Associations Between 24-Hour Time Use and Academic Achievement in Australian Primary School

Abstract

Background

Previous research has examined associations between individual activity behaviors and academic achievement. Yet activity behaviors should be analyzed together because they are codependent parts of the 24-hour day.

Aims

This study aims to explore the associations between all daily activity behaviors (sleep, sedentary time, light physical activity [LPA], and moderate-to-vigorous physical activity [MVPA]) and academic achievement using compositional data analysis.

Method

Participants for this study were drawn from two cohorts: the Australian arm of the cross-sectional International Study of Childhood Obesity, Lifestyle and the Environment (n = 452; mean age 10.7 years (SD = 0.4); 54% female) and CheckPoint (n = 1278; mean age 12 years [SD = 0.4]; 50% female), a cross-sectional study nested between Waves 6 and 7 of the Longitudinal Study of Australian Children. Objective daily activity behavior data (sleep, sedentary time, LPA, and MVPA) were collected using 8-day, 24-hour accelerometry. Academic achievement was assessed using a nationally administered standardized test in literacy (spelling, grammar and punctuation, writing and reading) and numeracy. Compositional models (adjusted for age, sex, socioeconomic position, and pubertal development) regressed academic scores against isometric log ratios of activity behaviors. We used the models to estimate academic achievement for observed daily activity mixes. Estimated outcomes were plotted against time spent in each individual activity domain, and loess curves were fitted.

Discussion and conclusion

In two different cohorts using two different accelerometers, lower LPA was related to better numeracy and literacy and higher sedentary time to better literacy (relative to time spent in other domains). Discussion and conclusion. LPA likely “drains time” from other movement behaviors, which are beneficial for academic achievement.

Keywords

academic achievement activity compositions compositional analysis physical activity sedentary behavior sleep

While children’s movement behaviors (physical activity, sedentary time, and sleep duration) have been linked with their academic achievement (Alvarez-Bueno et al., 2017; Carson et al., 2016; Short et al., 2018), most previous studies do not account for the confounding influence of time spent in other behaviors, for example, sedentary time or physical activity. Some research has evaluated the relationship between a combination of movement behaviors and academic achievement. For example, Maher et al. (2016) found that school-aged children with the combination of high sedentary time and high moderate-to-vigorous physical activity (MVPA) attained the highest academic achievement. More recently, research has focused on associations between academic achievement and meeting the new public health guidelines for 24-hour movement behaviors, which integrate sleep, recreational screen time, and physical activity. Meeting recommendations from the 24-hour movement guidelines in combination (e.g., sleep and screen recommendations) were more strongly associated with better cognition (Walsh et al., 2018) and academic achievement (Faught et al., 2017) compared with meeting recommendations in isolation.

While such studies have included multiple daily behaviors, the statistical methodology has not respected the compositional nature of these behaviors. Daily movement behaviors (i.e., sleep, sedentary time, light physical activity [LPA], and MVPA) are mutually exclusive and exhaustive parts of a whole (the 24-h day). Each day only ever has 24 hours; therefore, the behavior parts are constrained to a constant sum. Consequently, when time in one movement behavior is changed, time spent in one or more of the remaining behaviors must undergo a corresponding compensatory change to maintain the total of 24 hours.

Our recent cluster analysis compared academic achievement among groups of Australian children with similar combinations of lifestyle behaviors (Dumuid et al., 2017). The analysis included compositional daily movement behaviors as well as noncompositional behaviors (diet quality and screen use). Highest academic achievement was found among children in clusters characterized by (1) high sedentary time or (2) the combination of healthy diet, low screen use, and moderate MVPA/sedentary time. The findings, albeit cross-sectional, suggested that movement behaviors might influence children’s academic achievement. As movement behaviors are potentially modifiable, a closer investigation of the relative and integrated relationships of daily movement behaviors with academic achievement is warranted.

This study aims to explore the associations between all daily activity behaviors (sleep, sedentary time, LPA, and MVPA) and academic achievement using compositional data analysis.

Method

Study Design and Participants

Participants for this study were drawn from the Australian arm of the cross-sectional International Study of Childhood Obesity, Lifestyle and the Environment (ISCOLE; Katzmarzyk et al., 2013) and CheckPoint (Department of Social Services et al., 2019), a cross-sectional study nested between Waves 6 and 7 of the Longitudinal Study of Australian Children (LSAC).

ISCOLE Design and Participants

ISCOLE is a multinational cross-sectional study of around 7,000 children from 12 countries. Full methodological details of ISCOLE have been published previously (Katzmarzyk et al., 2013). Briefly, participants were from 26 (43% uptake) randomly selected schools in the Greater Adelaide region. A list of schools in Adelaide and surrounding areas was compiled and sorted into tertiles, based on the Index of Community Socio-Educational Advantage scores. Schools were randomly selected from tertiles with the probability of being chosen proportional to the estimated enrollment in Year 5. All Year-5 children (age 9–11 years) at randomly selected schools were invited to participate. Of those invited, 528 (57% uptake) were recruited. Data collection took place from September 2011 to December 2012.

Checkpoint Design and Participants

The study design of CheckPoint has been published previously (Department of Social Services et al., 2019). CheckPoint was conducted between February 2015 and March 2016, between the LSAC Waves 6 and 7, when children were aged 11 to 12 years. LSAC used a two-stage cluster sample design to recruit 5,107 (57.2% uptake) infants (age 0–1 years, birth or B cohort). CheckPoint was a one-off physical health and biomarker module offered to all LSAC B cohort participants who participated in Wave 6 data collection (n = 3,764; 74% retention). During Wave 6, home visit families were invited to consent to be contacted by the CheckPoint team (n = 3,513 families; 76% of eligible participants). Of the families that agreed to be contacted, 1,874 (53% of eligible participants, 42% of Wave 6 cohort and 37% of the original sample) consented to participate in CheckPoint.

Ethics

Ethical approval for ISCOLE was obtained from the institutional review board of the Pennington Biomedical Research Center in Baton Rouge, Louisiana, USA. Local ethical approval was granted for the Australian protocol by the University of South Australia Human Research Ethics Committee, the South Australian Department of Child Development, and the Catholic Education Department of South Australia. ISCOLE was registered on ClinicalTrials.gov, Identifier: NCT01722500. Written informed parental consent and child assent were obtained.

Ethical approval for CheckPoint was obtained from The Royal Children’s Hospital Melbourne Human Research Ethics Committee (33225D) and the Australian Institute of Family Studies Ethics Committee. A parent or a guardian provided written consent for their child’s participation in the study (14–26).

Measurement

Academic Achievement

For both CheckPoint and ISCOLE participants, additional consent was requested from school principals and parents to gather children’s academic achievement data via the Department for Education and Child Development. The results of a nationwide standardized assessment were used: the National Assessment Program—Literacy and Numeracy (NAPLAN), Year 5 (ISCOLE; age approximately 10–11 years) and Year 7 results (CheckPoint; age approximately 11–12 years; Australian Curriculum Assessment and Reporting Authority, 2016). For the ISCOLE analyses, NAPLAN data collected in 2015 were referenced, and for CheckPoint analyses, NAPLAN data collected during 2015 to 2017 were referenced. NAPLAN evaluates achievement in five domains: (1) grammar, (2) reading, (3) writing, (4) spelling, and (5) numeracy. An achievement score is calculated for each domain based on the number of correct responses and then converted to a scale score (0–1,000), with a higher score representing greater academic achievement. For the present study, the five academic domains were collapsed into two achievement categories representing literacy and numeracy skills. Literacy achievement was composed of the arithmetic mean of achievement in the literacy-related domains (grammar, reading, writing, and spelling scores), and numeracy achievement was composed of achievement in the numeracy domain.

Compositional 24-Hour Movement Behaviors

Sedentary behavior, LPA and MVPA, and sleep duration were derived from 7-day, 24-hour accelerometry. For ISCOLE participants, waist-worn Actigraph GT3X+ (ActiGraph LLC, Pensacola, FL, USA) were used, and for CheckPoint participants, wrist-worn GENEActiv (Activinsights, UK) accelerometers were used. For both samples, the minimum amount of data considered acceptable for inclusion in the analyses was 4 days with ≥10 hours of wear time per day during waking hours (including at least one weekend day) and ≥160 minutes total sleep duration for at least 3 nights (including at least one weekend night).

To determine daily sedentary time, LPA, and MVPA, the data were converted to 15-second (ISCOLE) or 60-second epochs (CheckPoint) before being converted to Evenson (Evenson et al., 2008) or Phillips (Phillips et al., 2013) cut points, respectively. For both samples, to estimate nocturnal sleep duration, data were downloaded in 1-second epochs and then collapsed into 60-second epochs. The raw accelerometer data were processed using ActiLife software (ActiGraph, Pensacola, FL) or Cobra custom software. CheckPoint participants completed a paper diary to record reasons for device removal. If the reason was “sport,” the associated period of nonwear time was replaced with 50% MVPA, 30% LPA, and 20% sedentary time. Detailed information regarding raw accelerometry data processing for both ISCOLE (Tudor-Locke et al., 2015) and CheckPoint (Fraysse et al., 2019) has been published elsewhere. The daily average time spent in each behavior was weighted for weekdays: weekend days at 5:2. Daily movement behaviors were normalized to sum to 1,440 minutes for each participant.

Sociodemographic Covariates

Child sex, age, socioeconomic position (SEP), and biological maturity were proxy-reported by parents. For ISCOLE, parental education (a proxy for SEP) was estimated from parental report (Katzmarzyk et al., 2013) of the highest education achieved by either parent (1 = less than high school, some high school, or completed high school; 2 = some postsecondary; 3 = bachelor’s degree; 4 = postgraduate). For CheckPoint, a composite SEP score was derived from parent-reported occupation, household income, and highest parental education level (Gibbings et al., 2009).

In ISCOLE, biological maturity was estimated using maturity offset, predicted from age, sex, sitting height, stature, and body mass (Mirwald et al., 2002). In CheckPoint, biological maturity was estimated using the Pubertal Development Scale (Robertson et al., 1992), where children self-reported their stage of pubertal development based on a number of typical physical indicators, including body hair development, occurrence of growth spurt, and skin changes. A mean score was calculated and later categorized as: 1 = prepubertal, 2 = early pubertal, 3 = mid-pubertal, 4 = late pubertal, and 5 =postpubertal.

Data Analysis

Data analysis consisted of (1) modeling the relationship between children’s daily movement behavior composition and their academic achievement and (2) using the models from (1) to predict academic achievement for activity mixes observed within the samples, visualizing each movement behavior individually.

Compositional analysis was used to model the relationship between movement behavior composition and academic achievement outcomes. This analytical approach enables all the daily movement behavior variables (sleep, sedentary time, LPA, and MVPA) to be considered concurrently as codependent predictors, rather than as individual independent variables (Aitchison, 1982). This is achieved by expressing the movement behavior variables as a set of log ratios rather than in raw units of minutes/day before including them as predictors in the regression model. As a result, regression coefficients are never interpreted as independent effects associated with increasing time spent in one movement behavior but as the effect associated with increasing movement behavior(s), while decreasing other movement behavior(s) to compensate.

Analyses were conducted using R (R Development Core Team, n.d.), with the packages: compositions (van Den Boogaart & Tolosana-Delgado, 2008), robCompositions (Templ et al., 2011), and lme4 (Douglas et al., 2015). Multiple linear regression models for academic achievement were used, with predictors of the daily movement behavior composition (expressed as isometric log-ratio coordinates) and covariates (sex, age, biological maturity, and SEP). Prior to log-ratio transformation, the movement behavior variables were checked for zero values. Zero MVPA was detected for two CheckPoint participants. As per published recommendations (Martín-Fernández et al., 2012), these zeros were replaced by small values to enable log-ratio transformation. Following procedures described in previous studies, we used the pivot coordinate system to produce a log-ratio that captures all information about one part of the composition (e.g., sleep), relative to the remaining parts of the composition (sedentary time, LPA, and MVPA; Chastin et al., 2015; Dumuid et al., 2018; Hron et al., 2017). We created four pivot coordinate systems, each containing a log-ratio that included one behavior (in the numerator) relative to the geometric mean of all remaining behaviors (in the denominator). For each data source (CheckPoint and ISCOLE), four multiple linear regression models were used, one for each of the pivot coordinate systems: (1) sleep versus remaining behaviors, (2) sedentary versus remaining behaviors, (3) LPA versus remaining behaviors, and (4) MVPA versus remaining behaviors. The regression coefficient for each of the pivot coordinates above was presented in a table to describe the relationship between each behavior, relative to the remaining behaviors. The significance of the overall movement behavior composition (i.e., the set of log-ratios) was assessed with an analysis of variance of the multiple linear regression model with type II tests.

Because ISCOLE participants were sampled from schools, random intercepts for school were included in the ISCOLE models. Preliminary analyses included interaction terms between sex and the movement composition. These were not significant and consequently omitted from the final models.

The second part of the analysis used the compositional regression models from above to estimate academic outcomes for the empirical mix of activity behaviors observed in the two samples. Estimated academic scores were plotted against each individual movement behavior. Loess curves with 95% confidence intervals were fitted to indicate the shape of the relationships.

Results

Participant Characteristics

Participants with complete data for all variables were included in the analyses (ISCOLE: Literacy n = 294, Numeracy n = 290; and CheckPoint: Literacy n = 940, Numeracy n = 930). Participant characteristics are described in Table 1.

Table 1.

Participant Characteristics.

Characteristics	ISCOLE (Adelaide)	CheckPoint
n ^a	294	940
Age, in years, M (SD)	10.7 (0.4)	12.0 (0.4)
Sex = Female, n (%)	161 (54.8)	449 (47.8)
Maturity offset, M (SD)	−1.8 (0.9)
Pubertal status, n (%)
Prepubertal		100 (10.6)
Early pubertal		242 (25.7)
Mid-pubertal		483 (51.4)
Late pubertal		111 (11.8)
Postpubertal		4 (0.4)
Highest parental education, n (%)
High School or lower	69 (23.5)
TAFE	100 (34.0)
University	71 (24.1)
Postgraduate	54 (18.4)
SEP z score		0.25 (1.00)
Movement behaviors, M (SD)
Sleep	566 (43)	567 (45)
Sedentary	383 (65)	673 (70)
LPA	440 (62)	164 (46)
MVPA	32 (18)	29 (22)
Movement behavior compositional means^d, % (sleep; sedentary; LPA; MVPA)	40; 27; 31; 2	40; 47; 11; 1
Literacy, M (SD)	497^b (60)	559^c (61)
Numeracy, M (SD)	489^b (65)^{*n = 290}	580^c(70)^{*n = 930}

Note. SEP = socioeconomic position; LPA = light physical activity; MVPA=moderate-to-vigorous physical activity; TAFE, technical and further education.

number of participants with valid accelerometry data. ^bYear 5 academic data (2015). ^cYear 7 academic data (2015–2017). ^dCompositional means are the geometric mean of each behavior, linearly adjusted to collectively sum to 100%.

The movement behavior composition was associated with literacy (p < .001 in both samples) and numeracy (ISCOLE p = .008, CheckPoint p = .04). The log-ratio coefficients from Table 2 indicate that lower LPA (relative to the remaining behaviors) was beneficially associated with both academic outcomes (Literacy: ISCOLE p = .002 and CheckPoint p = .001, Numeracy: ISCOLE p = .001 and CheckPoint p = .01). Higher sedentary time (relative to the remaining behaviors) was beneficially associated with literacy (ISCOLE p = .03, CheckPoint p = .009) but not with numeracy.

Table 2.

Relationships Between Movement Behaviors and Academic Achievement: Results From Compositional Models.

Test type	ISCOLE (Adelaide)				CheckPoint
Literacy	n	χ^a	p		N	Sum²	p
Overall composition	294	12.1	<.001		940	71939	<.001
Pivot coordinates	Regression coefficient	SE	t	p	Regression coefficient	SE	t	p
Sleep vs. remaining	19.1	32.8	0.6	.56	−11.4	16.2	−0.7	.48
Sedentary vs. remaining	46.4	20.8	2.2	.03	36.2	13.9	2.6	.009
LPA vs. remaining	−69.7	22.3	−3.1	.002	−26.0	8.0	−3.3	.001
MVPA vs. remaining	4.2	6.8	0.6	.54	1.3	2.7	0.5	.65
Numeracy	N	χ	p		N	Sum²	p
Overall composition	290	11.8	.008		930	35675	.04
Pivot coordinates	Regression coefficient	SE	t	p	Regression coefficient	SE	t	p
Sleep vs. remaining	50.0	34.7	1.4	.15	12.1	19.1	0.6	.53
Sedentary vs. remaining	16.7	22.0	0.8	.45	10.2	16.3	0.6	.53
LPA vs. remaining	−76.3	23.5	−3.2	.001	−23.6	9.4	−2.5	.010
MVPA vs. remaining	9.6	7.3	1.3	.19	1.4	3.3	0.4	.68

Note. Bold text indicates statistical significance (p < .05). Models adjusted for sex, age, maturity/puberty, highest parental education/household socioeconomic status. Remaining = the geometric mean of the remaining movement behaviors; LPA = light physical activity; MVPA = moderate-to-vigorous physical activity.

Multilevel models with random intercept for school were used for ISCOLE data to account for potential clustering because of the sampling frame. Multiple regression coefficient for multilevel models is presented as ChiSq, whereas the multiple regression coefficient for linear models is presented as SumSq.

The relationships are visualized in Figure 1. Each data point represents the “average” child’s academic score at an empirical duration of sleep, sedentary time, LPA, or MVPA. A narrow vertical spread of data points indicates that the variation in estimated outcome is highly dependent on the plotted movement behavior.

Figure 1.

The relationship between daily movement behaviors and NAPLAN (National Assessment Program—Literacy and Numeracy) literacy and numeracy achievement as estimated by compositional linear regression for ISCOLE (International Study of Childhood Obesity, Lifestyle and the Environment) and CheckPoint participants.

The standardized mean difference in literacy and numeracy achievement is shown in Figure 2. For literacy achievement, an effect size of up to 0.5 can be seen for a reduction in LPA by approximately 2 SDs from the mean. Similarly, for numeracy achievement, an effect size of up to 0.5 can be seen for a reduction in LPA by approximately 3 SDs from the mean.

Figure 2.

The relationship between daily behaviors and literacy and numeracy achievement as estimated by compositional linear regression models for CheckPoint and ISCOLE (International Study of Childhood Obesity, Lifestyle and the Environment) participants.

Discussion

This is the first study to use compositional data analysis in order to explore associations between children’s daily activity behavior compositions and academic achievement. The parts of children’s daily movement behavior composition were related to their academic achievement, with a small to medium effect size. In two Australian samples, higher LPA (at the expense of time spent in other behaviors, e.g., sleep, sedentary time, and MVPA) was associated with worse literacy and numeracy achievement. Additionally, higher sedentary time (relative to other behaviors) was associated with better literacy achievement in both samples. Notably, these results were consistent across two different samples, each of different ages (10 and 12 years), with each completing different NAPLAN tests (Year 5 and Year 7). Adding further robustness to these findings was that these results were obtained using different accelerometers (GENEActiv and ActiGraph) that yielded different estimates of both LPA and sedentary time.

The finding that relative to time spent in other movement behaviors, children with higher levels of LPA (e.g., playing catch, slow walking, and doing chores such as putting away clothes, taking out the rubbish, and feeding pets) achieve poorer academic achievement in both literacy and numeracy domains is new to the literature. Only two studies have considered associations between LPA and academic achievement, and neither used a compositional approach. Results from the available evidence have been inconsistent (Marques et al., 2018), reporting negative associations among girls, but positive associations among boys (Haapala et al., 2017), or no associations between LPA and academic achievement (Kwak et al., 2009). However, those studies did not account for the confounding influence of time spent in other daily movement behaviors (i.e., sleep, sedentary time, and MVPA), providing a possible explanation for these inconsistent findings.

Our analyses estimated academic achievement for the increase in LPA at the expense of all of the remaining behaviors, and results were strikingly consistent across both samples. Poorer academic achievement is unlikely to be related to higher LPA per se but may be explained by the displacement of time from the remaining behaviors. Time spent in LPA represents one of the exhaustive and mutually exclusive components of an individual’s 24-hour day. The non-LPA time remaining within an individual’s day can be divided into MVPA, sleep, and sedentary time. Thus, when time in LPA is increased, there must be a subsequent and equal decrease in time spent in the remaining movement behaviors (sleep, MVPA, and sedentary time). Children with higher LPA have less time available for activities that benefit learning: MVPA, sleep, and sedentary study. In other words, LPA may be associated with poorer academic achievement not because of what children do but because of what they do not do.

The finding that sedentary time (relative to other behaviors) was beneficially associated with literacy is not altogether surprising. While some sedentary activities (e.g., television viewing; Syvaoja et al., 2013) have previously been shown to be negatively related to academic achievement, sedentary time also encompasses behaviors that have been positively related to academic achievement (e.g., reading, writing, drawing, and studying; Haapala et al., 2014). School-related sedentary time constitutes 25% of total sedentary time (Olds et al., 2010). It is likely that children with higher LPA and higher screen time have less time available for studious activities, which may develop academic skills.

MVPA (relative to other behaviors) and sleep (relative to other behaviors) were not associated with academic achievement. Meta-analyses have demonstrated positive effects of both sleep (Short et al., 2018) and MVPA (Alvarez-Bueno et al., 2017) on academic achievement. However, the included studies viewed sleep and MVPA as separate entities. It is important here to bear in mind how to interpret compositional analyses of time-use data. The lack of an association between time spent in MVPA, for example, and academic achievement does not mean that MVPA per se is not beneficial for academic achievement. It means that higher MVPA in association with the trade-offs in time from other domains (e.g., sleep and sedentary study) that children actually make to achieve higher MVPA is neutral in relation to academic achievement. The potentially beneficial effect of MVPA might be attenuated by sacrificing sedentary study time, for example. Similar considerations apply for sleep.

Although our results cannot advise how much time should be spent in each movement behavior to optimize academic achievement, our results are consistent with the 24-hour movement guidelines of ≥60 minutes MVPA/day, ≤2 hours/day of recreational screen use, and 9 to 11 hours of sleep per night. When assessing associations with academic achievement, it is important to consider children’s activity behaviors within the context of the 24-hour day, rather than individually. Similar to results from our study, Walsh et al. (2018) and Faught et al. (2017) showed that meeting the new 24-hour movement guideline for MVPA was not associated with cognition or academic achievement. However, meeting the sleep and screen time recommendations in combination were positively associated with cognition and academic achievement. It is important to note that our study assessed time in all sedentary behaviors, while the 24-hour movement guidelines include only recreational screen time. Combined results suggest limiting recreational screen time, while increasing time in other sedentary behaviors (related to learning) and/or sleep may benefit academic achievement. Interventions aimed at improving academic achievement may focus increasing sleep and/or sedentary behavior related to learning at the expense of recreational screen time.

The strengths of this study include the use of 24-hour accelerometry and a standardized academic assessment in two large Australian samples. The analysis respected the compositional properties of daily behavior data. The robustness of the findings is strengthened by the use of two different samples, using two different accelerometers, and analyzing both literacy and numeracy outcomes. Nonetheless, certain limitations must be acknowledged. First, the study was cross-sectional; therefore, the direction of causation is unknown. Second, accelerometers cannot readily provide information regarding the type of behaviors. Unfortunately, we could not distinguish between intellectually passive and intellectually engaging activities, which may have different relationships with academic outcomes (Haapala et al., 2014). For example, we could not determine how much sedentary time was accumulated playing video games compared with reading. This may be of importance to the interpretation of findings and implications for intervention studies.

Conclusion

In summary, this study indicates that children’s academic achievement is related to their daily movement behavior composition. In particular, higher time spent in LPA at the expense of all remaining behaviors is associated with the most deleterious estimates for both literacy and numeracy scores. Interventions aimed at improving academic achievement among children may focus on encouraging nonscreen-based sedentary time, sleep, and MVPA at the expense of LPA. The findings suggest that any movement behavior intervention should be constructed within a 24-hour context—for example, if an intervention aims to increase MVPA, it would appear to be important to ensure that sedentary time and sleep are not decreased to compensate, but that the time is reallocated from LPA. Future studies should distinguish between screen and nonscreen sedentary behaviors, for example, by administering use of time recalls. There is future scope to investigate the effect of the daily movement composition on academic achievement in longitudinal and intervention studies. Because the kinds of time trade-offs children actually make will be important for outcomes, there is a need to understand how these trade-offs are made in intervention studies.

Footnotes

Author Contributions

TO and AW conceived the study. AW drafted the manuscript. TO and DD contributed to drafting the manuscript. DD conducted all statistical analyses. All authors reviewed and appraised the manuscript and have read and approved the final manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: ISCOLE is funded by The Coca-Cola Company. The Child Health CheckPoint is funded by the Australian National Health and Medical Research Council (NHMRC) Project Grant 1041352, supplemented by the Murdoch Childrens Research Institute, Royal Children’s Hospital Foundation (2014-241), Foundation for Children (Grant 2014-005), The University of Melbourne, the National Heart Foundation of Australia [100660] and the Financial Markets Foundation for Children (2014-055, 2016-310). Research at the MCRI is supported by the Victorian Government’s Operational Infrastructure Support Program. AW was funded by NHMRC project grant APP1143379 (2018–2022). DD is supported by an NHMRC early career fellowship (APP1162166) and a Heart Foundation postgraduate fellowship (102084). The funding bodies played no role in the design, collection, analysis, and interpretation of data or in writing the manuscript.

ORCID iD

Amanda Watson

References

Aitchison

(1982). The Statistical Analysis of Compositional Data. Journal of the Royal Statistical Society: Series B (Methodological), 44(2), 139–160. https://doi.org/10.1111/j.2517-6161.1982.tb01195.x

Alvarez-Bueno

Pesce

Cavero-Redondo

Sanchez-Lopez

Garrido-Miguel

Martinez-Vizcaino

(2017). Academic achievement and physical activity: A meta-analysis. Pediatrics, 140(6). https://doi.org/10.1542/peds.2017-1498

Australian Curriculum Assessment and Reporting Authority. (2016). National Assessment Program–Literacy and Numeracy. http://www.nap.edu.au/naplan/naplan.html

Carson

Hunter

Kuzik

Gray

C. E.

Poitras

V. J.

Chaput

J. P.

Saunders

T. J.

Katzmarzyk

P. T.

Okely

A. D.

Gorber

S. C.

Kho

M. E.

Sampson

Lee

Tremblay

M. S.

(2016). Systematic review of sedentary behaviour and health indicators in school-aged children and youth: An update. Applied Physiology, Nutrition, and Metabolism, 41(6 Suppl. 3), S240–S265. https://doi.org/10.1139/apnm-2015-0630

Chastin

S. F.

Palarea-Albaladejo

Dontje

M. L.

Skelton

D. A.

(2015). Combined effects of time spent in physical activity, sedentary behaviors and sleep on obesity and cardio-metabolic health markers: A novel compositional data analysis approach. PLoS ONE, 10(10), e0139984. https://doi.org/10.1371/journal.pone.0139984

Department of Social Services, Australian Institute of Family Studies, & Australian Bureau of Statistics. (2019). Growing up in Australia: Longitudinal Study of Australian Children (LSAC) Release 7.2 (Waves 1-7). http://doi.org/10.26193/F2YRL5

Douglas

Martin

Ben

Steve

(2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67(1), 1–48. https://doi.org/10.18637/jss.v067.i01

Dumuid

Olds

Martín-Fernández

J.-A.

Lewis

L. K.

Cassidy

Maher

(2017). Academic performance and lifestyle behaviors in Australian school children: A cluster analysis. Health Education & Behavior, 44(6), 918–927. https://doi.org/10.1177/1090198117699508

Dumuid

Stanford

T. E.

Martin-Fernandez

J. A.

Pedisic

Maher

C. A.

Lewis

L. K.

Hron

Katzmarzyk

P. T.

Chaput

J. P.

Fogelholm

Lambert

E. V.

Maia

Sarmiento

O. L.

Standage

Barreira

T. V.

Broyles

S. T.

Tudor-Locke

Tremblay

M. S.

Olds

(2018). Compositional data analysis for physical activity, sedentary time and sleep research. Statistical Methods in Medical Research, 27(12), 3726–3738. https://doi.org/10.1177/0962280217710835

10.

Evenson

K. R.

Catellier

D. J.

Gill

Ondrak

K. S.

McMurray

R. G.

(2008). Calibration of two objective measures of physical activity for children. Journal of Sports Sciences, 26(14), 1557–1565. https://doi.org/10.1080/02640410802334196

11.

Faught

E. L.

Ekwaru

J. P.

Gleddie

Storey

K. E.

Asbridge

Veugelers

P. J.

(2017). The combined impact of diet, physical activity, sleep and screen time on academic achievement: A prospective study of elementary school students in Nova Scotia, Canada. International Journal of Behavioral Nutrition and Physical Activity, 14(1), 29. https://doi.org/10.1186/s12966-017-0476-0

12.

Fraysse

Grobler

A. C.

Muller

Wake

Olds

(2019). Physical activity and sedentary activity: Population epidemiology and concordance in Australian children aged 11–12 years and their parents. BMJ Open, 9(Suppl. 3), 136–146. https://doi.org/10.1136/bmjopen-2018-023194

13.

Gibbings

Blakemore

Strazdins

(2009). Measuring family socioeconomic position. Australian Social Policy, 8, 121–168.

14.

Haapala

E. A.

Poikkeus

A. M.

Kukkonen-Harjula

Tompuri

Lintu

Vaisto

Leppänen

P. H. T.

Laaksonen

D. E.

Lindi

Lakka

T. A.

(2014). Associations of physical activity and sedentary behavior with academic skills: A follow-up study among primary school children. PLoS One, 9(9), e107031. https://doi.org/10.1371/journal.pone.0107031

15.

Haapala

E. A.

Vaisto

Lintu

Westgate

Ekelund

Poikkeus

A.-M.

Brage

Lakka

T. A.

(2017). Physical activity and sedentary time in relation to academic achievement in children. Journal of Science and Medicine in Sport, 20(6), 583–589. https://doi.org/10.1016/j.jsams.2016.11.003

16.

Hron

Filzmoser

de Caritat

Fišerová

Gardlo

(2017). Weighted pivot coordinates for compositional data and their application to geochemical mapping. Mathematical Geosciences, 49(6), 797–814. https://doi.org/10.1007/s11004-017-9684-z

17.

Katzmarzyk

P. T.

Barreira

T. V.

Broyles

S. T.

Champagne

C. M.

Chaput

J. P.

Fogelholm

Johnson

W. D.

Kuriyan

Kurpad

Lambert

E. V.

Maher

Maia

Matsudo

Olds

Onywera

Sarmiento

O. L.

Standage

Tremblay

M. S.

. . . Church

T. S.

(2013). The International Study of Childhood Obesity, Lifestyle and the Environment (ISCOLE): Design and methods. BMC Public Health, 13, Article 900. https://doi.org/10.1186/1471-2458-13-900

18.

Kwak

Kremers

S. P.

Bergman

Ruiz

J. R.

Rizzo

N. S.

Sjostrom

(2009). Associations between physical activity, fitness, and academic achievement. Journal of Pediatris, 155(6), 914–918.e911. https://doi.org/10.1016/j.jpeds.2009.06.019

19.

Maher

Lewis

Katzmarzyk

P. T.

Dumuid

Cassidy

Olds

(2016). The associations between physical activity, sedentary behaviour and academic performance. Journal of Science and Medicine in Sport, 19(12), 1004–1009. https://doi.org/10.1016/j.jsams.2016.02.010

20.

Marques

Santos

D. A.

Hillman

C. H.

Sardinha

L. B.

(2018). How does academic achievement relate to cardiorespiratory fitness, self-reported physical activity and objectively reported physical activity: A systematic review in children and adolescents aged 6–18 years. British Journal of Sports Medicine, 52(16), 1039. https://doi.org/10.1136/bjsports-2016-097361

21.

Martín-Fernández

J. A.

Hron

Templ

Filzmoser

Palarea-Albaladejo

(2012). Model-based replacement of rounded zeros in compositional data: Classical and robust approaches. Computational Statistics and Data Analysis, 56(9), 2688–2704. https://doi.org/10.1016/j.csda.2012.02.012

22.

Mirwald

R. L.

Baxter-Jones

A. D.

Bailey

D. A.

Beunen

G. P.

(2002). An assessment of maturity from anthropometric measurements. Medicine & Science in Sports & Exercise, 34(4), 689–694. https://doi.org/10.1097/00005768-200204000-00020

23.

Olds

T. S.

Maher

C. A.

Ridley

Kittel

D. M.

(2010). Descriptive epidemiology of screen and non-screen sedentary time in adolescents: a cross sectional study. Internation Journal of Behavioral Nutrition and Physical Activity, 7, Article 92. https://doi.org/10.1186/1479-5868-7-92.

24.

Phillips

L. R.

Parfitt

Rowlands

A. V.

(2013). Calibration of the GENEA accelerometer for assessment of physical activity intensity in children. Journal of Science and Medicine in Sport, 16(2), 124–128. https://doi.org/10.1016/j.jsams.2012.05.013

25.

R Development Core Team. (n.d.). R: A language and environment for statistical computing. https://cran.r-project.org/

26.

Robertson

E. B.

Skinner

M. L.

Love

M. M.

Elder

G. H.

Conger

R. D.

Dubas

J. S.

Petersen

A. C.

(1992). The Pubertal Development Scale. Journal of Early Adolescence, 12(2), 174–186. https://doi.org/10.1177/0272431692012002003

27.

Short

M. A.

Blunden

Rigney

Matricciani

Coussens

S. C, M. R.

Galland

(2018). Cognition and objectively measured sleep duration in children: A systematic review and meta-analysis. Sleep Health, 4(3), 292–300. https://doi.org/10.1016/j.sleh.2018.02.004

28.

Syvaoja

H. J.

Kantomaa

M. T.

Ahonen

Hakonen

Kankaanpaa

Tammelin

T. H.

(2013). Physical activity, sedentary behavior, and academic performance in Finnish children. Medicine & Science in Sports & Exercise, 45(11), 2098–2104. https://doi.org/10.1249/MSS.0b013e318296d7b8

29.

Templ

Hron

Filzmoser

(2011). robCompositions: An R-package for robust statistical analysis of compositional data. https://doi.org/10.1002/9781119976462.ch25

30.

Tudor-Locke

Barreira

T. V.

Schuna

J. M.

Jr. Mire

E. F.

Chaput

J. P.

Fogelholm

Kuriyan

Kurpad

Lambert

E. V.

Maher

Maia

Matsudo

Olds

Onywera

Sarmiento

O. L.

Standage

Tremblay

M. S.

Zhao

. . . Katzmarzyk

P. T.

(2015). Improving wear time compliance with a 24-hour waist-worn accelerometer protocol in the International Study of Childhood Obesity, Lifestyle and the Environment (ISCOLE). International Journal of Behavioral Nutrition and Physical Activity, 12, Article 11. https://doi.org/10.1186/s12966-015-0172-x

31.

van Den Boogaart

K. G.

Tolosana-Delgado

. (2008). “compositions”: A unified R package to analyze compositional data. Computers and Geosciences, 34(4), 320–338. https://doi.org/10.1016/j.cageo.2006.11.017

32.

Walsh

J. J.

Barnes

J. D.

Cameron

J. D.

Goldfield

G. S.

Chaput

J. P.

Gunnell

K. E.

Ledoux

A.-A.

Zemek

R. L.

Tremblay

M. S.

(2018). Associations between 24 hour movement behaviours and global cognition in US children: A cross-sectional observational study. Lancet Child & Adolescent Health, 2(11), 783–791. https://doi.org/10.1016/S2352-4642(18)30278-5

Associations Between 24-Hour Time Use and Academic Achievement in Australian Primary School–Aged Children

Abstract

Background

Aims

Method

Discussion and conclusion

Keywords

Method

Study Design and Participants

ISCOLE Design and Participants

Checkpoint Design and Participants

Ethics

Measurement

Academic Achievement

Compositional 24-Hour Movement Behaviors

Sociodemographic Covariates

Data Analysis

Results

Participant Characteristics

Discussion

Conclusion

Footnotes

Author Contributions

Declaration of Conflicting Interests

Funding

ORCID iD

References