Received: 4 November 2020 Revised: 10 April 2021 Accepted: 10 April 2021

DOI: 10.1111/obr.13263

PEDIATRIC OBESITY/BEHAVIOUR

Age-related change in sedentary behavior during childhood

and adolescence: A systematic review and meta-analysis

Elli Kontostoli1 | Andy P. Jones2 | Natalie Pearson3 | Louise Foley4 |

5
Stuart J. H. Biddle | Andrew J. Atkin1

1
School of Health Sciences, Faculty of
Medicine and Health Sciences, University of Summary
East Anglia, Norwich, UK Sedentary behaviors are highly prevalent in youth and may be associated with
2
Norwich Medical School, Faculty of Medicine
markers of physical and mental health. This systematic review and meta-analysis
and Health Sciences, University of East Anglia,
Norwich, UK aimed to quantify the age-related change in sedentary behavior during childhood
3
School of Sport, Exercise and Health and adolescence. Ten electronic databases were searched. Inclusion criteria speci-
Sciences, Loughborough University,
Loughborough, UK fied longitudinal observational studies or control group from an intervention; par-
4
Centre for Diet and Activity Research at the ticipants aged ≥5 and ≤18 years; a quantitative estimate of the duration of SB;
MRC Epidemiology Unit, University of
and English language, peer-reviewed publication. Meta-analyses summarized
Cambridge, Cambridge, UK
5
Centre for Health Research, University of
weighted mean differences (WMD) in device-assessed sedentary time and
Southern Queensland, Toowoomba, questionnaire-assessed screen-behaviors over 1-, 2-, 3-, or more than 4-year
Queensland, Australia
follow-up. Effect modification was explored using meta-regression. Eighty-five
Correspondence studies met inclusion criteria. Device-assessed sedentary time increased by
Elli Kontostoli, School of Health Sciences,
Faculty of Medicine and Health Sciences,
(WMD 95% confidence interval [CI]) 27.9 (23.2, 32.7), 61.0 (50.7, 71.4), 63.7
University of East Anglia, Norwich Research (53.3, 74.0), and 140.7 (105.1, 176.4) min/day over 1-, 2-, 3-, and more than
Park, Norwich NR4 7TJ, UK.
Email: e.kontostoli@uea.ac.uk
4-year follow-up. We observed no effect modification by gender, baseline age,
study location, attrition, or quality. Questionnaire-assessed time spent playing
Funding information
University of East Anglia, Faculty of Medicine
video games, computer use, and a composite measure of sedentary behavior
and Health Sciences; Academy of Medical increased over follow-up duration. Evidence is consistent in showing an age-
Sciences/the British Heart Foundation/the
Government Department of Business, Energy
related increase in various forms of sedentary behavior; evidence pertaining to
and Industrial Strategy/the Wellcome Trust variability across socio-demographic subgroups and contemporary sedentary
Springboard Award, Grant/Award Number:
SBF003\1015
behaviors are avenues for future research.

KEYWORDS
adolescents, change, children, sedentary behavior, systematic review

1 | I N T RO DU CT I O N higher risk of obesity and low cardiorespiratory fitness.2,3 Total seden-

tary time may also be associated with poor academic achievement
There is evidence that sedentary behavior during childhood may be and social interactions,2 depression, and low self-esteem.4–6 This evi-
associated with several adverse health outcomes, independently of dence is reflected in public health guidelines suggesting that seden-
physical activity.1 Device-measured total sedentary time and self- tary behavior be limited, though there remain important
reported screen-based sedentary behavior may be associated with a inconsistencies and limitations of the existing research,7–11 and

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium,
provided the original work is properly cited.
© 2021 The Authors. Obesity Reviews published by John Wiley & Sons Ltd on behalf of World Obesity Federation.

Obesity Reviews. 2021;1–11. wileyonlinelibrary.com/journal/obr 1

2 KONTOSTOLI ET AL.

further high-quality longitudinal and experimental research is required CRD42018106948). The review is reported in accordance with The
to better understand the links between sedentary behavior and health Preferred Reporting Items for Systematic Reviews and Meta-
in this population. Analyses (PRISMA) guidelines (PRISMA checklist is available in
Several studies have shown that sedentary behavior is highly Table S1) and Meta-analysis Of Observational Studies in Epidemiol-
prevalent in young people. Global surveillance data showed that ogy (MOOSE).23,24
approximately half of children and adolescents spend more than 2 h a
day in screen-based activities.12 A systematic review found that
device-measured sedentary time accounted for almost half of the 2.1 | Search strategy
after-school time in children and over half of the after-school time in
adolescents.13 Another review found that device measured sedentary Ten electronic databases were searched (PsycINFO, CINAHL, Web of
time accounted for 6.4 h a day in children and 7.3 h a day in adoles- Science, MEDLINE, Embase, Scopus, LILACS, Cochrane Library, Allied
cents.14 Informed by this evidence, public health recommendations and Complementary Medicine Database [AMED], and Applied Social
advise that children and adolescents should minimize the amount of Sciences Index and Abstracts [ASSIA]) in September 2018 with no
time they spend sedentary11 or limit the duration of specific sedentary chronological limits set. Searches were re-run in June 2020. Manual
15
behaviors, such as recreational screen time. searches of the reference lists of published systematic reviews and
In public health surveillance and epidemiological studies, seden- related articles were also completed to identify potentially relevant
tary behaviors are typically measured using body-worn devices articles. The searches were focused on three groups of keywords:
(e.g., accelerometry) or self- or proxy-reported questionnaires. Despite sedentary behavior, study design, and study population. Key terms
some overlap in content, the correlation between device- and were used in combination with relevant MeSH headings. The search
questionnaire-assessed sedentary behavior is typically low, and they strategy was developed in conjunction with an academic librarian. An
appear to be differentially associated with health markers.16,17 It is of example search strategy is provided in Table S2. The search was
interest, therefore, to examine both methods of measurement when conducted by EK.
exploring changes in sedentary behavior over time. A recent system-
atic review of longitudinal studies reported an increase of 10–20 min/
day/year in device-measured sedentary time and screen-based 2.2 | Inclusion and exclusion criteria
sedentary behavior during the transition from primary to secondary
education.18 This is consistent with cross-sectional data from the Studies were included if they (1) used an observational study design
International Children's Accelerometry Database (ICAD), which or a provided data for the control arm in an experimental study;
showed that device-measured sedentary time increased progressively (2) provided a quantitative estimate of duration of at least one seden-
from the age of 5 years.19 Previous research has shown that seden- tary behavior with data collected at ≥2 time points (minimum of
tary behavior may be higher in non-White children, those with a 1-year between baseline and follow-up); (3) included children and/or
higher body mass index (BMI),14,18 and those from families of lower adolescents aged ≥5 and ≤18 years at baseline and follow-up; and
socio-economic position,20,21 suggesting that age-related change in (4) were published in an English language peer-reviewed journal.
sedentary behavior may also vary in these subgroups. Understanding Commentaries, conference papers, qualitative studies, pilot studies,
of social and demographic variation in sedentary behavior change will and trials without a no-treatment control group were excluded, as
help with the targeting of behavior change interventions. were studies in clinical populations.
It is understood that health behaviors in childhood and adoles- When the same study was reported in multiple papers, the fol-
cence may persist into adulthood,22 highlighting the need to estab- lowing prioritization was applied to select papers for inclusion: (1) the
lish the timing of changes in sedentary behavior during this period, article with the most follow-up assessment points; (2) the article with
as well as the population groups that may be most at risk. To our a variety of activities (i.e., most sedentary behaviors) for self-reported
knowledge, there is no published systematic review that quantifies data; (3) the paper with the biggest sample size; and (4) stratification
the age-related change in sedentary behavior during childhood and for boys/girls, week/weekend days.
adolescence. Therefore, the aim of this review was to synthesize
existing evidence on age-related changes in sedentary behavior
during childhood and adolescence. A secondary aim was to 2.3 | Identification of relevant studies
examine whether the magnitude of change varied across social or
demographic population groups. Covidence review management software (www.covidence.org) was
used for the screening and selection of records retrieved from
electronic and manual searches, including the removal of duplicates.
2 | METHODS Articles were selected by screening the titles and the abstracts, and if
abstracts were not available or did not provide enough data, the entire
The review protocol was registered with the International article was retrieved and screened to determine whether it met the
Prospective Register of Systematic Reviews ([PROSPERO] inclusion criteria. Articles that were not available through open access
KONTOSTOLI ET AL. 3

publication were obtained through a university library subscription, device-measured sedentary time and self- or proxy-reported screen
email request to the author or inter-library loan as appropriate. Screen- behaviors were synthesized by meta-analysis. We opted to meta-
ing of titles, abstracts, and full-texts was undertaken by the lead author analyze screen-based behaviors due to prominence of those in chil-
(EK). A second reviewer (AJA) independently screened 10% of titles dren and adolescents and inclusion in public health guidelines. In order
and abstracts with disagreements resolved by discussion. Ninety-two to prepare data for meta-analysis, conversion for reports of device-
percent agreement was achieved at this stage. Ten percent of full texts based and self- or proxy-reported data were undertaken as follows.
were also screened by a second reviewer (NP). There was an agreement The metric chosen was the original unit reported in most of the
of 96% at this stage. Disagreements were solved by discussion and studies (i.e., minutes per day during the week [Monday to Sunday]).
when uncertainties were raised, adjudication was made by AJA. When studies reported sedentary time separately for Saturday and
Sunday, the mean and standard deviation (SD) of those values were
calculated to provide mean sedentary time for the weekend. Conver-
2.4 | Data extraction sions were also made for studies reporting minutes per day separately
on a weekday and minutes per day on a weekend; in those cases, the
Data were extracted on forms developed specifically for this mean value was calculated to provide mean weekly sedentary time
review. Extracted data included (1) author name, year of publica- ((5  weekday) + (2  weekend)/7). For studies reporting hours per
tion, country, and study name (if applicable); (2) study design; day or hours per week, data were converted to minutes per day. For
(3) aim(s) of the study; (4) follow-up duration; (5) sample size; studies reporting data in medians, interquartile range (IQR), and stan-
(6) baseline age, gender, ethnicity, socioeconomic position, weight dard error, data were converted to mean and SD following published
status, BMI, and BMI z-score; (7) age at follow-up; (8) methods uti- methods.28 For self or proxy methods, studies were grouped
lized for device-based (counts, epochs, time, and days needed for according to whether they reported on a single sedentary behavior
inclusion) and self- or proxy-reported assessments; (9) duration of (e.g., TV viewing only) or a composite of multiple behaviors in various
sedentary behavior for each assessment or change between assess- combinations (e.g., TV viewing, computer use, and video games).
ments; and (10) attrition rates. Data were extracted for the We opted not to meta analyze data on non-screen-based behav-
smallest reported independent subsample (k). Data extraction was iors due to limited number of studies providing this data, the hetero-
conducted by EK, and extracted data for 10% of papers were geneity in questionnaire content and the limited evidence of
checked for accuracy by AJA. associations with health and well-being. None of the studies tested
statistically for change over time. Findings are summarized in the table
but omitted from the synthesis.
2.5 | Methodological quality assessment

Included studies were appraised for methodological and reporting 2.7 | Statistical analysis
quality using a scale adapted from previous reviews of observational
longitudinal research.25–27 The following domains were assessed: Data on change in sedentary behavior were combined using random
study population and participation rate (two items); study attrition effect meta-analysis, conducted in STATA 16.0 (Stata Corporation,
(three items); data collection (three items); and data analysis (one Texas, USA). Data included in the meta-analysis were converted to a
item). An additional item, pertaining to report of cut-point used in data common metric, non-standardized weighted mean differences
processing, was included in appraisal of studies that assessed seden- (WMD). Studies were meta-analyzed according to the duration of
tary time by accelerometer (Table S3). Published methods papers follow-up (e.g., 1, 2, 3, and 4+ years) except for video games and
were reviewed alongside included studies where necessary. The lead computer use for which meta-analysis was conducted for 1, 2, and 3
author (EK) undertook quality appraisal. A second reviewer + years of follow-up due to the limited number of studies that
(LF) conducted duplicate quality appraisal in a 10% subsample of assessed change over 4 or more years (video games N = 2 and com-
papers and disagreements were resolved by discussion. Each item for puter use N = 3). Heterogeneity was quantified using the I2 statis-
the included studies was assessed with a 1 or 0 score. The overall tic.29 Meta-regression was used to explore the impact of possible
quality of a study was determined by the sum of positively scored effect modifiers (gender, age span, study's location, and quality). Can-
items and by converting to a percentage. Studies were rated high didate moderators were selected based on the data extracted and
quality if score was ≥71%, moderate quality if score was ≥41% and potential to inform behavior change interventions. Age range
≤70%, and low quality if score was ≤40%. referred to either childhood (age from 5 to 10 years old) or adoles-
cence (age 11 to 18 years old) at baseline. Study location was sum-
marized as Europe, North America, South America, Australia and
2.6 | Data synthesis New Zealand, Africa, or Asia. Annual change in sedentary behavior
(minutes/day/year) was estimated by subtracting baseline sedentary
The unit of analysis was independent subsample (k), defined as the behavior from follow-up and dividing by duration of follow up
smallest subsample for which relevant data were reported. Data on (years).29 SD of annual change was calculated according to methods
4 KONTOSTOLI ET AL.

described by Higgins et al., assuming a correlation of 0.5, consistent instruments to measure TV-viewing, video games, computer use,
with previous research.30 Eggers test for publication bias was con- doing homework, or traveling by car/bus, either separately or in com-
ducted for all meta-analyses. 31
bination. Thirty studies (k = 52) used device-based methods to assess
total sedentary time. Eight papers reported data for both device-
based and self-assessment or proxy assessment. Methodological char-
3 | RESULTS acteristics of included studies are summarized in Table 1, stratified by
method of measurement. A study-level summary of included studies
The literature search returned 17,296 references (Figure 1). After is presented in Tables S4a and S4b.
removal of duplicates, 14,341 titles and abstracts were screened, Methodological quality scores for each study are provided in
from which 834 full-text papers were assessed for eligibility. Of Table S5. An 84% agreement was achieved on bias scoring between
those, 722 were excluded due to not meeting the inclusion criteria. reviewers, and discrepancies were resolved via discussion. Of the
Subsequently, a further 27 papers were excluded as they included 85 included studies, 63% were rated high quality, 32% were rated
duplicate data available in other papers. Eighty-five papers were moderate quality, and 5% were rated low quality.
included in the review, of which 10 were identified in the updated
search in June 2020.
Most studies were conducted in Europe (N = 36) or in North 3.1 | Device-measured sedentary time
America (N = 23), had a sample size of >1000 participants (N = 33),
and had a follow-up duration of ≤3 years (N = 51). The majority Meta-analysis indicated that sedentary time increased by (WMD [95%
(N = 63; k independent samples = 129) used self-reported CI]) 27.9 (23.2, 32.7), 61 (50.7, 71.4), 63.7 (53.3, 74), and 140.7

F I G U R E 1 Literature search and article

screening process
KONTOSTOLI ET AL. 5

T A B L E 1 Descriptive characteristics of the included studies, Findings are summarized narratively only for studies that reported
stratified by method of sedentary behavior measurement change in academic related activities and travel by car/bus
Device Self-reported or proxy (Table S6).
measured reported

N = 30 N = 55
3.2.1 | Single sedentary behaviors
Sample size
<100 2 (6.6) 5 (9)
Meta-analysis indicated that changes in duration of TV viewing were
100–499 14 (46.6) 10 (18.1)
nonsignificant at 1 year (WMD [95% CI]) ( 0.6 [ 5.0, 3.7]), 2 years
500–999 7 (23.3) 14 (25.4)
(7 [ 0.1, 14.2]), and 3 years (0 [ 4.8, 4.8]) of follow-up. Based on
>1000 7 (23.3) 26 (47.2) 16 independent samples, an increase in TV viewing was reported in
Duration of follow-up those studies that reported change over 4+ years of follow-up (26.1
1 year 5 (16.6) 12 (21.8) [0.9, 51.3]). In all cases, heterogeneity was high (≥93.7%) and statisti-
2 years 12 (40) 18 (30.9) cally significant (Figure 3). Time spent playing video games increased
3 years 4 (13.3) 10 (16.3) by (WMD [95% CI]) 12.4 (4.8, 19.9), 5.7 (0.3, 11), and 15.3 (4.8,
4+ years 9 (30) 15 (27.2) 25.8) min/day over 1, 2, and 3+ years of follow-up, respectively. In all
Region cases, heterogeneity was high (≥92.2%) and statistically significant

Europe 18 (60) 18 (32.7) (Figure 4). Computer use increased by (WMD [95% CI]) 18.4 (5.3,
31.5), 28.7 (16.8, 40.5), and 35.5 (19.4, 51.6) min/day over 1, 2, and 3
Australia and NZ 5 (16.6) 12 (21.8)
+ years of follow-up, respectively. Heterogeneity was high (≥68%)
N. America 5 (16.6) 18 (32.7)
and statistically significant (Figure 5). Using Egger's test, there was no
S. America n/s 3 (5.4)
evidence for publication bias for single sedentary behaviors over 1, 2,
Asia 2 (6.6) 3 (5.4)
3, or 4+ years of follow-up. Meta-regression indicated no statistically
Africa n/s 1 (1.8)
significant effect modification by gender, baseline age, or study attri-
Age at baseline tion rate or quality (p > 0.05). Compared to Europe, studies conducted
Children only 14 (46.6) 26 (47.2) in South America reported larger increases in video game use over
Adolescents only 12 (40) 28 (50.9) 1 year of follow-up (p = 0.002) and those conducted in Asia reported
Children and 4 (13.3) 1 (1.8) larger increases in computer use over 2 years of follow-up (p = 0.03).
adolescents Estimated annual changes (minutes per day) in TV viewing, video
Data are presented N (%). game, and computer use were (ES [95% CI]) 0.6 ( 0.1, 1.4), 0.6 (0.2,
Abbreviations, NZ: New Zealand, n/s: no studies. 1.1), and 2 (1, 3), respectively (Figures S2–S4).

3.2.2 | Composite measures

(105.1, 176.4) min/day over 1, 2, 3, and 4+ years of follow-up,
respectively. In all cases, heterogeneity was high (≥96%) and statisti- Meta-analysis indicated that combined TV viewing, video game play,
cally significant (Figure 2). Meta-regression indicated no statistically and computer use increased by (WMD [95% CI]) 20.8 (9.9, 31.8), 19.9
significant effect modification by gender, baseline age or study's loca- (14.1, 25.6), 40 (16.3, 63.7), and 42.6 (21.1, 64.1) min/day over 1, 2, 3,
tion, attrition rate, or quality (p > 0.05). Using Egger's test, there was and 4+ years of follow-up, respectively. In all cases, heterogeneity
no evidence for publication bias in 1, 3, and 4+ years of follow-up, was high (≥97.3%) and statistically significant (Figure 6). Using Egger's
but there was some evidence for publication bias for 2-year duration test, there was no evidence for publication bias over 1, 2, 3, and 4
of follow-up (p = 0.04). + years of follow-up. Meta-regression indicated no statistically signifi-
Meta-analysis indicated an annual change in sedentary time cant effect modification by gender, baseline age or study's location,
(minutes per day) of (ES [95% CI]) 7.8 (6.4, 9.1) (Figure S1). The I2 attrition rate, or quality (p > 0.05).
value was 80.9%, indicating high heterogeneity. Estimated annual change (minutes per day) in TV viewing, video
game play, and computer use was (ES [95% CI]) 0.3 (0.2, 0.5)
(Figure S5).
3.2 | Self- or proxy-reported sedentary behavior

Studies reporting data collected by questionnaire presented data 4 | DI SCU SSION

for single behaviors (such as TV viewing, video games, computer
use, homework, and travel by car or bus) and/or behaviors aggre- To our knowledge, this is the first systematic literature review to sum-
gated in various combinations to create composite measures. marize and meta-analyze longitudinal data on changes in sedentary
6 KONTOSTOLI ET AL.

F I G U R E 2 Change in device-measured sedentary time over (A) 1-, (B) 2-, (C) 3-, and (D) 4- to 10-year duration. Abbreviations, B: boys, G:
girls, BG: boys and girls, y: year

behavior during childhood and adolescence. For device-based mea- evidence that sedentary behavior tracks moderately from childhood
sures of sedentary behavior, meta-analysis indicated that sedentary to adulthood,35,36 age-related increases in overall sedentary time, as
time increases over time, with larger increases seen over longer captured by device-based measurement, likely reflect changes in
durations of follow-up. The meta-regression indicated no statistically behavior in a number of domains and settings over time. The need for
significant differences in sedentary time change according to age, behavior change interventions to limit such changes will require
gender, study location, quality, or attrition. For self- or proxy- clearer evidence on the specific nature of these changes, accompanied
reported sedentary behavior, our synthesis indicated increases in by stronger epidemiological evidence on how specific behaviors are
time spent in video game play, computer use, and a composite linked with health and well-being.
marker of screen-based behavior, but TV viewing appeared rela- A key finding of this review was that change in sedentary
tively stable and increased only over the longest durations of behavior did not differ according to age at first assessment. Our
follow-up. meta-regression showed that changes in this behavior were similar in
The meta-analysis indicated that device-measured daily sedentary children (≥5 and <10 years old) and adolescents (≥10 and <18 years
time increased as children and adolescents age, by approximately old) for either device-measured or self- or proxy-reported sedentary
28 min over 1 year, 61 min over 2 years, 64 over 3 years, and 141 min behavior, supporting the view that, where appropriate, interventions
over 4 years of follow-up. Findings are consistent with cross-sectional to limit the age-related increase in sedentary behavior may need to
data from the ICAD study, which showed that sedentary time be implemented throughout the childhood period. There is substan-
increased in an approximately linear manner from the age of 5 years tial evidence that adolescents engage in higher levels of sedentary
onward, though the magnitude of change was not quantified in behavior than children,14,18 but this is the first study to our knowl-
19
minutes. Similarly, a recent study using pan-European harmonized edge that has examined whether changes in sedentary time within
accelerometer data showed a linear increase in sedentary time with these periods differ. Findings are consistent with recent evidence
age; at age 4/5 years, children accumulated approximately 250 min/ that the age-related decline in physical activity may start during
day of sedentary time increasing to around 450 min/day at age childhood, rather than being limited to the adolescent period.37
32
14/15. Changes in sedentary behavior mirror the well-documented Further information on how the accumulation of device-assessed
reduction in physical activity during childhood.19,30,33,34 Given sedentary time changes with age, including bout length and
KONTOSTOLI ET AL. 7

F I G U R E 3 Change in self- or proxy-reported TV viewing over (A) 1-, (B) 2-, (C) 3-, and (D) 4- to 10-year duration. Abbreviations, B: boys, G:
girls, BG: boys and girls, y: year

F I G U R E 4 Change in self- or proxy-reported video games over (A) 1-, (B) 2-, and (C) 3- to 4-year duration. Abbreviations, B: boys, G: girls,
BG: boys and girls, y: year
8 KONTOSTOLI ET AL.

F I G U R E 5 Change in self- or proxy-reported computer use over (A) 1-, (B) 2-, and (C) 3- to 5-year duration. Abbreviations, B: boys, G: girls,
BG: boys and girls, y: year

F I G U R E 6 Change in self- or proxy-reported composite screen-based behaviors over (A) 1-, (B) 2-, (C) 3-, and (D) 4- to 7-year duration.
Abbreviations, B: boys, G: girls, BG: boys and girls, y: year
KONTOSTOLI ET AL. 9

frequency of breaks, would be beneficial, as such factors may have is a need for qualitative studies to explore how adolescents'
important implications for health. Moreover, further evidence attitudes and preferences for different screen- and non-screen based
describing the social and environmental factors that influence behaviors change over time.
sedentary behavior and how these evolve over time is also required The data for non-screen-based behaviors showed that academic-
to inform intervention design. related activities and travel by car/bus increased over time, while time
We found no evidence that change in device-measured sedentary spent reading for school declined, but this was assessed in only three
time and self- or proxy-reported sedentary behavior differed between studies. Change in car/bus perhaps reflects greater engagement in
boys and girls over time. This is in line with the results from a recent social and recreational activities away from home as children age.44,45
study using pan-European accelerometer data, which suggested that, The concurrent increase in academic activities and the decline in
while girls accumulated more sedentary time than boys, the pattern of school-related reading appears contradictory. These contradictions
32
change with increasing age was similar. A recent systematic review may reflect the transition from reading being an academic activity in
that focused on sedentary behavior change across the primary to its own right to been a routine activity required to fulfill other school-
secondary school transition also found no evidence for a gender related tasks. The lack of studies reporting age-related changes in
difference, but this was not tested statistically.18 Additionally, a these behaviors is a clear gap in the evidence, and further research
systematic review of tracking of sedentary behavior from childhood would provide a richer picture of changes in young people's sedentary
to adolescence found little evidence of a gender difference.38 Results behavior patterns and preferences over time. In particular, only two
from an analysis in ICAD showed that boys were less sedentary and studies were identified that assessed time spent in academic-related
more active than girls at all ages, though the change in sedentary time activities with and without a computer or tablet. As we seek to further
19
appeared similar for both boys and girls over time. Despite the disentangle the detrimental and beneficial associations of sedentary
apparent consistency of existing evidence, changes in sedentary time behavior with physical and mental health, this topic in particular
between both genders over time have attracted little attention. Of would be worthy of further study.
the 85 studies that were included in the review, only 39 (42%) A key strength of this review is the inclusion of studies that used
reported data separately for boys and girls. Although trajectories of either device-based or questionnaire-based methods of measurement
overall sedentary time may be similar, it remains unclear whether and use of meta-analysis to synthesize the data. In addition, we used
changes in time spent in specific sedentary behaviors differ between broad search criteria to identify relevant articles across 10 electronic
boys and girls; further information on this topic would be informative databases and the manual searches without publication date restric-
for intervention design. tions. Our protocol was registered with PROSPERO and the review is
The meta-analyses indicated that time spent in video game play, reported in accordance with PRISMA guidelines23 and MOOSE.24 We
computer use, and a composite measure of screen-based behavior included and summarized evidence from studies that measured a
increased over all durations of follow-up examined; however, time broad range of sedentary behaviors, both individually and in combina-
spent in TV viewing did not change for up to 3 years but it tion, providing a comprehensive overview of the published literature
increased for more than 4-year duration of follow-up. Our findings and highlighting gaps to be addressed in future research. Limitations
are similar to recent studies that showed that time spent in a com- of this work include the restriction to English language publications in
posite marker of screen-based behavior and also computer use and peer-reviewed journals, which may have resulted in the exclusion of
12,39–41
video game play increased over time. Interestingly, our find- relevant articles. In addition, we deviated from our published protocol
ings on TV viewing partially contrasts with prior research reporting a by not searching the Global Health database (not available in our
decrease by a relatively small amount in traditional TV viewing over institution). We did duplicate appraisal of study quality for 20% of
time,12,42 but this was not tested statistically. In contradiction with studies. We found a high level of agreement when duplicate screening
earlier findings, a previous review of longitudinal studies looking at papers for inclusion, with no evidence of high levels of discrepancy
TV viewing found increases in boys and girls over time but the for particular items. As a result, we deemed it necessary to only
results were mixed in boys and girls according to weight status.18 duplicate screen 20% of papers, but an implication of this is that there
Most studies in the current review focused on traditional sedentary may have been some discrepancies in those not duplicate screened.
behaviors, such as TV viewing and video games, with very few Finally, we report selected accelerometer data collection and
describing changes in contemporary behaviors, such as tablet and processing criteria in our summary tables but did not include/exclude
phone use. The number of devices through which young people may papers from the meta-analysis on the basis of these factors. Variability
access the internet and/or audiovisual media has expanded rapidly in data processing methods and compliance with study protocols may
in recent years. Recent data showed that the proportion of children have contributed to heterogeneity in the estimates of change that
and adolescents aged 5–15 years old watching TV programs on were synthesized.
tablets increased from 27% in 2015 to 43% in 2019 and on mobiles The current study highlights several areas that would benefit
from 15% in 2015 to 26% in 2019.43 Further research is needed to from further research. Few of the included studies conducted
examine how the duration of time spent in newer screen-based stratified analyses to examine whether change in sedentary behavior
behaviors changes over time and whether this is displacing time varied according to social, demographic, or anthropometric factors,
previously spent watching broadcast television. Alongside this, there such as socio-economic position, ethnicity, or BMI. This information
10 KONTOSTOLI ET AL.

would aid in the identification of at-risk populations for intervention. RE FE RE NCE S

Additionally, observational longitudinal studies are needed to collect 1. Carson V, Hunter S, Kuzik N, et al. Systematic review of sedentary
data on the wide range of electronic media devices used by young behaviour and health indicators in school-aged children and youth: an
update. Appl Physiol Nutr Metab. 2016;41(3):240-265. https://doi.
people, moving beyond simplistic assessments of computer use or
org/10.1139/apnm-2015-0630
broadcast television. One in four young people (5–15 years old) do 2. Tremblay MS, LeBlanc AG, Kho ME, et al. Systematic review of sed-
not watch live broadcast TV at all and smartphone ownership entary behaviour and health indicators in school-aged children and
increased by 10% from 2015 to 2019.43 Further research is also youth. Int J Behav Nutr Phys Act. 2011;8(1):98. https://doi.org/10.
1186/1479-5868-8-98
warranted to examine multi-tasking of portable devices and the con-
3. Mitchell JA, Byun W. Sedentary Behavior and Health Outcomes in
text in which such devices are used, which may moderate how these Children and Adolescents. Am J Lifestyle Med. 2013;8(3):173-199.
behaviors influence health. This may necessitate the development https://doi.org/10.1177/1559827613498700
and validation of new tools to capture the diversity of electronic 4. Hoare E, Milton K, Foster C, Allender S. The associations between
sedentary behaviour and mental health among adolescents: A system-
media devices being used, in combination with information on
atic review. Int J Behav Nutr Phys Act. 2016;13(1):108. https://doi.
content and context.
org/10.1186/s12966-016-0432-4
5. Vancampfort D, Stubbs B, Firth J, van Damme T, Koyanagi A. Seden-
tary behavior and depressive symptoms among 67,077 adolescents
5 | C O N CL U S I O N aged 12-15years from 30 low- and middle-income countries. Int J
Behav Nutr Phys Act. 2018;15(1):1-9. https://doi.org/10.1186/
s12966-018-0708-y
This is the first systematic review to summarize published evidence 6. Rodriguez-Ayllon M, Cadenas-Sanchez C, Estévez-Lo  pez F, et al. Role
on age-related change in sedentary behavior in children and adoles- of Physical Activity and Sedentary Behavior in the Mental Health of
cents. Our findings show that device-measured sedentary time Preschoolers, Children and Adolescents: A Systematic Review and
Meta-Analysis. Sport Med. 2019;49(9):1383-1410. https://doi.org/10.
increases with age; with no evidence that the magnitude of change
1007/s40279-019-01099-5
varied by gender or age, though few studies provided the required 7. Hashem R, Pablo J, Lόpez R, et al. Associations between objectively
data for these analyses. Synthesis of data on screen-based sedentary assessed and questionnaire-based sedentary behaviour with body
behavior assessed by questionnaire also indicated an increase with mass index and systolic blood pressure in Kuwaiti adolescents. BMC
Res Notes. 2019;12(1):8-13. https://doi.org/10.1186/s13104-019-
age. Although the evidence base linking sedentary behavior with
4626-0
mental and physical health outcomes requires further development, 8. Skrede T, Steene-Johannessen J, Anderssen SA, Resaland GK,
our findings suggest that the development and evaluation of interven- Ekelund U. The prospective association between objectively
tions to limit age-related increases in specific sedentary behaviors measured sedentary time, moderate-to-vigorous physical activity and
cardiometabolic risk factors in youth: a systematic review and meta-
may be appropriate. Further research into patterns of contemporary
analysis. Obes Rev. 2019;20(1):55-74. https://doi.org/10.1111/obr.
sedentary behavior use and to identify population sub-groups that 12758
may accumulate higher amounts of sedentary behavior with age 9. Biddle SJH, García Bengoechea E, Wiesner G. Sedentary behaviour
would be beneficial for the targeting of behavior change programs. and adiposity in youth: A systematic review of reviews and analysis of
causality. Int J Behav Nutr Phys Act. 2017;14(1):1-21. https://doi.org/
10.1186/s12966-017-0497-8
ACKNOWLEDGMENTS
10. van Ekris E, Altenburg TM, Singh AS, Proper KI, Heymans MW,
With thanks to Dr. Kirsten Corder and Dr. Kathryn Richardson for Chinapaw MJM. Erratum to: An evidence-update on the prospective
their input on search methodology and data analysis. Elli Kontostoli is relationship between childhood sedentary behaviour and biomedical
partially supported by an Academy of Medical Sciences/the British health indicators: a systematic review and meta-analysis: Childhood
sedentary behaviour and future health. Obes Rev. 2017;18(6):
Heart Foundation/the Government Department of Business, Energy
712-714. https://doi.org/10.1111/obr.12526
and Industrial Strategy/the Wellcome Trust Springboard Award 11. Deapartment for Health. UK Chief Medical Officers Physical
[SBF003\1015], held by Dr. Andrew Atkin. Additional funding from Activity Guidelines 2019;(9):1–65. https://www.gov.uk/government/
the University of East Anglia, Faculty of Medicine and Health Sciences publications/physical-activity-guidelines-uk-chief-medical-officers-
report
is gratefully acknowledged.
12. Thomas G, Bennie JA, De Cocker K, et al. A Descriptive Epidemiology
of Screen-Based Devices by Children and Adolescents: a
CONF LICT OF IN TE RE ST Scoping Review of 130 Surveillance Studies Since 2000. Child Indic
The authors declare no conflicts of interest. Res. 2019;13(3):935-950. https://doi.org/10.1007/s12187-019-
09663-1
13. Arundell L, Fletcher E, Salmon J, Veitch J, Hinkley T. A systematic
ORCID review of the prevalence of sedentary behavior during the after-
Elli Kontostoli https://orcid.org/0000-0003-4285-4808 school period among children aged 5-18 years. Int J Behav Nutr Phys
Andy P. Jones https://orcid.org/0000-0002-3130-9313 Act. 2016;13(1):1-9. https://doi.org/10.1186/s12966-016-0419-1
14. Pate RR, Mitchell JA, Byun W, Dowda M. Sedentary behaviour in
Natalie Pearson https://orcid.org/0000-0003-2060-5966
youth. Br J Sports Med. 2011;45(11):906-913. https://doi.org/10.
Louise Foley https://orcid.org/0000-0003-3028-7340 1136/bjsports-2011-090192
Stuart J. H. Biddle https://orcid.org/0000-0002-7663-6895 15. Tremblay MS, Carson V, Chaput J-P, et al. Canadian 24-hour move-
Andrew J. Atkin https://orcid.org/0000-0002-3819-3448 ment guidelines for children and youth: an integration of physical
KONTOSTOLI ET AL. 11

activity, sedentary behaviour, and sleep. Appl Physiol Nutr Metab. 33. Dumith SC, Gigante DP, Domingues MR, Kohl HW. Physical activity
2016;41(6 [Suppl. 3]):S311-S327. https://doi.org/10.1139/apnm- change during adolescence: a systematic review and a pooled analy-
2016-0151 sis. Int J Epidemiol. 2011;40(3):685-698. https://doi.org/10.1093/ije/
16. Dowd KP, Szeklicki R, Minetto MA, et al. A systematic literature dyq272
review of reviews on techniques for physical activity measurement in 34. Chong KH, Parrish AM, Cliff DP, Kemp BJ, Zhang Z, Okely AD.
adults: a DEDIPAC. Int J Behav Nutr Phys Act. 2018;15(1):15. https:// Changes in physical activity, sedentary behaviour and sleep across
doi.org/10.1186/s12966-017-0636-2 the transition from primary to secondary school: a systematic review.
17. Guo W, Key TJ, Reeves GK. Accelerometer compared with question- J Sci Med Sport. 2019;23(5):498-505. https://doi.org/10.1016/j.
naire measures of physical activity in relation to body size and com- jsams.2019.12.002
position: a large cross-sectional analysis of UK Biobank. BMJ Open. 35. Biddle SJH, O'Connell S, Braithwaite RE. Sedentary behaviour inter-
2019;9(1):e024206. https://doi.org/10.1136/bmjopen-2018-024206 ventions in young people: a meta-analysis. Br J Sports Med. 2011;45
18. Pearson N, Haycraft E, Johnston JP, et al. Sedentary behaviour across (11):937-942. https://doi.org/10.1136/bjsports-2011-090205
the primary-secondary school transition: a systematic review. Prev 36. Larouche R, Laurencelle L, Shephard RJ, Trudeau F. Life transitions in
Med. 2017;94:40-47. https://doi.org/10.1016/j.ypmed.2016.11.010 the waning of physical activity from childhood to adult life in the
19. Cooper AR, Goodman A, Page AS, et al. Objectively measured physi- Trois-Rivières study. J Phys Act Heal. 2012;9(4):516-524. https://doi.
cal activity and sedentary time in youth: The International Children's org/10.1123/jpah.9.4.516
Accelerometry Database (ICAD). Int J Behav Nutr Phys Act. 2015;12 37. Reilly JJ. When does it all go wrong? Longitudinal studies of changes
(1):113. https://doi.org/10.1186/s12966-015-0274-5 in moderate-to-vigorous-intensity physical activity across childhood
20. Fairclough SJ, Boddy LM, Hackett AF, Stratton G. Associations and adolescence. J Exerc Sci Fit. 2016;14(1):1-6. https://doi.org/10.
between children's socioeconomic status, weight status, and sex, 1016/j.jesf.2016.05.002
with screen-based sedentary behaviours and sport participation. 38. Biddle SJH, Pearson N, Ross GM, Braithwaite R. Tracking of seden-
Int J Pediatr Obes. 2009;4(4):299-305. https://doi.org/10.3109/ tary behaviours of young people: a systematic review. Prev Med
17477160902811215 (Baltim). 2010;51(5):345-351. https://doi.org/10.1016/j.ypmed.2010.
21. Coombs N, Shelton N, Rowlands A, Stamatakis E. Children's and ado- 07.018
lescents' sedentary behaviour in relation to socioeconomic position. 39. Kann L, McManus T, Harris WA, et al. Youth risk behavior
J Epidemiol Community Health. 2013;67(10):868-874. https://doi.org/ surveillance—United States, 2017. Surveill Summ. 2018;67(8):1-114.
10.1136/jech-2013-202609 https://doi.org/10.15585/mmwr.ss670
22. Malina RM. Adherence to physical activity from childhood to 40. Iannotti RJ, Wang J. Trends in physical activity, sedentary behavior,
adulthood: a perspective from tracking studies. Quest. 2001;53(3): diet, and BMI among US adolescents, 2001–2009. Pediatrics. 2013;
346-355. https://doi.org/10.1080/00336297.2001.10491751 132(4):606-614. https://doi.org/10.1542/peds.2013-1488
23. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for 41. Ryu S, Kim H, Kang M, Pedisic Z, Loprinzi PD. Secular trends in sed-
reporting systematic reviews and meta-analyses of studies that evaluate entary behavior among high school students in the United States,
health care interventions: explanation and elaboration. J Clin Epidemiol. 2003 to 2015. Am J Health Promot. 2019;33(8):1174-1181. https://
2009;62(10):1-34. https://doi.org/10.1016/j.jclinepi.2009.06.006 doi.org/10.1177/0890117119854043
24. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational 42. Barnett TA, Kelly CAS, Young DR, et al. Sedentary behaviors in
studies in epidemiology: a proposal for reporting. JAMA. 2000;283 today's youth: approaches to the prevention and management of
(15):2008-2012. https://doi.org/10.1001/jama.283.15.2008 childhood obesity a scientific statement from the American Heart
25. Jones RA, Hinkley T, Okely AD, Salmon J. Tracking physical activity and Association. Circulation. 2018;138(11):e142-e159. https://doi.org/10.
sedentary behavior in childhood: a systematic review. Am J Prev Med. 1161/CIR.0000000000000591
2013;44(6):651-658. https://doi.org/10.1016/j.amepre.2013.03.001 43. Office of Communications. Children and parents: media use and atti-
26. Tooth L, Ware R, Bain C, Purdie DM, Dobson A. Quality of reporting tudes. Ofcom. 2019;(10):220. http://stakeholders.ofcom.org.uk/
of observational longitudinal research. Am J Epidemiol. 2005;161(3): binaries/research/media-literacy/october-2013/research07Oct2013.
280-288. https://doi.org/10.1093/aje/kwi042 pdf. Accessed February 2, 2020.
27. Tanaka C, Reilly JJ, Huang WY. Longitudinal changes in objectively 44. Auhuber L, Vogel M, Grafe N, Kiess W, Poulain T. Leisure activities of
measured sedentary behaviour and their relationship with adiposity in healthy children and adolescents. Int J Environ Res Public Health.
children and adolescents: Systematic review and evidence appraisal. 2019;16(12). https://doi.org/10.3390/ijerph16122078
Obes Rev. 2014;15(10):791-803. https://doi.org/10.1111/obr.12195 45. Evenson KR, Arredondo EM, Carnethon MR, et al. physical activity
28. Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance and sedentary behavior among US Hispanic/Latino Youth: The SOL
from the median, range, and the size of a sample. BMC Med Res Met- youth study. Med Sci Sports Exerc. 2019;51(5):891-899. https://doi.
hodol. 2005;5(1):13. https://doi.org/10.1186/1471-2288-5-13 org/10.1249/MSS.0000000000001871
29. Higgins JPT, Green S. Cochrane handbook for systematic reviews of
interventions Version 5.1.0 [updated March 2011]. Cochrane Collab
2011. Available from. https://www.handbook.cochrane.org SUPPORTING INF ORMATION
30. Farooq A, Martin A, Janssen X, et al. Longitudinal changes in Additional supporting information may be found online in the
moderate-to-vigorous-intensity physical activity in children and
Supporting Information section at the end of this article.
adolescents: a systematic review and meta-analysis. Obes Rev. 2019;
21(1):1-15. https://doi.org/10.1111/obr.12953
31. Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis
detected by a simple, graphical test. BMJ. 1997;315(7109):629-634. How to cite this article: Kontostoli E, Jones AP, Pearson N,
https://doi.org/10.1136/bmj.315.7109.629 Foley L, Biddle SJH, Atkin AJ. Age-related change in sedentary
32. Steene-Johannessen J, Hansen BH, Dalene KE, et al. Variations in behavior during childhood and adolescence: A systematic
accelerometry measured physical activity and sedentary time across
review and meta-analysis. Obesity Reviews. 2021;1–11.
Europe—harmonized analyses of 47,497 children and adolescents. Int
J Behav Nutr Phys Act. 2020;17(1):38. https://doi.org/10.1186/ https://doi.org/10.1111/obr.13263
s12966-020-00930-x