You are on page 1of 9

Psychological Medicine (2016), 46, 3329–3337.

© Cambridge University Press 2016 OR I G I N A L A R T I C L E


doi:10.1017/S0033291716001689

Prospective associations between televiewing at


toddlerhood and later self-reported social
impairment at middle school in a Canadian
longitudinal cohort born in 1997/1998

L. S. Pagani1*, F. Lévesque-Seck1 and C. Fitzpatrick2,3


1
School of Psycho-Education and Sainte-Justine’s Hospital Research Center (Brain Diseases Axis), Université de Montréal, Montréal, Québec,
Canada
2
Psychology Department, Université Ste-Anne, Church Point, Nova Scotia, Canada
3
Exercise Science Department, Concordia University, Montréal, Québec, Canada

Background. Using a large Canadian population-based sample, this study aimed to verify whether televiewing in
toddlerhood is prospectively associated with self-reported social impairment in middle school.

Method. Participants are from a prospective–longitudinal birth cohort of 991 girls and 1006 boys from the Quebec
Longitudinal Study of Child Development. Child self-reported ratings of relational difficulties at age 13 years were
linearly regressed on parent-reported televiewing at age 2 years while adjusting for potential confounders.

Results. Every additional 1 h of early childhood television exposure corresponded to an 11% S.D. unit increase in
self-reported peer victimization [unstandardized β = 0.03, 95% confidence interval (CI) 0.02–0.04], a 10% S.D. unit increase
in self-reported social isolation (unstandardized β = 0.04, 95% CI 0.03–0.05), a 9% S.D. unit increase in self-reported
proactive aggression (unstandardized β = 0.02, 95% CI 0.01–0.03) and a 6% S.D. unit increase in self-reported antisocial
behavior (unstandardized β = 0.01, 95% CI 0.01–0.01) at age 13 years. These results are above and beyond pre-existing
individual and family factors.

Conclusions. Televiewing in toddlerhood was prospectively associated with experiencing victimization and social
withdrawal from fellow students and engaging in antisocial behavior and proactive aggression toward fellow students
at age 13 years. Adolescents who experience relational difficulties are at risk of long-term health problems (like depres-
sion and cardiometabolic disease) and socio-economic problems (like underachievement and unemployment). These
relationships, observed more than a decade later, and independent of key potential confounders, suggest a need for
better parental awareness of how young children invest their limited waking hours.

Received 24 November 2015; Revised 16 June 2016; Accepted 21 June 2016; First published online 13 September 2016

Key words: Aggression, antisocial behavior, child televiewing, social impairment, victimization.

Introduction appropriate. Almost half of children exceed those limits,


especially if they attend child care facilities (Tandon et al.
Television viewing is a common pastime in very young
2012). Many parents seem unaware or unconcerned
children (Cespedes et al. 2014). In toddlerhood, its role
about these guidelines (Funk et al. 2009; Cespedes et al.
is more than entertainment as some parents depend
2014). What is more, adults do not always actively
on children’s screen time to keep them safe and busy
supervise the television remote, thus not being able to
while they engage in adult activities (Funk et al.
ensure optimal home viewing content (Funk et al.
2009). Current guidelines recommend that, beyond
2009; Fitzpatrick et al. 2012). As many as a third of chil-
age 2 years, children should minimize their teleview-
dren eat breakfast or supper in front of the television
ing time to no more than 2 h per day (American
and many children have a screen device in their bed-
Academy of Pediatrics, 2013). This is assuming that
room (Sisson et al. 2014). Both characteristics diminish
the content is of quality and developmentally
social exchanges and child monitoring in the home en-
vironment, which is daunting given that family inter-
action remains the primary vehicle of socialization in
* Address for correspondence: L. S. Pagani, Ph.D., École de
psychoéducation, Université de Montréal, C.P. 6128, succursale
childhood (Richert et al. 2011; Gingold et al. 2014).
Centre-ville, Montréal, Québec, H3C 3J7, Canada. From birth through the preschool years, exponential
(Email: Linda.s.pagani@umontreal.ca) brain maturation is characterized by increasingly

Downloaded from https:/www.cambridge.org/core. University of Arizona, on 07 May 2017 at 23:09:54, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
. https://doi.org/10.1017/S0033291716001689
3330 L. S. Pagani et al.

effective cognitive problem-solving, communication is generally associated with peer rejection and victimiza-
and interpersonal skills (Kuhl, 2007; Meltzoff et al. tion (Rubin & Barstead, 2014). It also predicts deviant
2009). During this developmental period, children peer affiliation in adolescence (Rudolph et al. 2014).
are particularly sensitive to environmental input Through social learning processes, victims also risk be-
through live social interaction and active play with coming reactively antisocial themselves (Arseneault
persons and objects across different social contexts et al. 2010). Although not based on a birth cohort data,
(Christakis, 2009; Richert et al. 2011). Rapid and exten- one previous study on American children from the
sive brain growth in the first several years of life also National Longitudinal Survey of Youth suggests that
implies considerable vulnerability in less stimulating televiewing at age 4 years predicted parent-reported
or neurotoxic home environments (Cillero & Jago, bullying between the ages of 6 and 11 years
2010; Shonkoff, 2011). (Zimmerman et al. 2005).
Social skills, such as sharing, valuing and being To greater extremes, antisocial behavior violates
respected by others, are rooted in early development major age-appropriate societal norms and the basic
and eventually play a key role in personal and eco- rights of others and is linked with social disruption in
nomic success (Heckman, 2006). Because there are the school, neighborhood and home (Byrd et al. 2014).
only so many waking hours in a day, more time It includes behaviors such as vandalism, risk-taking,
spent televiewing typical content might leave less stealing, lying and passive aggression. Although any
time for creative play, interactive endeavors and foun- such behaviors can occur from time to time in typical
dational experiences that hone basic socio-cognitive elementary school-aged children, their persistence
skills (Pagani et al. 2010). Televiewing does not require across development forecasts unhealthy lifestyle, men-
much cognitive effort, making it a more passive rather tal health, judiciary and socio-economic risks (Moffitt,
than active cognitive activity in daily preschool life 1993). Antisocial behavior is often accompanied by pro-
(Christakis, 2009). Consequently, it might serve as a active aggression, which refers to unprovoked, premedi-
means of distraction from endeavors that require tated and apathetic harmful behavior in order to obtain
more effortful attention (Barkley, 2012), thus interfer- or coerce a desired object from or outcome on others
ing with the development of the brain’s executive sys- (Hubbard et al. 2010). Longitudinally observing a
tem, which is important for emotional regulation and, Canadian birth cohort, Fitzpatrick et al. (2012) found
ultimately, social interaction (Nathanson et al. 2014). that preschool televiewing habits are also associated
Although it does sometimes help receptive vocabu- with antisocial behavior in later childhood. In children
lary in toddlers (Richert et al. 2011), excessive teleview- from the 1972–1973 Dunedin birth cohort (New
ing could be detrimental to learning skills associated Zealand), excessive televiewing from the ages of 5 to
with social competence and management of interper- 15 years is associated with increased chances of a crim-
sonal conflict (Mistry et al. 2007; Fitzpatrick et al. inal conviction, a diagnosis of antisocial personality dis-
2012). This could place children at later risk of experi- order, and more aggressive personality traits in
encing relational difficulties and, ultimately, poor men- emerging adulthood compared with those who viewed
tal health (Arseneault et al. 2010). In one observational less television in that age range (Robertson et al. 2013).
birth cohort study, Canadian toddlers who watched How youth relate to each other is of major clinical con-
more television were later rated by kindergarten cern in adolescence, as it is more resistant to treatment
teachers as more likely to be victimized than their and also charts a psychopathological developmental
counterparts who watched less (Pagani et al. 2013). course in adulthood (Moffitt, 1993; Arseneault et al.
Furthermore, higher amounts of televiewing at age 2010; Byrd et al. 2014).
2 years by the same children predicted an increased Very little is known about whether childhood tele-
risk of being victimized by classmates at age 10 years, viewing habits during early exuberant brain develop-
as reported by fourth-grade teachers (Pagani et al. ment in areas that regulate emotional regulation and
2010). In both studies, victimization at school was intelligence pose any long-term risks for relational im-
defined as experiencing repeated negative interactions pairment. Moreover, the prospective–longitudinal
with one (or more) classmate(s) who intentionally studies available have been either on externalizing or
aimed to inflict harm, alienate or cause discomfort to internalizing behavioral pathologies, but not both. An
a fellow classmate. Whether direct or indirect, such pro- inclusive design with consistent control over potential
active relational aggression is associated with a number confounders would be clearly advantageous for estab-
of short- and long-term mental health and academic lishing associations, if any, with subsequent relational
consequences for their targets as they grow into adoles- impairment. Thus, in the present study, we use a
cence and adulthood, including an increased risk of so- population-based birth cohort to verify whether tele-
cial withdrawal and even depression (Arseneault et al. viewing at age 2 years is prospectively associated
2010; Rudolph et al. 2014). Solitude, preferred or not, with self-reported difficulties in social relations at age

Downloaded from https:/www.cambridge.org/core. University of Arizona, on 07 May 2017 at 23:09:54, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
. https://doi.org/10.1017/S0033291716001689
Early television and social difficulties 3331

13 years, which corresponds to the end of the first year insulted by peers; actively excluded from being a
of middle school. Internalizing and externalizing part of a group; told mean words; laughed at; robbed;
behavioral outcomes include victimization and social and pushed, shoved, or hassled; (2) social isolation (α =
isolation, and proactive aggression and antisocial be- 0.72): tends to work on his own; low interest in activities
havior, respectively. It is expected that children with with others; and prefers playing alone; (3) proactive ag-
higher levels of televiewing during toddlerhood will gression (α = 0.74): threatened to hit others; encouraged
be at greater risk of social impairment in adolescence. others to pick on someone; sought to dominate others;
and used fear to obtain something; and (4) antisocial be-
havior (α = 0.75): broken his own belongings; broken
Method belongings of others; stolen from home; stolen outside
Participants of home; lied or cheated; committed vandalism; and
tried to exclude members of a group when angry at
The participants in this institutional review board (IRB)- them. Item responses ranged on a Likert scale ranging
approved study were from the Quebec Longitudinal from 1 (never) to 3 (often). Factor scores were then
Study of Child Development (www.jesuisjeserai.stat. rescaled from 0 to 10, with higher scores corresponding
gouv.qc.ca/etude_an.htm) which originates from a ran- to a higher level the factor of interest.
domly selected stratified sample of 2837 newborns
born in 1997 and 1998 in Quebec, Canada. From this Pre-existing and concurrent control variables
birth registry selection: 93 were deemed ineligible; 172
were untraceable due to incorrect coordinates; 14 were These include child and family factors which could
untraceable; and 438 refused participation. The baseline confound the relationship between early televiewing
sample, representing 82% of the eligible target popula- and later social impairment (Pagani et al. 2010, 2013;
tion, comprises 2120 infants followed up annually Fitzpatrick et al. 2012; Watt et al. 2015). Individual char-
from 5 months throughout childhood. Of these, 39% acteristics include temperament (using the sum of both
were firstborn. For each follow-up, informed consent parent-reports of difficult and unpredictable tempera-
was obtained from parents, teachers and children ment at 17 months), early cognitive skills (measured
when applicable. Predictor variable data were collected using the Imitation Sorting Task which assesses atten-
for 991 girls and 1006 boys at age 2 years (specifically, 29 tion and working memory at 29 months), concurrent
months), thus creating our subsample for analyses. screen time at age 13 years (television, DVD, and
Outcome variable data were collected at age 13 years videogame exposure during both week and weekend)
(158 months). and child gender (boys = 1 and girls = 0). Family char-
acteristics at the 17-month assessment include family
Predictor variable: daily televiewing (age 2 years) configuration (two-parent = 1 and single-parent = 0),
parental education (mother and father graduated
During the early childhood follow-up, parents answered high school = 2, mother or father graduated = 1, and
the following questions: ‘How much time per day does neither parent graduated = 0) and parental antisocial
your child spend watching TV on a typical weekday?’ behavior during adolescence and adulthood. This last
and ‘How much time per day does your child spend measure was assessed using a composite score from
watching TV on a typical weekend?’ Scores reflect the mother and father responses to the National Institute
total hours of weekly television exposure (weekday view- of Mental Health-Diagnostic Interview Schedule
ing multiplied by 5 plus weekend viewing multiplied by (Robins et al. 1981). Higher scores correspond to
2), which were then divided by 7 to give an estimate of more parental antisocial behavior.
typical daily viewing. This measure is comparable or
similar to previous population-based assessments of chil- Data analytic strategy
dren’s screen viewing in the home (Christakis et al. 2004;
Mistry et al. 2007), including our own (Pagani et al. 2010, In this study, we aimed to examine long-term prospect-
2013; Fitzpatrick et al. 2012; Watt et al. 2015). ive linear associations using multiple regression, in
which indicators of social impairment (SIi@age13, repre-
sented by victimization, social isolation, proactive ag-
Outcome variables: indicators of social impairment
gression and antisocial behavior) at age 13 years are
(age 13 years)
regressed on televiewing at age 2 years (TVi@age 2) for
Participants self-reported on social impairment over each individuali child. To reduce the possibility of omit-
the previous 6 months using four factors from the ted variable bias, we estimate this linear relationship
Social Behavior Questionnaire (Tremblay et al. 1994; while simultaneously controlling for pre-existing and
Dobkin et al. 1995; Pagani et al. 2010, 2013; concurrent child (CHILDi) and family (FAMILYi) char-
Fitzpatrick et al. 2012): (1) victimization (α = 0.81): acteristics that statistically or theoretically correlate with

Downloaded from https:/www.cambridge.org/core. University of Arizona, on 07 May 2017 at 23:09:54, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
. https://doi.org/10.1017/S0033291716001689
3332 L. S. Pagani et al.

our predictor and/or outcome variables (Christakis, Table 1. Descriptive statistics of predictor, outcome and control
2009; Arseneault et al. 2010; Byrd et al. 2014). Our results variables
refer to the adjusted model, where ‘a’ and ‘e’ corres-
pondingly represent the intercept and stochastic error: Mean
(standard
SIi@age13 = a1 + β1 TVi@age2 + γ1 CHILDi + γ2 FAMILYi
Independent variables deviation)
+ eit
Early televiewing, age 29 months, h/week 8.82 (6.17)
Concurrent televiewing, age 13 years, h/week 4.33 (1.46)
Statistical analysis
Child gender, boys = 1
This population-based longitudinal study required Boys, % 50
data from several collection waves and sources. Over Temperament, 17 months 0.47 (0.62)
a third of the data were incomplete either for control Cognitive skills, age 29 months 1.15 (0.79)
or outcome variables (37.4%). Consequently, we con- Family configuration, 17 months, %
One parent 24
ducted an attrition analysis to compare the 1314 cases
Two parents 76
with complete televiewing data at age 2 years with
Parental education, 17 months, %
the 823 cases from the original sample (n = 2120) with
Both no high school 11
incomplete control variable data, using independent- One high school 25
sample t tests. Some differences were found. Children Both high school 64
with complete televiewing data were less likely to be Parental antisocial behavior, 17 months 0.09 (0.29)
raised by single parents (mean = 0.80 v. 0.81, t2039 = Proactive aggression, age 13 years 0.61 (1.30)
2.42, p = 0.04). An attrition analysis also compared the Social isolation, age 13 years 2.33 (2.39)
children with complete v. incomplete outcome data at Antisocial behavior, age 13 years 0.63 (0.95)
age 13 years and control variable data. More girls had Victimization, age 13 years 1.41 (1.55)
complete outcome data than boys (mean = 0.45 v.
0.59, t1312 = −4.912, p < 0.001). Between-group differ-
ences for temperament problems at 17 months were
developmentally appropriate. Nevertheless, 11% of
marginally significant (mean = 0.493 v. 0.427, t1312 =
children in our subsample surpassed the AAP tele-
1.883, p = 0.060). The NORM multiple imputation pro-
viewing recommendations at age 2 years.
gram was used to impute missing control and outcome
Bivariate associations between the predictor, out-
data (Schafer, 1999). By drawing values from the con-
comes and confounding variables are detailed in
ditional variable distributions, NORM uses an iterative
Table 2. Televiewing at age 2 years is associated with
method based on the EM algorithm to impute missing
all four outcome variables at age 13 years. Table 3
data, depending on the available and valid observa-
reports standardized regression coefficients reflecting
tions from the original dataset (http://www.stat.psu.
the relationship between televiewing at age 2 years
edu/~jls/misoftwa.html).
and pre-existing individual and family factors, which
were implemented as control variables. As expected
from the bivariate relationships in Table 2, children
Results
with less educated parents watched more television
Table 1 provides descriptive statistics of predictor, out- (11.62 v. 8.65 h/week, t157.136 = 4.196, p < 0.001) and chil-
come and control variables. Boys comprised 50% of the dren from single-parent families watched more televi-
sample and 24% lived in single-parent homes when the sion (9.93 v. 8.65 h/week, t444.71 = −2.83, p < 0.01) at
child was 17 months. In toddlerhood, participants age 2 years, compared with children from more edu-
watched on average 1.26 h/day (S.D. = 0.88 h/day), cated and two-parent families, respectively.
based on 8.82 h for the entire week (S.D. = 6.17 h). Table 4 reports regression coefficients which reflect
These quantities are similar to averages found with the relationship between early televiewing and social
American children at this age (Christakis et al. 2004; impairment in early adolescence. At age 2 years,
Mistry et al. 2007). At age 13 years, participants television habits predicted increases in self-reported
watched on average 4.33 h for the entire week (S.D. = victimization, social isolation, proactive aggression and
1.46 h). Televiewing mostly occurred on the weekend, antisocial behavior at age 13 years. More specifically,
rather than daily, and this is probably due to home- every 1 h increase in daily televiewing was prospective-
work and extracurricular activities. These quantities ly associated with an 11% S.D. unit increase in victimiza-
are within current American Academy of Pediatrics tion [unstandardized β = 0.03, 95% confidence interval
(AAP) (2013) limitations of not more than 2 h/day be- (CI) 0.02–0.04], a 10% S.D. unit increase in social isolation
yond 2 years of age, assuming that the content is (unstandardized β = 0.04, 95% CI 0.03–0.05), a 9% S.D.

Downloaded from https:/www.cambridge.org/core. University of Arizona, on 07 May 2017 at 23:09:54, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
. https://doi.org/10.1017/S0033291716001689
Early television and social difficulties 3333

12
Table 3. Standardized regression coefficients which reflect the

_
_
_
_
_
_
_
_
_
_
_
1
relationship between child televiewing at toddlerhood and
pre-existing individual and family characteristics

0.09**
11

_
_
_
_
_
_
_
_
_
_
1
Independent variables Early televiewing

−0.37***
0.8***
Child gender 0.04 (0.33)
_
_
_
_
_
_
_
_
_
1
Temperament, 17 months −0.01 (0.27)
10

Cognitive skills, age 29 months −0.01 (0.21)


Family configuration, 17 months 0.03 (0.69)**

0.10**
−0.02

0.00
Parental education, 17 months −0.20 (0.27)***
_
_
_
_
_
_
_
_
1
Parental antisocial behavior, 17 months 0.02 (0.56)
9

Adjusted R2 0.05
0.05
−0.02
0.00
−0.01
_
_
_
_
_
_
_
1

Data are given as standardized regression coefficient


8

(standard error).
* p 4 0.05, ** p < 0.01, *** p < 0.001.
−0.01
−0.04
0.01
−0.01
0.05
_
_
_
_
_
_
1
7

unit increase in proactive aggression (unstandardized


−0.05
−0.02
−0.02
0.00
−0.02
−0.01

β = 0.02, 95% CI 0.01–0.03) and a 6% S.D. unit increase


_
_
_
_
_
1

in antisocial behavior (unstandardized β = 0.01, 95% CI


6

0.01–0.01). Increases in concurrent televiewing were


0.12**

0.10**

associated with additional independent risks of victim-


0.06
0.05
−0.05
0.05
−0.02

ization (unstandardized β = 0.07, 95% CI0.04–0.10),


_
_
_
_
1
5

proactive aggression (unstandardized β = 0.09, 95% CI


0.07–0.11) and antisocial behavior (unstandardized β =
0.14***

0.08*
−0.08*

0.08, 95% CI 0.06–0.10). Coefficients in Tables S1 and


0.02
0.02
0.00
0.03

−0.02

S2 are not corrected for attrition bias.


_
_
_
1
4

0.13***
0.47***
0.10***
0.12***

−0.08**

Discussion
0.07*
0.07*
0.05

0.03
Table 2. Bivariate associations between the predictor, outcomes and confounding variables

_
_
1

Identifying early modifiable factors that foretell later


3

risks for child well-being represents an important


goal for public health and social pediatrics (American
0.35***
0.30***
0.35***

0.13***

−0.13***
0.13***
0.08*

−0.08*
0.08*
−0.01

Academy of Pediatrics, 2013). Developing solid peer


_
1

relationships, getting along with others and establish-


2

ing a positive social identity within a group represent


0.13***

0.11***
−0.21***

essential elements of a successful transition from child-


0.10**
0.10**
0.08*
0.06
0.04
−0.01
−0.04

0.04

hood to adolescence. We sought to examine the pro-


1
1

spective influence of early childhood televiewing on


four key indicators of social impairment in typically
12. Parental antisocial behavior, age 17 months

developing youth at age 13 years. Remarkably, our


self-reported findings revealed long-term risks for
10. Family configuration, age 17 months
6. Concurrent televiewing, age 13 years

developmental psychopathology. Moreover, save for


9. Early cognitive skills, age 29 months

11. Parental education, age 17 months

* p 4 0.05, ** p < 0.01, *** p < 0.001.


3. Proactive aggression, age 13 years

social isolation, televiewing at age 13 years also


1. Early televiewing, age 29 months

5. Antisocial behavior, age 13 years

showed additional concurrent risks for social impair-


8. Temperament, age 17 months
4. Social isolation, age 13 years
2. Victimization, age 13 years

ment. This suggests independent dose–response asso-


ciations operating over the lifetime of the participants.
7. Child gender, boys = 1

Approximately every additional daily hour of tele-


viewing at age 2 years predicted an increased risk of
later victimization and socially isolative behavior.
Victimization represents a traumatic social learning ex-
Variables

perience and its relationship with early televiewing is


detectable around school entry (Pagani et al. 2013).
Adolescent experiences such as being teased, assaulted

Downloaded from https:/www.cambridge.org/core. University of Arizona, on 07 May 2017 at 23:09:54, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
. https://doi.org/10.1017/S0033291716001689
3334 L. S. Pagani et al.

Table 4. Standardized regression coefficients which reflect the relationship between televiewing at toddlerhood and self-reported victimization,
social withdrawal, proactive aggression and antisocial behavior at age 13 years

Independent variable Victimization Social isolation Proactive aggression Antisocial behavior

Early televiewing, age 29 months 0.11 (0.01)*** 0.10 (0.01)*** 0.09 (0.01)** 0.06 (0.00)*
Concurrent televiewing, age 13 years 0.07 (0.03)* 0.03 (.05) 0.10 (0.02)*** 0.13 (0.02)***
Child gender 0.11 (0.08)*** −0.02 (0.13) 0.11 (0.07)*** 0.01 (0.05)
Temperament, 17 months 0.02 (0.07) −0.01 (0.11) 0.06 (0.06)* 0.04 (0.04)
Cognitive skills, age 29 months −0.07 (0.05)* 0.04 (0.08) −0.07 (0.05)* −0.08 (0.03)**
Family configuration, 17 months 0.07 (0.17)* 0.04 (0.27) 0.04 (0.15) 0.06 (0.07)
Parental education, 17 months −0.09 (0.07)** −0.02 (0.11) −0.02 (0.06) 0.01 (0.04)
Parental antisocial behavior, 17 months 0.07 (0.14)* −0.01 (0.23) 0.06 (0.12)* 0.10 (0.09)**
Adjusted R2 0.06 0.01 0.04 0.04

Data are given as standardized regression coefficient, corrected for attrition bias (standard error).
* p 4 0.05, ** p < 0.01, *** p < 0.001.

or insulted by other classmates and personal prefer- interaction frequency which practises socio-emotional
ences toward solitude for typical leisure activities skills (Pagani et al. 2013). Related to all of this is that
were associated with higher levels of watching televi- more early televiewing is linked to developmental
sion in toddlerhood. These social characteristics place deficits in executive functions, which drive interper-
adolescents at risk for long-term negative psychologic- sonal problem solving, emotional regulation, socially
al and interpersonal sequelae such as underachieve- competent peer play and positive social contact
ment, loneliness, lower self-esteem and depression (Nathanson et al. 2014). Like a coordinated orchestra,
(Gini & Pozzoli, 2009; Schreier et al. 2009; Reijntjes the executive functions play a key role in developing
et al. 2010; Rudolph et al. 2014; Van Geel et al. 2014). one’s cognitive and emotional processing (Barkley,
The consequences of childhood victimization show 2012). Time spent televiewing represents choosing
developmental stability into adulthood (Arseneault less effortful endeavors in lieu of partaking in activities
et al. 2010; Brunstein et al. 2010). Victimization experi- and playing games that require more sustained atten-
ences often developmentally precede predispositions tion and working memory (Nathanson et al. 2014).
toward antisocial behavior, which includes bullying Habitually elevated televiewing predicts increased ex-
(Arseneault et al. 2010). posure to fast-paced cartoon content, which is common
Approximately every additional 1 h of early child- in preschool programming and undermines executive
hood television exposure at age 2 years also forecasted functioning (Lillard & Peterson, 2011).
an increased risk of engaging in deception, theft, van- As for microscopic mechanisms, eye contact is at the
dalism, and threatening and instilling fear on others, heart of positive social development (Zeidner et al.
and encouraging social exclusion of others as an 2009). In the first years of life, attention and self-
aggressive means of social domination. Such proactive assertion skills recruit common sensory systems and
antisocial behaviors were self-reported at age 13 years, cortical structures (Marsh et al. 2008). Throughout de-
which is disconcerting from a clinical perspective be- velopment, a shared gaze between people functions
cause externalizing childhood psychopathology pre- beyond mere visual detection, as it remains a vital
dicts Axis I and II mental health disorders, unhealthy cue system for mutual understandings between people
lifestyle habits, criminality and physical disease in (Beier & Spelke, 2012). Eye contact is a necessary con-
adulthood (Morcillo et al. 2012; Loth et al. 2014). dition for human empathy and appropriate social
There are macroscopic mechanisms that can eluci- exchanges, as it modulates the activation of the social
date the link between early childhood televiewing brain as it processes socially relevant sensory informa-
and subsequent victimization and such extreme disre- tion and internal states of partners in social interaction
gard for the well-being of others. Excessive televiewing (Farroni et al. 2002; Senju & Johnson, 2009). Although
habits create a time debt for effortful interpersonal wired prenatally, this inherent skill is honed by opportun-
interactions that sharpen the critical foundations of ities for increasingly complex social interactions during
self-regulation and emotional intelligence. That is, early childhood (Beier & Spelke, 2012). Children who de-
higher amounts of televiewing indicate a higher velop asynchronous gaze behavioral patterns, such as
proportion of passive rather than active daily cognitive looking down or away from others, tend to evoke more
activity in toddlerhood, ultimately affecting social negative social responses from others throughout

Downloaded from https:/www.cambridge.org/core. University of Arizona, on 07 May 2017 at 23:09:54, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
. https://doi.org/10.1017/S0033291716001689
Early television and social difficulties 3335

development (Zeidner et al. 2009; Beier & Spelke, 2012) self-reports of specific social outcomes within an inclu-
and risk having callous–unemotional traits (Dadds sive study, not only replicate previous singular
et al. 2012). Very few televised programs have scenes findings based on adult observations but also to exam-
where the central character speaks directly to the view- ine whether habitual cognitive and physical sedentari-
er, which both parallels a real-life social interaction and ness in toddlerhood influences social relationship
fosters a virtual social relationship with the character patterns in adolescence. This represents an improve-
(Cummins & Cui, 2014). More time televiewing ment on prior research since adults may not always
means less time being spoken to directly, either in be fully aware of the extent to which a child is experi-
person or from the screen. This adds support to our encing or causing psychosocial distress. By statistical
speculative hypothesis regarding disrupted gaze de- standards, the results only show modest associations,
velopment. It is plausible that a time debt for complex yet small adjusted effect sizes can conservatively trans-
social interaction created by excessive televiewing at late into important social costs when projected over a
age 2 years could interfere with progressive inter- lifespan and across an entire population. Because social
dependence in overlapping cortical structures which impairment risks can translate into a general propen-
eventually support higher-order cognitive functions sity for morbidity and mortality in adulthood (Teo
and cortico-thalamic sensory relay processes involved et al. 2013; Non et al. 2014), excessive early childhood
in social interaction and emotional intelligence television exposure could pose epigenetic risks to
(Zeidner et al. 2009). Using functional imaging with population health. Our findings thus suggest a need
116 Japanese children aged 5–18 years, Takeuchi et al. for better parental awareness of how young children
(2015) remains the only study to our knowledge that invest their time in what remains of their 24-h day.
has documented cross-sectional and longitudinal cor-
relations between televiewing time and compromised
regional gray/white matter volume. Taken together, Supplementary material
these neuroscientific explanations could translate into The supplementary material for this article can be
plausible risks for impaired social and emotional de- found at http://dx.doi.org/10.1017/S0033291716001689
velopment and warrant further direct investigation in
relation to excessive early televiewing.
This study is not without limitations. Foremost is Acknowledgements
that epigenetic influences cannot be denied their direct
and even more plausible indirect explanatory role in All authors have had full access to all data in the study
how environmental quality either activates or sup- and take responsibility for its integrity and the accur-
presses phenotypic expression. Second, our measure acy of its analysis. This IRB-approved study was con-
of television exposure did not comprise a finely ducted without specific funding to any of the authors.
tuned, objective scale of quantity and content of televi-
sion exposure. Moreover, parents using child care may
Declaration of Interest
have underestimated the true amount of total screen-
time. Despite such a limited television measure, we None.
still found robust net effects. This is statistically re-
markable and points to a stronger associations than
suggested by the coefficients. The question remains: References
influence of what? We have argued that too much tele- American Academy of Pediatrics (2013). Policy statement:
viewing means less time left for other developmental American Academy of Pediatrics: children, adolescents, and
enriching activities during waking hours. Framing it the media. Pediatrics 132, 958–961.
this way suggests that, beyond genetic influence, exces- Arseneault L, Bowes L, Shakoor S (2010). Bullying
sive early childhood screen time represents a marker for victimization in youths and mental health problems: ‘Much
deficient socialization experiences. From this perspec- ado about nothing’? Psychological Medicine 40, 717–729.
tive, campaigns and guidelines that aim to discourage Barkley RA (2012). Executive Functions: What They Are, How
screen time could instead be reconceptualized to en- They Work, and Why They Evolved. Guilford Press:
New York.
courage the value-added of socialization as an enriching
Beier JS, Spelke ES (2012). Infants’ developing
alternative.
understanding of social gaze. Child Development 83,
The use of a prospective longitudinal design from 486–496.
birth onward represents a major strength, as it does Brunstein KA, Sourander A, Gould M (2010). The association
not depend on retrospective recall and monitors of suicide and bullying in childhood to young adulthood: a
growth and development without researcher bias or review of cross-sectional and longitudinal research
interference. The findings, which capitalize on child findings. Canadian Journal of Psychiatry 55, 282–288.

Downloaded from https:/www.cambridge.org/core. University of Arizona, on 07 May 2017 at 23:09:54, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
. https://doi.org/10.1017/S0033291716001689
3336 L. S. Pagani et al.

Byrd AL, Loeber R, Pardini DA (2014). Antisocial behavior, Marsh R, Gerber AJ, Peterson BS (2008). Neuroimaging
psychopathic features, and abnormalities in reward and studies of normal brain development and their relevance
punishment processing in youth. Clinical Child and Family for understanding childhood neuropsychiatric disorders.
Psychology Review 17, 125–156. Journal of the American Academy of Child and Adolescent
Cespedes EM, Gillman MW, Kleinman K, Rifas-Shiman SL, Psychiatry 47, 1233–1251.
Redline S, Taveras EM (2014). Television viewing, Meltzoff AN, Kuhl PK, Movellan J, Sejnowski TJ (2009).
bedroom television, and sleep duration from infancy to Foundations for a new science of learning. Science 325,
mid-childhood. Pediatrics 133, e1163–e1171. 284–288.
Christakis DA (2009). The effects of infant media usage: what Mistry KB, Minkovitz CS, Strobino DM, Borzekowski DL
do we know and what should we learn? Acta Paediatrica 98, (2007). Children’s television exposure and behavioral and
8–16. social outcomes at 5.5 years: does timing of exposure
Christakis DA, Zimmerman FJ, DiGiuseppe DL, McCarty matter? Pediatrics 120, 762–769.
CA (2004). Early television exposure and subsequent Moffitt TE (1993). Adolescence-limited and life-course-
attentional problems in children. Pediatrics 113, 708–713. persistent antisocial behavior: a developmental taxonomy.
Cillero IH, Jago R (2010). Systematic review of correlates of Psychological Review 100, 674.
screen-viewing among young children. Preventive Medicine Morcillo C, Duarte CS, Sala R, Wang S, Lejuez CW, Kerridge
51, 3–10. BT, Blanco C (2012). Conduct disorder and adult psychiatric
Cummins RG, Cui B (2014). Reconceptualizing address in diagnoses: associations and gender differences in the US adult
television programming: the effect of address and affective population. Journal of Psychiatric Research 46, 323–330.
empathy on viewer experience of parasocial interaction. Nathanson AI, Aladé F, Sharp ML, Rasmussen EE, Christy
Journal of Communication 64, 723–742. K (2014). The relation between television exposure and
Dadds MR, Allen JL, Oliver BR, Faulkner N, Legge K, Moul executive function among preschoolers. Developmental
C, Scott S (2012). Love, eye contact, and the developmental Psychology 50, 1497.
origins of empathy v. psychopathy. British Journal of Non AL, Rewak M, Kawachi I, Gilman SE, Loucks EB,
Psychiatry 200, 191–196. Appleton AA, Román JC, Buka SL, Kubzansky LD (2014).
Dobkin PL, Tremblay RE, Mâsse LC, Vitaro F (1995). Childhood social disadvantage, cardiometabolic risk, and
Individual and peer characteristics in predicting boys’ early chronic disease in adulthood. American Journal of
onset of substance abuse: a seven‐year longitudinal study. Epidemiology 180, 263–271.
Child Development 66, 1198–1214. Pagani LS, Fitzpatrick C, Barnett TA (2013). Early childhood
Farroni T, Csibra G, Simion F, Johnson MH (2002). Eye television viewing and kindergarten entry readiness.
contact detection in humans from birth. Proceedings of the Pediatric Research 74, 350–355.
National Academy of Sciences 99, 9602–9605. Pagani LS, Fitzpatrick C, Barnett TA, Dubow E (2010).
Fitzpatrick C, Barnett T, Pagani LS (2012). Early exposure to Prospective associations between early childhood television
media violence and later child adjustment. Journal of exposure and academic, psychosocial, and physical well-
Developmental and Behavioral Pediatrics 33, 291–297. being by middle childhood. Archives of Pediatrics and
Funk JB, Brouwer J, Curtiss K, McBroom E (2009). Parents of Adolescent Medicine 164, 425–431.
preschoolers: expert media recommendations and ratings Reijntjes A, Kamphuis JH, Prinzie P, Telch MJ (2010). Peer
knowledge, media-effects beliefs, and monitoring practices. victimization and internalizing problems in children: a
Pediatrics 123, 981–988. meta-analysis of longitudinal studies. Child Abuse and
Gingold JA, Simon AE, Schoendorf KC (2014). Excess screen Neglect 34, 244–252.
time in US children: association with family rules and Richert RA, Robb MB, Smith EI (2011). Media as social
alternative activities. Clinical Pediatrics 53, 41–50. partners: the social nature of young children’s learning
Gini G, Pozzoli T (2009). Association between bullying and from screen media. Child Development 82, 82–95.
psychosomatic problems: a meta-analysis. Pediatrics 123, Robertson LA, McAnally HM, Hancox RJ (2013). Childhood
1059–1065. and adolescent television viewing and antisocial behavior
Heckman JJ (2006). Skill formation and the economics of in early adulthood. Pediatrics 131, 439–446.
investing in disadvantaged children. Science 312, 1900–1902. Robins LN, Helzer JD, Croughan J, Ratcliff KS (1981). The
Hubbard JA, McAuliffe MD, Morrow MT, Romano LJ National Institute of Mental Health Diagnostic Interview
(2010). Reactive and proactive aggression in childhood and Schedule: its history, characteristics, and validity. Archives
adolescence: precursors, outcomes, processes, experiences, of General Psychiatry 38, 381–389.
and measurement. Journal of Personality 78, 95–118. Rubin KH, Barstead MG (2014). Gender differences in child
Kuhl PK (2007). Is speech learning ‘gated’ by the social brain? and adolescent social withdrawal: a commentary. Sex Roles
Developmental Science 10, 110–120. 70, 274–284.
Lillard AS, Peterson J (2011). The immediate impact of Rudolph KD, Lansford JE, Agoston AM, Sugimura N,
different types of television on young children’s executive Schwartz D, Dodge KA, Pettit GS, Bates JE (2014). Peer
function. Pediatrics 128, 644–649. victimization and social alienation: predicting deviant
Loth AK, Drabick DA, Leibenluft E, Hulvershorn LA (2014). peer affiliation in middle school. Child Development 85,
Do childhood externalizing disorders predict adult 124–139.
depression? A meta-analysis. Journal of Abnormal Child Schafer JL (1999). Multiple imputation: a primer. Statistical
Psychology 42, 1103–1113. Methods in Medical Research 8, 3–15.

Downloaded from https:/www.cambridge.org/core. University of Arizona, on 07 May 2017 at 23:09:54, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
. https://doi.org/10.1017/S0033291716001689
Early television and social difficulties 3337

Schreier A, Wolke D, Thomas K, Horwood J, Hollis C, Teo AR, Choi H, Valenstein M (2013). Social relationships
Gunnell D, Harrison G (2009). Prospective study of peer and depression: ten-year follow-up from a nationally
victimization in childhood and psychotic symptoms in a representative study. PLOS ONE 8, e62396.
nonclinical population at age 12 years. Archives of General Tremblay RE, Pihl RO, Vitaro F, Dobkin PL (1994).
Psychiatry 66, 527–536. Predicting early onset of male antisocial behavior from
Senju A, Johnson MH (2009). The eye contact effect: preschool behavior. Archives of General Psychiatry 51, 732–739.
mechanisms and development. Trends in Cognitive Sciences Van Geel M, Vedder P, Tanilon J (2014). Relationship between
13, 127–134. peer victimization, cyberbullying, and suicide in children and
Shonkoff JP (2011). Protecting brains, not simply stimulating adolescents: a meta-analysis. JAMA Pediatrics 168, 435–442.
minds. Science 333, 982–983. Watt E, Fitzpatrick C, Derevensky JL, Pagani LS (2015). Too
Sisson SB, Sheffield-Morris A, Spicer P, Lora K, Latorre C much television? Prospective associations between early
(2014). Influence of family structure on obesogenic childhood televiewing and later self-reports of victimization
behaviors and placement of bedroom TVs of American by sixth grade classmates. Journal of Developmental and
children: National Survey of Children’s Health 2007. Behavioral Pediatrics 36, 426–433.
Preventive Medicine 61, 48–53. Zeidner M, Matthews G, Robert RD (2009). What We Know
Takeuchi H, Taki Y, Hashizume H, Asano K, Asano M, About Emotional Intelligence: How it Affects Learning,
Sassa Y, Kawashima R (2015). The impact of television Work, Relationships, and Our Mental Health. MIT Press:
viewing on brain structures: cross-sectional and Cambridge, MA.
longitudinal analyses. Cerebral Cortex 25, 1188–1197. Zimmerman FJ, Glew GM, Christakis DA, Katon W (2005).
Tandon PS, Zhou C, Christakis DA (2012). The frequency Early cognitive stimulation, emotional support, and
of outdoor play for preschool age children cared for at television watching as predictors of subsequent bullying
home-based child care settings. Academic Pediatrics 12, among grade-school children. Archives of Pediatrics and
475–480. Adolescent Medicine 159, 384–388.

Downloaded from https:/www.cambridge.org/core. University of Arizona, on 07 May 2017 at 23:09:54, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
. https://doi.org/10.1017/S0033291716001689

You might also like