ARTICLE IN PRESS

THE JOURNAL OF PEDIATRICS • www.jpeds.com ORIGINAL

ARTICLES
Excessive Screen Time and Psychosocial Well-Being: The Mediating Role
of Body Mass Index, Sleep Duration, and Parent-Child Interaction
Jin Zhao, PhD1,2,*, Yunting Zhang, PhD2,*, Fan Jiang, PhD3,4, Patrick Ip, PhD5, Frederick Ka Wing Ho, PhD5,
Yuning Zhang, PhD6, and Hong Huang, PhD1

Objectives To examine the relationship between excessive screen time and psychosocial well-being in pre-
school children, and the potential mediating role of body mass index, sleep duration, and parent-child interaction.
Study design A cross-sectional survey was conducted in Shanghai, China using stratified random sampling design.
A representative sample of 20 324 children aged 3-4 years old from 191 kindergartens participated in this study.
Parents completed the Strengths and Difficulties Questionnaire and reported the child’s time spent on screen ex-
posure, sleep duration, height, weight, and parent-child interactive activities.
Results Preschool children in Shanghai were exposed to 2.8 (95% CI 2.7, 2.9) hours/day of screen time, with
78.6% (95% CI 77.8,79.3) exceeding 1 hour/day and 53% (95% CI 52.0,53.9) exceeding 2 hours/day. Every ad-
ditional hour of screen time was associated with increased risk for poor psychosocial well-being. Body mass index,
sleep duration, and parent-child interaction mediated the effect of excessive screen time on children’s psychoso-
cial well-being, among which parent-child interaction contributed most. Parent-child interaction could explain 28.1%
of the effect on total difficulties and 58.6% on prosocial behavior.
Conclusions Excessive screen time during early childhood exists in Shanghai preschool children. Excessive screen
exposure was associated with poor psychosocial well-being in preschool children via a number of mediators, mostly
by reducing parent-child interaction. (J Pediatr 2018;■■:■■-■■).

creen-viewing among children increases rapidly in the first 3 years of life,1 and gradually stabilizes in preschool as a

S routine behavior persisting into school age and adolescence, and even into adulthood.2-5 International statistics indi-
cated that over 60% of children were spending more than 2 hours daily watching electronic media in 27 countries from
6 continents.6 Approximately 3.6 hours/day of screen time was detected in American preschool children, which occupied a large
majority of their leisure time.7 Brain development during preschool is characterized by its “blossoming” nature8 and is mostly
determined by external stimulation.9 Limited studies have reported negative associations between screen time and emotional
symptoms,10-12 conduct problems,10,12 hyperactivity/inattention,10,13-15 peer problems,10-12 prosocial behaviors problems,10,12,16 ex-
ternalizing behavior,17 aggression,18 and self-control problems.14
The aims of this study were to study time of screen exposure in Shanghai preschool children; examine whether excessive
screen time is associated with psychosocial well-being in preschool children; and explore the potential mediating role of body
mass index (BMI), sleep duration, and parent-child interaction.

Methods
A representative sample of newly enrolled preschoolers in Shanghai kindergar-
tens was selected in September 2016 by stratified random sampling design. In Shang-
From the 1Shanghai Key Laboratory of Children’s
Environmental Health, Xinhua Hospital, Shanghai Jiao
Tong University School of Medicine, Shanghai, China;
2Child Health Advocacy Institute, Shanghai Children’s

Medical Center, Shanghai Jiao Tong University School of

hai, all 3-year-old children are required to enter kindergarten (preschool). Children
are registered in the system of Shanghai Education Committee upon enroll-
ment. A total of 167 597 children enrolled in 2016 according to the database. All
16 districts of Shanghai were set as primary sampling units. In each primary sam-
pling unit, ultimate sampling units (USUs) were defined according to the ranking
and ownership of the kindergartens. The rankings of kindergartens were rated by
Medicine; 3Department of Developmental and Behavioral
Pediatrics, Shanghai Children’s Medical Center, Shanghai
Jiao Tong University School of Medicine; 4Ministry of
Education-Shanghai Key Laboratory of Children’s
Environmental Health, Shanghai, China; 5Department of
Pediatrics and Adolescent Medicine, The University of
Hong Kong, Hong Kong; and 6Institute of Psychiatry,
Psychology, and Neuroscience, King’s College London,
London, United Kingdom

Shanghai Education Committee, reflecting a comprehensive capacity on space, fa- *Contributed equally.
Supported by Chinese National Natural Science
cilities, teaching personnel, and etc. Ownership of kindergartens is either state owned Foundation (81773443, 81602868, 81601162, 81602870);
the fourth round of Three-Year Public Health Action Plan
(2015-2017) (GWIV-36); Shanghai Municipal Education
Commission (D1502); Shanghai Science and Technology
Commission (17411965300, 14441904004); and the
BMI Body mass index Shanghai Municipal Commission of Health and Family
RMB Renminbi Planning (20164Y0095). The authors declare no conflicts
of interest.
SES Socioeconomic status
SDQ Strengths and Difficulties Questionnaire 0022-3476/$ - see front matter. © 2018 Elsevier Inc. All rights
USU Ultimate sampling unit reserved.
https://doi.org10.1016/j.jpeds.2018.06.029

FLA 5.5.0 DTD ■ YMPD10065_proof ■ August 9, 2018

THE JOURNAL OF PEDIATRICS • www.jpeds.com
or private owned. In total, there are 9 combinations of rankings
and ownership for all the kindergartens. In some USUs, there
is only 1 kindergarten all together. Therefore, we randomly se-
lected up to 2 kindergartens in each USUs within all districts
to maintain heterogeneity of all Shanghai kindergartens. In-
ternational kindergartens were excluded as children were mostly
foreigners. Initial sample size is 20 899 of 3- to 4-year-old chil-
dren from 191 kindergartens.
An exclusive login code linking personal information re-
trieved from the Shanghai Kindergarten Registry Database in-
cluding name, birthdate, sex, and attending kindergartens was
generated for each sample child. An online questionnaire along
with a login code was distributed by head teachers to the parents
of each participant. The investigation was open for 2 weeks,
and parents of 20 324 children consented to participate and
completed the survey, a response rate of 97%. Sampling weights
were computed using inverse probability weighting, which rep-
resented the inverse of the combined selection probability in
each sampling stage, including the nonresponse rate.
The study was approved by the Institutional Review Board
of the Shanghai Children’s Medical Center, Shanghai Jiao Tong
University (SCMCIRB-K2016022-01). All parents of chil-
dren who participated in the study gave informed consent at
the beginning of the online survey.
Time spent on video programs, electronic games, and
browsing the web via screen (including television, computer,
cellphone, iPad, etc) on weekdays and weekends in the latest
month was reported by parents. Daily average screen time
was calculated using the formula: ([weekday screen
time×5] + [weekend screen time×2])/7. We capped the
maximum amount of screen time per day at 16 hours for 150
parents (0.74 % of sample size).19 To investigate the relation-
ship between screen time and psychosocial well-being, screen
time was categorized into 5 groups: ≤1 hours/day, 1-2 hours/
day, 2-3 hours/day, 3-4 hours/day, and >4 hours/day.
The Strengths and Difficulties Questionnaire (SDQ) re-
ported by parent was used to assess the psychosocial well-
being status of the child.20 The SDQ is an internationally
recognized tool. There is good support for its validity in
Chinese children 3-17 years old by convergent and discrimi-
nate validity analysis. The reliability has been tested by
internal consistency, and the Cronbach a coefficient range
from 0.30 to 0.83 in subscales. The SDQ contains 25 attri-
butes, each of which was scored on a 3-point scale (0 = not
true, 1 = somewhat true, 2 = certainly true). The 25 items
were divided into 5 scales: emotional symptoms, conduct
problems, hyperactivity/inattention, peer relationship prob-
lems, and prosocial behavior. The first 4 scales (excluding
prosocial behavior) were combined to generate a total diffi-
culties score. The scores of each scale were dichotomized
into healthy (normal category) or at-risk (borderline and
abnormal categories) groups using the recommended cut-
off values in China, which had been previously validated for
children aged 3-16 years.21 The scores of each scale were used
as a continuous variable in the mediation analysis. High
scores represent high risk in all scales except the prosocial
behavior, which is the opposite.
2 Zhao et al
FLA 5.5.0 DTD ■ YMPD10065_proof ■ August 9, 2018
Volume ■■ • ■■ 2018
Parent-Child Interaction
Parent-child interaction was assessed using the Chinese Parent-
Child Interaction Scale, a standardized questionnaire, which
has been well-validated using Rasch analysis.22 It has also been
shown to have good person-separation reliability (0.82). Parents
were asked to indicate the frequency of 12 parent-child inter-
active activities they engaged in during the past month on a
scale of 0-5 (0 = not occurring, 5 = very frequent). The ques-
tionnaire measures the quantity of the most common activi-
ties of learning related, reading, recreation, and interaction with
environment. The total scores added from all activities were
used in the analysis.
Sleep Duration
Sleep onset and wake up time in the past weekdays and week-
ends were collected from a parent reported Children Sleep
Habits Questionnaire23 and sleep duration in weekdays and
weekends was calculated separately. Daily average sleep dura-
tion was calculated using the formula: ([weekday sleep dura-
tion×5] + [weekend sleep duration×2])/7.
BMI
Height and weight were reported by parents. By dividing the
weight in kilograms by the square of the height in meters, BMI
was calculated. BMI-for-age z scores were calculated using the
child growth standards of World Health Organization24 and
were used in the analysis.
Demographics
Age and sex of all participants were retrieved from the Shang-
hai Kindergarten Registry Database of the Shanghai Educa-
tion Committee and further confirmed by parents at the
beginning of the investigation.
Socioeconomic Status
Maternal education and household income were self-reported.
Maternal education was categorized into 4 groups: high school
and below, junior college, undergraduate, master, and above.
Household income per year was divided into 4 categories:
<100 000 Renminbi (RMB), 100 000-150 000 RMB, 150 000-
300 000 RMB, and ≥300 000 RMB.
Family Structure
The presence of other children in the family, the primary care-
giver of children (mother, father, grandparents, or others), and
the marital status (divorced or not) were reported by parents.
Statistical Analyses
Descriptive analysis was conducted to calculate the time that
children spent on screen exposure. Linear regression was run
to examine relationships between screen time and demo-
graphic, socioeconomic status (SES), and family structure. Lo-
gistic regression was performed to explore the association
between screen exposure and psychosocial well-being. To in-
vestigate the mechanism underneath the relationship, we used
the multiple mediator model. This model involves investigat-
ing the total indirect effect; and determining the relative
■■ 2018 ORIGINAL ARTICLES

Table I. Descriptive characteristics of the Shanghai population* (n = 20 234)

Screen time(h/d)
Coefficient (95% CI)
Mean (95% CI)/
Characteristics proportion (95% CI) Mean (95% CI) Unadjusted Adjusted†
Age 44.3 (44.2, 44.4) 2.8 (2.7, 2.9) .01‡ (.00, .02) .01§ (.00, .02)
Sex
Boys 52.7 (51.7, 53.6) 2.9 (2.9, 3.0) Ref. Ref.
Girls 47.3 (46.4, 48.3) 2.7 (2.7, 2.8) −0.13§ (−020, −.07) −0.13§ (−0.19, −.07)
Maternal education
High school and below 26.2 (25.2, 27.1) 4.0 (3.9, 4.1) Ref. Ref.
Junior college 23.4 (22.6, 24.2) 2.9 (2.8, 3.0) −0.96§ (−1.05, −0.87) −0.78§ (−0.88, −0.68)
Undergraduate 40.9 (39.9, 41.8) 2.3 (2.2, 2.4) −1.57§ (−1.65, −1.49) −1.29§ (−1.39, −1.20)
Master and above 9.5 (9.0, 10.1) 1.8 (1.7, 1.9) −2.07§ (−2.18, −1.96) −1.69§(−1.82, −1.56)
Household income¶
<100 000 26.9 (25.9, 27.8) 3.7 (3.6, 3.8) Ref. Ref.
100 000~150 000 17.9 (17.2, 18.7) 3.0 (2.9, 3.1) −0.54§ (−0.64, −0.44) −0.23§ (−0.34, −0.13)
150 000~300 000 34.1 (33.2, 35.1) 2.6 (2.5, 2.6) −1.01§ (−1.09, −0.92) −0.44§ (0.53, −0.35)
≥ 300 000 21.1 (20.3, 21.8) 2.1 (2.1, 2.2) −1.42§ (−1.51, −1.33) −0.72§ (−0.83, −0.63)
Siblings
One or more 28.9 (28.0, 29.8) 3.2 (3.1, 3.3) Ref. Ref.
None 71.1 (70.2, 72.0) 2.7 (2.6, 2.7) −0.35§ (−0.42, −0.28) −0.17§ (−0.24, −.09)
Parents' divorce
No 96.9 (96.5, 97.2) 2.8 (2.8, 2.9) Ref. Ref.
Yes 3.1 (27.9, 34.7) 3.3 (2.9, 3.6) 0.56§ (0.38, 0.74) 0.30§ (0.12, 0.48)
Primary caregiver
Mother 57.6 (56.7, 58.5) 2.9 (2.8, 2.9) Ref. Ref.
Father 6.0 (5.5, 6.5) 3.3 (3.1, 3.5) 0.37§ (0.23, 0.52) 0.35§ (0.21, 0.49)
Grandparents 35.5 (34.6, 36.4) 2.7 (2.7, 2.8) .01 (−.06, 0.73) 0.23§ (0.16, 0.30)
Others 0.9 (0.7, 1.1) 2.6 (2.2, 2.9) −0.11 (−0.44, 0.21) .04 (−0.30, 0.37)

Ref, reference.
Boldface indicates statistical significance.
*Shanghai population estimates were acquired by using sample weights.
†Adjusted by other demographic variables in Table I.
‡P < .05
§P < .001.
¶Household income was calculated by years and in RMB.

magnitude of each specific indirect effect. The coefficients and was observed in older children, those with lower SES family,
SE of each path was estimated by the Sobel test.25,26 Although 1 or more siblings, parental divorce, or father acting as the
it is reasonable to expect the sampling distribution to be ap- primary caregiver.
proximately normal in large sample size, bootstrapping was Table II reports the OR of the association between screen
used to provide the most powerful and reasonable method of time and children being classified as at risk for psychosocial
obtaining CI for specific indirect effect. Percentile and bias- problems, adjusted by age, sex, number of siblings, family SES
corrected CI were both presented by 5000 bootstrap re- status, primary caregiver, maternal education level, and house-
sampling according to the recommendation.27 Descriptive and hold income. It indicates that every additional hour of screen
regression analyses were conducted using Stata v 14.2 time was associated with increased risk for poor psychoso-
(StataCorp LP, College Station, Texas). The multiple media- cial well-being. Compared with less than 1 hour/day, the ORs
tor model was operated in Mplus 8 (Muthen & Muthen, Los
Angeles, California). Sampling weights were used in all analy-
ses to ensure representativeness. Table II. Association between screen time and risk for
psychosocial problems (n = 18 562)
Results Prosocial
Total behavior
difficulties problems
The mean age of children in our study was 44.3 (95% CI 44.2,
OR 95% CI OR 95% CI
44.4) months adjusted by sampling weight. On average, pre-
Screen time 0-1 h Ref. Ref. Ref. Ref.
school children in Shanghai were exposed to 2.8 (95% CI 2.7, 1-2 h 1.2* (1.1, 1.3) 1.1 (1.0, 1.2)
2.9) hours/day of screen time, with 78.6% (95% CI 77.8, 79.3) 2-3 h 1.5* (1.3, 1.6) 1.2* (1.1, 1.3)
exceeding 1 hour/day and 53.0% (95% CI 52.0, 53.9) exceed- 3-4 h 1.6* (1.5, 1.8) 1.3* (1.2, 1.5)
>4 h 2.1* (1.9, 2.3) 1.4* (1.3, 1.6)
ing 2 hours/day. Table I presents the weighted proportion of
demographic factors and the differences of screen time in each Adjusted by age, sex, only-child, family status, primary caregiver, maternal education level, and
household income.
characteristic. Boys spent more hours on screen than girls (2.9 Boldface indicates statistical significance.
vs 2.7 hours/day, P < .001). Significant increased screen time *P < .001.

Excessive Screen Time and Psychosocial Well-Being: The Mediating Role of Body Mass Index, 3
Sleep Duration, and Parent-Child Interaction
FLA 5.5.0 DTD ■ YMPD10065_proof ■ August 9, 2018
THE JOURNAL OF PEDIATRICS • www.jpeds.com Volume ■■

Figure 1. Multiple mediator model between screen time and total difficulties through BMI, sleep duration, and parent-child interaction.

for total difficulties/prosocial behavior problems were 1.2 (95%
CI 1.1,1.3)/1.1 (95% CI 1.0,1.2) for 1-2 hours/day of screen
time, 1.5 (95% CI 1.3,1.6)/1.2(95% CI 1.1,1.3) for 2-3 hours/
day, 1.6 (95% CI 1.5,1.8)/1.3(95% CI 1.2,1.5) for 3-4 hours/
day, and 2.1 (95% CI 1.9,2.3)/1.4(95% CI 1.3,1.6) for more
than 4 hours/day.
As illustrated in Figures 1 and 2, screen time was related to
all 3 mediators (BMI, sleep duration, and parent-child inter-
action), which in turn was significantly related to total diffi-
culties and prosocial behavior.
The direct and total indirect effect of screen time on total
difficulties was separately 0.294 (95% CI 0.261,0.328; P < .001)
and 0.123 (95% CI 0.112,0.135; P < .001), which were added
up to 0.417 as the total effect. The effects of screen time on
total difficulties were mediated through parent-child interac-
tion (28.1%), BMI (1.2%), and sleep duration (0.5%).
The direct and total indirect effect of screen time on prosocial
behavior was separately -.033 (95% CI -.047, -.020; P < .001)
and -.054 (95% CI -.059, -.049; P < .001), which added up to
-.087 as the total effect. The effects of screen time on prosocial
behavior were mediated through parent-child interaction
(58.6%), BMI (2.3%), and sleep duration (1.1%).
Finally, there were no significant contrasts of the effect sizes
between the mediator of BMI and sleep duration. However,
parent-child interaction had a significantly greater indirect effect
than BMI and sleep duration as evidenced by CIs that did not
contain zero (Tables III and IV; available at www.jpeds.com).
Discussion
The guideline from the American Academy of Pediatrics speci-
fied preschoolers should have at most 1 hour of screen time

Figure 2. Multiple mediator model between screen time and prosocial behavior problems through BMI, sleep duration, and parent-
child interaction.

4 Zhao et al

FLA 5.5.0 DTD ■ YMPD10065_proof ■ August 9, 2018

■■ 2018
per day.28A large majority of preschoolers in Shanghai have ex-
ceeded this limit. The prevalence of screen exposure is similar
to Western countries, where 74%-85% of preschool children
spend more than 1 hour/day on screen,29,30 and 53%-75% spend
more than 2 hours/day.31-34 The study found that older age, boys,
from lower SES families, with parental divorce, with siblings,
and/or the father as primary caregiver were risk factors for ex-
cessive screen time, which is consistent with studies in Western
countries.35-40
In our study, we found a negative association between screen
time and psychosocial well-being in preschool children, which
is consistent with previous studies.10-16,18 Screen time was usually
dichotomized by more or less 2 hours/day in previous studies.
However, in our study, we categorize screen time by incre-
ments of 1 hour/day. The result of logistic regression analysis
indicated that preschool children face increasing risk for psy-
chosocial problems with every additional hour of screen time.
It provides evidence for shortening the limitation of screen time
from 2 hours/day to 1 hour/day for preschool children as rec-
ommended in the updated AAP guideline.
In our analysis, BMI, sleep duration, and parent-child in-
teraction appear to be significant mediators accounting for
the effect of screen time on psychosocial well-being, among
which parent-child interaction contributed most. According
to the mediation model, parent-child interaction can explain
the effect of screen time on total difficulties with 28.1% and
on prosocial behavior problem with 58.6%. The contribu-
tion of each mediator was determined by its effect size of as-
sociation with both predictor and outcome. Excessive exposure
to media may have the strongest effect on parent-child inter-
action, which in turn is a key risk factor for child psychoso-
cial problems. The social withdrawal hypothesis has been raised
by several studies,10,11,41 but has not yet been tested in large
population research. In our study, we found that the detri-
mental effect of excessive screen time might be partly coun-
teracted by improving parent-child interaction. Many parents
consider media as an “electronic babysitter” and omit the im-
portance of parent-child interaction in the development of
children. Especially in the families with more than 1 child,
parents might prefer using media to calm them or separate
squabbling brothers and sisters.42 We also found that the pres-
ence of siblings significantly contributed to the time pre-
school children spent on screen in this study, which is consistent
with other studies.36,37,39,43 However, parent should be aware
that the negative impact brought by excessive screen time will
actually increase the burden for parents in later life. We also
tested the mediation role of sleep duration, and the effect size
is much smaller than that of parent-child interaction. The
timing of exposure for preschool children may be different
from the adverse effect of screen time on sleep duration in
school-aged children. Screen exposure is also related to child
obesity, yet in our analysis obesity has little impact on psy-
chosocial problems. Obesity has been shown to have an impact
on psychological problems. However, in traditional Chinese
culture, overweight young children, in contrast to adolescents/
adults, may be considered cute rather than mocked by others
in this age.44,45
Excessive Screen Time and Psychosocial Well-Being: The Mediating Role of Body Mass Index,
Sleep Duration, and Parent-Child Interaction
FLA 5.5.0 DTD ■ YMPD10065_proof ■ August 9, 2018
ORIGINAL ARTICLES
Several limitations to the present study are worthy of at-
tention. First, our cross-sectional study could only investi-
gate relationships rather than causality, as reverse causation
could occur. For example, the relationship between screen time
and parent-child interaction might be explained by parents with
fewer parenting skills allowed children more screen-based be-
havior. The direction of this relationship should be tested in
further longitudinal studies. Second, although most measure-
ment tools used in this survey were standardized question-
naires with well reliability and validity, direct observation would
be an alternative means to collect the data. Third, the specific
content of the screen viewing has not been analyzed. Differ-
ent components of screen time might have different effects on
psychosocial well-being. Further evaluation of specific com-
ponents of screen time would help elucidate whether there are
specific aspects associated with psychosocial well-being.
In an era in which electronic devices have become almost
universal, it is important to make parents aware of the poten-
tial hazard of excessive screen exposure, and guide parents to
improve the quantity and quality of nurturing through parent-
child interaction. ■
We thank the participating kindergartens and Shanghai Municipal Edu-
cation Committee for supporting us collect data. We also thank Profes-
sor Myron Belfer and Professor Mark Simms for revision of this
manuscript.
Submitted for publication Feb 15, 2018; last revision received May 10, 2018;
accepted Jun 4, 2018
Reprint requests: Hong Huang, PhD, Shanghai Key Laboratory of Children’s
Environmental Health, Xinhua Hospital, Shanghai Jiao Tong University School
of Medicine, 1665 Kongjiang Rd, Shanghai 200092, China. E-mail:
huanghong@smhb.gov.cn
References
1. Certain LK, Kahn RS. Prevalence, correlates, and trajectory of television
viewing among infants and toddlers. Pediatrics 2002;109:634-42.
2. Gebremariam MK, Totland TH, Andersen LF, Bergh IH, Bjelland M,
Grydeland M, et al. Stability and change in screen-based sedentary
behaviours and associated factors among Norwegian children in the tran-
sition between childhood and adolescence. BMC Public Health 2012;12:1-
9.
3. Janz KF, Burns TL, Levy SM. Tracking of activity and sedentary behav-
iors in childhood: the Iowa Bone Development Study. Am J Prev Med
2005;29:171-8.
4. Christakis DA, Zimmerman FJ. Early television viewing is associated with
protesting turning off the television at age 6. Medgenmed 2006;8:63.
5. Smith L, Gardner B, Hamer M. Childhood correlates of adult TV viewing
time: a 32-year follow-up of the 1970 British Cohort Study. J Epidemiol
Commun Health 2015;69:309-13.
6. Tremblay MS, Barnes JD, Gonzalez SA, Katzmarzyk PT, Onywera VO, Reilly
JJ, et al. Global Matrix 2.0: report card grades on the physical activity of
children and youth comparing 38 countries. J Phys Act Health 2016;13(11
Suppl 2):S343-66.
7. Tandon PS, Zhou C, Lozano P, Christakis DA. Preschoolers’ total daily
screen time at home and by type of child care. J Pediatr 2011;158:297-
300.
8. Brown TT, Jernigan TL. Brain development during the preschool years.
Neuropsychol Rev 2012;22:313-33.
9. Kieling C, Baker-Henningham H, Belfer M, Conti G, Ertem I, Omigbodun
O, et al. Child and adolescent mental health worldwide: evidence for action.
Lancet 2011;378:1515-25.
5
THE JOURNAL OF PEDIATRICS • www.jpeds.com
10. Wu X, Tao S, Rutayisire E, Chen Y, Huang K, Tao F. The relationship
between screen time, nighttime sleep duration, and behavioural prob-
lems in preschool children in China. Eur Child Adolesc Psychiatry
2016;26:541-8.
11. Hinkley T, Verbestel V, Ahrens W, Lissner L, Molnar D, Moreno LA, et al.
Early childhood electronic media use as a predictor of poorer well-
being: a prospective cohort study. JAMA Pediatr 2014;168:485-92.
12. Shiue I. Duration of daily TV/screen watching with cardiovascular, re-
spiratory, mental and psychiatric health: Scottish Health Survey, 2012-
2013. Int J Cardiol 2015;186:241.
13. Christakis DA, Zimmerman FJ, Digiuseppe DL, McCarty CA. Early tele-
vision exposure and subsequent attentional problems in children. Pedi-
atrics 2004;113:917-8.
14. Mistry KB, Minkovitz CS, Strobino DM, Borzekowski DL. Children’s tele-
vision exposure and behavioral and social outcomes at 5.5 years: does
timing of exposure matter? Pediatrics 2007;120:762-9.
15. Miller CJ, Marks DJ, Miller SR, Berwid OG, Kera EC, Santra A, et al. Brief
report: television viewing and risk for attention problems in preschool
children. J Pediatr Psychol 2007;32:448-52.
16. Christakis DA, Zimmerman FJ. Violent television viewing during pre-
school is associated with antisocial behavior during school age. Pediat-
rics 2007;120:993-9.
17. Tomopoulos S, Dreyer BP, Valdez P, Flynn V, Foley G, Berkule SB, et al.
Media content and externalizing behaviors in Latino toddlers. Ambul
Pediatr 2007;7:232-8.
18. Manganello JA, Taylor CA. Television exposure as a risk factor for ag-
gressive behavior among 3-year-old children. Arch Pediatr Adolesc Med
2009;163:1037-45.
19. Radesky JS, Silverstein M, Zuckerman B, Christakis DA. Infant self-
regulation and early childhood media exposure. Pediatrics 2014;133:e1172-
8.
20. Goodman A, Goodman R. Strengths and difficulties questionnaire as a
dimensional measure of child mental health. J Am Acad Child Adolesc
Psychiatry 2009;48:400-3.
21. Du Y, Kou J, Coghill D. The validity, reliability and normative scores of
the parent, teacher and self report versions of the Strengths and Diffi-
culties Questionnaire in China. Child Adolesc Psychiatr Ment Health
2008;2:8.
22. Ip P, Tso W, Rao N, Ho FKW, Chan KL, Fu KW, et al. Rasch validation
of the Chinese parent-child interaction scale (CPCIS). World J Pediatr
2018:1-9.
23. Li S, Jin X, Shen X, Wu S, Jiang F, Chou Y. Development and psycho-
metric properties of the Chinese version of Children’s Sleep Habits Ques-
tionnaire. Chin J Pediatr 2007;45:176-80.
24. Espejo MR. WHO child growth standards: methods and development. J
R Stat Soc 2010;170:512.
25. Sobel ME. Asymptotic confidence intervals for indirect effects in struc-
tural equation models. Sociol Methodol 1982;13:290-312.
26. Sobel ME. Some new results on indirect effects and their standard errors
in covariance structure models. Sociol Methodol 1986;16:159-86.
27. Preacher KJ, Hayes AF. Asymptotic and resampling strategies for assess-
ing and comparing indirect effects in multiple mediator models. Behav
Res Methods 2004;40:879-91.
28. Council On C, Media. Media and young minds. Pediatrics
2016;138:e20162591.
6 Zhao et al
FLA 5.5.0 DTD ■ YMPD10065_proof ■ August 9, 2018
Volume ■■
29. Schranz NK, Olds T, Boyd R, Evans J, Gomersall SR, Hardy L, et al. Results
from Australia’s 2016 report card on physical activity for children and
youth. J Phys Act Health. 2016;13(11 Suppl 2):S87-94.
30. Barnes JD, Cameron C, Carson V, Chaput JP, Faulkner GE, Janson K, et al.
Results from Canada’s 2016 ParticipACTION report card on physical ac-
tivity for children and youth. J Phys Act Health. 2016;13(11 Suppl 2):S110-
6.
31. Tammelin TH, Aira A, Hakamaki M, Husu P, Kallio J, Kokko S, et al. Results
from Finland’s 2016 report card on physical activity for children and youth.
J Phys Act Health. 2016;13(11 Suppl 2):S157-64.
32. Roman-Vinas B, Marin J, Sanchez-Lopez M, Aznar S, Leis R, Aparicio-
Ugarriza R, et al. Results from Spain’s 2016 report card on physical ac-
tivity for children and youth. J Phys Act Health. 2016;13(11 Suppl 2):S279-
83.
33. Katzmarzyk PT, Denstel KD, Beals K, Bolling C, Wright C, Crouter SE,
et al. Results from the United States of America’s 2016 report card on physi-
cal activity for children and youth. J Phys Act Health 2016;13(11 Suppl
2):S307-13.
34. Tyler R, Mannello M, Mattingley R, Roberts C, Sage R, Taylor SR, et al.
Results from Wales’ 2016 report card on physical activity for children and
youth: is wales turning the tide on children’s inactivity? J Phys Act Health.
2016;13(11 Suppl 2):S330-6.
35. Carson V, Spence JC, Cutumisu N, Cargill L. Association between
neighborhood socioeconomic status and screen time among pre-school
children: a cross-sectional study. BMC Public Health 2010;10:1539-
58.
36. Lindsay AC, Greaney ML, Wallington SF, Mesa T, Salas CF. A review of
early influences on physical activity and sedentary behaviors of preschool-
age children in high-income countries. J Spec Pediatr Nurs 2017;22.
37. Anand S, Krosnick JA. Demographic predictors of media use among
infants, toddlers, and preschoolers. Am Behav Sci 2005;48:539-61.
38. Matarma T, Koski P, Löyttyniemi E, Lagström H. The factors associated
with toddlers’ screen time change in the STEPS Study: a two-year follow-
up. Prev Med 2015;84:27-33.
39. Duch H, Fisher EM, Ensari I, Harrington A. Screen time use in children
under 3 years old: a systematic review of correlates. Int J Behav Nutr Phys
Act 2013;10:1-10.
40. Hoyos Cillero I, Jago R. Systematic review of correlates of screen-
viewing among young children. Prev Med 2010;51:3-10.
41. Pagani LS, Fitzpatrick C, Barnett TA, Dubow E. Prospective associations
between early childhood television exposure and academic, psychoso-
cial, and physical well-being by middle childhood. JAMA Pediatr
2010;164:425.
42. Rideout V, Hamel E, Foundation KF. The media family: Electronic media
in the lives of infants, toddlers, preschoolers and their parents. 2006.
43. Carson V, Janssen I. Associations between factors within the home setting
and screen time among children aged 0-5 years: a cross-sectional study.
BMC Public Health 2012;12:539.
44. McElroy SL, Kotwal R, Malhotra S, Nelson EB, Keck PE, Nemeroff CB.
Are mood disorders and obesity related? A review for the mental health
professional. J Clin Psychiatry 2004;65:634.
45. Scott KM, Bruffaerts R, Simon GE, Alonso J, Angermeyer M, de Girolamo
G, et al. Obesity and mental disorders in the general population: results
from the world mental health surveys. Int J Obes (Lond) 2008;32:192-
200.
■■ 2018 ORIGINAL ARTICLES

Table III. Mediating effect of screen time on total difficulties through BMI, sleep duration, and parent-child interac-
tion (n = 20 075)
Bootstrapping
Product of Percentile
coefficients 95% CI BC 95% CI
Point
estimate SE Z Lower Upper Lower Upper
Indirect effects
BMI .005 * .001 3.056 .002 .008 .002 .008
Sleep duration .002† .001 2.127 0 .003 0 .004
Parent-child interaction 0.117‡ .005 21.437 0.107 0.128 0.107 0.128
Total 0.123‡ .006 21.69 0.112 0.135 0.112 0.135
Contrasts
BMI vs sleep duration .003 .002 1.637 −.001 .006 −.001 .006
BMI vs parent-child Interaction −0.113‡ .006 −19.83 −0.124 −0.102 −0.124 −0.102
Sleep duration vs Parent-child interaction −0.115‡ .006 −20.882 −0.126 −0.105 −0.126 −0.105

BC, bias corrected.

5000 bootstrap samples.
*P < .01
†P < .05
‡P < .001.

Table IV. Mediating effect of screen time on prosocial behavior through BMI, sleep duration, and parent-child inter-
action (n = 20 075)
Bootstrapping
Product of Percentile
coefficients 95% CI BC 95% CI
Point
estimate SE Z Lower Upper Lower Upper
Indirect Effects
BMI −.002*† .001 −2.631 −.003 0 −.003 0
Sleep duration −.001† 0 −2.115 −.002 0 −.002 0
Parent-child interaction −.051‡ .002 −21.418 −.056 −.047 −.056 −.047
Total −.054‡ .003 −21.338 −.059 −.049 −.059 −.049
Contrasts
BMI vs sleep duration −.001 .001 −1.264 −.002 0 −.002 0
BMI vs parent-child integration .05‡ .002 20.076 .045 .055 .045 .055
Sleep duration vs parent-child Interaction .051‡ .002 20.961 .046 .056 .046 .056

5000 bootstrap samples.

*P < .01
†P < .05
‡P < .001.

Excessive Screen Time and Psychosocial Well-Being: The Mediating Role of Body Mass Index, 6.e1
Sleep Duration, and Parent-Child Interaction
FLA 5.5.0 DTD ■ YMPD10065_proof ■ August 9, 2018