Achieving the Recommended Levels of Physical Activity and Sleep among American Adolescent Athletes.

Background. Few adolescents are getting sucient levels of physical activity and sleep despite the well understood health benets of achieving sucient levels of both, concurrently. This may be related to the limited understanding on the utility of programs to concurrently improve physical activity and sleep in adolescent populations. This study aimed to determine whether sports participation was associated with achieving the recommended levels of sleep and physical activity among a representative sample of American adolescents. Methods. Participants included a representative sample (n=58,008) of American adolescents who self-reported sport-participation (0, 1, 2, 3 or more teams yearly), and recommended levels of physical activity (at least 60-minutes per day of moderate or vigorous intensity physical activity [MVPA], and at least 3-days per week of muscle-strengthening activities [MSA]) and sleep ( ≤ 12-13 years: 9 to 10+ hours; 14-17 years: 8-10+ hours; ≥ 18 years: 7-9 hours). Multivariable logistic regressions were used to estimate the relative odds of meeting the physical activity (MVPA and MSA) and sleep recommendations (independently and in combinations) in sex stratum among those participating in sports versus non-participation. Results. Overall, 49.2% (95%CI: 47.1%-51.4%) of males and 38.6% (95%CI: 37.3%-40.0%) of females report not participating in any sport. Sport-participation differed by sex, age, ethnicity (Hispanic), and weight status. Those who play 1-sport team had higher odds (females: OR=1.95 [95% CI=1.59-2.40]; males: OR=2.20 [1.91-2.52] of concurrently achieving recommendations for sleep and physical activity (MVPA and MSA) compared to those playing no sports teams, after adjustment. Measures of effect were larger for more sports teams played, indicating a possible dose-response relation.


Introduction
Physical activity and sleep are integral components of healthy development during childhood and adolescence.(1) Aside from sharing many of the same health-related effects, physical activity and sleep are related behaviors. Children who are more physically active tend to sleep better, (2) and those who sleep better tend to be more active. (3) Unfortunately, recent data suggest few children/adolescents are getting su cient levels of physical activity and sleep. Roman-Viñas et al. (4) found that 44% of children/adolescents across various sites meet physical activity recommendations (≥ 60 minutes/day of moderate intensity aerobic activity) and only 42% meet sleep duration recommendations (9)(10)(11) hours/night). When considered together, only 18% of children/adolescents spend the optimal amount of time physically active and sleeping. Among American participants, only 10% met recommendations for both physical activity and sleep. (5) Numerous studies have examined the interrelation between sleep and physical activity among adolescents. (6)(7)(8) However, few studies have examined programs to concurrently improve physical activity and sleep. Notably, a school-based intervention improved physical activity levels and sleep quality over time for students in the intervention group but not in the control group. (9) In addition, research among adolescent athletes suggest they experience better sleep compared to non-athletes.(10) Despite these promising results, most studies examining physical activity programs among adolescents have not included sleep as an outcome. (11) Given the potential role of sports among adolescents, more research is necessary to better understand the relation between sport participation, physical activity, and sleep among this group.
The purpose of this analysis is to determine whether participation in high school sports is associated with concurrently and independently achieving recommended levels of sleep and physical activity among a nationally representative sample of American adolescents. We hypothesize that adolescents who participate in sports will be signi cantly more likely to achieve the recommended levels of physical activity and sleep, both independently and concurrently.

Methods
Data for this analysis are from multiple waves (2011, 2013, 2015, and 2017)

of the Youth Behavioral Risk
Factor Surveillance System (YRBS). YRBS is a serial cross-sectional, school-based national survey of health-risk behaviors among American adolescents. (12) Data are collected bi-annually by the Centers for Disease Control and Prevention using trained data collectors to administer a self-reported questionnaire to students during regular school hours. Participation is voluntary and requires parental consent. All data are anonymous and are freely and publicly available. For this analysis, the unweighted number of participants in each year of data collection were as follows: 2011 (n = 15,425), 2013 (n = 13,583), 2015 (n = 15,624), 2017 (n = 14,765). Participants were in grades 9-12 and are a representative sample of US high-school age students. More details on YRBS methodology appear elsewhere. (12) All variables used in this analysis were participant-reported. The primary exposure of interest was sport participation within the past 12-months and included any sport team run by school or community groups.
The primary outcomes were meeting the recommended time for sleep and physical activity, which were examined independently and concurrently. For sleep, participants reported the number of hours they sleep on an average school night based on an ordered categorical response options in one-hour increments from 4 or less hours to 10 or more hours. Using this, the prevalence of meeting the age speci c sleep recommendations was de ned as 9 to 10 or more hours for those aged 12 and under to 13 years, 8 to 10 or more hours for those aged 14 to 17 years, and 7-9 hours for those aged 18 years and over, consistent with the National Sleep Foundation recommendations. (13) For physical activity, participants reported the number of days, over the past 7-days, they were physically active at least a moderate intensity level (e.g., "increased your heart rate and made you breathe hard some of the time") for at least 60 minutes per day. Those who reported 7-days were classi ed as meeting the recommendation for moderate-to vigorousintensity aerobic physical activity. Participants additionally reported the number of days during the past 7-

Results
The prevalence estimates with 95% con dence intervals (CI) of students who participated in sports across all years of observation are presented in Table 1 (female) and Table 2 (male). Overall, 49.2% (95% CI: 47.1%-51.4%) of females and 38.6% (95% CI: 37.3%-40.0%) of males report not participating in any school or community-based sport. These sex differences (males had a greater levels of participation) were consistent across characteristics of interest. The oldest aged participants (18 years  days that they did "exercises to strengthen or tone [their] muscles, such as push-ups, sit-ups, or weight lifting?" Those who reported at least 3-days were classi ed as meeting the recommendation for muscle strengthening physical activity. Potential covariates included in the analysis were demographics (e.g., sex, age, race/ethnicity [non-Hispanic white, non-Hispanic black, Hispanic/Latino, other]), weight status (e.g., underweight [< 5th percentile], healthy weight [5th to < 85th percentile], overweight [85th to < 95th percentile], obese [ > = 95th percentile]), screen use (television, computer/video game), and comorbidities (e.g., asthma diagnosis, depression symptoms).

Statistical analysis.
To minimize bias associated with missing data, multiple imputation with chained equations (MICE) was used to impute missing values. (14) Categorical variables (e.g., sex, race/ethnicity) were imputed using multinomial logistic regression, ordinal variables (e.g., sport participation, grade, age, days of physical activity, sleep time, TV time, and video game/computer time) were imputed using predictive mean matching, dichotomous variables (asthma diagnosis, depression symptoms) were imputed using logistic regression, and continuous estimates (BMI percentile) were imputed using linear regressions. The survey sampling weight was included as a predictor for missing values and imputations were conducted separately by survey year. Data were imputed using 10, 15, 20, and 25 iterations, to determine if the larger number of iterations stabilized the estimates. It was found that 10 iterations were su cient, and were therefore used in all further analyses. YRBS survey weights were used in all analyses.
The proportion and 95% con dence intervals of sport participation were estimated in sex strata by potential covariates (i.e., age, race/ethnicity, weight status, television/computer screen-use, asthma diagnosis, and the presence of symptoms of depression). The relative odds of meeting the physical activity and sleep recommendations (independently and in combinations) were calculated using logistic regressions in sex stratum and were adjusted by the potential covariates.  Abbreviation: CI, con dence interval; MSA, muscle-strengthening activity; MVPA, moderate-to-vigorous intensity aerobic activity.
a Includes Native Hawaiian/Paci c Islander and non-Hispanic multiple race/ethnicities; b healthy weight is de ned as a body mass index (BMI) percentile from the 5th to less than the 85th percentile, underweight is classi ed as a BMI < 5th percentile, overweight is de ned as a BMI percentile from the 85th to less than the 95th percentile, obese is de ned as a BMI percentile equal or greater than the 95th percentile; c Depressive symptoms is de ned as reporting being sad for at least 2-weeks within the last year. 11.3 (9.6-13.0) Abbreviation: CI, con dence interval; MSA, muscle-strengthening activity; MVPA, moderate-to-vigorous intensity aerobic activity.
a Includes Native Hawaiian/Paci c Islander and non-Hispanic multiple race/ethnicities; b healthy weight is de ned as a body mass index (BMI) percentile from the 5th to less than the 85th percentile, underweight is classi ed as a BMI < 5th percentile, overweight is de ned as a BMI percentile from the 85th to less than the 95th percentile, obese is de ned as a BMI percentile equal or greater than the 95th percentile; c Depressive symptoms is de ned as reporting being sad for at least 2-weeks within the last year. Abbreviation: CI, con dence interval; MSA, muscle-strengthening activity; MVPA, moderate-to-vigorous intensity aerobic activity.
a Includes Native Hawaiian/Paci c Islander and non-Hispanic multiple race/ethnicities; b healthy weight is de ned as a body mass index (BMI) percentile from the 5th to less than the 85th percentile, underweight is classi ed as a BMI < 5th percentile, overweight is de ned as a BMI percentile from the 85th to less than the 95th percentile, obese is de ned as a BMI percentile equal or greater than the 95th percentile; c Depressive symptoms is de ned as reporting being sad for at least 2-weeks within the last year. Abbreviation: CI, con dence interval; MSA, muscle-strengthening activity; MVPA, moderate-to-vigorous intensity aerobic activity. a Includes Native Hawaiian/Paci c Islander and non-Hispanic multiple race/ethnicities; b healthy weight is de ned as a body mass index (BMI) percentile from the 5th to less than the 85th percentile, underweight is classi ed as a BMI < 5th percentile, overweight is de ned as a BMI percentile from the 85th to less than the 95th percentile, obese is de ned as a BMI percentile equal or greater than the 95th percentile; c Depressive symptoms is de ned as reporting being sad for at least 2-weeks within the last year.
Among those who did not report participating in any sport, 2.5% (95% CI = 2.2-2.9%) of females and 6.3% (95% CI = 5.6%-7.0%) of males met the guidelines for sleep and physical activity (aerobic and musclestrengthening) (  The relative odds of meeting the sleep and physical activity recommendations among American high school students participating in sports are presented by sex in Figs. 1 (female) and 2 (male). Results indicate that among both females and males, sports participation is associated with signi cantly higher odds of meeting the recommendations for time spent sleeping and engaging in physical activity concurrently. Speci cally, among females, those who play 1 sport are 1.95 (95% CI = 1.59-2.40) times more likely to achieve the recommendations for sleeping and physical activity (aerobic and musclestrengthening) compared to those playing no sports, after controlling for age, race/ethnicity, weight status, screen-use, and comorbidities (asthma, and depressive symptoms). Measures of effect were consistently larger for the greater number of sports played, indicating a possible dose-response relation.
Meeting the recommended time sleeping was not associated with sports participation among females, however. Results were similar for males, except sport participants were signi cantly more likely to meet sleep guidelines compared to those that play no sports. Additionally, the measures of effect were generally greater for males than females. Interestingly, the strongest associations across all the recommendations for females and males were for muscle-strengthening physical activity both independently, and in combination with aerobic physical activity. For females, those who participated in 3 or more sports, were 7.82 (95% CI = 6.68-8.36) times more likely to meet the recommendations for physical activity (aerobic and muscle-strengthening) compared to those who do not play sports, after controlling for the covariates of interest (results not shown in Fig. 1). For males, those who participated in 3 or more sports, were 6.94 (95% CI = 6.20-7.76) times more likely to meet the recommendations for physical activity (aerobic and muscle-strengthening) compared to those who did not play sports, after controlling for the covariates of interest (results not shown in Fig. 2).

Discussion
The purpose of this paper was to determine whether participation in high school sports was associated with concurrently and independently achieving the recommended levels of sleep and physical activity among a nationally representative sample of American adolescents. Signi cant associations were found between sports participation and meeting sleep and physical activity (aerobic and muscle-strengthening) guidelines. Furthermore, a dose-response relation was seen for the number of sports played in the past year and the relative odds of meeting all three guidelines, with a strong association between sports participation and meeting physical activity guidelines.
The signi cant associations for sports participation and meeting guidelines align with those from a study by Mäkelä et al, who compared physical activity, sleep time, and screen time between Finnish sports club participants and non-participants (ages [14][15][16]. (15) They found that sports participants (versus nonparticipants) had signi cantly higher odds of meeting aerobic physical activity guidelines, sleeping ≥ 9 hours on the weekends, and having < 2 hours daily screen time. Sports provide an opportunity for youth to be physically active, so it follows that students who participate in sports have been found to be more likely to meet aerobic physical activity guidelines than non-participants. (16) Further, it has been found that adolescents with higher (versus lower) physical activity levels are more likely to have better sleep quality and quantity,(6) including among athletes versus non-athletes, (17)(18)(19)(20) which also aligns with the presented ndings. The strong association between sports participation and meeting musclestrengthening guidelines may be due to more opportunities for athletes (versus non-athletes) to engage in muscle-strengthening activities within both training and game sessions. Muscle-strengthening activities (e.g., push-ups, sit-ups, weight-lifting) are often incorporated into sports-related training, but are less frequently seen in the other common physical activities that youth engage in outside of sports (e.g., walking, cycling, running). (21) The association between sports participation and meeting sleep and physical activity guidelines had not previously been examined with consideration for the number of sports played, as done in the present study. The dose-response relationship that emerged for all guidelines among males and all but sleep among females is a valuable contribution to the literature. These ndings suggest there are more bene ts with playing multiple sports during the year, versus single-sport or non-participation, although additional research is needed to examine whether this dose-response relation holds when adolescents play multiple sports concurrently, versus at different times throughout the year. These ndings also suggest that the binary categorization of sports participation (yes/no) often seen in research may not adequately capture this construct, and future research should take this into consideration.
The ndings for racial/ethnic and sex differences, with a higher prevalence of sport participation among non-Hispanic whites (versus all other racial/ethnic groups) and among males (versus females), align with those often seen for physical activity levels and sports participation among these sociodemographic groups. (15,22) However, future research should further examine why such disparities persist within the associations of sports participation and meeting guidelines. For example, although a signi cant association between sports participation and meeting sleep guidelines was found among males, this association was not signi cant for females. Increasing the understanding of these sex and racial/ethnic disparities, as well as reducing barriers to sports participation, could help efforts to equitably increase the prevalence of youth sports participation and the associated health bene ts.
Strengths and Limitations.
The strengths of this study include the inclusion of a large, population-based sample, which provides external study validity, as well as the use of multiple imputation to address potential bias due to missing data. This study also addresses numerous gaps in the literature. Few studies have examined the association of sports participation with sleep and physical activity, both independently and concurrently.
Furthermore, the number of sports played was included in the measure of sports participation, allowing for a greater understanding of its relationship with meeting guidelines -namely, the dose-response relation that emerged. Such ndings highlight sports participation as a potential method for addressing the low prevalence of youth meeting sleep and physical activity guidelines. For example, although sleep interventions have been implemented among elite athletes,(23) few school-based policies aiming to improve sleep among students exist. However, schools could consider sports participation as a method for increasing sleep quantity, as well as reducing screen time, among this population.
The limitations of this study must also be noted. First, the YRBS is a cross-sectional survey, which limits the ability to make causal inferences between the exposures and outcomes under study. Second, the selfreported items in YRBS are prone to recall bias and potential social desirability bias. Third, the YRBS does not contain a measure of socioeconomic status (SES) so it was not possible to examine potential differences by SES. Fourth, the time-frame recall was different for physical activity (past 7 days), sleep (average for school night), and sports participation (past 12 months), and it was not possible to determine if any of the sports were played concurrently. However, although some respondents classi ed as sports participants may not have been actively engaged in sports at the time of the survey, signi cant associations were found. This could suggest that sports participants are engaging in healthy sleep and physical activity behaviors during non-sport periods as well, although additional research is needed to explore this hypothesis. Finally, neither the sport type nor competitive level (e.g., recreational, club, etc.) were included within the YRBS, and ndings may differ across these groups due to the variations in intensity. (24) This may have also played a role in the disparities seen for sex, considering the distribution of males and females across sports played and level of participation. (22) Future research could expand on the few studies examining youth demographics by type of sport, with additional consideration for its role in the association of sports participation and meeting guidelines.
The reported ndings demonstrate signi cant associations of sports participation and achieving the recommended levels of sleep and physical activity (both independently and concurrently). Efforts to increase the prevalence of youth meeting guidelines should consider promoting sports participation while reducing barriers to participate, particularly among females and racial/ethnic minorities. Practitioners can respond to concerns from parents and youth about potential risk of injury by emphasizing the bene ts of sports participation, including those related to screen time and mental and physical health.

Declarations
Ethics approval and consent to participate.