T H E B Ä (L IT Kl P R© M© T 0d

Weight Control

Effects of TV in the Bedroom on Young

Hispanic Children
Du Feng, PhD; Debra B. Reed, PhD, RD, LD; M. Christina Esperat, RN, PhD, FAAN;
Mitsue Uehida, PhD

Abstract factors and an obesogenic environ-

Purpose. The main purposes of this study were to assess TV vietving among Hispanic young ment that encourages sedentary life-
children and to examine effects of having a ly in the child's bedroom (lYIB). styles and excessive food intake con-
Design and Setting. A quasi-experimental design was used to evaluate an intervention tribute to the increased prevalence of
program that is collecting longitudinal data in West Texas. Hoiuever, the current report uses overweight.^ Furthermore, Hispanics
only the baseline data of the ongoing study. may have other risk factors that con-
Subjects. Predominantly loiv-income and Hispanic parents/guardians (N = 315) and their tribute to childhood overweight, in-
children of 5 to 9 years (TV = 597). cluding parental obesity, low income,
Measures. Children's anthropométrie measures were obtained. Thär overweight status was acculturation to the U.S. lifestyle, and
determined based on age- and gender-adjusted body mass index. A demographic questionnaire, limited health care access.'' Various
acculturation scale (brief version of Acculturation Rating Scale for Mexican Americans II), government and health professionals'
and family survey were used among parents. groups have promoted a multicompo-
Analysis. Descriptive statistics, t-tests, x^ tests, and logistic regressions were used. nent approach to address childhood
Results. Most children (10%) have TVIB; more than 30% were or were at risii of overweight including increasing fruit
overweight. Demographic characteristics did not significantly predict TVIB. Children with and vegetable intake, increasing phys-
TVIB spent .93 hours more daily watching TV/DVD (t = 3.07; df= 283; p = .003), and ical activity, decreasing sedentary ac-
children (at one site) ate more fast food (x^ = 5.46; df= l;p= . 019), compared with children tivity, decreasing fast food intake, and
without TVIB whose parertts better supported physical activity (t = 2.11; df= 275; p = .039). increasing family meals together.''-''
Conclusion. Most children in this bw-income and Hispanic sample have 'TVIB, which is Sedentary activity may include "screen
associated with unhealthy behaviors. (Am J Health Promot 2011;25[5]:310-318.) time" activities such as watching TV
Key Words: Obesity, Overweight, Television, Child, Hispanic. Manuscript format: and videos or using the computer for
research; Research purpose: relationship testing, descriptive; Study design: quasi- games and homework.*"
experimental; Outcome measure: cognitive, behavioral, biométrie; Setting: family,
school; Health focus: nutrition, physical activity; Strategy: education, skill building/ TV Viewing and Its Effects on Children's
behavior change; Target population age: youth; Target population circumstances: Weight Status
low income level. West Texas, Hispanic A significant association between
screen time and prevalence of obesity
in children has been documented by
numerous studies.'"'"^ Some interven-
tions aimed at reducing TV viewing
Du Feng, PhD; Debra B. Reed, PhD, RD, LD; M. Christina Esperat, RN, PhD, FAAN; Mitsue have been effective. For instance, a
Uehida, PhD, are from Texas Tech University, Lubbock, Texas.
randomized, controlled trial found
Send reprint requests to Du Feng, PhD, Texas Tech University, Department of Human Development that third- and fourth-graders in the
and Family Studies, PO Box 41230, Lubbock, TX 79409-1230; [email protected]. intervention group had statistically
This manuscript was submitted September 30, 2008; revisions were requested September 30 and November 14, 2009; the significant decreases in TV viewing and
manuscript was accepted for publication November IS, 2009. in body mass index (BMI) relative to
Copyright © 2011 by American Joumal of Health Promotion Inc controls.'^ Most research has agreed
0890-1171/ll/$?.00 + 0 on the persuasive evidence for a
DOI: 10.42 78/ajhp. 080930-QUAN-228 relationship between heavy television
INTRODUCTION viewing and childhood overweight.
overweight among children, 2 to 19 Four hypothesized mechanisms
years, from all ethnic/racial groups was through which TV viewing may lead to
Childhood overweight has reached 15.5%. For Mexican-American boys childhood overweight have been iden-
epidemic proportions, with Hispanic and girls in this age range, the preva- tified: (1) lowering resting energy
youth disproportionately affected.' lence was 23.2% and 18.5%, respec- expenditure, (2) displacement of
In 2005 to 2006, the prevalence of tively. The interaction of genetic physical activity, (3) increased energy

310 American Journal of Health Promotion May/June 2011, Vol. 25, No. 5
intake from exposure to food advertis- simple effect' on obesity, as well as bigher risk of overweight and are
ing, and (4) eating while viewing interaction witb poverty and sedentary heavier consumers of TV at tbe ages of
leading to increased energy intake." lifestyle in affecting obesity. 9 years and older or under 5 years.
TV/video viewing was correlated with A longitudinal study based on a However, tbere are gaps in informa-
fast food consumption,'* as well as cobort of 3- to 4-year-old Anglo-Ameri- tion about the critical age between 6
energy intake and obesity.'® Evaluation can, African-American, and Hispanic and 8 years. In addition, altbougb
of food consumption during TV view- children found physical activity, TV individual studies have looked at a few
ing among third- and fifth-graders viewing, and dietary factors to be • mediators to obesity and TV viewing,
showed that 17% to 18% and approx- significant predictors of BMI.^'' Specifi- no study was found that simultaneously
imately 26% of total daily energy was cally, physical activity was negatively included other potentially important
consumed during TV viewing on associated with and TV viewing was mediators (fruit and vegetable intake,
weekdays and weekend days, respec- positively associated with BMI, and tbe pbysical activity, fast food intake, and
tively.'^" A cross-sectional Australian associations became stronger as tbe family meals) in Hispanic cbildren.
study showed that after adjustments for cbildren aged. Findings of this study
suggest that 6 to 7 years may be a critical Purpose of the Current Study
background and health behavior vari-
time wben TV viewing and physical In light of tbe gaps identified above
ables, children who watched TV for
activity may affect BMI. in previous research, tbe objectives of
more than 2 hours per day were
tbis study were to (1) assess tbe level of
significantly more likely to have one or Acculturation to the American life-
daily screen time among predominantly
more daily servings of high-energy style has been associated witb an
Hispanic young children of 5 to 8 years
drinks and to have one or more daily increase in less bealthy dietary and
in West Texas; (2) examine tbe effects
servings of savory snacks. Also, these physical activity bebaviors.^^'^^ Accul-
of baving TVIB on TV viewing and total
children were less likely to have two or turation was found to be a risk factor
screen time, sweetened beverage intake,
more daily servings of fruit or to for obesity-promoting bebaviors in-
vegetable and fruit intake, support for
participate in any organized physical cluding fast food consumption and
pbysical activity, family meals, fast food,
activity.'^' sedentary activities in a study witb
and tbe cbild's age- and gender-adjust-
sixth- and seventh-grade Asian-Ameri-
Ethnic Differences ed BMI status; and (3) identify signifi-
can and Hispanic youth living in
Ethnic differences in TV viewing and cant demograpbic covariates of TVIB.
California.^'' Similarly, a national study
its relationship to energy intake and Tbe main hypothesis for this study was
of Hispanic adolescents reported that
energy expenditure have been docu- that Hispanic children witb TVIB have
first-generation Mexican, Puerto Ri-
mented. For example, sixth- and sev- an increased TV viewing and total
can, and Cuban adolescents ate
enth-grade Latino students reported screen time, bigber sweetened beverage
bealthier foods compared with second-
more dme watching television and intake, lower vegetable and fruit intake,
generation adolescents and that ado-
more consumption of soft drinks than lower parental pbysical activity support,
lescents who were born in the United
did non-Hispanic white or Asian stu- fewer family meals, and more fast food.
States engaged in more sedentary
dents.^^ Another study found that the It was also bypotbesized tbat these
activities compared with their foreign-
average time 1- to 4-year-old children children witb TVIB are more likely to be
born counterparts.'^'^
spent watching TV/video was bigber overweigbt or at risk of overweight
among black children and Hispanic TV Set in the Bedroom based on BMI adjusted for age and
children compared with white chil- Witb tbe average family owning four gender.
dren.^'' The 2003 National Survey of working TV sets, more than two-thirds
Children's Health using preteenage of school-aged cbildren have a TV in METHODS
children and adolescents reported the bedroom (TVIB)." Several studies
higher obesity prevalence among those suggest that children witb TVIB watch Design
wbo bad an ethnic minority status, significantly more television than those Transformación Para Salud is an
nonmetropolitan residence, lower so- witbout.^^'^*"™ Furthermore, baving ongoing intervention study begun in
cioeconomic status, higher television TVIB seems to impact a cbild's BMI 2007 tbat is aimed at prevention and
viewing, and higher physical inactivity status independently. For instance, an control of childhood overweigbt
levels.^'' Specifically, tbe odds of obesity evaluation study sbowed that tbe odds among predominantly Hispanic young
were 2.7, 1.9, and 3.2 times bigber for ratio of having a BMI bigher than tbe cbildren of 5 to 8 years in two urban
tbe poor Hispanic, white, and black 85tb percentile was 1.31 among pre- cities in West Texas. Tbis multisite and
cbildren, respectively, wben compared scbool cbildren who had TVIB, com- multicomponent longitudinal study
with white children of bigber income. pared with those with no TVIB, even used a quasi-experimental design, in
Hispanic, wbite, and black cbildren after statistical adjustment for the wbicb the intervention and compari-
wbo watched television 3 bours or child's age, gender, and TV/video son groups were matched on school-
more per day bad 1.8, 1.9, and 2.5 viewing hours per week, as well as the level demographic cbaracteristics. Five
times bigher odds of obesity, respec- mother's BMI, education, and race/ data collections were conducted at five
tively, than white children who ethnicity.^"* scbools before, during, and after an 18-
watched television less than 1 hour per Thus, previous research has identi- month intervention program that in-
day. Thus, ethnic minority status has a fied that Hispanic cbildren may be at cluded nutrition education, physical

American Journal of Health Promotion May/June 2011, Vol. 25, No. 5 311
 activity, and gardening was imple- Family Survey. Demographic and so- children's physical activity was not
mented. cioeconomic data such as age, gender, determined. Instead, based on the
race, occupation status, marital status, significant positive relationship re-
and highest achieved educational de- ported for parental support of physical
Sample
gree were collected at baseline from a activity and children's physical activi-
The sample of the current study is self-administered family survey com-
based on the baseline data for both ty,''"'" parental support of physical
pleted by participating parents. It is activity was used to assess the home
intervention and comparison schools. generally recognized that children
Briefly, anthropométrie data were col- environment, which can encourage or
under 8 years cannot accurately report discourage physical activity. A scale
lected from 597 kindergarten through food intake or participate in stirvey
second-grade students in the interven- score was calculated as the mean of
research'^'*'^''; thus, for the current parent's responses to eight items of the
tion and comparison schools (n = 332 study, parents provided the informa-
[56%] for intervention children; n = family survey that relate to parental
tion related to all reported variables. support of physical activity. These
265 for comparison children), and 315
In the development of the family items were adapted from the Cirls
completed self-report surveys were re-
survey (which included questions re- Health Enrichment Multi-Site Studies"^
turned by parents of these children. The
lated to TV, parental support for and included such questions as "When
study sample for main hypothesis test- physical activity, and food intake), brief
ing consisted of 314 parent-child dyads the weather is nice, I go for a walk with
targeted questions previously used my child" and "I do something phys-
in which the parent returned the survey and/or validated in other studies were
and self-reported to be Hispanic. More ically active each week with my child."
used based on recommendations that Items were rated on a four-point Likert
than 70% of eligible students partici- longer questionnaires may not be valid
pated in the study, which was approved scale (0 = never, 3 = always); there-
in Hispanic audiences.^" Specifically, fore, a higher scale score indicated a
by the Internal Review Board of Texas the TV/screen-related variables,^''"'"'
Tech University Health Sciences Center. higher level of parental support of
family meals,''^ and fast food'** ques- physical activity. In the study sample,
Consent to participate was obtained tions for the current study were pat-
from all parents, and assent from the the internal consistency reliability
terned after those studies cited. (Cronbach ot) was .73, and a single-
children. See the Results section for The family survey was reviewed for
detailed demographic characteristics of factor solution was supported by a
face validity by university faculty in principle component analysis (37% of
the current sample. nutrition, psychology, and human de- variance was explained by the single
velopment and by teachers who were factor, and the factor loadings ranged
parent coordinators in their schools. from .39 to .73).
Measures
To improve face validity with the target
Anthropométrie Measures. Standardized population, the family survey was pre- Food Intake. To capture total sweetened
anthropométrie measurements were tested with 37 Hispanic low-income beverage intake and thus overcome the
performed on students without shoes mothers whose children were similar in lack of completeness in other studies
or jackets. Subjects' heights were mea- age to the sample in this study.''*' Based that collected just soda intake,'^''^''''•''''
sured to the nearest .1 cm in triplicate on the pretesting, some wording/for- the family survey asked "How many
using a stadiometer. Weight was mea- mat changes were made, and examples ounces of sweetened beverages does
sured on an electronic scale accurate related to the response options were your child drink in a typical weekday
to .1 kg (Tanita Body Composition provided. (Monday, Tuesday, Wednesday,
Analyzer TBF 300A, Tanita Corpora- Thursday, Friday)?" and the same
tion of America, Inc, Arlington TVIB and TV Viewing. The presence of a questions for "a typical weekend day."
Heights, Illinois). The height and TV in the participating child's bed- Parents recorded the number of
weight values were used to calculate room, as well as the presence of a TV in ounces separately for soda, fruit drink,
BMI.^' Age- and gender-adjusted BMI someone else's bedroom, was mea-
sports drink, tea, and lemonade, and
percentiles were obtained using a sttred by a family survey completed by
these values were then stimmed to
standardized online calculator.^^ BMI the parents. Parents also reported on
yield intake for either weekend or
percentiles from the 2000 growth the participating child's screen time,
weekday.
charts from the Centers for Disease based on which an average daily TV/
DVD time and an average daily total One question each was used to
Control and Prevention were used to
screen time (including computer, In- estimate intake of fruits and vegetables
determine children who were normal
ternet, and video games) were com- based on previously validated questions
(BMI percentile <84tb), at risk of
puted. in a multiethnic sample of youth (N =
overweight (BMI percentile falls be-
138) who were 23% Hispanic."'^ For our
tween 85th and 94th), and overweight
Parental Support of Physical Activity. study, we changed the word "servings"
(BMI percentile >95th).'''' BMI status
Because of the difficulty associated to "cups" to reflect the newer MyPyr-
was further collapsed into two catego-
with children's physical activity assess- amid terminology.""^ Parents were
ries—normal vs. BMI percentile great-
ment when reported by either children asked to provide the number of cups
er than or equal to 85th percentile—to
or their parents'*'' and the respondent eaten each day. These two questions
obtain an adequate subgroup size in
burden already present in this multi- included written tips to increase accu-
the multivariate analyses.
component study, direct assessment of racy; for example, one piece of fresh

312 American Journal of Healfh Promofion May/June 2011, Vol. 25, No. 5
fruit equals 1 cup. Parents were also
prompted to include fresh, frozen,
dried, canned, and 100% juice in their Table 1
responses. Characteristics of the Children (N = 314)*
Based on the participating parent's
Characteristic Value
answers, the average daily sweetened
beverage intake, average daily fruit and Children's sex. No. of girls (%) 175 (56)
vegetable intake, frequency of having Children's age, y, mean ± SD (range) 6.61 ± 0.889 (5-9)
family meals (dummy coded into ev-
Children's grade. No. (%)
eryday vs. not every day), and number
Kindergarten 62 (20)
of times the child eats food from a fast
First grade 156 (50)
food restaurant per week were calcu- Second grade 96 (31)
lated.
BMI status. No. (%)
Acculturation. The Acculturation
<85th 210(68.4)
Rating Scale for Mexican Americans
85th-94th 38 (12.4)
II (ARSMA-II)''' is a well-known >95th 59 (19.2)
measure to assess acculturation. An
abbreviated version of this scale. Brief * BMI indicates body mass index.
ARSMA-II,'"* which has been shown to
have adequate reliability and validity,
was used in the current study. Twelve Analysis and education level, child's average
items measuring language, ethnic Our analysis plan included basic daily intake of sweetened beverages
identity, and ethnic interaction descriptive statistics of demographic and fruits and vegetables, family meals,
formed two subscales; a Mexican characteristics of the participating fast food, and parental support for
orientation subscale (MOS) and an children and their parents, as well as physical activity. All analyses were
Anglo orientation subscale (AOS). children's BMI status and parents' conducted using SPSS version 16.0
Subjects responded on a five-point response to various family survey vari- (SPSS Inc, Chicago, Illinois)."^
Likert scale (1 = not at all, 5 = ables regarding home environment,
almost always/extremely often). food behaviors, and parental support
Examples of items on the MOS are "I for physical activity. Independent t- RESULTS
speak Spanish," "I enjoy Spanish tests were used to compare group
language TV," "My thinking is done means of the TVIB group with children Descriptive and Preliminary Analyses
in Spanish." Examples of items on who do not have TVIB in terms of the Descriptive statistics were obtained
the AOS are "I speak English," "I family survey variables (e.g., screen for all background variables. Table 1
write letters in English," "My friends time, sweetened beverage intake, fruit shows the demographic characteristics
are of Anglo origin." For each and vegetable intake, and parental of the children. There were slightly
subject, an MOS subscale score and support for physical activity). Bivariate more girls and more children in first
an AOS subscale score were correlations (Pearson's r) between and second grades than kindergarten
calculated by averaging the six these family survey variables were in the sample. BMI assessment showed
corresponding items. A linear calculated. Chi-square tests were per- that 32% of the children were either at
acculturation score was derived by formed to evaluate the association risk of overweight or overweight.
subtracting the MOS mean from the between TVIB and key categorical Table 2 presents descriptive statistics
AOS. Furthermore, acculturation variables such as children's BMI status about demographic characteristics of
typologies were generated by and to-assess the significance of asso- the participating parents. Most parents
comparing respondents' MOS and ciations between TVIB and children's responding were female and married,
the AOS subscale scores with and parents' background characteris- did not have college educations, had
respective recommended points.''^ tics such as gender, marital status, and very low income, participated in at least
Each respondent was assigned into acculturation type. Finally, two binary one food/nutrition assistance pro-
one of the five acculturative logistic regressions were used to test gram, and were "traditional" in accul-
categories; traditional, high integrated potential covariates and effects of turation status. Most families ate to-
bicultural, marginal (low bicultural), TVIB. The first logistic regression used gether everyday, and more than 67% of
assimilated, and unidentified. TVIB as the dependent variable and families ate fast food at least once per
Consistent with the literature, a two- child's gender and age and parent's week. Other descriptors of the sample
factor structure (MOS and AOS) was marital status, income, education level, population, not included in Table 2,
found in the current study sample and acculturation type as covariates. are 71% of children had TVIB and
(70% of variance was explained by The second logistic regression used a more than 86% of children watched
the two factors extracted, and target dichotomized variable for children's TV in a bedroom (either their own or
factor loadings ranged from .64 to BMI status as the dependent variable someone else's bedroom) at home.
.94). Cronbach a was .90 for both and TVIB as the main predictor, while Before the hypotheses were tested
MOS and AOS subscales. controlling for parents' marital status regarding effects of TVIB on behav-

American Journal of Health Promotion May/June 2011, Vol. 25, No. 5 313
 significantly associated with the TV
viewing variables nor were the dietary
Table 2 variables significantly correlated with
Characteristics Of the Participating Parent (N = 314)
each other. Interestingly, in a series of
correlations between tbe TV viewing
Characteristic Value
variables and dietary variables, the
Parent/guardian gender. No. of females (%) 277 (91.4) daily intake of sweetened beverages was
Parent/guardian age, y, mean ± SD (range) 32.49 ± 7'.14(15-61) significantly associated with daily TV/
DVD time (r = .15; p = .012) and with
Relationship to child. No. (%)
the average TV/DVD time per weekday
Mother 281 (89.5) (r = .18; p = .003) but not with the
Father 27 (8.6) average TV/DVD time per weekend
Grandmother 4 (1.3)
day, after correction for the hypothesis-
Other relative 2 (0.6)
wise type I error rate. Additionally, the
Marital status. No. (%) daily intake of fruits and vegetables was
Single, never married 53 (17.7) positively correlated with the physical
Married 182 (60.7) activity support variable (r = .29; p <
Divorced/separated 60 (20.0) .001) after the correction; however, it
Widowed 5 (1.7) was not significantly correlated with TV
Education, No. (%) viewing variables, although the r be-
Some high school or less 132 (44.6) tween fruits and vegetables intake and
High school diploma 83 (28.0) the average TV/DVD time per week-
Some college or higher 81 (27.4) end day was significant (r = .14; p =
Annual household income. No. (%) .019) before the Rom procedure was
$0-$15,000 113 (40.4) used.
$15,001-$30,000 107 (38.2)
$30,001-$45,000
Next, potential covariates of having a
24 (8.6)
$45,001-$60,000
TVIB were examined. Chi-square sta-
13 (4.6)
$60,001-$75,000 4 (1.4)
tistics obtained in testing whether
>$75,000 19 (6.8) child's gender, parents' marital status
(married vs. not currently married),
Acculturation type. No. (%)
and highest achieved educational de-
Traditional 117 (39.1) gree were associated with either having
Low bicultural 48 (16.1)
High bicultural
a TVIB or TV in someone else's
41 (13.7)
Assimilated bedroom were not significant. Similar-
38 (12.7)
Unclassified 55 (18.4) ly, the binary logistic regression used to
identify covariates of TVIB did not
Family meal. No. (%)
yield significant findings, regardless of
Every day 176 (61.3) whether the parent's acculturation
Few times per week 90 (31.4) scores (i.e., AOS and MOS) or the
Once per week 12 (4.2)
acculturation category ("assimilated"
Less than weekly 5 (1.7)
Never 4 (1.4) was the referent category) were in-
cluded as one of the independent
Fast food. No. (%)
variables (restilts not shown).
Every day 1 (0.3)
Few times per week 59 (20.6)
Once per week 134 (46.7) Hypothesis Testing
Less than weekly 76 (26.5)
Never 17 (5.9) Effects of TVIB on Diet and Parental
Participating in food/nutrition assistance program(s) 223 (72.6) Support of Physical Activity. Activity-
related and dietary variables from the
family survey were examined by having
TVIB and not having TVIB status.
ioral and physiologic outcomes, bivar- variables) and in testing correlations Table 4 presents the group means and
iate correlations (Pearson's r) were between the two categories of variables. standard deviations, as well as group
examined for the various family survey Not surprisingly, correlations between mean differences and ¿-test results on
variables (Table 3). Because multiple the average daily TV/DVD time over these study variables. As seen in
pairwise correlations were computed, the week, per weekday, and per week- Table 4, significant mean differences
the Rom procedure^" was used to end day were positive and significant were found for the activity-related
control the hypothesis-wise type I error (r = .54-.92; p < .001) after the Rom variables (i.e., children witb TVIB tend
rate within each variable category (i.e., procedure was used. However, parental to spend more time watching TV/
activity-related variables and dietary support for physical activity was not DVDs on a daily basis and have less

314 American Journal of Health Promotion May/June2011, Vol. 25, No. 5

 Table 3
Bivariate Correlation (/) Among Family Survey Variables (N = 314)

2. Average 3. Average 4. Parentai 5. Daiiy 6. Daily Fruit

TV/DVD Time TV/DVD Time Support of Sweetened and Vegetable
per Weekday, per Weekend Physlcai Beverages, intake.
h Day, h Activity Ounces Cups
Activity-related variables
1. Daily TV/DVD time, h 0.92*** 0.82*** 0.02 0.15** 0.12
2. Average TV/DVD time per weekday, h 0.54*** -0.04 0.18** 0.06
3. Average TV/DVD time per weekend day, h 0.06 0.11 0.14*
4. Parental support of physical activity 0.01 0.29**
Dietary variables
5. Daily sweetened beverages, ounces -0.04
6. Daily fruit and vegetable intake, cups
* p < 0.05.
**p< 0.01.
***p< 0.001.

support from their parents for physical bock, Texas, with TVIB ate fast food whether the child's BMI percentile was
activity when compared with children meals more frequently than peers with at or below the 84th percentile or was
without TVIB) but not for the dietary no TVIB (84% among the former and higher. This dichotomous variable was
variables. For instance, children with 59% among the latter, respectively, then used in a binary logistic regres-
TVIB had an average of 3.51 hours of had fast foods at least once per week; sion, testing whether TVIB, parents'
daily TV/DVD time, compared vnih X' = 5.46; df=\,p= .019). marital stattis, parent's education level,
2.58 hours of daily TV/DVD time parent's acculturation scores, parent's
among children without TVIB (t = Effects of TVIB on Children's report on the child's average daily
3.07; df= 283; p = .003). The four t- Wei^t Status. The x^ test for the intake of sweetened beverages and
tests on activity-related variables were association between having TVIB and fruits and vegetable, having at least one
significant after correction for multiple the participating child's BMI status family meal everyday, having at least
tests using the Rom procedure,^" but (normal, at risk of overweight, and one fast food meal per week, and
the dietary variables were not signifi- overweight) was not significant (results physical activity support were signifi-
cant (Table 4). not shown). Because of the relatively cant correlates of the child's dummy-
Chi-square statistics obtained for the small sample size of the at risk of coded BMI status. The omnibus tests of
association between having TVIB and overweight and overweight groups, the model coefficients was not significant
frequency of family meals did not yield child's BMI status was dummy-coded (p = .283), and no individual inde-
significant findings (results not into two categories, normal vs. at risk of pendent variables included in this
shown). However, children in Lub- overweight/overweight, indicating analysis were significant.

Table 4
Means and SD of the Family Survey Variabies (N = 314)t

Fuil Sampie, TViB, No TViB, iVIean

mean (SD) mean (SD) mean (SD) Difference f P
Activity-related variables
Daily TV/DVD time, h 3.37(2.16) 3.51 (2.21) 2.58(1.69) 0.93 3.07** 0.003
Average TV/DVD time per weekday, h 2.87 (2.05) 2.97(2.10) 2.29 (1.70) 0.67 2.21* 0.031
Average TV/DVD time per weekend day, h 4.98 (3.40) 5.15 (3.46) 3.81 (2.73) 1.34 2.65* 0.011
Parental support of physical activity 1.42 (0.54) 1.40 (0.56) 1.55 (0.42) -0.16 -2.11* 0.039
Dietary variables
Daily sweetened beverages, ounces 19.71 (15.79) 20.06(16.25) 18.12 (12.87) 1.94 0.84 0.406
Daily fruit and vegetables intake, cups 2.82 (1.71) 2.79(1.68) 3.05(1.85) -0.26 -0.83 0.411
* p < 0.05.
* * p < 0.01.
t TVIB indicates TV in child's bedroom.

American Journal of Health Promotion May/June 2011, Vol. 25, No. 5 315
DISCUSSION Tbe finding tbat cbildren witbout dramatically increased recently.^®
TVIB bad greater parental support of Comparing 2004 witb 2007, tbere were
Tbe purpose of tbis study was to physical activity bas not been reported 135% more ads for fruit and vegetable
assess TV viewing among Hispanic previously. This measure merits more juices and more tban 400% more ads
cbildren in kindergarten tbrougb sec- investigation because of its ease of use for fruits and vegetables on children's
ond grades and to examine tbe effects and its potential contribution as an TV.'''' A recent report on European
of baving TVIB on BMI and dietary intermediate variable (i.e., does pa- scbool cbildren sbowed tbat exposure
and pbysical activity support variables. rental support for pbysical activity to TV ads for fruits and vegetables was
Tbe bypothesis tbat Hispanic cbildren need to be present before cbildren's associated witb tbeir intake.'^'^
witb TVIB bave an increased TV pbysical intake is increased?). Although the negative effect of TVIB
viewing and total screen time, more Tbe association found in our study on overweight risk has been reported
sweetened beverage intake, less vege- between baving TVIB and fast food in existing literature,^^'^^ no association
table and fruit intake, lower parental intake bas not been previously report- was found in our study. Tbe lack of
pbysical activity support, fewer family ed. More tban 67% of cbildren in tbis association of TVIB, BMI, and accul-
meals, and more fast food was only study ate fast food meals at least once turation refiects tbe complexity of tbe
partially supported, with significant per week (80% in Lubbock, Texas; issue of acculturation.''*'*'' Accultura-
results shown for increased TV and 69% in El Paso, Texas) compared witb tion may be protective of overweigbt if
screen time, lower parental support of 58% reported for Mexican-American acculturated cbildren are more pbysi-
physical activity, and increased fast children 8 to 18 years in California.'''' A cally active (e.g., participate more on
food intake (at one site). The hypoth- recent study with Mexican-American sports teams).'''' However, increased
esis tbat cbildren witb TVIB are more children 4 to 7 years showed tbat acculturation may be detrimental in
likely to be overweigbt or at risk of families who ate most often at fast food terms of increased intake of fast food.^^
overweigbt was not supported. restaurants were more likely to be at As tbe prevalence of overweigbt
risk of overweigbt (odds ratio, 2.2; 95% increases, tbe need to reduce seden-
Because results in publisbed studies
confidence interval, 1.2-4.3).'*' How- tary behaviors and to promote a more
are not uniformly reported for various
ever, our study did not find a signifi- active lifestyle becomes essential. Re-
age, gender, and racial/etbnic groups,
cant association between fast food and stricting TV is considered to be one of
and different measures are used, com-
overweigbt. tbe most easily modifiable bebaviors to
parisons across studies are difficult. In
a national sample of Mexican-Ameri- In our study, cbildren with TVIB reduce overweigbt in cbildren,'''''''^ al-
can cbildren, 32.6% of children 2 to 5 consumed more sweetened beverages tbougb it is recognized tbat TV watcb-
years and 42.9% of children 6 to 11 and fewer fruits and vegetables tban ing sbould not be tbe only targeted
years were at risk of overweight/over- cbildren without TVIB, but results were bebavior. Tbe effect of TV in general
weight. In a recent study witb Hispanic not statistically significant (Table 3). on pbysical activity is considered to be
kindergarteners to second-graders in Consistent witb otber studies,'''''^^ small and may lack meaningful clinical
California, 27% were overweigbt and sweetened beverage intake was corre- relevance, based on a meta-analysis of
46% were at risk of overweight/over- lated to TV/DVD time using several 52 studies.'''' Single markers of inactiv-
weight."*' Thus, the lower prevalence variable formats, total and weekday, ity, sucb as TV viewing or video/
found in tbis study of 32% at risk of but not weekend time. computer game use, may be limited in
overweight/overweigbt may bave con- Tbe relatively bigb correlation be- tbeir ability to sbow a relationsbip
tributed to tbe limited number of tween fruit and vegetable intake and between sedentary bebavior and
significant relationships found among parental support of pbysical activity bealtb. Even so, a recent expert review
study variables. possibly refiects a bome environment of the evidence about tbe role tbat
For our sample, 70% of the children tbat encourages bealthy bebavior clus- media, specifically television, plays in
had TVIB compared with 43% of 4- to ters. Healtby bebavior clustering of tbe prevalence of overweight among
6-year-oldsä2 and 69% of 8- to 10-year- pbysical activity and intake of fruits and children recommended tbe elimina-
olds^* in national samples. As expect- vegetables bas been reported in ado- tion of TV from children's bedrooms.'^
ed, having TVIB was associated with an lescents^'' and young adults'"'' and In tbis study of young Hispanic
increase in daily total screen time. Tbe reinforces tbe recommendation tbat cbildren, recommendations to remove
mean total daily screen time of family efforts to prevent/reduce cbild- TVs from cbildren's bedrooms is war-
3.4 bours (3.5 for TVIB and 2.6 for no bood overweight include a multicom- ranted as an important strategy for tbe
TVIB) in tbis study is bigber tban tbe ponent intervention witb botb diet and prevention and control of overweigbt,
2.8 bours of TV time reported by a pbysical activity."^ owing to its association witb increased
Mexican-American sample of youth 8 Tbe low positive bivariate r between total TV/DVD time and decreased
to 18 years in California.'*'' It is similar average TV/DVD time on weekend parental support for pbysical activity.
to the approximately 4 hours of daily days and fruit and vegetable intake was Witb the change in technology to
screen time reported for a national significant before, but not after, tbe high-definition TVs, parents may be
sample of Hispanic youth 8 to 18 Rom adjustment for multiple tests. replacing their current TVs with more
years''^*' and is well above the 2 bours or Nonetheless, tbis relationsbip sbould updated models. Thus, now would be
less a day recommended by the Amer- be explored furtber because TV ads an opportune time to educate parents
ican Academy of Pediatrics (AAP) .^^ featuring fruits and vegetables bave on the value of having only one TV per

316 American Journal of Health Promotion May/June2011, Vol. 25, No. 5

household (this would save money and The generalizability of our study 4. Barlow SE, Expert Committee. Expert
possibly contribute to a healthier home results can only extend to kindergarten committee recommendations regarding the
prevention, assessment, and treatment of
environment). The education message through second-grade students of His- child and adolescent over'weight and
also should include the AAP's recom- panic ethnicity. Further, the cross-sec- obesity: summary report. Pediatrics.
mendation for 2 hours or less of TV tional design does not allow for deter- 2007;120:164-192.
per day'"'' as well as nutrition mination of causation but does provide 5. American Dietetic Association. Position of
the ADA: individual-, family-, school-, and
recommendations for children's fruit some insight into factors that may be community-based interventions for
and vegetable intake,''® strategies for important to include in experimental pédiatrie overweight. J Am Diet Assoc.
increasing the quality and quantity of studies related to TV/screen time and 2006;106:925-945,
family meals,'''' and activities for health variables and outcomes. 6. Must A, Tybor DJ. Physical activity and
sedentary behavior: a review of longitudinal
families to do instead of watching studies of weight and adiposity in youth,
So What? Implications for Health IntJ Obes Retat Metab Disord. 2005;29:84-96,
Some strengths of this study are that Promotion Practitioners 7. Gortmaker SL, Must A, Sobol AM, et al.
the variable of TVIB was explored in Television viewing as a cause of increasing
and Researchers obesity among children in the United
relation to diet and parental support of What is already known on this topic? States, 1986-1990. Arch Pediatr Adoks Med.
physical activity in a high-risk sample of The relationship between chil- 1996;150:356-362,
young, low-income Hispanic children dren's TV/other screen time activi- 8. Gable S, Chang Y, Krull JL. Television
for whom little information is available. watching and frequency of family meals are
ties, diet, physical activity, accultur- predictive of overweight onset and
BMI was measured, not self-reported, ation, and weight is a complex issue persistence in a national sample of school-
and data on several aspects of diet and where conclusive results across aged children. / A m Diet Assoc.
physical activity support were collected studies are not always found. 2007;107:53-61,
in recognition of the multifaceted 9. Vandewater EA, Shim M, Caplovitz AG.
What does this article add? Linking obesity and activity level with
nature of childhood overweight. This study found that having a TV children's television and video game use,
There are several limitations to in tbe bedroom was related to fast JAdotesc. 2004;27:71-85,
consider when interpreting the results. food intake and lack of parental 10. Utter J, Neumark-Sztainer D,Jeffery R, Story
First, with the exception of BMI, all M, Couch potatoes or French fries: are
support of physical activity, but not sedentary behaviors associated with body
other measures were self-reported. body mass index, in" young low- mass index, physical activity, and dietary
Parents with children who have TVTB income Hispanic children for whotn behaviors among adolescents?/Am Diet
are known to underreport.'® A general little information has been previ- Assoc. 2003;103:1298-1305.
limitation of the field at this time is 11. Kaur H, Choi WS, Mayo MS, Hards KJ,
ously reported. Duration of television watching is associated
that there are no standardized mea- What are the implications for health with increased body mass index, / Pediatr.
sures for most of the mediating vari- promotion practice or research? 2003;143:506-511,
ables (sweetened beverage intake, fast Health intervention messages 12. Mendoza JA, Zimmerman FJ, Christakis DA.
food intake, and family meals). For Television viewing, computer use, obesity,
should encourage parents to elimi- and adiposity in US preschool children,
example, some studies use "soft nate TV/screen activities in chil- Intf Behav Nutr Phys Act. 2007;4:44.
drink" as the only indicator of sweet- dreri's bedrooms, and provide 13. ReillyJJ, Armstrong J, Dorosty AR, et al.
ened beverages and thus miss other strategies for parents that support Early life risk factors for obesity in
sweetened beverage contributors such childhood: cohort study. BMf.
children's physical activity with in- 2005;330:1357.
as sweet tea and fruit-type drinks. Some creased opportunities to meet daily 14. Hancox RJ, Poulton R. Watching television
studies have asked respondents to recommendations. Preliminary is associated with childhood obesity: but is it
report how many cans of soft drink study findings of a low positive clinically important? Intf Obes.
they consumed,^^'''' thus not account- association between intake of fruits 2006;30:171-175.
15. Coon KA, Goldberg J, Rogers BL, Tucker
ing for the intake of soft drinks poured and vegetables and TV watching on KL. Relationships between use of television
from large containers or consumed as weekends needs to be further during meals and children's food
fountain drinks. This makes explored as TV ads for fruits and consumption patterns. Pediatrics.
comparison across studies difficult. vegetables have increased. 2001;107:e7,
16. Robinson JL, Winiewicz DD, Fuerch JH, et
Another limitation of this study was al. Relationship between parental estimate
that data were not collected related to and an objective measure of child television
the accessibility of TV during meals watching. IntJ Behav Nutr Ptiys Act.
References 2006;3:43.
and snacks. Four in 10 families have a
1. Ogden CL, Carroll MD, Flegal KM. High 17. Jordan AB, Robinson TN. Children,
TV in the dining room or kitchen,'''' body mass index for age among US television viewing, and weight status:
and one-third to two-thirds of chil- children and adolescents, 2003-2006. summary and recommendations from an
dren's snacks and one-third of their JAAÍA. 2008;299:2401-2405. expert panel meeting. Ann Am Acad Pot Soc
dinners are eaten while watching tele- 2. Allison DB, Matz PE, Pietrobelli A, et al. Sd. 2008;615:l 19-132,
Genetic and environmental influences on 18. Taveras EM, Sandora TJ, Shih MC, et al.
vision.^" Moreover, children in house- obesity. In: Bendich A, Deckelbaum RJ, eds. The association of television and video
holds who consume their family meals Primary and Secondary Preventive Nutrition. viewing with fast food intake by preschool-
with the TV on report eating more red Totowa, NJ: Humana Press; 2001:147. age children. Obesity (Sitver Spring).
meat, pizza, snack foods, and soft 3. Kumanyika SK. Obesity treatment in 2006;14:2034-2041.
minorities. In: Wadden T, Stunkard A, eds. 19. Crespo CJ, Smit E, Troiano RP, et al.
drinks, and fewer fruits and Handbook of Obesity Treatment. New York, NY: Television watching, energy intake, and
vegetables.'® Gullford Press; 2002:416-446. obesity in US children: results from the

American Journal of Healfh Promofion May/June 2011, Vol. 25, No. 5 317
 Third National Health and Nutrition children and adolescents. Brf Nutr. 51. Duerksen SC, Elder JP, Arredondo EM, et
Examination Survey, 1988-1994. Arch 2004;92:213-222. al. Family restaurant choices are associated
Pediatr Adolesc Med. 2001 ; 155:360-365. 35. Borgers N, de Leeuw E, Hox J. Children as with child and adult overweight status in
20. Matheson DM, Killen JD, Wang Y, et al. respondents in survey research: cognitive Mexican-American families. / Am Diet Assoc.
Children's food consumption during development and response quality. Bull 2007;107:849-853.
television viewing. AmJ Clin Nutr. Methodot Sociol. 2000;66:60-75. 52. Henry J. Kaiser Foundation. The media
2004;79:1088-1094. 36. Wakimoto P, Block C, Mandel S, Medina N. family: electronic media in the lives of
21. Salmon J, Campbell KJ, Crawford DA. Development and reliability of brief dietary infants, toddlers, preschoolers and their
Television viewing habits associated with assessment tools for Hispanics. Prêt) Chronic parents. Available at: http://www.kff.org/
obesity risk factors: a survey of Melbourne Dis, 3:A95. entmedia/7500.cfm. Accessed April 24,
schoolchildren. MedJAust. 2006;184:64-67. 37. Feldman S, Eisenberg ME, Neumark- 2009.
22. Giammattei J, Blix C, Marshak HH, et al. Sztainer D, Story M. Associations between 53. American Academy of Pediatrics.
Television watching and soft drink watching TV dtiring family meals and Committee on Public Education. American
consumption: associations with obesity in dietary intake among adolescents./Nuir Academy of Pediatrics: children,
11- to 13-year-old schoolchildren. Arch Educ Behav. 2007;39:257-263. adolescents, and television. Pediatrics.
Pediatr Adolesc Med. 2003;157:882-886. 38. Reed DB, Feng D, Chyu M, et al. Early 2001;107:423-426.
23. Dennison BA, Erb TA, Jenkins PL. lessons learned in the development of a 54. Lowry R, Lee SM, McKenna ML, et al.
Television viewing and television in childhood overweight prevention program Weight management and fruit and
bedroom associated with overweight risk in west Texas using community based vegetable intake among US high school
among low-income preschool children. participatory research. Texas Public Health students./ScA Health. 2008;78:417-424.
Pediatrics. 2002;109:1028-10S5. Assoc J. 2008;60:4-8. 55. Serpico RA, Behrens TK, Nanney MS.
24. Singh CK, Kogan MD, Van Dyck PC, 39. Deforche B, De Bourdeaudhuij I, D'hondt Vigorous physical activity and fruit and
Siapush M. Racial/ethnic, socioeconomic, E, Cardon C. Objectively measured physical vegetable consumption in a large cohort of
and behavioral determinants of childhood activity, physical activity related personality college students. Res (¿Exerc Sport.
and adolescent obesity in the United States: and body mass index in 6- to 10-yr-old 2007;78:A-36.
analyzing independent and joint children: a cross-sectional study. IntJ Behav 56. Food, Beverage, and Restaurant Advertising in
associations. Ann Fpidemiol. Nutr Phys Act. 2009;6:25. 2007; Chitdren's Impressions and Expenditures
2008;18:682-695. 40. Sallis JF, Nader PR, Broyles SL, et al. on Children's Programs. Washington, DC:
25. Jago R, Baranowski T, Baranowski JC, et al. Correlates of physical activity at home in Georgetown Economic Services; 2008.
BMI from 3-6 y of age is predicted by TV Mexican-American and Anglo-American 57. Klepp K, Wind M, de Bourdeaudhuij I, et al.
viewing and physical activity, not diet. preschool children. Health Psychol. Television viewing and exposure to food-
IntJ Obes. 2005;29:557-564. 1993;12:390-398. related commercials among European
26. Unger JB, Reynolds K, Shakib S, et al. school children, associations with fruit and
41. Hoefer WR, McKenzie TL, Sallis JF, et al.
Acculturation, physical activity, and fast- vegetable intake: a cross-sectional study.
Parental provision of transportation for
food consumption among Asian-American IntJ Behav Nutr Phys Act. 2007;4:46.
adolescent physical activity. AmJ Prev Med.
and Hispanic adolescents. / Commîtrai)) 58. Ariza AJ, Chen EH, Binns HJ, Christoffel
2001;21:48-51.
Health. 2004;29:467-481. KK. Risk factors for overweight in five- to
42. Story M, Sherwood NE, Himes JH, et al. An
27. Cordon-Larsen P, Harris KM, Ward DS, six-year-old Hispanic-American children: a
after-school obesity prevention program for
Popkin BM. Acculturation and overweight- pilot study. J Urban Health. 2004;81:150-161.
African-American girls: the Minnesota
related behaviors among Hispanic 59. Yeh M, Viladrich A, Bruning N, Roye C.
CEMS pilot study. Ethn Dis.
immigrants to the US: the National Determinants of Latina obesity in the
2003;13:SI54-SI64.
Longitudinal Study of Adolescent Health. United States. J Transcult Nurs.
Soc Sd Med. 2003;57:2023-2034. 43. Warner ML, Harley K, Bradman A, et al.
2008;20:105-115.
28. Henry J. Kaiser Family Foundation. Soda constimption and overweight stattis of
60. Singh CK, Yu SM, Siahpush M, Kogan MD.
Generation M: media in the lives of 8-18 2-year-old Mexican-American children in
High levels of physical inactivity and
year-olds. Available at: http://www.kff.org/ California. Obesity. 2006;14:1966-1974.
sedentary behaviors among US immigrant
entmedia/7250.cfm. Accessed April 24, 44. Ayala CX, Baquero B, Arredondo EM, et al.
children and adolescents. Arch Pediatr
2009. Association between family variables and
Adolesc Med. 2008;162:756-763.
29. Delmas C, Platat C, Schweitzer B, et al. Mexican American children's dietary
61. Robinson TN. Television viewing and
Association between television in bedroom behaviors./Mt/r£á«cBehav. 2007;39:62-69.
childhood obesity. Pediatr Clin North Am.
and adiposity throughout adolescence. 45. ProchaskaJJ, Sallis JF. Reliability and 2001;48:1017-1025.
Obesity. 2007;15:2495-2503. validity of fruit and vegetable screening 62. Texas Department of State Health Services.
30. Barr-Anderson DJ, van den Berg P, measure for adolescents. y/ldofe.si; Health. Strategic Plan for the Prevention of Obesity
Neumark-Sztainer D, Stoiy' M. 2004;34:163-165. in Texas: 2005-2010. Available at: h t t p : / /
Characteristics associated with older 46. United States Department of Agriculture. www.dshs.state.tx.us/obesity/pdf/
adolescents who have a television in their MyPyramid for Kids. Available at: h t t p : / / strategic_plan.pdf. Accessed April 23, 2009.
bedrooms. Pediatrics. 2008;121:718-724. www.mypyramid.gov/kids/index.html. 63. Marshall SJ, Biddle SJH, Gorely T, et al.
31. Coran Ml, Driscoll P, Johnson R, et al. Accessed April 27, 2009. Relationships between media use, body
Cross-calibration of body composition 47. Cuellar 1, Arnold B, Maldonado R. fatness and physical activity in children and
techniques against dual energy x-ray Acctilturation rating scale for Mexican youth: a meta-analysis. IntJ Obes.
absorptiometry in young children. AmJ Clin Americans-ll: a revision of the original 2004;28:1238-1246.
Nutr. ]996;6S:299-305. ARSMA scale. HispJ Behav Sd. 64. Fulkerson JA, Story M, Neumark-Sztainer D,
32. Centers for Disease Control and Prevention. 1995;17:275-304. Rydell S. Family meals: perceptions of
BMI percentile calculator for child and 48. Bauman S. The reliability and validity of the benefits and challenges among parents of 8-
teen. Available at: http://apps.nccd.cdc. Brief Acculttiration Rating Scale for to 10-year-old children.//Im Wei A^joc.
gov/dnpabmi/. Accessed April 27, 2009. Mexican Americans-II for children and 2008;108:706-709.
33. Kuczmarski RJ, Ogden CI, Cuo SS, et al. adolescents. HispJ Behav Sd. 65. Center for Screen-Time Awareness,
2000 CDC growth charts for the United 2005;27:426-441. Available at: http://www.screentime.org.
States: Methods and development. National 49. SPSS for Windows. Version 16.0. Chicago, 111: Accessed April 23, 2009.
Center for Health Statistics. Vit Health Stat. SPSS Inc; 2007. 66. Jordan AB, HerseyJC, McDivittJA, Heitzler
2002;l 1:1-190. 50. Rom DM. A sequentially rejective test CD. Reducing children's television-viewing
34. Livingstone MB, Robson PJ, Wallace JM. procedtire based on a modified Bonferroni time: a qualitative sttidy of parents and their
Issties in dietary intake assessment of inequality. Biometrika. 1990;77:663-665. children. Pediatrics. 2006;l 18:1303-1310.

318 American Journal of Health Promotion May/June 2011, Vol. 25, No. 5
Copyright of American Journal of Health Promotion is the property of American Journal of Health Promotion
and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright
holder's express written permission. However, users may print, download, or email articles for individual use.