Physiology & Behavior 163 (2016) 161–166

Contents lists available at ScienceDirect

Physiology & Behavior

journal homepage: www.elsevier.com/locate/phb

Relationships among fitness, obesity, screen time and academic

achievement in Japanese adolescents
Noriteru Morita a,⁎, Toshihiro Nakajima b, Koichi Okita c, Toru Ishihara d, Masato Sagawa a, Koji Yamatsu e
a
Department of Sport Cultural Studies, Hokkaido University of Education, Iwamizawa, Hokkaido, Japan
b
Hokkaido University of Science, Hokkaido, Japan
c
Department of Sports Education, Hokusho University, Hokkaido, Japan
d
Department of Physical Fitness Science, Graduate School of Education, Hokkaido University, Hokkaido, Japan
e
Faculty of Education, Saga University, Saga, Japan

H I G H L I G H T S

• Poor fitness interfered with satisfactory academic performance in Japanese boys.

• Obesity interfered with satisfactory academic performance in Japanese girls.
• The influences of fitness and obesity were independent of time spent on video games.
• The influences of fitness and obesity were independent of attending cram school.

a r t i c l e i n f o a b s t r a c t

Article history: Purpose: Students who study intensively in one of Japan's ‘cram schools’ and/or spend excess time on electronic
Received 21 July 2015 devices such as video games are in a sedentary state much of the time, and this may affect their physical fitness.
Received in revised form 14 April 2016 We investigated whether there are relationships among obesity, physical fitness and academic achievement in
Accepted 30 April 2016 Japanese students after controlling for socioeconomic and behavioral confounding factors.
Available online 3 May 2016
Methods: The data of 315 students (152 females [48%], 163 males [52%]; 12–13 yrs old) were analyzed. Academic
achievement was assessed by the total grade points on eight school subjects (GP8). Students with a body mass
Keywords:
Academic performance
index at or above the 85th percentile of each gender were classified as the overweight/obese group. Physical fit-
Children ness was evaluated by the total score on eight fitness tests. Socioeconomic and behavioral confounders including
Exercise capacity the mother's educational background, household income, cram school utilization and time spent on video games/
Overweight mobile phones were used as covariates.
Physical inactivity Results: The GP8 of the overweight/obese students was significantly lower than that of the normal weight stu-
Sedentary dents (27.2 vs. 29.0 points, respectively). After adjusting for the confounders, the physical fitness score was
found to be a significant factor for determinants of GP8 in boys (β = 0.324), but not in girls. The obesity status
was a factor for GP8 in the girls (β = −0.160) but not in the boys.
Conclusion: These results suggest that physical fitness in boys and obesity status in girls could be important fac-
tors not only for health status but also for academic achievement, independent of socioeconomic and behavioral
backgrounds.
© 2016 Elsevier Inc. All rights reserved.

1. Introduction
Obesity and poor physical fitness among children are critical public
health problems in developed countries including Japan, because they
lead to an increased risk of cardiovascular disease and premature mor-
tality in adulthood [1–5]. It has been demonstrated that aerobic capaci-
ty, an indicator of physical fitness, is related to reading and arithmetic
⁎ Corresponding author at: Department of Sport Cultural Studies, Hokkaido University
of Education, 2-34-1 Midorigaoka, Iwamizawa, Hokkaido 068-8642, Japan.
E-mail address: morita.noriteru@i.hokkyodai.ac.jp (N. Morita).
ability in children [6–21] and that acute exercise temporarily enhances
cognitive function [22,23]. In addition, childhood obesity, which is asso-
ciated with poor fitness, has been reported to have a negative influence
on cognitive function and academic achievement [24–31].
Many children in Japan and other East Asian countries attend cram
schools (test preparation centers) or have private teachers [32] with
the goal of enhancing their academic achievement, but the cram
school/private tutoring time may limit the children's available time for
physical activity and sports participation. Compared with previous find-
ings in European countries and the United States [6–23], there may be
different relationships among physical fitness, obesity, and academic

http://dx.doi.org/10.1016/j.physbeh.2016.04.055
0031-9384/© 2016 Elsevier Inc. All rights reserved.

Descargado para Anonymous User (n/a) en Universidad de San Martin de Porres de ClinicalKey.es por Elsevier en octubre 12, 2018.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2018. Elsevier Inc. Todos los derechos reservados.
162 N. Morita et al. / Physiology & Behavior 163 (2016) 161–166
achievement in children in these East Asian countries. In Japan, children
spent a lot of time watching television (whether on a TV or on the Inter-
net), playing video games and/or using mobile phones; it was reported
that 43% of Japanese children engaged in over 2 h of daily screen time
[5]. During these electronic device-based activities, the children are sed-
entary, and this may negatively affect both their physical fitness and ac-
ademic achievement. As a matter of public health, it is necessary to
clarify the relationships among physical fitness, obesity, and academic
achievement after adjusting for social and lifestyle confounders in
children.
The purpose of the present study was to determine whether there
are relationships among physical fitness, obesity and academic achieve-
ment in Japanese first-year junior high school students. To better under-
stand the relationships among obesity, low physical fitness and
academic achievement, we adjusted for potential confounders such as
household income, mother's educational background, attendance at a
cram school or use of a private teacher, and time spent on electronic
devices.
2. Methods
2.1. Participants
During the autumn of 2012, 315 first-year junior high school stu-
dents (equal to seventh graders in the United States; male, n = 163; fe-
male, n = 152; age range, 12–13 yrs and born during the same year)
from three public schools in Hokkaido Prefecture participated in this
study. To recruit participants, all students of these schools were given
a pack containing a leaflet and questionnaires for themselves, a letter
of explanation and request for cooperation with this study to their par-
ents/guardians, and questionnaires for their parents/guardians (n =
493). In the explanation letter, they were informed that all data collect-
ed from the students and their parents/guardians would be anonymous;
thus, personal information such as students' names and birth dates was
not obtained. Only those students whose parents/guardians consented
to their participation based on the provided explanation completed
and returned the questionnaires (n = 344; male/female = 180/164),
thereby consenting to participate. A total of 29 students were excluded
due to missing data or because they did not complete the questionnaires
or fitness tests (incomplete questionnaires, n = 4; missing grade
points, n = 1; incomplete physical fitness tests, n = 20; and missing
body weight/height data, n = 4). The data of a total of 315 students
were thus analyzed in the present study. This study was approved
by the institutional review board of Hokkaido University of
Education.
2.2. Academic achievement
Academic achievement was assessed by the total grade point (GP) of
school subjects evaluated by school teachers; individual grades were re-
ported for 8 school subjects (Japanese, Mathematics, Social Studies, Sci-
ences, English, Music, Arts, and Home Economics/Vocational
Technology). Students earn 1 to 5 credit points for each subject; thus,
a student can receive a maximum of 40 grade points for the 8 subjects
in question. The total GPs of all 8 school subjects (GP8) and of a subset
of 5 subjects (GP5; Japanese, Mathematics, Social Studies, Sciences,
and English) were used as continuous variables. The 5 subjects included
in GP5 are commonly used for the entrance examination to high school
in Japan.
2.3. Physical fitness
Physical fitness was assessed based on the 8 tests shown in the Sup-
plemental Fig. 1 in the online version at http://dx.doi.org/10.1016/j.
physbeh.2016.04.055. (50-m sprint, standing broad jump, repeated
side-steps, sit and reach, sit-ups, hand grip strength, handball throw,
Descargado para Anonymous User (n/a) en Universidad de San Martin de Porres de ClinicalKey.es por Elsevier en octubre 12, 2018.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2018. Elsevier Inc. Todos los derechos reservados.
and 20-m shuttle run). These fitness tests were adopted by the Ministry
of Education, Culture, Sports, Science and Technology (MEXT) of Japan
and are conducted nationwide. Students perform the fitness tests dur-
ing physical education classes wearing light clothes and rubber-soled
shoes. Prior to the measurements, the students stretched and did
some warm-up exercises. The rest interval between tests is at least
3 min. The physical education teachers of each school must follow strict-
ly the measurement protocol stipulated in the test manual [33].
Performance-to-score conversion tables made by the MEXT are used
to standardize the results into 10 grades of score for each gender (see
Supplemental Tables 1 and 2 in the online version at http://dx.doi.org/
10.1016/j.physbeh.2016.04.055.). The score classifications of each fit-
ness test in Supplemental Tables 1 and 2 in the online version at
http://dx.doi.org/10.1016/j.physbeh.2016.04.055. are based on the
mean and standard deviation values of Japanese children [33]. Levels
of comprehensive physical fitness were evaluated based on total physi-
cal fitness test scores (PFS; the PFS points ranged from 8 to 80). Detail
measurement procedures for the physical tests were in Supplemental
method because of words number limit.
2.4. Body mass index (BMI)
Height and body mass were measured by using stadiometers and
scales. BMI was calculated as body mass (kg) / height (m)2. In order to
maintain anonymity, we did not obtain the birth dates of the students,
and were therefore unable to calculate BMI z-score based on the refer-
ence values for Japanese children. Each student was classified into either
the normal weight group, defined as a BMI below the 85th percentile, or
the overweight/obese group, which was defined as a BMI at or above the
85th percentile according to the definition of the U.S. Centers for Disease
Control and Prevention [34].
2.5. Socioeconomic and lifestyle confounders
The students and their parents completed a questionnaire
concerning their socioeconomic background. The household income,
mother's education, use of a cram school or private teacher and time
spent on video games/mobile phones were determined. The household
income and mother's education were reported by the parents, and the
use of a cram school or private teacher and the time spent using elec-
tronic devices were reported by the students. Household income was
assessed with one questionnaire item (“How much does your family
earn annually?”) on a 5-point scale (i: b ¥ 2,000,000; ii: ¥ 2,000,000–
¥ 3,999,999; iii: ¥ 4,000,000–5,999,999; iv: ¥ 6,000,000–¥ 7,999,999;
v: ¥ 8,000,000 or more).
The highest level of the mother's education was assessed with one
item (“What is the highest level of education attempted by the moth-
er?”) on a 6-point scale (i: completed junior high school; ii: completed
or dropped out of high school; iii: completed or dropped out of a voca-
tional school; iv: completed or dropped out of a junior college; v: com-
pleted or dropped out of undergraduate studies; vi: completed or
dropped out of graduate studies). A dichotomous item (“Do you use a
cram school/private teacher?” with 0: No; 1: Yes) was used to assess
the use of a cram school/private teacher. Time spent using electronic de-
vices was assessed with one item (“How much time do you spend daily
playing video games or using a mobile phone on school days?”) on a
5-point scale (i: ≥2 h; ii: 1–2 h; iii: 30–60 min; iv: 15–30 min; v: almost
not use).
The following were used as dichotomous items because these vari-
ables were used as covariates in the multiple regression analysis: integrat-
ed scales of household income (0: b ¥ 6,000,000; 1: ≥ ¥ 6,000,000),
mother's education (0: high school graduate or less; 1: completed or
dropped out from a vocational school, a junior college, or undergradu-
ate or graduate studies) and excess time spent on electronic devices
(0:b 2 h/day; 1: ≥2 h/day).
 N. Morita et al. / Physiology & Behavior 163 (2016) 161–166 163

2.6. Statistical analysis
Prior to analysis, we assessed the normality of the variables using the
Kolmogorov-Smirnov test. BMI was a non-normally distributed vari-
able, and the other variables showed normal distributions. Pearson
and Spearman correlation analyses were used for normally and non-
normally distributed variables, respectively. Multiple regression analy-
ses for each gender were used for GP8 or GP5 as a dependent variable,
and household income, mother's education, utilization of a cram
school/private teacher, excess time spending for electronic devices, obe-
sity status and PFS as independent variables. To avoid multicollinearity
between the obesity status and PFS, we used a dichotomous variable
as overweight/obesity (1) or normal weight (0) in the multiple regres-
sion analyses.
Comparisons between two groups were analyzed using the
Student's t-test for normally distributed variables and the Mann-
Whitney U test for non-normally distributed variables. Our comparison
of numbers in the two categories was performed using the chi-square
test.
After comparing the non-obese and overweight/obese groups, we
calculated effect size (Cohen's d) [35] and statistical power (1-β)
using the G*Power statistical package (version 3.1) [36]. Statistical anal-
yses were performed by PASW statistics version 18 (SPSS Inc., Chicago,
IL, USA), except for statistical power analysis. All data were presented as
mean ± standard deviation (SD), unless otherwise noted. The level of
statistical significance was set at p b 0.05.
significantly lower GP8 scores than the normal weight students (effect
size d = 0.332; 1-β = 0.560). However, no significant difference in
GP5 was revealed between the overweight/obesity students and the
normal weight students.
The results of the present univariate correlation analyses are shown
in Table 2. In male students, there were significantly positive correla-
tions between PFS and GP5/GP8. In female students, no significant cor-
relation was found between PFS and GP5/GP8. Overweight/obesity
status was negatively correlated with PFS and GP5/GP8 in female
students.
Table 3 shows the results of the present multiple regression analy-
ses, which were carried out to examine the relationships among aca-
demic achievement, obesity status and comprehensive physical fitness
levels. After adjusting for socioeconomic and behavioral confounding
factors such as mother's education, household income, utilization of a
cram school/private teacher and excess time spending on electronic de-
vices, PFS was found to be a significant correlating factor for GP5 and
GP8 in male students. In female students, obesity status, but not PFS,
was significantly associated with both GP5 and GP8 after adjusting for
the confounders. Fig. 1 shows the relationships between PFS and the ac-
ademic achievement measures (Fig. 1A, GP5; B, GP8) after adjusting for
the socioeconomic and behavioral confounding factors. As shown in Fig.
2, after adjusting for these confounding factors, the overweight/obese
female students showed lower adjusted academic achievement mea-
sures (Fig. 2A, GP5; B, GP8) compared to the normal-weight female
students.

3. Results
4. Discussion
As shown in Table 1, there were no statistically significant differ-
ences in BMI between the two genders. Female students showed higher 4.1. Major findings and implications
scores in PFS, and GP8 compared with male students. Although house-
hold income, mother's educational background, and excess time spend- The major findings of the present study are twofold. First, the aca-
ing on electronic devices were not different between genders, the rate of demic achievement among the overweight/obese students was lower
cram school utilization in male students was higher than that in female than that of the non-obese students. In the female students, obesity
students (50.9% in males vs. 36.8% in females, p b 0.05). was a significant determinant for academic achievement after control-
A total of 48 students were classified into the overweight/obesity ling for socioeconomic and lifestyle factors. Second, we also found a pos-
group (male/female = 25/23). The overweight/obese students had itive relationship between academic achievement and physical fitness
levels in all students and in male students after adjusting for the con-
Table 1 founding factors. These results suggest that physical fitness in boys
Basic characteristics of 315 students (163 males and 152 females). and obesity in girls could be important factors not only for health status
Male Female p value but also for academic achievement, independent of socioeconomic and
behavioral backgrounds.
Number, n 163 152
Height, cm 153.4 ± 7.8 152.5 ± 6.3 – Our present findings support the results of previous studies
Weight, kg 46.4 ± 10.2 44.9 ± 9.1 – [6–18,20,21,24–31] and also provide the new information that
BMI, kg/m2 19.6 ± 3.2 19.2 ± 3.1 0.321⁎ physical fitness in boys and obesity status in girls remain significant
Grip strength, kg 24.4 ± 6.1 22.3 ± 3.9 – determinants for academic achievement, after controlling for excess
Sit-ups, reps 23.8 ± 6.7 18.4 ± 5.9 –
Sit-and-reach, cm 36.0 ± 9.2 39.0 ± 9.2 –
time spent on video games/mobile phones and using a cram school/
Repeated side-steps, reps 45.1 ± 7.9 39.4 ± 7.8 – private teacher, in early adolescents in Japan. Approximately 48% of
20-m shuttle run, reps 63.2 ± 25.0 42.3 ± 18.7 – first-year junior high school students in Japan attend a cram school
50-m sprint, sec 8.8 ± 1.0 9.4 ± 0.8 – or see a private teacher [32]. Indeed, in the present study, we found
Standing broad jump, cm 178.4 ± 27.7 155.4 ± 24.0 –
that 44% of the students utilized such extra learning instruction.
Handball throw, m 19.0 ± 5.1 11.1 ± 3.6 –
PFS 31.9 ± 9.8 38.0 ± 10.3 b0.001 When students spend time learning in cram schools, they might
GP5 17.6 ± 4.0 18.2 ± 3.7 0.141 enhance their academic performance, but they are more sedentary.
GP8 28.0 ± 5.6 29.5 ± 5.3 0.017 Our present findings indicate that the sedentary time spent by
Household income children and adolescents on video games and mobile phones not
b ¥ 6,000,000, % 60.1% 58.6% 0.819 only adds to their sedentary lifestyle but could also negatively affect
their academic performance. To increase physical fitness and prevent
Mothers' education
Below high school graduate, % 56.4% 59.9% 0.569 obesity, sedentary behaviors should be discouraged, and moderate-
Cram school utilization, % 50.9% 36.8% 0.013† to-vigorous physical activity such as exercise and sports should be
encouraged.
Excess time spent for electronic devices
Over 2 h, n 40 38 1.000 A recent report indicates that increasing the time and intensity of
physical education improves academic achievement [19]. The results
PFS, total physical fitness scores; GP5, total grade points of 5 school subjects; GP8, total
grade points of 8 school subjects.
of the present study suggest that among junior high school students,
⁎ Mann-Whitney U test. moderate-to-vigorous physical activity would not be a negative factor
†
Chi-square test. for academic achievement and that it could support efficient learning.

Descargado para Anonymous User (n/a) en Universidad de San Martin de Porres de ClinicalKey.es por Elsevier en octubre 12, 2018.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2018. Elsevier Inc. Todos los derechos reservados.
164 N. Morita et al. / Physiology & Behavior 163 (2016) 161–166

Table 2
Univariate correlation analyses among the study variables for each gender.

1 2 3 4 5 6 7 8

Male 1 Household income 1.000

2 Mothers' education −0.008 1.000
3 Cram school utilization −0.123 −0.120 1.000
4 Excess time spent for electronic devices −0.144 0.017 0.125 1.000
5 Overweight/obesity 0.105 −0.065 0.059 0.153 1.000
6 PFS 0.079 0.031 −0.052 −0.057 −0.102 1.000
7 GP5 0.193⁎ 0.293⁎⁎ −0.186⁎ −0.195⁎ −0.021 0.357⁎⁎ 1.000
8 GP8 0.169⁎ 0.264⁎⁎ −0.155⁎ −0.241⁎⁎ −0.051 0.354⁎⁎ 0.979⁎⁎ 1.000
Female 1 Household income 1.000
2 Mothers' education 0.238⁎⁎ 1.000
3 Cram school utilization −0.222⁎⁎ −0.080 1.000
4 Excess time spent for electronic devices 0.039 −0.039 −0.100 1.000
5 Overweight/obesity −0.057 −0.009 0.012 0.180⁎ 1.000
6 PFS 0.061 −0.063 −0.062 −0.230⁎⁎ −0.182⁎ 1.000
7 GP5 0.394⁎⁎ 0.243⁎⁎ −0.064 −0.013 −0.187⁎ 0.113 1.000
8 GP8 0.378⁎⁎ 0.239⁎⁎ −0.049 −0.036 −0.199⁎ 0.158 0.975⁎⁎ 1.000

PFS, total physical fitness scores; GP5, total grade points of 5 school subjects; GP8, total grade points of 8 school subjects.
⁎ p b 0.05.
⁎⁎ p b 0.01.

4.2. Obesity and academic achievement
A previous report demonstrates that percent body fat and BMI score
were negatively correlated with academic performance in 2nd- and
4th-grade children [29]. Additionally, a recent systematic review
found relationships between obesity and cognitive function or academic
performance in children and adolescents [31]. In the present study, the
obese students showed significantly lower academic achievement com-
pared to non-obese students. These results support previous reports
that have shown a relationship between obesity and academic achieve-
ment, and also suggest that exceeding the cutoff point of adult obesity
during early adolescence has important academic consequences.
In contrast, other reports have found that obesity did not affect aca-
demic performance in children [10,27,28,30]. Although the reasons for
this inconsistency are not obvious, the differences may reflect assess-
ment methods for child obesity such as skinfold thickness [10], partici-
pant characteristics including race and growth stages [30], and/or
analysis procedure, e.g., some reports included a relatively small sample
size of obese children [27,28]. On the other hand, to the best of our
knowledge, only a few previous studies demonstrate effect size for the
influence of obesity on academic achievement in children. Based on
the effect size categories established by Cohen [35], the present levels
of effect size were classified as “small,” even after adjusting for socioeco-
nomic and behavioral confounding factors. It appears that inconsistent
results among previous studies may be at least partly responsible for
the “not large” influence of child obesity on academic achievement/cog-
nitive performance. Further study is needed to clarify the influence of
overweight/obesity on cognitive function and academic achievement
based on effect size classification.
The mechanisms behind the influence of childhood obesity on aca-
demic achievement were not clear in the present study. Academic
achievement is thought to be affected by the socioeconomic, physiolog-
ical and behavioral characteristics of each individual. Some previous
studies have shown that an undesirable daily lifestyle has a negative in-
fluence on academic achievement in overweight/obese students [14,
26]. Shore et al. [26] demonstrate that obese children tend to be absent
from school more often than non-obese children, and it has been shown
that childhood obesity is related to short sleeping time [31]. It is known
that sleep duration is related to cognitive performance [37]. It seems
that an undesirable daily lifestyle in children could lead to obesity and
less attention in classes, which then leads to low academic performance.
Additionally, obese adolescents have been shown to have a lower hip-
pocampal volume and reduced cognitive function compared with
non-obese counterparts [38]. Although the age and obesity levels of
the participants in Yau's study [38] were higher than those of the pres-
ent participants, it is possible that the early stages of the neuronal re-
sponses induced by obesity may have affected academic achievement
in the obese students in the present study.

Table 3
Summary of multiple regression analyses for variables predicting academic achievement.

Variables GP5 GP8

B SEB β p value R2 B SEB β p value R2

Male Household income 1.04 0.57 0.129 0.068 1.18 0.81 0.104 0.145
Mother's education 2.21 0.55 0.277 b0.001 2.82 0.78 0.251 b0.001
Cram school utilization −0.82 0.55 −0.104 0.140 −0.79 0.79 −0.071 0.315
Excess time spent for electronic devices −1.44 0.65 −0.156 0.028 −2.67 0.92 −0.206 b0.001
Overweight/obesity 0.52 0.77 0.047 0.503 0.35 1.10 0.023 0.748
PFS 0.13 0.03 0.329 b0.001 0.19 0.04 0.324 b0.001
Constant 13.39 1.34 0.272 22.18 1.90 0.258
Female Household income 2.58 0.58 0.347 b0.001 3.51 0.83 0.330 b0.001
Mother's education 1.27 0.57 0.169 0.028 1.84 0.82 0.172 0.026
Cram school utilization 0.27 0.58 0.036 0.640 0.49 0.83 0.045 0.557
Excess time spent for electronic devices 0.26 0.65 0.030 0.694 0.25 0.93 0.021 0.791
Overweight/obesity −1.61 0.77 −0.157 0.040 −2.33 1.11 −0.160 0.037
PFS 0.03 0.03 0.083 0.283 0.06 0.04 0.127 0.103
Constant 15.23 1.61 0.213 24.37 2.30 0.214

GP5, total grade points of 5 school subjects; GP8, total grade points of 8 school subjects; B, non-standardized regression coefficients; SEB, standard error of coefficient; P, standardized re-
gression coefficients; R2, coefficient of determination; PFS, total physical fitness test scores.

Descargado para Anonymous User (n/a) en Universidad de San Martin de Porres de ClinicalKey.es por Elsevier en octubre 12, 2018.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2018. Elsevier Inc. Todos los derechos reservados.
 N. Morita et al. / Physiology & Behavior 163 (2016) 161–166 165

Fig. 1. Relationships between adjusted PFS and adjusted academic achievement measures (panel A, GP5; panel B, GP8) after controlling for the mother's educational background,
household income, use of private teacher/cram school, obesity status, excess time spent on electronic devices. Both panel A and B show residual scatterplots by predicted values. pr,
partial correlations. ●: Male students; □: Female students.

4.3. Fitness and academic achievement
Most of the previous reports, which were carried out in European
countries and the United States, demonstrate positive correlations be-
tween academic achievement and physical fitness [7–10,12,13,16,20,
21]. The present study showed a similar association in early adolescents
in Japan. Thus, our results provide support for the concept that physical
fitness influences academic achievement in school children, and also
suggest that this concept would apply to school children in East Asian
countries. In contrast, two reports have found no significant association
between physical fitness and academic achievement [17,29]. We are un-
able to explain the reasons for this discrepancy. It is possible that gen-
der, age and/or racial differences may to some extent account for
these inconsistent results.
It is difficult to determine the possible mechanisms influencing the
relationship between fitness and academic achievement. One possible
explanation is that potential confounding factors other than the adjust-
ed confounders in the present study, such as behavioral and psycholog-
ical factors, affect individual academic achievement. In addition, brain
hippocampal volume and the brain activity of children with high aero-
bic capacity have been shown to be greater than those of children
with low aerobic capacity [11,23]. It is possible that such neurological
differences may have had an effect on the results of the more and less
fit children in the present study.
4.4. Limitations
One of the limitations of the present study is its cross-sectional re-
search design. No conclusion can be drawn from the present study
with respect to whether weight loss, daily physical activity or exercise
training might lead to improvements in academic achievement. Second,
the present research made use of socioeconomic data (parents' educa-
tion and household income) from only the approximately 67.3% of par-
ents/guardians who agreed to the use of their data. This data availability
rate was not necessarily lower than those of previous studies [10,12,18,
30], however it is possible that the results would be different if the data
of the remaining 33% of students were included. Third, although the
participants of the present study were born in the same year, not only
daily physical activity but also differences in growth could have influ-
enced their physical fitness levels as well as their academic
achievement.

Fig. 2. Comparison of adjusted academic achievement measures (A, GP5; B, GP8) in the normal weight and overweight/obese students after controlling for the mother's educational
background, household income, use of private teacher/cram school, excess time spent on electronic devices. Data are means ±standard error of mean. □: Normal weight students. ■:
Overweight/obese students.

Descargado para Anonymous User (n/a) en Universidad de San Martin de Porres de ClinicalKey.es por Elsevier en octubre 12, 2018.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2018. Elsevier Inc. Todos los derechos reservados.
166 N. Morita et al. / Physiology & Behavior 163 (2016) 161–166
5. Conclusion
In conclusion, overweight/obese Japanese junior high school stu-
dents showed lower academic achievement than those at normal
body weight. After adjusting for social and lifestyle factors such as
household income and excess time spent on electronic devices, the
physical fitness levels in boys and the obesity status in girls was associ-
ated with academic achievement. These results suggest that low physi-
cal fitness in boys and obesity in girls could be a negative factor for
academic achievement, independent of social and behavioral
backgrounds.
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.physbeh.2016.04.055.
Conflict of interest
None.
Acknowledgments
The present study was supported in part by JSPS KAKENHI Grant
Number JP24650379, JP23500737, JP26282188, JP15K12723 from the
Ministry of Education, Culture, Sports, Science and Technology of Japan.
References
[1] J.L. Baker, L.W. Olsen, T.I. Sørensen, Childhood body-mass index and the risk of cor-
onary heart disease in adulthood, N. Engl. J. Med. 357 (23) (2007) 2329–2337.
[2] P.W. Franks, et al., Childhood obesity, other cardiovascular risk factors, and prema-
ture death, N. Engl. J. Med. 362 (6) (2010) 485–493.
[3] F.B. Ortega, et al., Muscular strength in male adolescents and premature death: co-
hort study of one million participants, BMJ 345 (2012), e7279.
[4] Ministry of Education, Culture, Sports, Science and Technology of Japan, (a). Annual
Report of School Health Statistics Research of Japan. 2014 edAvailable at: http://
www.e-stat.go.jp/SG1/estat/List.do?bid=000001014499&cycode=0 Accessed July
15, 2015.
[5] Ministry of Education, Culture, Sports, Science and Technology of Japan, (b).
Annual Report of Physical Fitness and Motor Performance Statistics Research
of Japan. 2014 edAvailable at: http://www.e-stat.go.jp/SG1/estat/List.do?bid=
000001055014&cycode=0 Accessed July 15, 2015.
[6] D.M. Castelli, et al., Physical fitness and academic achievement in third- and fifth-
grade students, J. Sport Exerc. Psychol. 29 (2) (2007) 239–252.
[7] L.A. Cottrell, K. Northrup, R. Wittberg, The extended relationship between child car-
diovascular risks and academic performance measures, Obesity 15 (12) (2007)
3170–3177.
[8] C.H. Hillman, K.I. Erickson, A.F. Kramer, Be smart, exercise your heart: exercise ef-
fects on brain and cognition, Nat. Rev. Neurosci. 9 (1) (2008) 58–65.
[9] M.A. Aberg, et al., Cardiovascular fitness is associated with cognition in young adult-
hood, Proc. Natl. Acad. Sci. U. S. A. 106 (49) (2009) 20906–20911.
[10] L. Kwak, et al., Associations between physical activity, fitness, and academic achieve-
ment, J. Pediatr. 155 (6) (2009) 914–918.e1.
[11] L. Chaddock, et al., A neuroimaging investigation of the association between aerobic
fitness, hippocampal volume, and memory performance in preadolescent children,
Brain Res. 1358 (2010) 172–183.
[12] C.K. Roberts, B. Freed, W.J. McCarthy, Low aerobic fitness and obesity are associated
with lower standardized test scores in children, J. Pediatr. 156 (5) (2010) 711–718,
718.e1.
Descargado para Anonymous User (n/a) en Universidad de San Martin de Porres de ClinicalKey.es por Elsevier en octubre 12, 2018.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2018. Elsevier Inc. Todos los derechos reservados.
[13] C.L. Davis, S. Cooper, Fitness, fatness, cognition, behavior, and academic achievement
among overweight children: do cross-sectional associations correspond to exercise
trial outcomes? Prev. Med. 52 (1) (2011) S65–S69.
[14] J.U. Edwards, L. Mauch, M.R. Winkelman, Relationship of nutrition and physical ac-
tivity behaviors and fitness measures to academic performance for sixth graders in a
midwest city school district, J. Sch. Health 81 (2) (2011) 65–73.
[15] R.A. London, S. Castrechini, A longitudinal examination of the link between youth
physical fitness and academic achievement, J. Sch. Health 81 (7) (2011) 400–408.
[16] D.P. van Dusen, et al., Associations of physical fitness and academic performance
among schoolchildren, J. Sch. Health 81 (12) (2011) 733–740.
[17] M.T. Kantomaa, et al., Physical activity and obesity mediate the association between
childhood motor function and adolescents' academic achievement, Proc. Natl. Acad.
Sci. U. S. A. 110 (5) (2013) 1917–1922.
[18] H.J. Syväoja, et al., Physical activity, sedentary behavior, and academic performance
in Finnish children, Med. Sci. Sports Exerc. 45 (11) (2013) 2098–2104.
[19] D.N. Ardoy, et al., A physical education trial improves adolescents' cognitive perfor-
mance and academic achievement: the EDUFIT study, Scand. J. Med. Sci. Sports 24
(1) (2014) e52–e61.
[20] I. Esteban-Cornejo, et al., Independent and combined influence of the components
of physical fitness on academic performance in youth, J. Pediatr. 165 (2) (2014)
306-12.e2.
[21] D.M. Hansen, et al., Linear/nonlinear relations of activity and fitness with children's
academic achievement, Med. Sci. Sports Exerc. 46 (12) (2014) 2279–2285.
[22] M.B. Pontifex, et al., The effect of acute aerobic and resistance exercise on working
memory, Med. Sci. Sports Exerc. 41 (4) (2009) 927–934.
[23] L.B. Raine, et al., The influence of childhood aerobic fitness on learning and memory,
PLoS One 8 (9) (2013), e72666.
[24] A. Datar, R. Sturm, Childhood overweight and elementary school outcomes, Int. J.
Obes. 30 (9) (2006) 1449–1460.
[25] Y. Li, et al., Overweight is associated with decreased cognitive functioning among
school-age children and adolescents, Obesity 16 (8) (2008) 1809–1815.
[26] S.M. Shore, et al., Decreased scholastic achievement in overweight middle school
students, Obesity 16 (7) (2008) 1535–1538.
[27] C. Barrigas, I. Fragoso, Obesity, academic performance and reasoning ability in Por-
tuguese students between 6 and 12 years old, J. Biosoc. Sci. 44 (2) (2012) 165–179.
[28] L.J. Chen, et al., A longitudinal study of childhood obesity, weight status change, and
subsequent academic performance in Taiwanese children, J. Sch. Health 82 (9)
(2012) 424–431.
[29] K. Kamijo, et al., The relation of adiposity to cognitive control and scholastic achieve-
ment in preadolescent children, Obesity 20 (12) (2012) 2406–2411.
[30] S.D. Baxter, et al., There is no relationship between academic achievement and body
mass index among fourth-grade, predominantly African-American children, J. Acad.
Nutr. Diet. 113 (4) (2013) 551–557.
[31] J. Liang, et al., Neurocognitive correlates of obesity and obesity-related behaviors in
children and adolescents, Int. J. Obes. 38 (4) (2014) 494–506.
[32] Ministry of Education, Culture, Sports, Science and Technology of Japan, Actual Con-
ditions Survey of Learning Activity of Children after School of Japan, 2008.
[33] Ministry of Education, Culture, Sports, Science and Technology of Japan, (c). Stan-
dard operating procedure for physical fitness and exercise performance tests 12–
19 yr. edAvailable at: http://www.mext.go.jp/a_menu/sports/stamina/05030101/
002.pdf Accessed July 15, 2015.
[34] R.J. Kuczmarski, et al., 2000 CDC Growth Charts for the United States: methods and
development. Vital Health Stat 11, 246 (2002) 1–190.
[35] J. Cohen, Statistical Power Analysis for the Behavioral Sciences, second ed. Lawrence
Erlbaum Associates, Hillsdale (NJ), 1988 567 pp.
[36] F. Faul, et al., Statistical power analyses using G*Power 3.1: tests for correlation and
regression analyses, Behav. Res. Methods 41 (2009) 1149–1160.
[37] F.B. Ortega, et al., Sleep duration and cognitive performance in adolescence. The
AVENA study, Acta Paediatr. 99 (3) (2010) 454–456.
[38] P.L. Yau, et al., Obesity and metabolic syndrome and functional and structural brain
impairments in adolescence, Pediatrics 130 (4) (2012) e856–e864.