The associations between TV viewing, food intake, and BMI. A prospective analysis of data from the Longitudinal Study of Australian Children

Abstract

a b s t r a c t Objective: Despite cross-sectional evidence of a link between TV viewing and BMI in early childhood, there has been limited longitudinal exploration of this relationship. The aim of the present study was to explore the potential bi-directionality of the relationship between TV viewing and child BMI. A second-ary aim was to evaluate whether this relationship is mediated by dietary intake. Study design: Parents of 9064 children (4724 recruited at birth, 4340 recruited at age 4) from the Longi-tudinal Study of Australian Children (LSAC) completed measures of their child's dietary intake and TV viewing habits at three equidistant time points, separated by 2 years. Objective measures of height and weight were also obtained at each time point to calculate BMI. Cross-lagged panel analyses were con-ducted to evaluate potential bi-directional associations between TV viewing and child BMI, and to eval-uate mediation effects of dietary intake for this relationship. Results: Our longitudinal findings suggest that the relationship between TV viewing and BMI is bi-direc-tional: Individuals who watch TV are more likely to gain weight, and individuals who are heavier are also more likely to watch TV. Interestingly, dietary intake mediated the BMI-TV viewing relationship for the older children, but not for the birth cohort. Conclusions: Present findings suggest that sedentary behaviours, particularly when coupled with unhealthy dietary habits, constitute a significant risk factor for excessive weight gain in early childhood. Interventions targeted at helping parents to develop healthy TV viewing and eating habits in their young children are clearly warranted.

Full text

Page 1

Research report
The associations between TV viewing, food intake, and BMI. A prospective
analysis of data from the Longitudinal Study of Australian Childrenq
Matthew Fuller-Tyszkiewicza,⇑, Helen Skouterisa, Louise L. Hardyb, Christine Halsec
aSchool of Psychology, Deakin University, Victoria, Australia
bMedical School, University of Sydney, New South Wales, Australia
cEducation, Deakin University, Victoria, Australia
a r t i c l ei n f o
Article history:
Received 10 August 2012
Accepted 11 September 2012
Available online 18 September 2012
Keywords:
Child BMI
TV viewing
Dietary intake
Longitudinal design
Obesity
a b s t r a c t
Objective: Despite cross-sectional evidence of a link between TV viewing and BMI in early childhood,
there has been limited longitudinal exploration of this relationship. The aim of the present study was
to explore the potential bi-directionality of the relationship between TV viewing and child BMI. A second-
ary aim was to evaluate whether this relationship is mediated by dietary intake.
Study design: Parents of 9064 children (4724 recruited at birth, 4340 recruited at age 4) from the Longi-
tudinal Study of Australian Children (LSAC) completed measures of their child’s dietary intake and TV
viewing habits at three equidistant time points, separated by 2 years. Objective measures of height
and weight were also obtained at each time point to calculate BMI. Cross-lagged panel analyses were con-
ducted to evaluate potential bi-directional associations between TV viewing and child BMI, and to eval-
uate mediation effects of dietary intake for this relationship.
Results: Our longitudinal findings suggest that the relationship between TV viewing and BMI is bi-direc-
tional: Individuals who watch TV are more likely to gain weight, and individuals who are heavier are also
more likely to watch TV. Interestingly, dietary intake mediated the BMI-TV viewing relationship for the
older children, but not for the birth cohort.
Conclusions: Present findings suggest that sedentary behaviours, particularly when coupled with
unhealthy dietary habits, constitute a significant risk factor for excessive weight gain in early childhood.
Interventions targeted at helping parents to develop healthy TV viewing and eating habits in their young
children are clearly warranted.
? 2012 Elsevier Ltd. All rights reserved.
Introduction
Increased sedentariness is ubiquitous in Australia and other
developed countries (Healy, Matthews, Dunstan, Winkler, & Owen,
2011). Of all sedentary activities, screen time (ST; i.e., watching TV/
DVDs/videos and recreational computer use) is the most popular
among young children, adolescents and adults in Australia, the
US, UK, and Europe (Baxter & Hayes, 2007; Bertrais et al., 2005;
Hardy, Dobbins, Denney-Wilson, Okely, & Booth, 2006; Jakes
et al., 2003; Marsh, 2005). In 2004, national Australian guidelines
were established which recommend children aged 5–18 spend
no more than 2 h a day on ST (Commonwealth Department of
Health & Ageing, 2004) and, more recently, that for children aged
2–5 years ST should be limited to <1 h per day (Commonwealth
of Australia, Department of Health and Ageing, 2010). In adults
there is strong evidence that prolonged TV viewing is associated
independently with obesity, type 2 diabetes, and biomarkers of
metabolic and cardiovascular disease (Healy et al., 2008). Recent
findings have also revealed that there is an increased risk of insulin
resistance among adolescent boys who exceed ST guidelines, inde-
pendent of body mass index (BMI), diet, fitness, and pubertal status
(Hardy, King, Kelly, Farrell, & Howlett, 2010). Furthermore, 11–
15 year olds who exceed the guidelines have significantly lower
cardio respiratory fitness (Hardy, Dobbins, Denney-Wilson, Okely,
& Booth, 2009).
A recent systematic literature review on the impact of TV on
preschoolers’ weight status revealed that of 26 relevant studies,
23 reported a positive association between hours of TV and child
adiposity(Cox, Skouteris,Rutherford,
2011). Five cross-sectional studies (Dubois, Farmer, Girard, & Pet-
erson, 2008; Jackson, Djafarian, Stewart, & Speakman, 2009; Janz
et al., 2002; Manios et al., 2009; Vandebosch & Van Cleemput,
2007) and four prospective studies (Epstein et al., 2008; Jago,
Baranowski, Baranowski, Thompson, & Greaves, 2005; Viner &
Cole, 2005; Zimmerman & Bell, 2010) assessed whether the effect
of TV on preschool child weight status is related to physical activity
and/or diet. Taken together, the findings of these studies suggest
&Fuller-Tyszkiewicz,
0195-6663/$ - see front matter ? 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.appet.2012.09.009
qCompeting interests: The authors declare that they have no competing interests.
⇑Corresponding author. Address: School of Psychology, Deakin University,
Victoria, Australia.
E-mail address: matthewf@deakin.edu.au (M. Fuller-Tyszkiewicz).
Appetite 59 (2012) 945–948
Contents lists available at SciVerse ScienceDirect
Appetite
journal homepage: www.elsevier.com/locate/appet

Page 2

that TV viewing is conducive to unhealthy eating, and that the rela-
tionship between TV viewing and weight status reduces in magni-
tude once eating habits – but not physical activity – are controlled
for. The only study to formally test for mediation (Jackson et al.,
2009), demonstrated that physical activity is not a reliable media-
tor of the TV viewing-BMI relationship. The status of dietary habits
as a potential mediator requires empirical evaluation. Likewise, the
possibility that BMI instead/also promotes sedentary behaviours
warrants attention in light of the consistent link between the
two (Dubois et al., 2008; Epstein et al., 2008; Jackson et al., 2009;
Jago et al., 2005; Janz et al., 2002; Manios et al., 2009; Vandebosch
& Van Cleemput, 2007; Viner & Cole, 2005; Zimmerman & Bell,
2010). Given that there was no Australian research among the
studies included in this systematic review, the aim of the current
study was to evaluate the longitudinal relationships between TV
viewing, child BMI, and food intake in a cohort sample of Austra-
lian children.
Method
Participants
Participants were from the Longitudinal Study of Australian
Children (LSAC) (n = 9064), and comprised 4724 children who were
recruited at age 0–1 years (Cohort B) and 4340 children who were
recruited at age 4–5 years (Cohort K) from Waves 2–4. Wave 1 data
were excluded from the present study as information on children’s
TV viewing habits and BMI were not reported. Data were collected
from the primary caregiving parent via face-to-face interviews in
the home. Furthermore details about the LSAC design, methodol-
ogy and demographic details of the sample are available online
(The Longitudinal Study of Australian Children; LSAC, 2010).
Measures
Child BMI
Trained researchers measured children’s weight (kg) and height
(m2), from which BMI was calculated (kg/m2). Children’s BMI was
categorised according to international criteria (Cole, Bellizzi, Flegal,
& Dietz, 2000; Cole, Flegal, Nicholls, & Jackson, 2007).
Dietary intake and television viewing
Parents provided information on their child’s dietary intake.
Dietary scores were derived from the sum of scores from seven die-
tary items: total high fat food including full cream milk (five items:
pie, hot chips, potato chips, biscuits, and full-cream milk) and high
sugar drinks (two items: juice and soft drink). Parents indicated the
number of daily serves for each of these items. Parents also esti-
mated daily weekday and weekend TV viewing (in minutes) for
their children. Total weekly TV viewing was derived by summing
weekday and weekend viewing.
Data analytic strategy
An autoregressivecross-lagged panelmodel approach
(MacKinnon, 2008) was used to evaluate the possibility of recipro-
cal influences between child BMI and TV viewing. Models were
estimated using Mplus Version 6.1 (Muthén & Muthén, 2010), with
full information maximum likelihood estimation and robust stan-
dard errors. Missing data were imputed using the intention to treat
approach, which is considered the gold standard for longitudinal
assessment (Heritier, Gebski, & Keech, 2003). Unlike other ap-
proaches to missing data imputation which typically fit the
existing patterns in the data set (e.g., regression, expectation max-
imisation, etc.), the intention to treat approach provides a more
conservative estimate of change in variables over time as missing
data is imputed with values obtained from the preceding wave of
data collection (Heritier et al., 2003).
Adequacy of model fit was examined using the following crite-
ria: Comparative Fit Index (CFI) and Tucker–Lewis Index (TLI),
where; >.95 = good fit, >.90 = adequate fit), Root Mean Square Error
of Approximation (RMSEA) where 6.05 = good fit, <.08 = adequate
fit), and Standardised Root Mean Square Residual (SRMR) where
<.05 = good fit, <.08 = adequate fit) (Byrne, 2011; Hu & Bentler,
1999).
In the event that the longitudinal associations between TV
viewing and child BMI (and child BMI ? TV viewing) were found
to be significant, subsequent autoregressive models were con-
ducted to evaluate whether either of these relationships were
mediated by the child’s dietary intake. Mplus’ indirect effect com-
mand, with bootstrapping (n = 1000), was used to formally evalu-
ate the longitudinal mediational relationships between Wave 2
IV, Wave 3 MV, and Wave 4 DV in these models.
Results
Descriptive statistics and correlations
ThemeanageofchildreninCohortK(51%boys)andB(51%boys)
atWave2–4were2.29(SD=.45),4.25(.43),and6.32(.47)forCohort
B,and6.29(.46),8.26(.44),and10.31(.47)forCohortK.Tables1and
2 provide means, standard deviations, and inter-correlations for TV
viewing (in minutes/week) and child BMI by Wave for Cohort B
and Cohort K, respectively. TV viewing and child BMI showed mod-
erate temporal stability across waves for both cohorts, rs P .59 for
BMI and rs P .40 for TV viewing. Inter-correlations between child
BMIandTVviewingweresmall,butsignificantforbothcohorts,with
slightly higher correlations for the Cohort K.
Cross-lagged panel analyses
The cross-lagged model provided an adequate fit of the data for
both Cohort B [v2(df=17)= 184.13, p < .001, CFI = .96, SRMR = .04,
RMSEA = .10] and Cohort K [v2(df=17)= 215.18, p < .001, CFI = .95,
SRMR = .04, RMSEA = .09]. As shown in Fig. 1, TV viewing was a
small but significant positive predictor of child BMI both cross-sec-
tionally and longitudinally for both cohorts. Likewise, child BMI
predicted TV viewing within and across waves for Cohort B and Co-
hort K, with the exception of a non-significant association between
BMI and TV at Wave 2 for the younger cohort.
Tests of mediation models
Additional auto-regressive analyses were conducted to evaluate
the possibility that longitudinal associations between TV viewing
and child BMI were mediated by unhealthy eating habits. The
hypothesised paths in these models (as depicted in Fig. 2) provided
a reasonable fit to the data overall. The fit statistics for the TV view-
ing ? food intake ? BMI models were as follows for Cohort B;
v2(df=17)= 607.85, p < .001, CFI = .95, SRMR = .04, RMSEA = .09, and
for Cohort K; v2(df=17)= 450.12, p < .001, CFI = .97, SRMR = .04,
RMSEA = .07. The following fit statistics were obtained for the
BMI ? food intake ? TVviewing
v2(df=17)= 621.32, p < .001, CFI = .95, SRMR = .05, RMSEA = .09, and
for Cohort K; v2(df=17)= 439.17, p < .001, CFI = .97, SRMR = .04,
RMSEA = .08.
With respect to the mediation component of the models, the TV
viewing-BMI relationship was not mediated by food intake for
model forCohortB;
946
M. Fuller-Tyszkiewicz et al./Appetite 59 (2012) 945–948

Page 3

Cohort B (indirect effect = .001, p = .20) or Cohort K (indirect ef-
fect = .001, p = .75). However, food intake was a significant media-
tor of the longitudinal relationship between child BMI and TV
viewing in Cohort K (indirect effect = ?.01, p < .05) but not in Co-
hort B (indirect effect = .001, p = .40). The significant mediation
model for Cohort K is shown in Fig. 2. All non-significant models
may be obtained from the first author upon request.
Discussion
The present study evaluated the associations between TV
viewing, dietary intake, and child BMI in two separate samples of
Australian children. Consistent with prior research (Dubois et al.,
2008; Epstein et al., 2008; Jackson et al., 2009) small but significant
associations were found between TV viewing and BMI at each of
three waves of data collection. A longitudinal relationship between
these variables was also found. Whereas it is often assumed that
TV viewing habits influence BMI, key novel findings from our
cross-lagged autoregressive modelling suggest that the relation-
ship between TV viewing and BMI is bi-directional. This latter find-
ing may be driven by a general tendency towards sedentary
behaviour: general sedentariness may predispose children to lar-
ger BMIs which in turn may lead to greater engagement in specific
sedentary behaviours such as watching TV, and then to continued
excessive weight gain, and so on. Future research is needed to
assess the tenability of this assumption as the LSAC surveys did
not include a global index of child sedentariness.
Interestingly,thereweresomenoteworthydifferencesinthepat-
tern of findings for the two cohorts. First, the correlations between
Table 1
Means, standard deviations, and correlations among child BMI (kg/m2) and TV viewing (mins/week) for Cohort K (n = 4724).
Variable
MSD
Child BMITV viewing
Wave 2 Wave 3Wave 4 Wave 2 Wave 3 Wave 4
BMI (kg/m2)
Wave 2
Wave 3
Wave 4
16.82
16.38
16.56
1.61
1.75
2.20
–
.68
.59
–
.75–
TV viewing (minutes/week)
Wave 2
Wave 3
Wave 4
799.25
853.67
786.21
536.33
570.00
470.03
.01
.04
.04
.03
.03
.05
.04
.04
.08
–
.45
.40
–
.50–
All correlations r > .01 are significant.
Table 2
Means, standard deviations, and correlations among child BMI (kg/m2) and TV viewing (minutes/week) for Cohort B (n = 4340).
Variable
M SD
Child BMITV viewing
Wave 2Wave 3Wave 4Wave 2 Wave 3Wave 4
BMI (kg/m2)
Wave 2
Wave 3
Wave 4
16.51
17.55
18.95
2.15
2.81
3.72
–
.87
.75
–
.85–
TV viewing (minutes/week)
Wave 2
Wave 3
Wave 4
772.05
822.39
892.61
440.03
499.47
501.78
.10
.10
.08
.12
.12
.11
.11
.11
.13
–
.47
.41
–
.46–
All correlations r > .01 are significant.
Cohort B
Cohort K
Fig. 1. Co-efficients from auto-regressive cross lag analysis on the reciprocal
influences between child BMI and TV viewing.
presented in these models. All coefficients above .01 are significant in this figure.
?Standardised coefficients are
Fig. 2. Coefficients for longitudinal associations between TV viewing and child BMI,
mediated by unhealthy eating habits (Cohort K only). Notes: Standardised coeffi-
cients are presented in this model. All coefficients above .01 are significant in this
Figure.
M. Fuller-Tyszkiewicz et al./Appetite 59 (2012) 945–948
947

Page 4

TV viewing and child BMI were stronger for the older subgroupwho
startedthestudyatage4(CohortK)thanforthesubgroupwhocom-
menced at birth (Cohort B). Second, the longitudinal association be-
tween BMI and TV viewing was mediated by dietary habits for
Cohort K, but this effect was not found for Cohort B. While it is pos-
sible that these group differences are simply an artefact of sampling
fluctuations, the direction of these age-related effects may be more
meaningfullyexplainedintermsofdifferencesduetolevelofauton-
omy. As children age, they are more likely to be given freedom in
what they choose to eat and watch. In this light, it would not be sur-
prisingtofindthatTVviewingisabetterpredictorofBMI,orthateat-
ing habits better account for the relationship between BMI and TV
viewing habits, in the older cohort. Several studies have demon-
strated that increased energy intake, particularly during television
viewing, may contribute to overweight or obesity in preschool chil-
dren (Dubois et al., 2008; Epstein et al., 2008; Manios et al., 2009).
However, few studies have examined the mediating effects on the
longitudinal association between BMI and TV viewing over years
as we have done with the current dataset.
Strengths and limitations
Our findings should be placed within the context of strengths
and limitations of the current design. First, the present study uti-
lised a large sample (n = 4340 for Cohort B and n = 4724 for Cohort
K) that is nationally representative, with respect to gender distri-
bution and socioeconomic status of participants. To the authors’
knowledge, it is the first study to longitudinally analyse the rela-
tionship between TV viewing, BMI, and dietary intake of children
in an Australian population. It is also the first to statistically eval-
uate the possibility that dietary intake mediates the TV viewing-
BMI relationship (cf. Dubois et al., 2008; Epstein et al., 2008; Jack-
son, Djafarian, Stewart, & Speakman, 2009; Jago et al., 2005; Janz
et al., 2002; Manios et al., 2009; Vandebosch & Van Cleemput,
2007; Viner & Cole, 2005; Zimmerman & Bell, 2010). However,
while child BMI was measured objectively by trained experts, TV
viewing and dietary habits were reported retrospectively by the
child’s parents or guardians. This introduces potential error in
measurement that may serve to under-estimate or over-estimate
the true strength of association between variables in our study. Re-
lated to this, it would be ideal to include a separate measure of eat-
ing habits and types of food eaten while watching television, rather
than simply relying on estimates for daily intake. Additionally, the
present study did not control for child physical activity level. It is
possible that the effects of TV viewing on BMI may be reduced
for children who are highly physically active.
Conclusions
Present findings highlight the potential impact of TV viewing on
child BMI, and suggest that dietary intake mediates this association
for older children (aged 4+ years). Given the increasing prevalence
of, and health consequences associated with, obesity, these find-
ings of the co-occurrence of sedentary behaviours (TV viewing
and eating) in such a young population are alarming. Interventions
targeted at successfully alerting parents to the fact that TV viewing
and eating may be contributing to excessive weight gain when
combined, and helping parents to develop healthy TV viewing
and eating habits in their young children, are clearly warranted.
References
Baxter, J., & Hayes, A. (2007). How four year-olds spend their days. Insights into the
caring contexts of young children. Family Matters, 76, 34–43.
Bertrais, S., Beyeme-Ondona, J., Czernichow, S., Galan, P., Hercberg, S., & Oppert, J.
(2005). Sedentary behaviours, physical activity, and metabolic syndrome in
middle-aged French subjects. Obesity Research, 13, 936–944.
Byrne, B. M. (2011). Structural equation modeling in Mplus. Basic concepts,
applications, and programming. New York, NY: Routledge.
Cole, T. J., Bellizzi, M. C., Flegal, K. M., & Dietz, W. H. (2000). Establishing a standard
definition for child overweight and obesity worldwide. International survey.
British Medical Journal, 320, 1–6.
Cole, T. J., Flegal, K. M., Nicholls, D., & Jackson, A. A. (2007). Body mass index cut offs
to define thinness in child and adolescents. International survey. British Medical
Journal, 335, 1–8.
Commonwealth Department of Health and Ageing (2004). Physical activity
recommendations for children and young people. Canberra: Commonwealth of
Australia.
Commonwealth of Australia, Department of Health and Ageing (2010). National
physicalactivityrecommendations
www.health.gov.au/internet/main/publishing.nsf/Content/phd-physical-activity-
0-5-pdf-cnt.htm> (16.10.10).
Cox, R., Skouteris, H., Rutherford, L., & Fuller-Tyszkiewicz, M. (2011). The association
between television viewing and preschool child body mass index. A systematic
review of English papers published from 1995 to 2010. Journal of Children and
Media, 6(2), 198–220.
Dubois, L., Farmer, A., Girard, M., & Peterson, K. (2008). Social factors and television
use during meals and snacks is associated with higher BMI among pre-school
children. Public Health Nutrition, 11, 1267–1279.
Epstein, L. H., Roemmich, J. N., Robinson, J. L., Paluch, R. A., Winiewicz, D. D., Fuerch,
J. H., et al. (2008). A randomized trial of the effects of reducing television
viewing and computer use on body mass index in young children. Archives of
Pediatric and Adolescent Medicine, 162, 239–245.
Growing up in Australia. The Longitudinal Study of Australian Children (LSAC).
<http://www.aifs.gov.au/growingup/data/index.html> (updated 1.12.10).
Hardy, L., Dobbins, T., Denney-Wilson, E., Okely, A., & Booth, M. (2006). Descriptive
epidemiology of small screen recreation among Australian adolescents. Journal
of Paediatrics and Child Health, 42, 709–714.
Hardy, L., Dobbins, T., Denney-Wilson, E., Okely, A., & Booth, M. (2009).
Sedentariness, small-screen recreation, and fitness in youth. American Journal
of Preventive Medicine, 36, 120–125.
Hardy, L., King, L., Kelly, B., Farrell, L., & Howlett, S. (2010). Munch and Move.
Evaluation of a preschool healthy eating and movement skill program. The
International Journal of Behavioral Nutrition and Physical Activity, 7, 80.
Healy, G. N., Dunstan, D. W., Salmon, J., Shaw, J. E., Zimmet, P. Z., & Owen, N. (2008).
Television time and continuous metabolic risk in physically active adults.
Medicine & Science in Sports & Exercise, 40, 639–645.
Healy, G. N., Matthews, C. E., Dunstan, D. W., Winkler, E. A. H., & Owen, N. (2011).
Sedentary time and cardio-metabolic biomarkers in US adults. NHANES 2003–
06. European Heart Journal, 32, 590–597.
Heritier, S. R., Gebski, V. J., & Keech, A. C. (2003). Inclusion of patients in clinical trial
analysis. The intention-to-treat principle. The Medical Journal of Australia, 179,
438–440.
Hu, L., & Bentler, P. M. (1999). Cut-off criteria for fit indexes in covariance structure
analysis. Conventional criteria versus new alternatives. Structural Equation
Modeling, 6, 1–55.
Jackson, D. M., Djafarian, K., Stewart, J., & Speakman, J. R. (2009). Increased
television viewing is associated with elevated body fatness but not with lower
total energy expenditure in children. American Journal of Clinical Nutrition, 89,
1031–1036.
Jago, R., Baranowski, T., Baranowski, J. C., Thompson, D., & Greaves, K. A. (2005). BMI
from 3–6 years of age is predicted by TV viewing and physical activity, not diet.
International Journal of Obesity, 29, 557–564.
Jakes, R. W., Day, N. E., Khaw, K.-T., Luben, R., Oakes, S., Welch, A., Bingham, S., &
Wareham, N. J. (2003). Television viewing and low participation in vigorous
recreationare independently associated
cardiovascular disease risk. EPIC-Norfolk population-based study. European
Journal of Clinical Nutrition, 57, 1089–1096.
Janz, K. F., Levy, S. M., Burns, T. L., Torner, J. C., Willing, M. C., & Warren, J. J. (2002).
Fatness, physical activity, and television viewing in children during the
adiposity rebound period. The Iowa bone development study. Preventive
Medicine, 35, 563–571.
MacKinnon, D. P. (2008). Introduction to statistical mediation analysis. New York:
Lawrence Erlbaum and Associates.
Manios, Y., Kourlaba, G., Kondaki, K., Grammatikaki, E., Anastasiadou, A., & Roma-
Giannikou, E. (2009). Obesity and television watching in preschoolers in Greece.
The GENESIS study. Obesity, 17, 2047–2053.
Marsh, J. (Ed.). (2005). Popular culture, new media and digital technology in early
childhood. London: Routledge.
Muthén, B., & Muthén, L. (2010). Mplus user’s guide (sixth ed.). Los Angeles, CA:
Muthén & Muthén.
Vandebosch, H., & Van Cleemput, K. (2007). Television viewing and obesity among
pre-school children. The role of parents. Communications, 32, 417–446.
Viner, R. M., & Cole, T. J. (2005). Television viewing in early childhood predicts adult
body mass index. The Journal of Pediatrics, 147, 429–435.
Zimmerman, F. J., & Bell, J. F. (2010). Associations of television content type and
obesity in children. American Journal of Public Health, 100, 334–340.
forchildren0–5 years.<http://
with obesity and markers of
948
M. Fuller-Tyszkiewicz et al./Appetite 59 (2012) 945–948