71

ORIGINAL

Effects of lifestyle habits and eating meals together with

the family on the prevalence of obesity among school
children in Tokushima, Japan : a cross-sectional ques-
tionnaire-based survey
1 1 2 1 1
Kyoko Yuasa , Masako Sei , Eiji Takeda , Ashraf A. Ewis , Hokuma Munakata ,
1 1
Chiemi Onishi , and Yutaka Nakahori
1
Department of Human Genetics and Public Health, and 2Department of Clinical Nutrition, Insti-
tute of Health Biosciences, The University of Tokushima Graduate School, Tokushima, Japan

Abstract : Obesity in children has become a major global public health concern. The pre-
vention of obesity must start from early childhood in order to establish sound lifestyle
habits and promote healthy adulthood. In this study, we evaluated factors associated with
the prevention of obesity and the development of healthy lifestyle habits in children. A
cross-sectional, questionnaire-based survey was performed in elementary and junior high
school students in Tokushima Prefecture, Japan, during the summer of 2004. The ques-
tionnaire consisted of 30 items such as physique, sleep, eating habits, diet, exercise, free
time, and attending after-school lessons. Our study revealed that eating meals as a fam-
ily every day is associated with a lower rate of obesity as well as getting good lifestyle hab-
its such as eating balanced meals and getting enough sleep. Of the 3,291 students who
responded to the questionnaire, 2,688 (81.7% %) reported that they eat meals with their fam-
ily every day. The percentage of students who eat meals with their family every day de-
creased with increasing school grade, with the lowest percent in the junior high school
students. However, the results regarding female junior high school students revealed a
marked association between eating meals with the family every day and good lifestyle hab-
its. We recommend that parents and school teaching staff encourage the establishment
of sound, healthy lifestyle habits in children from early childhood as an effective meas-
ure for the prevention of obesity. J. Med. Invest. 55 : 71-77, February, 2008

Keywords : obesity, children, questionnaire survey, eating with family, lifestyle habits

INTRODUCTION been stated that there could be further increases

in the prevalence of obesity among U.K. children,
Recent publications predict that the severity of with about 19% of boys and 24% of girls aged 10
obesity will continue to rise and its prevalence will being predicted to be obese by 2010 (1). Many stud-
increase globally ; in the United Kingdom, it has ies demonstrated that overweight children tend to
become overweight and obese adults. Also, there
Received for publication October 22, 2007 ; accepted December is substantial evidence that childhood obesity is a
7, 2007.
major risk factor for adulthood morbidity, especially
Address correspondence and reprint requests to Yutaka Nakahori,
M.D., Ph. D., Department of Human Genetics and Public Health,
cardiovascular disease. Therefore, in recent years,
Institute of Health Biosciences, The University of Tokushima the prevention of obesity in childhood has been con-
Graduate School, Kuramoto-cho, Tokushima 770-8503, Japan and sidered crucial for the prevention of lifestyle-related
Fax : +81-88-633-7453.

The Journal of Medical Investigation Vol. 55 2008

 72 K. Yuasa, et al. Childhood obesity : Effects of lifestyle habits

diseases in adults (2, 3). questions and return the questionnaires to us. For
Previous studies have shown the association be- elementary school children, parents were requested
tween the prevention of obesity in children and their to answer the questionnaire for their children, while
lifestyle habits, such as the contents of meals, exer- junior high school children responded to the ques-
cise, and how they spend their leisure time (such tionnaire by themselves. Obesity was checked in
as watching TV) (4-7). The prevention of obesity every child according to the standard methods fol-
should start from childhood and must be combined lowed in determining obesity in Japan. Obesity in
with health education strategies and other measures children is defined as those whose relative body
for the encouragement of health-promoting habits, weight is more than 120% of the standard weight for
which is necessary for the healthy growth of the their sex and height. The standard weight based on
bodies and minds of children (8). the median weight for each height (per cm) by sex
In this study, we aimed at evaluating the factors was calculated from Tokushima-based data obtained
associated with the prevention of obesity and the between 2000 and 2002 (9). We achieved a 100%
formation of healthy lifestyle habits in school chil- response rate for our questionnaire-based survey.
dren in Tokushima Prefecture. However, nine children missed answering some
questions. The number of students was 1144 (569
SUBJECTS AND METHODS males and 575 females), 1,020 (513 males and 507
females), and 1,137 (581 males and 556 females) for
To evaluate the factors associated with childhood the first and fourth grade elementary and first grade
obesity and the methods for its prevention among junior high schools, respectively.
children in Tokushima Prefecture, we included chil-
dren from three different age groups. Therefore, we Statistical analysis
selected three different grades of school children Questionnaires were collected, the data were
to represent the desired age groups eligible for our coded and entered into a computer, and then ana-
study : first and fourth grade elementary and first lyzed using the SPSS (Statistical Package for Social
grade junior high school children. In Tokushima Sciences) version 11. The chi-square test, Mann-
Prefecture, there are 237 elementary schools and 93 Whitney test, Mantel-Haenszel analysis, and analy-
junior high schools, from which 88 and 39 schools sis of covariance (ANCOVA) were used to examine
were randomly selected, respectively. From each the differences between groups with and without
one of the 127 randomly selected schools, only one adjustments for school years.
class was included in our study. Thus, 45 classes of
first grade and 43 classes of fourth grade elementary RESULTS
school and 39 classes of junior high school children
constituted the total sample size of 3,301 children A total of 1,336 males (80.6%) and 1,352 females
who participated in our study. We conducted this (82.8%) replied that they eat meals as a family every
cross-sectional, questionnaire-based survey study day. According to school years, the percentage of
in Tokushima Prefecture during a two-month period students who eat meals as a family every day de-
from June to July 2004. creased with increasing school years in both males
We prepared a 30 item questionnaire that focused and females : 522 males (92.1%) and 533 females
on personal information, e.g., age, sex, weight, (93.0%) in the first year of elementary school, 456
height, etc., and information on meals, physical ac- males (89.1%) and 452 females (89.5%) in the fourth
tivity, sleeping, and how the children spend their lei- year of elementary school, and 358 males (61.8%)
sure time. Questionnaires were distributed to all par- and 367 females (66.0%) in the first year of junior
ticipating children, who were asked to answer the high school (Table 1).

Table 1 Gender and eating with one’s family of the children distributed according to their school grade.
Boys Girls
Eat with family Eat with family
Every day Not every day p Every day Not every day p
Number(%)
First grade of elementary school 522(92.1) 45(7.9) 533(93.0) 40(7.0)
Fourth grade of elementary school 456(89.1) 56(10.9) !0.001 452(89.5) 53(10.5) !0.001
First grade of junior high-school 358(61.8) 221(38.2) 367(66.0) 189(34.0)
Total 1336(80.6) 322(19.4) 1352(82.7) 282(17.3)
 The Journal of Medical Investigation Vol. 55 February 2008 73

1st-year of elementary school 1st-year of junior high school

Table 2 shows the analysis of data regarding obe- In males in the first year of junior high school,
sity and life style among school children in the first the percentages of children who eat breakfast every
year of elementary school. In males, the percent- day and those who eat vegetables every day were
ages of children who eat breakfast every day and significantly higher in the group who eat meals as
those who eat vegetables every day were signifi- a family every day than in their counterparts. In the
cantly higher in the group who eat meals as a fam- females, the girls who eat meals as a family every
ily every day compared to those who do not, while, day showed a significantly higher percentage of eat-
in the females, the percentage of girls who eat vege- ing breakfast every day, eating vegetables every day,
tables every day was significantly higher in the and of getting a longer sleep time than their coun-
group who eat meals as a family every day than in terparts. Those girls also showed a significantly
the other group. lower frequency of eating precooked noodles, and
a significantly shorter time spent indoors or watch-
4th-year of elementary school ing TV (Table 4).
In males in the fourth year of elementary school, After adjustments for school years, data were
the percentage of children who eat breakfast every separately re-analyzed for boys and girls (Table 5).
day was significantly higher in the group who eat In males, the percentages of children who eat break-
meals as a family every day compared to those who fast every day and those who eat vegetables every
do not. However, significance was not observed in day were significantly higher in the group who eat
females of the same grade. The percentage of obese meals as a family every day than in their counter-
children was significantly lower among females of parts. In addition, the group who eat meals as a fam-
the fourth grade who eat meals as a family every ily every day showed a significantly longer sleep
day than the group who do not (Table 3). time than the others.

Table 2 Comparison between first grade elementary school children who eat with their family every day and those who do not as
regards obesity and different lifestyle habits.
Boys Girls
Eat with family Eat with family
Every day Not every day p Every day Not every day p
Number(%) 522 (92.1) 45 (7.9) 533 (93.0) 40 (7.0)
Physique
Obese 43 (9.5) 6 (15.4) 0.241a) 30 (6.6) 2 (5.9) 0.876a)
Non-obese 409 (90.5) 33 (84.6) 427 (93.4) 32 (94.1)
Feeding on awakening in morning
Good 182 (35.0) 12 (26.7) 0.259a) 178 (33.5) 18 (45.0) 0.140a)
Not good 338 (65.0) 33 (73.3) 353 (66.5) 22 (55.0)
Eating breakfast
Every day 463 (88.7) 35 (77.8) 0.032a) 462 (86.7) 33 (82.5) 0.457a)
Not every day 59 (11.3) 10 (22.2) 71 (13.3) 7 (17.5)
Eating vegetables
Every day 206 (39.6) 11 (24.4) 0.045a) 218 (41.0) 9 (22.5) 0.021a)
Not every day 314 (60.4) 34 (75.6) 314 (59.0) 31 (77.5)
Frequency of eating instant noodles
Once a week 447 (86.1) 37 (82.2) 0.471a) 473 (89.1) 32 (80.0) 0.083a)
More once a week 72 (13.9) 8 (17.8) 58 (10.9) 8 (20.0)

Average hours spent !SD

Sleeping 9.3!0.58 9.3!0.58 0.543b) 9.3!0.65 9.1!0.58 0.165b)
In room 3.2!1.55 3.4!1.55 0.394b) 2.9!1.44 3.1!1.89 0.896b)
Watching television 1.5!0.91 1.6!1.06 0.904b) 1.5!0.93 1.6!0.99 0.797b)
a)χ2 test
b)Mann-Whitney test
 74 K. Yuasa, et al. Childhood obesity : Effects of lifestyle habits

Table 3 Comparison between fourth grade elementary school children who eat with their family every day and those who do not
as regards obesity and different lifestyle habits.
Boys Girls
Eat with family Eat with family
Every day Not every day p Every day Not every day p
Number(%) 456 (89.0) 56 (10.9) 452 (89.5) 53 (10.5)
Physique
Obese 52 (12.3) 10 (19.6) 0.145a) 36 (8.4) 13 (27.7) !0.001a)
Non-obese 370 (87.7) 41 (80.4) 392 (91.6) 34 (72.3)
Feeding on awakening in morning
Good 132 (29.0) 11 (20.0) 0.160a) 140 (31.0) 12 (22.6) 0.211a)
Not good 323 (71.0) 44 (80.0) 312 (69.0) 41 (77.4)
Eating breakfast
Every day 407 (89.3) 41 (73.2) !0.001a) 404 (89.6) 47 (88.7) 0.840a)
Not every day 49 (10.7) 15 (26.8) 47 (10.4) 6 (11.3)
Eating vegetables
Every day 190 (41.7) 17 (30.4) 0.104a) 201 (44.5) 21 (39.6) 0.501a)
Not every day 266 (58.3) 39 (69.6) 251 (55.5) 32 (60.4)
Frequency of eating instant noodles
Once a week 379 (83.5) 43 (76.8) 0.211a) 378 (83.8) 45 (84.9) 0.838a)
More once a week 75 (16.5) 13 (23.2) 73 (16.2) 8 (15.1)

Average hours spent !SD

Sleeping 9.3!0.58 8.6!0.87 0.030b) 8.8!0.66 8.8!0.69 0.501b)
In room 3.3!1.55 3.1!1.60 0.278b) 3.4!1.60 3.4!1.94 0.800b)
Watching television 1.6!0.99 1.4!0.93 0.110b) 1.8!1.10 1.9!1.17 0.901b)
a)χ2 test
b)Mann-Whitney test

Table 4 Comparison between first grade junior high-school children who eat with their family every day and those who do not as
regards obesity and different lifestyle habits.
Boys Girls
Eat with family Eat with family
Every day Not every day p Every day Not every day p
Number(%) 358 (61.8) 221 (38.2) 367 (66.0) 189 (34.0)
Physique
Obese 53 (15.3) 41 (19.3) 0.218a) 40 (11.6) 19 (10.9) 0.811a)
Non-obese 293 (84.7) 171 (80.7) 304 (88.4) 155 (89.1)
Feeding on awakening in morning
Good 71 (19.8) 39 (17.6) 0.515a) 72 (19.7) 24 (12.8) 0.042a)
Not good 287 (80.2) 182 (82.4) 294 (80.3) 164 (87.2)
Eating breakfast
Every day 290 (81.0) 155 (70.1) 0.003a) 314 (85.6) 122 (64.9) !0.001a)
Not every day 68 (19.0) 66 (29.9) 53 (14.4) 66 (35.1)
Eating vegetables
Every day 189 (52.8) 91 (41.2) 0.007a) 219 (59.7) 75 (40.1) !0.001a)
Not every day 169 (47.2) 130 (58.8) 148 (40.3) 112 (59.9)
Frequency of eating instant noodles
Once a week 208 (58.1) 125 (56.6) 0.716a) 271 (73.8) 112 (60.2) !0.001a)
More once a week 150 (41.9) 96 (43.4) 96 (26.2) 74 (39.8)

Average hours spent !SD

Sleeping 9.3!0.58 7.4!1.28 0.193b) 7.4!1.07 7.0!1.15 !0.001b)
In room 5.0!2.48 5.0!2.58 0.735b) 5.1!2.51 5.9!2.72 0.003b)
Watching television 1.9!1.50 1.9!1.50 0.703b) 2.1!1.60 2.7!2.07 0.003b)
a)χ2 test
b)Mann-Whitney test
 The Journal of Medical Investigation Vol. 55 February 2008 75

Table 5 Comparison of obesity and different lifestyle habits in all the studied children after adjusting for their school year.
Boys Girls
Eat with family Eat with family
Every day Not every day p Every day Not every day p
Number(%) 1336 (80.6) 322 (19.4) 1352 (82.8) 282 (17.3)
Percent
Physique*
Obese 12.4 18.1 0.049a) 8.9 14.3 0.114a)
Non-obese 87.6 81.9 91.1 85.7
Feeding on awakening in morning*
Good 27.9 21.5 0.103a) 28.0 27.1 0.197a)
Not good 72.1 78.5 72.0 72.9
Eating breakfast*
Every day 86.2 73.7 !0.001a) 87.2 78.4 !0.001a)
Not every day 13.8 26.3 12.8 21.6
Eating vegetables*
Every day 44.9 32.1 !0.001a) 48.4 33.8 !0.001a)
Not every day 55.1 67.9 51.6 66.2
Frequency of eating instant noodles*
Once a week 78.0 71.6 0.332a) 82.3 74.8 0.002a)
More once a week 22.0 28.4 17.7 25.2

Average hours spent !SD

Sleeping** 8.6!0.56 8.3!0.41 !0.001b) 8.5!0.51 8.1!0.40 !0.001b)
In room** 3.8!1.22 4.0!0.89 0.337b) 3.8!1.21 4.4!0.94 !0.001b)
Watching television** 1.7!0.70 1.7!0.52 0.937b) 1.8!0.79 2.2!0.62 !0.001b)
a)Mantel-Haenszel test
b)ANCOVA test
* : rate after direct adjustment for the children school year (standard population : total)
** : average hour time and SD after adjustment for the children school year

For females, the percentages of girls who eat swer was from one of five choices : 1 : almost every
breakfast every day and those who eat vegetables day of the week, 2 : four or five times / week, 3 : two
every day were significantly higher in the group or three times / week, 4 : once / week, 5 : almost
who eat meals as a family every day than in their never. Based on the answers, we categorized stu-
counterparts. In addition, females who eat meals as dents into two main groups : those who eat almost
a family every day showed a significantly lower fre- everyday with their families (answer No.1) as one
quency of eating precooked noodles than the other group, and those who gave a different answer (an-
group. Moreover, the girls who eat meals as a fam- swer Nos. 2-5) as the other. In fact, the question
ily every day showed a significantly longer mean did not specify which members of the family to eat
sleeping time and a significantly shorter time spent with or the meal that is shared with the family. In
indoors and watching TV. Japan, eating lunch usually takes place in schools,
while breakfast and supper are eaten at home.
Also, we do not know if the respondents under-
DISCUSSION stood whether the question meant eating with the
family while sitting together at the table to eat, or
Our study evaluated factors associated with obe- just to eat at home in the presence of one or more
sity among school children in Tokushima, aiming to of the family members who may be either eating to-
identify lifestyle habits that can affect the occur- gether or busy doing something else.
rence of obesity in order to prevent them. In this Children’s lifestyle habits (such as contents of
cross-sectional study, we evaluated the issue of eat- meals and eating habits) are affected by their par-
ing meals as a family every day to clarify if it is as- ents. Therefore, many studies have suggested the
sociated with obesity and other lifestyle habits. The necessity for interventions in parents for the pre-
question was : “Do you eat together with your fam- vention of obesity and the formation of good life-
ily at least once a day?” The corresponding an- style habits in their children (10-16). The question-
 76 K. Yuasa, et al. Childhood obesity : Effects of lifestyle habits

naire item that asked about eating meals as a fam- tionally unstable, so it is good for girls to eat meals
ily every day was included to reflect the lifestyles of as a family, show a longer sleep time, have pleas-
both children and their parents. ant feelings on awakening, and eat breakfast every
Therefore, our analysis was based on compari- day. The second reason is that obesity and lifestyle
sons between those who eat with their family eve- habits in children have been reported to be associ-
ryday and those who do not. Our results demon- ated with those in adults. Therefore, it is very im-
strated that eating meals as a family every day was portant for children to acquire good lifestyle habits
associated with the absence of obesity and good life- during their childhood to continue such good life-
style habits such as the contents of meals and sleep. style practices into adulthood. These girls are fu-
Also, the results showed a significantly earlier sup- ture mothers, who will have a great influence on the
per time for girls who eat meals as a family every lifestyles of their children and subsequent genera-
day than others who do not (data not shown). As a tions.
parent-associated factor, this early supper time may Our findings are important for those who are con-
be associated with the home situation (for example, cerned with and planning adolescent health pro-
full-time housewives and working conditions such motion, and may provide clues to help in develop-
as part-time jobs) (2, 6), allowing parents (mostly ing and establishing good lifestyle habits among
mothers or grandmothers) to prepare supper rela- children. Parents and school teaching staff should
tively early to facilitate eating together as a family, promote sound, healthy lifestyle habits in children
though this situation was not surveyed in this study. from early childhood as an effective primary preven-
Though the number of family members eating to- tive measure against childhood and adult obesity.
gether was not included in this questionnaire, eat-
ing in a lively atmosphere may reduce the risk of
obesity, contributing to the development of good ACKNOWLEDGMENTS
lifestyle habits. This is supported by reports that
confirmed a lower risk of obesity (especially in fe- The authors would like to thank all members
male children) in the presence of brothers and sis- of the Committee for Prevention Strategy against
ters (17). Lifestyle-related Disease, and all teachers and par-
Concerning child-associated factors (cram school, ents of the children who responded to our question-
watching TV, games, etc.), the percentage of chil- naire survey.
dren who do not attend cram schools was signifi-
cantly higher in the group who eat meals as a fam-
ily every day (data not shown). Generally, the per- REFERENCES
centage of children who attend cram schools rises
with increasing school grade, a finding which may 1. Zaninotto P, Wardle H, Stamatakis E, Mindell
have reduced the percentage of children who eat J, Head J : Forecasting obesity to 2010. Lon-
meals as a family every day. don : National Centre for Social Research/De-
As for these results, two factors are thought to be partment of Health, 2006
involved : first, for children attending cram school, 2. Gibson LY, Byrne SM, Davis EA, Blair E,
there is a possibility their supper time changes. A Jacoby P, Zubrick SR : The role of family and
second reason is that, with advancing school grade, maternal factors in childhood obesity. Med J
children grow and become independent of their Aust 186 : 591-595, 2007
parents, so, mothers usually start to have jobs, and 3. Anderson SE, Cohen P, Naumova EN, Must
this may explain the later supper times. It is impor- A : Relationship of childhood behavior disor-
tant here to mention that among the female junior ders to weight gain from childhood into adult-
high school students, there were significant differ- hood. Ambul Pediatr 6 : 297-301, 2006
ences in getting good lifestyle habits among those 4. Sekine M, Yamagami T, Hamanishi S, Handa
who eat meals as a family every day compared to K, Saito T, Nanri S, Kawaminami K, Tokui N,
those who do not. Yoshida K, Kagamimori S : Parental obesity, life-
These data are of great importance for two rea- style factors and obesity in preschool children :
sons. One reason is that junior high school girls are results of the Toyama Birth Cohort study. J Epi-
undergoing or will soon be entering their second demiol 12 : 33-39, 2002
physical growth period, in which they are often emo- 5. Kagamimori S, Yamagami T, Sokejima S,
 The Journal of Medical Investigation Vol. 55 February 2008 77

Numata N, Handa K, Nanri S, Saito T, Tokui bese preadolescents. Int J Eat Disord 30 : 447-
N, Yoshimura T, Yoshida K : The relationship 453, 2001
between lifestyle, social characteristics and obe- 12. Seibold ES, Knafl K, Grey M : The family con-
sity in 3-year-old Japanese children. Child Care text of an intervention to prevent type 2 diabe-
Health Dev 25 : 235-247, 1999 tes in high-risk teens.Diabetes Educ 29 : 997-
6. Takahashi E, Yoshida K, Sugimori H, Miyakawa 1004, 2003
M, Izuno T, Yamagami T, Kagamimori S : In- 13. Borra ST, Kelly L, Shirreffs MB, Neville K,
fluence factors on the development of obesity Geiger CJ : Developing health messages : quali-
in 3-year-old children based on the Toyama tative studies with children, parents, and teach-
study. Prev Med 28 : 293-296, 1999 ers help identify communications opportunities
7. Saelens BE, Sallis JF, Nader PR, Broyles SL, for healthful lifestyles and the prevention of
Berry CC, Taras HL : Home environmental in- obesity. J Am Diet Assoc 103 : 721-728, 2003
fluences on children’s television watching from 14. Wang H, Sekine M, Chen X, Kanayama H,
early to middle childhood. J Dev Behav Pedi- Yamagami T, Kagamimori S : Sib-size, birth or-
atr 23 : 127-32, 2002 der and risk of overweight in junior high school
8. Chen X, Sekine M, Hamanishi S, Wang H, students in Japan : results of the Toyama Birth
Gaina A, Yamagami T, Kagamimori S : Life- Cohort Study. Prev Med 44 : 45-51, 2007
styles and health-related : quality of life in Japa- 15. Young KM, Northern JJ, Lister KM, Drummond
nese school children : a cross-sectional study. JA, O’Brien WH : A meta-analysis of family-
Prev Med 40 : 668-678, 2005 behavioral weight-loss treatments for children.
9. Sei M, Nakatsu T, Yuasa K, Tanaka H, Indriani, Clin Psychol Rev 27 : 240-249, 2007
Munakata H, Nakahori Y : Prevalence of meta- 16. Brunstrom JM, Mitchell GL, Baguley TS : Po-
bolic complication in children with severe obe- tential early-life predictors of dietary behaviour
sity : Pediatrics International 49 : 545-552, 2007 in adulthood : a retrospective study. Int J Obes
10. Laessle RG, Uhl H, Lindel B, Müller A : Pa- (Lond) 29 : 463-74, 2005
rental influences on laboratory eating behav- 17. Dubois L, Girard M, Potvin Kent M : Break-
ior in obese and non-obese children. Int J Obes fast eating and overweight in a pre-school popu-
Relat Metab Disord 1 : S60-62, 2001 lation : is there a link? Public Health Nutr 9 :
11. Laessle RG, Uhl H, Lindel B : Parental influ- 436-442, 2006
ences on eating behavior in obese and nono-