24) and in adults, is clinically meaningful (25). Prior research has similarly indicated a positive association of recess provision with children’s systolic blood pressure (3). This difference was not present in the analysis of children measured beteween May 1 and October 31, indicating an acute benefit of recess that may not be maintained over a school break. Research examining the association of systolic blood pressure with the amount of recess provided using longitudinal or intervention data is warranted.
Lack of association between recess provision and other outcomes may also be due to lack of information about potential moderators like schoolyard design, equipment provision, recess quality, or PA intensity or participation during recess. For example, we report no association of recess provision with grip strength, similar to prior research (13). However, grip strength has been associated with schoolyard size or equipment type (26, 27); an intervention that modified the outdoor space led to changes in grip strength (28). Similarly, where children play outdoors moderates the relationship with vitamin D level (29), while environmental features like greenery or air pollution may be important to capture for blood pressure or C-reactive protein, respectively (6, 30). Capturing PA type and intensity may be important as vigorous-intensity activity and unusual strain distributions that are afforded during recess are integral to the modification of bone mineral content (4), and anthropometrics have been related to moderate-to-vigorous intensity PA (31). Future research may aim to better understand the interplay between recess provision, the recess environment, and outcomes.
The results of the present study must be interpreted with the major caveat that we are unable to quantify the total volume of recess provided due to the response options included in the NHANES interview and lack of information about the number of recess periods per day. Specifically, most children with low recess provision had 10–15 min of daily recess (>75%) and it is possible that these children had multiple recess periods per day, albeit shorter in duration, that result in them having similar levels of overall recess provision as those with 30 min of recess five days∙week−1. While more recent data are unavailable, the majority of schools in 2005 (55–66% depending on grade) provided one recess period per day while 25–38% provided two or three recess periods per day (32). That being said, children in the high recess group definitively exceed the recommendation of 20 min of daily recess set forth by SHAPE, CDC, and NASPE (11, 12). While some children in other categories may also meet or exceed this recommendation due to multiple recess periods per day, findings of positive associations in the high recess provision group can be interpreted as a positive association from providing children with more recess than is currently being recommended. More detailed surveillance data are needed to understand the interplay between our outcomes and recess frequency, duration, and overall volume.
There are other limitations inherent to using NHANES data for these analyses. We have no information about physical education or other opportunities for PA during the school day (e.g., active classroom breaks) that may be important due to, for example, the inverse relationship between physical education and recess provision (33). For some outcomes, like accelerometry (n=340), the sample size was relatively small. The cross-sectional design and lack of inclusion of children over 11 years old may obscure relationships between recess and health outcomes that take years to develop. There may also be super acute (immediately after recess) or chronic (in adolescence or adulthood) benefits to cardiometabolic or bone health not captured by NHANES, which can only capture medium-term benefits in the days or weeks after recess exposure. Additionally, the disparity between parent/guardian-reported recess (which may be akin to amount of scheduled recess) and the actual amount of recess provided is unknown.
Focusing on the November 1 to April 30 time period may have weakened associations between recess provision and some outcomes; for example, outdoor time is not the primary source of vitamin D during this time period and shorter daylight hours reduce opportunity for outdoor time outside of school (29). This temporal restriction also reduced our sample size, and our inclusion of effect sizes revealed that recess provision may have only small associations which we would be underpowered to detect. We were unable to stratify statistical analyses by sex which is of interest because boys and girls have different PA levels and participate in different activities during recess (34). This is supported by a recent analysis of NHANES data that indicated girls with no recess had greater odds of obesity than those with 30 min∙day−1 of recess (35).
More than 30 min of daily recess is associated with increased parent/guardian-reported and device-based estimates of activity. If more than 30 min·day−1 of recess are needed to confer benefits such as increased PA participation, this has important implications for policy because only approximately 20% of children receive this level of recess provision. Only one state indicates schools should provide 30 min of recess while several others currently require or recommend 20 min·day−1 and only ~30% of school districts recommend or require this amount of recess (10). While future surveillance efforts should incorporate more precise measures of recess provision, these findings support the importance of providing sufficient recess for children’s PA.
Supplementary Material
.docx, Flow chart of participant inclusion/exclusion from 2013–2016
.docx, Flow chart of participant inclusion/exclusion from 2013–2014
.docx, Flow chart of participant inclusion/exclusion from 2015–2016
.docx, Participant characteristics for those measured May 1-October 31
.docx, Unadjusted means for those measured November 1-April 30
.docx, Marginal means for continuous outcomes for those measured May 1-October 31
.docx, Marginal means for binary outcomes for those measured May 1-October 31
.docx, Unadjusted means for those measured May 1-October 31
.docx, Unadjusted means for males measured November 1-April 30
.docx, Unadjusted means for females measured November 1-April 30
Acknowledgements
We would like to thank the NHANES participants. The results of the present study do not constitute endorsement by the American College of Sports Medicine.
Source of Funding
There was no funding for this project.
Footnotes
Conflicts of Interest
The authors have no conflicts of interest to declare.
References
- 1.Weaver RG, Crimarco A, Brusseau TA, et al. Accelerometry-derived physical activity of first through third grade children during the segmented school day. J Sch Health. 2016;86(10):726–33. [DOI] [PubMed] [Google Scholar]
- 2.Erwin H, Abel M, Beighle A, et al. The contribution of recess to children’s school-day physical activity. J Phys Act Health. 2012;9(3):442–8. [DOI] [PubMed] [Google Scholar]
- 3.Väistö J, Eloranta A-M, Viitasalo A, et al. Physical activity and sedentary behaviour in relation to cardiometabolic risk in children: cross-sectional findings from the Physical Activity and Nutrition in Children (PANIC) Study. Int J Behavl Nutr Phys Act. 2014;11(1):1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tan VP, Macdonald HM, Kim S, et al. Influence of physical activity on bone strength in children and adolescents: a systematic review and narrative synthesis. J Bone Miner Res. 2014;29(10):2161–81. [DOI] [PubMed] [Google Scholar]
- 5.Voortman T, van den Hooven EH, Heijboer AC, et al. Vitamin D deficiency in school-age children is associated with sociodemographic and lifestyle factors. J Nutr. 2015;145(4):791–8. [DOI] [PubMed] [Google Scholar]
- 6.Duncan MJ, Clarke ND, Birch SL, et al. The effect of green exercise on blood pressure, heart rate and mood state in primary school children. Int J Environ Res Public Health. 2014;11(4):3678–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Dills AK, Morgan HN, Rotthoff KW. Recess, physical education, and elementary school student outcomes. Econ Educ Rev. 2011;30(5):889–900. [Google Scholar]
- 8.Farbo D, Maler LC, Rhea DJ. The preliminary effects of a multi-recess school intervention: using accelerometers to measure physical activity patterns in elementary children. Int J Environ Res Public Health. 2020;17(23):8919. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Fernandes MM, Sturm R. The role of school physical activity programs in child body mass trajectory. J Phys Act Health. 2011;8(2):174–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Centers for Disease Control and Prevention. Results from the school health policies and practices study 2014 [Internet]. Washington: US Department of Health and Human Services; 2015. [cited 2021 Dec 4]. Availabe from: https://www.cdc.gov/healthyyouth/data/shpps/pdf/SHPPS-508-final_101315.pdf. [Google Scholar]
- 11.Centers for Disease Control and Prevention, SHAPE America. Strategies for Recess in Schools [Internet]. Atlanta (GA): Centers for Disease Control and Prevention, US Dept of Health and Human Services; 2017. [cited 2021 Dec 4]. Available from: https://www.cdc.gov/healthyschools/physicalactivity/pdf/2016_12_16_schoolrecessstrategies_508.pdf. [Google Scholar]
- 12.National Association for Sport and Physical Education. Recess for Elementary School Students. Position Statement [Internet]. Reston (VA): National Association for Sport and Physical Education; 2006. [cited 2021 Dec 4]. Available from: https://files.eric.ed.gov/fulltext/ED541609.pdf. [Google Scholar]
- 13.Clevenger KA, McNarry M, Mackintosh KA, Berrigan D. Association of recess provision with elementary school-aged children’s physical activity, adiposity, and cardiorespiratory and muscular fitness. SportRxiv: 10.51224/SRXIV.20 [Preprint]. 2021. [cited 2021 Dec 4]. Available from: 10.51224/SRXIV.20. [DOI] [PubMed] [Google Scholar]
- 14.Kelly TL, Wilson KE, Heymsfield SB. Dual energy X-Ray absorptiometry body composition reference values from NHANES. PloS One. 2009;4(9):e7038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Flynn JT, Kaelber DC, Baker-Smith CM, et al. Clinical practice guideline for screening and management of high blood pressure in children and adolescents. Pediatrics. 2017;140(3):e20171904. [DOI] [PubMed] [Google Scholar]
- 16.Heart National, Lung, and Blood Institute. Expert panel on integrated guidelines for cardiovascular health and risk reduction in children and adolescents: summary report. Pediatrics. 2011;128(Suppl 5):S213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Belcher BR, Wolff-Hughes DL, Dooley EE, et al. US Population-referenced Percentiles for Wrist-Worn Accelerometer-derived Activity. Med Sci Sports Exerc. 2021;53(11):2455–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Laurson KR, Saint-Maurice PF, Welk GJ, Eisenmann JC. Reference curves for field tests of musculoskeletal fitness in US children and adolescents: The 2012 NHANES National Youth Fitness Survey. J Strength Cond Res. 2017;31(8):2075–82. [DOI] [PubMed] [Google Scholar]
- 19.Kumar J, Muntner P, Kaskel FJ, et al. Prevalence and associations of 25-hydroxyvitamin D deficiency in US children: NHANES 2001–2004. Pediatrics. 2009;124(3):e362–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Reilly JJ, Johnston G, McIntosh S, Martin A. Contribution of school recess to daily physical activity: systematic review and evidence appraisal. Health Behav Policy Rev. 2016;3(6):581–9. [Google Scholar]
- 21.Dale D, Corbin CB, Dale KS. Restricting opportunities to be active during school time: do children compensate by increasing physical activity levels after school? Res Q Exerc Sport. 2000;71(3):240–8. [DOI] [PubMed] [Google Scholar]
- 22.Loprinzi P, Cardinal B, Crespo C, et al. Objectively measured physical activity and C-reactive protein: National Health and Nutrition Examination Survey 2003–2004. Scand J Med Sci Sports. 2013;23(2):164–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Belcher BR, Moser RP, Dodd KW, et al. Self-reported versus accelerometer-measured physical activity and biomarkers among NHANES youth. J Phys Act Health. 2015;12(5):708–16. [DOI] [PubMed] [Google Scholar]
- 24.Cesa CC, Sbruzzi G, Ribeiro RA, et al. Physical activity and cardiovascular risk factors in children: meta-analysis of randomized clinical trials. Prev Med. 2014;69:54–62. [DOI] [PubMed] [Google Scholar]
- 25.Hess NC, Carlson D, Inder J, et al. Clinically meaningful blood pressure reductions with low intensity isometric handgrip exercise. A randomized trial. Physiol Res. 2016;65(3):461–8. [DOI] [PubMed] [Google Scholar]
- 26.Grunseit AC, O’Hara BJ, Drayton B, et al. Ecological study of playground space and physical activity among primary school children. BMJ Open. 2020;10(6):e034586. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Heneghan C The effect of fixed playground equipment on primary school children’s fitness levels. Irish Teachers’ Journal. 2016:99. [Google Scholar]
- 28.Christiansen L, Toftager M, Boyle E, et al. Effect of a school environment intervention on adolescent adiposity and physical fitness. Scand J Med Sci Sports. 2013;23(6):e381–9. [DOI] [PubMed] [Google Scholar]
- 29.Pagels P, Wester U, Mårtensson F, et al. Pupils’ use of school outdoor play settings across seasons and its relation to sun exposure and physical activity. Photodermatol Photoimmunol Photomed. 2020;36(5):365–72. [DOI] [PubMed] [Google Scholar]
- 30.Clifford S, Mazaheri M, Salimi F, et al. Effects of exposure to ambient ultrafine particles on respiratory health and systemic inflammation in children. Environ Int. 2018;114:167–80. [DOI] [PubMed] [Google Scholar]
- 31.Tarp J, Child A, White T, et al. Physical activity intensity, bout-duration, and cardiometabolic risk markers in children and adolescents. Int J Obes. 2018;42(9):1639–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Parsad B, Lewis L. Calories In, Calories Out: Food and Exercise in Public Elementary Schools, 2005 (NCES 2006–057) [Internet]. Washington (DC): National Center for Education Statistics; 2006. [cited 2021 Dec 4]. Available from: https://nces.ed.gov/pubs2006/2006057.pdf. [Google Scholar]
- 33.Slater SJ, Nicholson L, Chriqui J, et al. The impact of state laws and district policies on physical education and recess practices in a nationally representative sample of US public elementary schools. Arch Pediatr Adolesc Med. 2012;166(4):311–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.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. [DOI] [PubMed] [Google Scholar]
- 35.Rogers CD, Richardson MR, Churilla JR. Recess and overweight and obesity in children 5–11 years of age: 2013–2016 National Health and Nutrition Examination Survey. J Sch Health. 2022;92(1):63–70. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
.docx, Flow chart of participant inclusion/exclusion from 2013–2016
.docx, Flow chart of participant inclusion/exclusion from 2013–2014
.docx, Flow chart of participant inclusion/exclusion from 2015–2016
.docx, Participant characteristics for those measured May 1-October 31
.docx, Unadjusted means for those measured November 1-April 30
.docx, Marginal means for continuous outcomes for those measured May 1-October 31
.docx, Marginal means for binary outcomes for those measured May 1-October 31
.docx, Unadjusted means for those measured May 1-October 31
.docx, Unadjusted means for males measured November 1-April 30
.docx, Unadjusted means for females measured November 1-April 30