ARTICLE IN PRESS

Appetite 49 (2007) 263–267

www.elsevier.com/locate/appet

Short communication

Beyond-brand effect of television (TV) food advertisements/

commercials on caloric intake and food choice of 5–7-year-old children
Jason C.G. Halforda,, Emma J. Boylanda, Georgina Hughesa,
Lorraine P. Oliveiraa, Terence M. Doveya,b
a
Kissileff Laboratory for the Study of Human Ingestive Behaviour, School of Psychology, Eleanor Rathbone Building, Bedford Street South,
University of Liverpool, Liverpool L69 7ZA, UK
b
Department of Psychology, Staffordshire University, College Road, Stoke-on-Trent ST4 2DE, UK
Received 9 October 2006; received in revised form 22 November 2006; accepted 11 December 2006

Abstract

Food advert exposure has been shown to influence calorie intake and food choice in 9–11 year olds. However, little is known about the
effect of food advertisements on feeding behaviour in younger children. Therefore, we conducted a study with 93 children aged 5–7 years,
28 of whom were over weight or obese. The children were exposed to 10 non-food adverts and 10 food adverts in a repeated measures
design. Their consumption of sweet and savoury, high and low fat snack foods, and fruit were measured following both sessions. Food
advert exposure produced a significant increase in total food intake in young children. The collection of recognition data was incomplete.
These data replicate previous findings in that exposure to food adverts increases food intake in all children, but recognition of food
adverts is related to body mass index (BMI). Beyond their effects on brand choice, exposure to food advertisements (commercials)
promotes over-consumption in younger children.
r 2007 Elsevier Ltd. All rights reserved.

Keywords: Young children; Food advertisements; Food choice; Intake

Introduction behaviour and caloric intake. Data from older children

Recently, data from a UK-based longitudinal study have
shown that TV viewing at 5 years of age is associated with
later adult body mass index (BMI) at the age of 30 years
(Viner & Cole, 2005). For each additional hour of weekend
TV watched, the risk of adult obesity was significantly
increased. These data are consistent with the results from
other prospective studies in children (e.g. Proctor et al., 2003).
Notably, TV viewing time is also associated with the number
of food items requested by children (Chamberlain, Wang, &
Robinson, 2006; Marquis, Filion, & Dafenais, 2005).
Numerous studies have shown that adverts are effective
at changing food brand preference in young children
(Borzekowski & Robinson, 2001; Brody, Stoneman, Lane,
& Sanders, 1981). However, few studies have examined the
consequence of food advert exposure on children’s feeding
Corresponding author. Tel.: +44 151 794 2952; fax: +44 151 794 2945.
E-mail address: j.c.g.halford@liverpool.ac.uk (J.C.G. Halford).
suggest that exposure to food adverts not only affects
product preference but has ‘beyond-brand’ influences on
food choice and total caloric intake (Halford, Gillespie,
Brown, Pontin, & Dovey, 2004). In our previous study we
found that exposure to TV food adverts exaggerated
already distinct patterns of food choice in children aged
9–11 years. After viewing food adverts the children
increased their intake of most but not all food types. The
ability to correctly recognise these food adverts was
significantly associated with higher food intake following
food advert exposure. It was also found that obese children
recognised a greater number, and a greater proportion, of
food TV adverts compared to non-food TV adverts. They
also recognised more TV food adverts than the normal
weight children.
In a recent study, researchers found that eating in front
of the TV did not produce over consumption in young
children aged 3–5, and in fact could inhibit intake in those
not used to simultaneously eating and viewing (Francis &

0195-6663/$ - see front matter r 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.appet.2006.12.003
 ARTICLE IN PRESS
264 J.C.G. Halford et al. / Appetite 49 (2007) 263–267
Birch, 2006). However, the TV material shown to these
children was devoid of any food cues. In the current study,
we decided to specifically examine the effect of food adverts
on intake and food preferences in a younger population.
We were interested in ascertaining if recent exposure to TV
food adverts had the same effect on subsequent total food
intake and food choice in younger children. We were also
interested to ascertain if any difference existed in the ability
to recognise food adverts between normal weight and
overweight/obese children at this age.
It was hypothesised that (i) both exposure to food
adverts and the child’s weight status would influence the
amount of food consumed during this study, (ii) obese and
overweight children would recognise more food-related
adverts than their normal weight counterparts, (iii) an
increased ability to recognise food-related adverts would be
significantly associated with the amount consumed after
food advert exposure.
Methods
Participants
Ninety-three young children (39 male, 54 female) aged
between 5 years 2 months and 7 years 9 months (mean age
6 years 3 months) were recruited from a UK school to
participate in the study. Two groups were created; lean and
overweight/obese, based on BMI converted to a standard
deviation (s.d.) score using the revised 1990 reference
standards (Cole, Freemans, & Preece, 1995). Sixty-five
children were lean (NW), 28 were overweight (n ¼ 13) or
obese (n ¼ 15) (OW). Although the mean standardised
BMI score was lower in 7 year olds (0.239) than in the 5
and 6 year olds (0.610 and 0.704), this difference proved
insignificant.
Materials
Television advertisements and advert recognition
Three videos were used: a cartoon, a collection of non-
food-related adverts, and a collection of food-related
adverts. Advertisements, 10 in number, were recorded
from children’s and family programming and previously
used in Halford et al. (2004). Ability to recognise the
adverts was assessed by identification of the product from a
list which was read out to the children. The list of 20
consisted of products that were either shown in the
advertisements or not. A cartoon was shown after the
adverts to simulate the normal context in which children
see food advertisements.
Foods and food intake measurement
The children were given the opportunity to select and eat
from an assortment of foods. The foods offered were, low-
fat savoury (snack-a-jacks, cheese flavour), low-fat sweet
(haribo jelly sweets), high-fat sweet (chocolate buttons),
high-fat savoury (walker’s ready salted potato crisps), and
fruit (green seedless grapes). In order to measure the food
consumption of each child, the food was weighed before
presentation and the leftovers remaining were re-weighed
post-consumption.
Procedure
Two weeks before the study the children were asked if
they wished to take part in the study and consent forms
were given to their parents. The study was a within-
subjects, counterbalanced design and the children were
tested on two occasions separated by 2 weeks. On each
occasion, the children were told that there would be some
adverts at the beginning of the cartoon. The same cartoon
was used on both occasions. After the 10 min cartoon, the
tape was stopped. The children were then divided into
groups of four or five and each child was presented with
pre-weighed individual servings of the five foods on one
plate. They were instructed that they could eat as little or as
much food as they liked and if they ran out of food the
experimenter offered them more. There was no time
constraint, and once the children had finished eating, the
remaining uneaten food was re-weighed. The children were
then read a list of 20 advertisements, 10 of which they had
actually seen. The children were read out loud the brand
names in the 20 adverts and instructed to answer ‘yes’ if
they recognised it and ‘no’ if they did not. After the second
session, weight and height measurements were taken
individually.
Analysis
All intake data collected adhered to the assumptions for
parametric data. Therefore analysis took the form of
analysis of variance (ANOVA, and MANOVA where
appropriate) and t-tests. The total kcal intake in each
condition was analysed as were the intakes of each of the
five foods. Flavour preference (sweet versus savoury, and
low versus high fat) was assessed by adding the kcal intake
of the foods which fitted into each of those categories (high
fat: crisps and chocolate, low fat: snack-a-jacks, jellies and
grapes, sweet: chocolate, jellies and grapes, savoury: crisps
and snack-a-jacks). Post hoc comparisons used appropriate
repeated measures or independent sample t-tests. One way
comparisons were used for predicted changes such as the
effects of condition or weight status on food preference. All
other comparisons were two tailed. The same applied for
post hoc correlations. All the analysis was completed using
SPSS v14 for pc (SPSS Inc., Chicago, USA).
Results
Total food intake
A main effect of advert exposure on total kcal intake was
found in the two-way mixed ANOVA (F(1,91) ¼ 20.273;
po0.001). Total kcal intake was significantly higher after
 ARTICLE IN PRESS
J.C.G. Halford et al. / Appetite 49 (2007) 263–267 265

exposure to food adverts (FA) than after the non-food of the various foods (F(1,91) ¼ 20.273; po0.001) (see
adverts (NA). There was no main effect of weight status on Table 1). The children’s level of intake of each food was
intake despite the fact that mean intake in the overweight also significantly associated with the type of food
and the obese children was higher in both experimental (F(4,88) ¼ 126.467; po0.001), and the effect of advert
conditions, nor was there any significant interaction with condition on the types of foods chosen by the children was
weight status (see Fig. 1). There was a significant positive near significant (F(4,88) ¼ 2.265; p ¼ 0.069). The same
correlation between BMI s.d. score and the amount eaten effects were found even if the mass of foods consumed
after exposure to the food adverts (r (93) ¼ 0.185, rather than their energy content were used in the analysis.
p ¼ 0.038 one tailed). With regard to energy intake, children consumed signifi-
cantly more calories from chocolate and jellies than from
Intake of individual foods crisps, significantly more calories from crisps than from
grapes, and significantly more calories from grapes than
The MANOVA (violation of sphericity) examining the from snack-a-jacks (all tests po0.001). Exposure to FA
effects of adverts and weight status on food choice significantly increased the kcal intake of haribo jellies
produced a main effect of advert on the kcal intake (t(92) ¼ 3.136; p ¼ 0.002) and snack-a-jacks (t(92) ¼ 5.336;
po0.001).

Amount of Food Eaten After Presentation of

Adverts Taste preference
800
Foods were combined into categories based on whether
700 they were sweet or savoury. They were also combined
into categories based on whether they were high or low in
600 fat. Children preferred sweet foods over the savoury
foods (F(1,91) ¼ 343.821; po0.001). Exposure to the FA
Amount eaten in kcal

500 significantly increased the total intake of all foods

(F(1,91) ¼ 42.455; po0.001). However, the interaction
400 between advert and taste preference showed that the
increase in intake produced by the FA varied
300 food eaten after (F(1,91) ¼ 343.821; po0.001). Specifically, the increase in
non-food adverts the intake of high fat foods produced by food advert
200 food eaten after exposure was 13%, low fat foods was 28%, sweet foods
food adverts
was 15% and savoury foods was 44% (all tests po0.05).
100
Advert recognition
0
se
t

Experience using the verbal recognition test suggested

gh

be
ei

t/o
w

that the younger children found the procedure difficult to

gh
al
rm

ei

follow. Data were gained from 27 of the children before the

w
no

er

procedure was dropped. However, a significant difference

ov

was found between OW (8.670.6) and NW children

Fig. 1. Illustrates the mean (7 SE) amount of food eaten by the normal
weight children and the obese and overweight children in the two (6.070.36) in FA condition (t(25) ¼ 3.154, p ¼ 0.004).
advertisement conditions. There was a significant and positive correlation between

Table 1
Shows the effects of advisement condition in the kcal intake of each food items in the normal weight children, the overweight and obese children and in all
the children

Kcal Non food advert Food advert

NW OW and OB All NW OW and OB All

Grapes 47.874.4 44.578.2 46.873.9 48.174.9 57.978.2 50.974.22

Jellies 187.7712.8 223.7731.3 198.2712.9 235.7717.1* 237.2728.1 236.2714.5**
Chocolate 225.9 718.7 238.0725.1 229.4715.1 256.1718.8* 264.67a36.2 258.6716.9*
Snack-a-jacks 0.970.2 1.070.4 0.970.2 25.475.6* 28.479.7 26.274.9***
Crisps 82.479.3 87.7713.0 83.9.77.6 91.679.8 104.0714.8* 95.278.1*

p ¼ o0:05 one tailed, 
p ¼ o0:01, 
p ¼ o0:001.
 ARTICLE IN PRESS
266 J.C.G. Halford et al. / Appetite 49 (2007) 263–267
BMI s.d. score and the number of FA correctly recognised
(r(27) ¼ 0.503, p ¼ 0.007) but not the number of NA
correctly recognised.
Discussion
Exposure to the food advertisements produced signifi-
cant and substantial increases in energy intake in the group
as a whole. The increase in caloric intake was roughly of
the same magnitude in the two weight status groups, 17%
(112 kcal) in the normal weight children and 14% (97 kcal)
in the obese and overweight children, and proves sig-
nificant in both. This replicates our previous findings in
older children (Halford et al., 2004). As far as we are
aware, this is the first study to demonstrate that exposure
to advertisements increases food intake in children of this
age. Previously adverts have been shown to have an effect
on young children’s subsequent attempts to influence
grocery shopping toward foods advertised (Brody et al.,
1981), and to alter brand preference more immediately
after exposure (Borzekowski & Robinson, 2001). Despite
the fact that in younger children TV viewing may interfere
with consumption (Francis & Birch, 2006), possibly due to
the suppressive effects of TV viewing on behaviour
(Klesges, Shelton, & Klesgas, 1993), the content of TV
can still have potent effects on eating behaviour.
In the control condition of our study, children showed a
distinct pattern of preference by consuming more sweet
foods than savoury. In terms of amount (i.e. grams), the
food items the children consumed the most were grapes,
followed by jellies, chocolate buttons, then the crisps and
finally the low fat savoury snack. However, due to the
differing macronutrient contents and energy densities of
these foods, the children consumed a far greater number of
calories from the chocolate buttons, the jellies and crisps
rather than from the grapes. Children consumed three
times the amount of sweet food items than the savoury
ones (even with the grapes removed from the analysis).
Savoury foods only contributed 15% to the total energy
intake. The children also ate near three times as much of
the low fat foods as the high fat ones, mainly due to the
consumption of grapes. In our previous study, older
children showed greater preference for savoury items,
specifically high fat ones, but avoided the low fat savoury
item.
The differential effect of FA exposure on the consump-
tion of various food items was reflected in the one near
significant and two significant interactions observed in the
MANOVA and ANOVA models for food type and taste
preference. The increase in energy intake was associated
with significant increases in the amount of the jellies (23%)
and the low fat snacks (65%) consumed. The increases in
chocolate button (11%) and crisp (12%) consumption were
not significant on two-tailed tests. Interestingly, there was
no evidence for an advert-induced increase in the only non-
branded food item, grapes (o1% difference between
means). It is notable that the intake of the snack-a-jacks
was very low in this study, especially in the control
condition. It is difficult to find low fat savoury snack foods
that children will consume a substantial amount of when
presented alongside sweet and/or high fat food items
(Halford et al., 2004). Moreover, while older children have
come across this particular food (and we have successfully
used this item in other studies in older children), younger
children are probably not familiar with them. There are
obvious limitations in using only four or five food items to
assess children’s habitual taste preferences and the effect of
advert exposure on them.
There were no effects of weight status on food intake in
this study. Moreover, exposure to adverts produced similar
changes in consumption in both weight status groups.
Many far larger studies have failed to demonstrate clear
weight status based differences in young children’s habitual
food intake (Rodriguez & Moreno, 2006). This study did
not control for the socio-economic status of the children
involved. Nor did it examine any relationship between
response to food adverts and the children’s habitual TV
viewing behaviour. The amount and type of TV program-
ming watched, and the amount and the differing types of
advert exposure these children regularly received is
probably critical in determining their responsiveness to
food advert exposure-induced increases in caloric intake.
To conclude, these data clearly demonstrate that
exposure to food adverts increases energy intake in young
children regardless of their weight status. While the
greatest increase in calories came from the intake of sweet
foods, advert exposure significantly increased the intake of
all study foods, with the exception of fruit. Critically, as
none of the study foods were the brands shown in the
adverts and were presented without any packaging, in this
situation, advert exposure clearly produced effects on
children’s feeding behaviour ‘beyond-brand’, producing a
net increase in consumption.
Acknowledgements
The authors would like to acknowledge Mrs. Lisa D.M.
Richards–Halford for help with this manuscript and Dr.
Clare Lawton (Institute for Psychological Studies, Uni-
versity of Leeds) for extensive comments on the draft.
References
Borzekowski, D. L. G., & Robinson, T. N. (2001). The 30-second effect:
An experiment revealing the impact of television commercials on food
preferences of preschoolers. Journal of the American Diabetes
Association, 101(1), 42–46.
Brody, G. H., Stoneman, Z., Lane, T. S., & Sanders, A. K. (1981).
Television food commercials aimed and children, family grocery
shopping, and mother–child interactions. Family Relations, 30,
435–439.
Chamberlain, L. J., Wang, Y., & Robinson, T. N. (2006). Does children’s
screen time predict requests for advertised products? Cross sectional
and prospective analysis. Archives of Pediatrics and Adolescent
Medicine, 160, 363–368.
 ARTICLE IN PRESS
J.C.G. Halford et al. / Appetite 49 (2007) 263–267
Cole, T. J., Freemans, J. V., & Preece, M. A. (1995). Body-mass index
reference curves for the UK, 1990. Archives of Disease in Childhood, 73,
25–29.
Francis, L. A., & Birch, L. L. (2006). Does eating during television viewing
affect preschool children’s intake? Journal of the American Dietetic
Association, 106, 598–600.
Halford, J. C. G., Gillespie, J., Brown, V., Pontin, E. E., & Dovey, T. M.
(2004). The effect of television (TV) food advertisements/commercials
on food consumption in children. Appetite, 42(2), 221–225.
Klesges, R. C., Shelton, M. L., & Klesgas, L. M. (1993). Effects of
television on metabolic rate—potentials implications for childhood
obesity. Pediatrics, 91, 281–286.
267
Marquis, M., Filion, Y., & Dafenais, F. (2005). Does eating while
watching television influence children’s food-related behaviours?
Canadian Journal of Dietetic Practice and Research, 66, 12–18.
Proctor, M. H., Moore, L. L., Cupples, L. A., Bradlee, Hood, M. Y., &
Ellison, R. C. (2003). Television viewing and change in body fat from
preschool to early adolescence: The Framingham Children’s Study.
International Journal of Obesity, 27, 287–833.
Rodriguez, G., & Moreno, L. A. (2006). Is dietary intake able to explain
difference in body fatness in children and adolescents? Nutrition,
Metabolisms and Cardiovascular Diseases, 16, 294–301.
Viner, R. M., & Cole, T. J. (2005). Television viewing in early childhood
predicts adult body mass index. Journal of Pediatrics, 147, 429–435.