I
c
AMERICAN JOURNAL OF HUMAN BIOLOGY 10:385396 (1998)
Physical Fitness and Body Composition in Relation to Physical
Activity in Prepubescent Senegalese ChiIdren
ERIC BÉNÉFICE*
Laboratoire de Nutrition tropicale, Institut Français de Recherche
Scientifique pour le Développement en Coopération (ORSTOM), BP 5045,
34032 Montpellier, France
ABSTRACT
Y
z
9,
3
u-
!J
Q
@
n
-
Relationships among estimated body composition, habitual
physical activity, and physical fitness were considered in Senegalese children
8.5-13.5 years of age. Anthropometric dimensions (arm and calf circumferences, trunk, and extremity skinfolds, body mass index), four motor performances (dash, standing long jump, throw, grip strength), a step-test (cardiorespiratory fitness), and heart rate (HR) monitoring (physical activity) were
collected in 140 children (66 boys and 74 girls). Age and sex had a major effect
on indicators of body composition and physical fitness. Height stunting used
as an indicator of chronic undernutrition had a remarkable effect on body
composition but only a limited influence on physical fitness. Physical activity,
represented by percentage of time above the flex-HR (%fHR), did not vary
with sex, age, and nutritional status. However, there was a low-to-moderate
correlation between %fHR and several body composition indices, grip
strength, and cardiorespiratory fitness. Comparisons of children in the upper
and lower quartiles of %fHRindicated that better indices of body composition
in boys, and better strength and cardiorespiratory fitness in girls were positively associated with a higher level of physical activity. Am. J. Hum. Biol.
10:385-396, 1998. O 1998 Wiley-Liss, Inc.
Body size and proportions of growing children are factors affecting physical performance (Malina, 1994). In developed countries, there is a special concern about fatness because of its negative influence on
fitness (Parizkova, 1996). In the U.S., subcutaneous fat has increased over time (Bar
Or and Malina, 1995).Physical inactivity is
often considered a significant factor in overweight; however, inactivity is a relative concept and the level of activity needed by children t o achieve and maintain fitness is not
exactly known (Bar Or and Malina, 1995).
Paradoxically, obese children expend more
energy in physical activity per task than
normal weight children, though not M e r ing in the daily volume of physical activity
(Gazzaniga and Burns, 1993). Results of
studies on physical activity and adiposity
are equivocal. However, it seems that obese
adolescents are less active t h a n their
peers although their total energy expenditure is equal or even greater (Bar-Or and
O 1998 Wiley-Liss, Inc.
- _ .
'I
I
Baranowski, 1994). Further, the negative
effect of adiposity on fitness is more apparent at the extremes ranges of fatness; a
moderate increase of subcutaneaous fat
may have only a marginal effect on physical
performance (Beunen et al., 1993; Malina
et al., 1995).
The preceding are based on observations
of children in developed countries. The majority of the world's children, however, live
under marginal nutritional and economic
circumstances. Chronic undernutrition
leads to stunted growth and reduced body
size. Further, a high level of habitual physical activity may be needed for subsistence in
many parts of the world. Competition between the physiological requirements of the
*Correspondence to: Eric Benefice, Laboratoire de Nutrition
Tropicale, ORSTOM, BP 5045, 34032 Montpellier, France.
E-mail [email protected]
Received 24 February 1997; Accepted 21 May 1997
nos
386
E. BÉNÉFICE
children during growth and maturation,
and the social pressures for domestic or economic tasks may be assumed.
Studies of undernutrition and motor performance of children in developed countries
have been done in Mexico (Malina and BusChang, 1985; Malina and Little, 1985; Malina et al., 1991), Brazil (Rocha da Ferreira
et al., 1991; Matsudo, 1993; Anjos and Boileau, 1992), Zaïre (Ghesquière et al., 1984),
and Senegal (Benefice, 1993). These studies
generally show absolute lower performances compared with well-nourished children, but also stress the role of small body
size in the outcomes. The same applies t o
maximi4 oxygen consumption, where absolute values are lower in malnourished children, but are quite equal after normalization for body weight or lean body mass
(Spurr and Reina, 1989).
Physical activity is also a factor of variation in physical performance, especially for
aerobic power. However, the effect of a high
level of physical activity, like sport training,
is less than in adults (Livingstone, 1994;
Morrow and Freedson, 1994). Further, it is
difficult to distinguish these effects from
those of maturation during puberty (Malina, 1989).
Studies of physical activity of children in
developing countries are scarce and the impact of malnutrition is not clear. The difference in physical activity levels between undernourished. and normal children, expressed as %VO,max, in Colombia is not
significant (Spurr and Reina, 1990). However, in a previous study, these scientists
showed that during a summer camp, activity was positively affected by a meal in undernourished children, but over a shorter
period than in normal children (Spurr and
Reina, 1988).Using a HR recording method,
it was shown that physical activity levels in
Bolivian children were more influenced by
socioeconomic factors and nutritional status
than by environmental factors such as altitude (Kemper et al., 1994; Slooten et al.,
1994). In Senegal, physical activity, recorded with a direct observation method, revealed low to moderate correlations with
performance and body composition (Benefice, 1993). In moderately undernourished
children, the relationship between physical
activity and performance is modest.
There are many gaps in the knowledge of
the influence of physical activity and nutritional factors on growth and motor develop-
ment of children from developing countries.
The aim of this study is to describe variation
in motor fitness of 8.5-13.5 year old Senegalese children in relation t o anthropometric
indices of muscle and fat mass, and estimated physical activity.
SUBJECTS AND METHODS
Sampling
The study was carried out in 3 villages of
the district of Lambaye in the center of Senegal known as the “peanut basin” (14’45
north latitude and 17”30 west longitude).
This area has a typical dry and hot Sahelian
climate. Inhabitants are Muslim farmers
cultivating millet and peanuts during an
unique and brief rainy season (June to October). The nutritional situation is mediocre
with an estimated per capita energy intake
less than FAO/WHO”U requirements,
and chronic undernutrition is widely prevalent (Chevassus and Ndiaye, 1981).
A sample of 140 children (66boys and 74
girls) with a mean age of 11.3 years (SD =
1.4), rangingfrom 8.5-13.5 years of age, was
drawn from a health and fitness survey in
the area. Children were recruited after a
preliminary home-to-home census was done
in the villages. They were selected on the
basis of 1)absence of clinically detectable
abnormalities, especially in the orthopedic,
neuromotor, or nutritional spheres; 2)
known age after birth roles or determined
with an accuracy of about 3 months using a
local calendar of Muslim ceremonies and
events; and 3) consent given by the parents.
As parents were illiterate, after complete information about the aim and methods t o be
used in the survey, oral consent was given
and the presence of a familiar person during
the tests was required. The study was authorized by national and regional administrative authorities of Senegal.
Anthropometry
Children were measured according t o
techniques recommended by the International Biological Program (Weiner and Lourie, 1981). In addition to height and weight,
arm circumference (AC, cm), calf circumference (CC, cm), and four skinfold thicknesses
(mm) measured with a Holtain caliper on
the left side: triceps (TSF), biceps, subscapular, and suprailiac are reported. Measures were done three times and the average value was recorded. Extremity (triceps
+ biceps, Ext skfl and trunk (subscapular +
387
PHYSICAL FITNESS OF SENEGALESE ADOLESCENTS
I
O
1
2
3
4
5
6
7
8
9
I
I
I
1 1 . 1
I I
1
1 1213141516
10
time (min)
-$-
Boys
*Girls
Fig. 1. Cardiac response to exercise in Senegalese children (mean i 95% IC).
suprailiac, Trunk skf) skinfolds were used
as proxies for fat mass and distribution. CC
and estimated arm muscle circumference
(AMC) calculated after Gurney and Jelliffe
(1975) as AMC = AC - T x TSF, were used
as a proxy for limb muscle mass. The body
mass index (BMI, weight/stature2) was also
calculated.
Cardiorespiratory fitness
Cardiorespiratory fitness was estimated
with a step-test with 3 two-step benches
with different heights: 17, 23, and 30 cm.
This allowed a gradual increase of effort. Mter 3 minutes rest in a sitting position, the
children started going up and down the
steps at a rhythm of 30 steps per minute,
given
by a metronome. They stepped for 3
Motor fitness
minutes at each bench height; the total exMotor fitness items included 3 trials of ercise time was thus of 9 minutes. They
the following tests: 1) softball throw for dis- then rested for 5 minutes. The whole time of
tance (m); 2)standing long jump (m); and 3) the test was 17 minutes (Figure 1).Heart
hand grip strength measured by squeezing rate frequency (HR, beat-per-minutes,bpm)
a rubber bulb connected t o a manometer was recorded every 5 secondes with a Sport
(MartinR, Tuttlingen). Values were pres- TesterR monitor (Polar Electro &y, Finsure expressed in kilos Pascal (Wal. In ad- land). Children were encouraged during the
dition, children had to run a distance of 33 test by an assistant. The difference between
m with time recorded in seconds (sec). They HR recorded at the end of the exercise and
ran bare feet on a levelled sandy track. The the 1st minute of recovery was used as an
children had no previous training. The day index of cardiorespiratory endurance (1stbefore testing, the tests were demonstrated min recovery). Children were not familiar
and the children had an opportunity to prac- with the step-test, but a previous trial using
tice. To estimate the reliability of the mea- an ergocycle show us that they were more
sures, test-retest correlations were calcu- comfortable with the stepping and more calated by measuring a subsample of 40 chil- pable to produce a regular effort.
dren after a one day interval. Coefficients
Physical activity
were moderate for the dash (0.671, jump
(0.73) and grip strength (0.711, and high for
Physical activity was estimated from HR
the throw (0.93). There was no significant recording during the day. Initially it was dedifference between sexes.
cided that children should wear the Sport
388
E.BÉNÉFICE
TABLE 1. HR values in different levels of activity
(mean f SD [range]. of prepubescent
Semgalese children
HR (bpm)
Activity
Lying
Sitting
Standine
u
Light activity
Moderate activity
,
Vigourous activity
76.0 f 9.0
(56-99)
94.6 f 12.7
(66-129)
106.1 f 14.6
(74-147)
138.3 f 13.6
(99.6-174.6)
152.5 f 14.9
(107-191)
162.8 2 15.2
(116.3-194.6)
Tester for at least 12 hours, but several
problems occured: electrode displacement,
discomfort due t o the heat, and excessive
dryness of the skin at certain periods. In
practice, only 7-8 hours of recorded time
were usable. Children were instructed not
to change their habitual activities. Recording occurred between 8-10 a.m. to 4-6 p.m.
The data were then down-loaded and interfaced with a PC. The Polar ElectroR program was not used to analyze the results.
Rather, the flex-HR (DIR) method (Spun: et
al., 1988) was used. The flex-HR is the critical cut-off value of KR between resting measurements and the lowest HR recorded during exercise. The rationale is that during
exercise there 5s a linear relationship between HR and VOZ,whereas during resting
activities the correlation is very low. In the
present study, several measurements of HR
at rest while lying and sitting, and while
standing were done after at least 5 minutes
of habituation time in a group of 40 chil:
dren. Values are indicated in Table 1. It was
estimated that fHR should be set between
HR sitting and HR standing. For simplicity,
fHR was calculated from the step-test as the
average of the highest HR during the 3 minutes-rest in sitting position and the lowest
HR at the beginning of stepping. Percentage
of time over fHR was used as an overall index of activity (%f€lR).To further explore
the level of activity over fHR, the length of
time spent in light-to-moderate, moderateto-vigorous, and vigorous activity was calculated using the mean HR recorded during
(light activity), 2nd
the step-test for the lSt
(moderate activity), and 3rd(vigorous activity) benches, as thresholds. Cut-off point
values are also indicated in Table 1.
Four girls had incomplete data and were
not included in this analysis. Forty children
attended an elementary school; i n these
cases, data were collected during holidays.
There were no differences in anthropometric indices, performances, or activity patterns between children attending or not attending school, and in consequence this factor was not taken into account in the
analysis.
Statistical ana/ysis
Data were analysed using the BMDP statistical software (BMDP statistical software
Inc.R, Los Angeles CA). Means, partial correlations, 3-way analysis of variance, and
covariance analysis were used. A p value of
.O5 was accepted as significant.
RESULTS
Compared with the World Health Organization reference (WHO, 1986), children
were short in stature : median height-forage (Wage) = -1.05 z-score (range -3.51.5) and underweighted : median weightfor-age (W/age) = -1.47 z-score (-2.861.29). About a quarter of the sample was
below 2 standard deviations and less than
10%were above the median in both nutritional indices. These deficits could be attributed to mild-to-moderate malnutrition.
Given the age of the subjects, this must be a
chronic form of malnutrition and to take
into account this effect, a height-for-age (W
age) less than 1 standard deviation (-1 zscore) was chosen as cut-off point. This divides the sample into undernourished (n =
74; mean Wage = -1.6 & 0.5 z-score) and
better nourished (n = 66; mean z-score =
-0.12 f 0.7 z-score) children, and allows for
a mean difference greater than -1.5 z-score
between the 2 groups. Children were also
grouped according to the median age of the
sample: over and below 11.3 years of age.
Comparisons were done by age, sex, and nutritional group.
Girls have on average, greater values of
the anthropometric body composition indices than boys (Table 2). There is a significant ( p < 0.051, though small, interaction between age and sex for AC, AMC, and CC.
Thus, comparisons were done again in the 2
age groups separately. Girls are still larger
than boys. There is no age effect in skinfolds, and girls have larger values than
boys. The truddextremity ratio is greater
for boys. In the >11.3 year group, there is no
significant trend for undernourished boys to
389
PHYSICAL FITNESS OF SENEGALESE ADOLESCENTS
I
TABLE 2. Anthropometric characteristics (mean and SE) of Senegalese children
I
Age
N
S11.3
>11.3
9
27
Boys
Undernourished
N
16.5 (1.2)
17.4(1.1)
18
12
Girls
Normal
N Undernourished N
Arm circumference (AC, em)
17.0 (1.4)
17
16.8 (1.3)
23
17.6 (1.6)
21
18.5 (1.3)
13
Normal
3-Way ANOVAI
pprobability
17.6 (1.5)
19.9 (1.2)
s < 0.0012
A < 0.001
N i 0.001
15.5 (1.0)
17.3 (1.2)
s < 0.0012
A < 0.001
N < 0.001
Calf circumference (CC, cm)
24.7 (1.7)
23.2 (1.8)
25.4 (1.6)
16.4 (1.6)
25.0 (2.2)
17.3 (1.5)
s < 0.0012
A < 0.001
N < 0.001
Arm muscle circumference (AMC, cm)
15.4 (1.2)
15.0 (1.1)
15.9 (1.0)
16.4 (1.4)
s11.3
>11.3
14.8 (0.8)
15.8 (1.0)
S 11.3
>IL3
23.5 (1.8)
25.0 (2.0)
S11.3
>11.3
9.8 (1.9)
9.4 (1.3)
Extremity skinfolds (EA skf, mm)
9.4 (1.5)
10.8 (2.1)
10.0 (4.3)
11.4 (2.1)
12.5 (4.0)
14.0 (4.1)
s < 0.001
A = ns
N < 0.02
S11.3
>11.3
9.6 (1.7)
9.3 (1.5)
Trunlr skinfold (!Ihr&skf, mm)
9.2 (1.3)
9.9 (1.8)
10.6 (3.8)
10.9 (1.4)
11.7 (4.4)
12.9 (0.7)
A
S11.3
>11.3
0.98 (0.07)
0.99 (0.13)
".&Extremity
0.98 (0.15)
1.07 (0.09)
0.94 (0.16)
0.95 (0.23)
A = ns
e11.3
>11.3
14.8 (1.2)
14.7 (0.7)
Body mass index (BMI, kg/m2)
14.9 (1.2)
14.4 (1.2)
14.8 (1.1)
15.6 (1.5)
15.1 (1.6)
16.0 (1.3)
S11.3
>11.3
-1.59 (0.19)
-1.75 (0.11)
Wage (2-score)
-0.17 (0.13)
-1.45 (0.14)
-0.59 (0.17)
-1.75 (0.12)
0.14 (0.12)
-0.10 (0.16)
ratio (mm)
0.93 (0.17)
0.97 (0.15)
s < 0.001
= ns
N < 0.01
S < 0.04
N = ns
'
S < 0.03
A < 0.04
N = ns
s < 0.02
A < 0.001
lEffects: S = sex; A = age; N = nutritional group; ns = not significant.
2Significant interaction between age and sex.
gain more trunk than extremity fat, as it is
the case for well-nourished boys. In general,
the undernourished children have lower indices than normal children, and girls .have
larger skinfolds than boys. The BMI is not
different between the two Wage categories.
However, Wage is different between sexes
and age groups: girls are less stunted in
stature than boys, and younger children are
less stunted than older children.
Variations in motor fitness are shown in
Table 3. Grip-strength values were logtransformed because the distribution was
non-normal. Performances are better in
older children (not significant for the cardiorespiratory endurance index), and boys
have higher values than girls, except in grip
strength and cardiorespiratory endurance.
However, detailed analysis of HR during the
step test shows that boys achieve significantly lower HR values than girls during
exercise and 1st-min recovery (Figure 1).
Nutritional group has only a limited effect
in the dash and throw, where normal children perform better than undernourished
children.
The average value of HR is lower in boys
(99.6 rl: 8.0 bpm) than in girls (105.0 A 8.8
bpm), p < 0.001.There is no variation in %
fHR by sex, age, and nutritional group : children spend 52.8% f 29.7 of their daytime
over the EO3 value. However, the range of
variation is wide (0.6-99.4%).Using the cutoff values of HR indicated in Table 1, it is
apparent that children spend 48.6 27.4%
(0.3-96.4%) of their activity time without
achieving 138.3 bpm and time spent in categories of activity over 138.3 bpm is low:
light-to-moderate activities (138.2 bpm <
HR < 152.5 bpm) = 2.5 3.8% (range: 030.6%);time spent in moderate-to-vigorous
activities (152.5bpm < HR < 162.8 bpm) =
*
390
E.BÉNÉFICE
TABLE 3. Motor and physical fitness indicators (mean and SE) of Senegalese children
AEe
N
S11.3
>11.3
9
27
Boys
Undernourished N
7.0 (0.51)
6.6 (0.54)
.
18
12
Girls
Normal
N Undernourished N
.
Dash (sec)
6.7 (0.52) 17
7.8 (1.28)
23
6.6 (0.39) 21
7.5 (1.00)
13
Normal
3-Way ANOVAI
D Drobabilitv
7.1 (0.61)
6.7 (0.82)
s < 0.001
A < 0.01
N < 0.001
>11.3
1.39 (0.11)
1.45 (0.17)
1.44 (0.23)
1.46 (0.22)
Jump (m)
1.29 (0.27)
1.38 (0.21)
C11.3
>11.3
13.8 (1.6)
16.8 (3.6)
15.6 (4.3)
17.2 (3.6)
Throw (m)
10.0 (3.0)
13.6 (4.5)
G11.3
>11.3
4.58 (0.39)
4.66 (0.25)
Log Grip Strength (Ik Pa)
4.60 (0.38)
4.46 (0.29)
4.74 (0.23)
4.73 (0.29)
4.45 (0.31)
4.78 (0.21)
S = ns
A < 0.001
N = ns
C11.3
>11.3
49.6 (20.2)
50.8 (15.9)
1Min Recovery (bpm)
40.5 (17.1)
44.8 (12.3)
51.0 (20.8)
49.7 (12.4)
43.8 (14.5)
44.9 (13.5)
S = ns
A = ns
N = ns
G11.3
1.29 (0.16)
1.44 (0.22)
12.2 (2.3)
.15.3 (3.0)
s < 0.02
A < 0.02
N = ns
s < 0.001
A 0.00
N < 0.01
lEffects: S = sex; A = age; N =nutritional group. ns = not significant.
1.2 & 2.0% (range: O-11.5%); and time spent
in vigourous activities (HR > 162.8 bpm) =
0.4 f 0.9% (range: 04.2%). There are no
differences according to age, nutritional status, and sex between the different categories of activity. This suggests that these
children spend about one-half of their daytime in resting activities, and most of the
remaining time in light activities in a standing position or while walking slowly.
body compoCorrelation's between %m,
sition, and motor fitness controlling for age
and sex, are indicated in Table 4. Correlations are low but reach statistical significance for AC, extremity skinfolds, Trunk/
Extremity ratio, grip strength, and 1st-min
recovery indice. Children in the lower and
the upper quartiles of %fHR were also compared. The lower quartile (n = 34) corresponded t o a mean %fHR of 13.2 & 6.8 and
the upper quartile (n = 34) t o a mean %f€€R
of 84.0 2 7.5. Comparisons were done after
adjusting for age. Figure 2a indicates that
active boys have generally higher anthropometric indices of body composition than inactive boys. In the case of girls (Figure 2b),
no values reach statistical significance. Better fitness is also observed in active children; differences are statistically significant
only for grip strength and the throw in boys
(Figure 3a), and grip strength and 1st-min
recovery index in girls (Figure 3b).
TABLE 4. Third-order partial correlations between
habitual physical activity (%fHR),body composition,
and physical fitness, controlling for age and sex
Variable
AC
AMC
cc
Ext skf
Trunk skf
" k / E x t ratio
BMI
Dash
JUP
Throw
Log Grip Strength
1st Min Recovery
r
0.17
0.09
0.07
0.19
0.08
-0.17
0.07
-0.10
0.12
0.10
0.26
0.18
P
0.04
ns
ns
0.02
ns
0.04
ns
ns
ns
ns
0.02
0.03
DISCUSSION
Anthropometric indices of body composition of the Senegalese children varied with
age, sex, and nutritional status. Age and sex
differences in motor and cardiorespiratory
fitness were also observed, but mild-tomoderate undernutrition had only a limited
effect in the run and throw.
Girls had higher indices of fat mass than
boys. It is known that at the beginning of
puberty, maturation is associated with sex
differences in muscle, fat mass, and relative
fat distribution. Girls gain more subcutaneous fat on the extremities than boys, and the
Trunmxtremity ratio increases more in
'
I
J
t
il
Activity and Body Composition
a) Boys
30
Indices of b o d y composition
.
25
T
p<0.05
20
15
10
5
O
AC
Ext s k f
AMC
Trunk s k f
BI
Activity and Body Composition
b) Girls
30
Indices o f b o d y composition
25
20
15
10
5
O
AC
(cml
AM C
(cm1
IIOW
cc
(cm)
activity
Fig. 2. Activity and body composition (mean
Ext s k f
(")
Trunk s k f
("1
BMI
(kg/m2)
high activity
1 SD): a) Senegalese boys; b) Senegalese girls.
Activity and Physical Performances
a) Boys
25
Physical performances
I
2ol
T
i-
10
5
- -,-Jump
(m/10)
O
33m-run
(sec)
O IOW
Throw
(m)
activity
Log strength
(IKpa)
Recovery
(bpm*lO)
high activity
Activity and Physical performances
b) Girls
20
Physical performances
T
-
15
I
T
T
10
p<0.02
5
o-
33m-run
(sec)
Jump
(m/10)
O IOW
Throw
(m)
activity
Log strength
. (IKpa)
Recovery
(bpm*lO)
high activity
Fig.3. Activity and motor fitness (mean f 1SD): a) Senegalese boys; b) Senegalese girls.
boys (Malina, 1996).Hormonal regulation of
growth differs between boys and girls and
accounts
. for differences
...
.. in fat
. mass. More.
1.
n
sensitive t o nutritional influence; higher
IGFl in girls allows them to be better prepared for
the metabolic demands of preg,11
.
1
d....”.
m
3
1
PHYSICAL FITNESS OF SENEGALESE ADOLESCENTS
acteristics may give girls an advantage compared with boys during nutritional stress
such as chronic undernutrition. This could
explain why Senegalese prepubescent girls
have better Wage than boys; however this
could also mean that these Senegalese girls
are more advanced in pubertal growth than
boys.
Girls had higher indices of muscle mass
than boys; this may explain the similarity of
grip strength in both sexes. Strength is proportional to the cross-sectional area of a
muscle (Astrand and Rodhal, 19771, represented in this study by AMC. However, such
relationships may not be observed in immature children, who are sometimes termed
“metabolic non-specialists” (Bar Or, 1983).
Actually, while having less arm muscle
mass boys were better than girls in most
motor fitness items, emphasizing that other
factors such as skill, training, physical activity, and motivation play an important
role.
The higher values of HRs of girls during
exercise are consistent with other reports
(Strong et al., 1990). They are sometimes
attributed t o the smaller lean body mass
and low level of hemoglobin of girls compared with boys. It is also said that girls
have a lower stroke volume and that autonomic cardiac regulation is different (Strong
et al., 1990). HRs of girls paralleled those of
boys and, as a result, there was no significant difference in the 1st-min recovery indice. However, the lower HRs in boys at the
end of exercise and during rest, suggested
better cardiorespiratory efficiency.
Interindividual variation in maturation
at the onset of puberty also needs t o be considered because, especially in the case of
boys; early maturers tend to be stronger and
better in motor performance than late maturers (Malina, 1989). This could explain
the limited effect of nutritional status on
motor fitness in this study. The occurrence
of puberty in this area of Senegal is delayed
i n comparison with Europe o r North
America (Simon et al., 1995). While the survey design did not permit the assessment of
pubertal status, some children of both sexes
were obviously more advanced than other.
Because of the interrelation of growth and
maturation events at the beginning of puberty, the use and interpretation of anthropometric indices of nutritional status like
Wage or the BMI can thus be misleading.
Physical activity is a factor frequently invoked in the explanation of sex differences
393
in motor and aerobic fitness. The length of
time spent in vigorous activity appears to be
important (Verschuur and Kemper, 1985;
Sunnegardh et al., 1987). In the present
study, no differences in physical activity between sexes and by age and nutritional status were observed. This may be due t o the
lack of accuracy and precision of the HR
method in evaluating physical activity in
this population. It should be emphasized
that children of this study were mildly or
moderately, but not severely, undernourished and the difference in nutritional status between subjects may be too weak to
affect levels of physical activity. Advantages
and limitations of the HR method had been
extensively reviewed (Saris, 1986; Verschuur and Kemper, 1985; Janz et al., 1992).
The method is not considered accurate
enough to estimate daily energy expenditure of individuals (Li et al., 1993). However, it is useful in group studies or for epidemiologic purposes (Kalkwarf et al., 1989;
Spurr et al., 1988; Livingstone et al., 1992).
Further, in West Africa, it is often difficult
to investigate physical activity without interfering with subject behavior. HR recording appears t o be an acceptable alternative
method. The main limitation of the present
study was the difficulty of monitoring HR
for long periods. In developed countries,
long periods of recording are necessary because there are substantial differences in
activity levels between school days and holidays, and from one day to another (Falgairette et al., 1996). In rural Senegal, patterns of habitual physical activity are quite
monotonous. In another study of physical
activity using an observational method, no
changes in intensity between days and also
between different seasons of the year were
observed, while nature of tasks changed
dramatically because children were involved in agricultural tasks during the
rainy season (Benefice, 1993). A popular
way to express physical activity from HR is
to determine fixed limits to separate mild,
moderate, and vigorous activity, for example 140 and 160 bpm (Amstrong and
Bray, 1991); in contrast, individually calibrated indicators have the advantage of taking into account the underlying cardiorespiratory fitness of the child (Panter Brick et
al., 1996). However, it must be emphasized
that %MR values do not provide direct estimates of physical activity; rather, the reflect individual adaptation t o physical activity (Panter Brick et al., 1996).
394
E. BÉNÉFICE
As a rule, the overall level of activity of
the Senegalese children was low. While they
spent 50% of their time over the fHR, they
had less than 10 min per day in moderate to
vigorous activity. This is consistent with
other observations from Senegal: in a previous study, children spent only 4% of their
time above 140 bpm, with differences between sexes (Benefice, 1992); Diaham and
colleagues (1992) noticed in a rural district
closed t o this area that children achieved
140 bpm during only 2.4% of the day for
girls and 2.8% for boys. The present study
was done at the end of the dry season where
only a few agricultural tasks were executed
and where heat was torrid. This could favor
a sedentary behavior. However, extreme
heat is not rare in this area and never prevents people to be normally active. Similar
low levels of HR patterns were observed in
children from other developing countries
(Panter Brick et al., 1996). A low level of
activity was also reported in industrialized
countries using the HR method; school children experienced relatively short periods of
intense activity (Amstrong and Bray, 1991;
Gilliam et al., 1981).This may reflect a bias
in the method because in children, vigorous
activity occurs in short bouts of time, perhaps less than 10 sec, and the habitual recording period of 1min. may be too long t o
detect small changes. Further, children
have the ability to recover quickly after exercise (Livingstone et al., 1992).
The low relationships between estimated
physical activity and motor fitness variables
may also be the consequence of a ceiling effect due to the lack of variation in physical
activity in this sample. As a rule however,
such relationships are weak during childhood, especially in the case of cardiorespiratory fitness (Livingstone, 1994). Interestingly, habitual physical activity seems to be
a greater predictor of cardiorespiratory fitness in obese children than in normal children (Taylor and Baranowski, 1991). When
Senegalese children were contrasted according t o the lower and higher quartiles of fHR,
relationships between activity and AC, extremity skinfolds, and the BMI were observed in boys but not in girls. More active
boys appeared t o have more subcutaneous
fat and estimated total fatness. This implies
perhaps that active boys were in better nutritional status than inactive boys. Conversely, a low level of activity may be a way
for undernourished boys t o meet their en-
ergy requirements. In the case of undernourished girls, it is likely that their important participation in domestic tasks does
not permit a reduction in habitual activity
to the same extent (Bénéfice, 1993).
In both sexes grip strength is greater in
active children. Cardiorespiratory fitness is
influenced by activity in girls, but not in
boys. This could mean that boys were fit
enough and their level of activity was insufficient to cause a further increase in cardiorespiratory performance. In contrast, lower
cardiorespiratory fitness of girls make them
more sensitive t o the training effect of
higher physical activity.
This study suggests that physical activity
in Senegalese children, as in children from
industrialized countries, is not an important determinant of cardiorespiratory fitness and motor fitness. However, better nutritional status in boys, and motor fitness in
both sexes are positively associated with a
higher level of activity.
ACKNOWLEDGMENTS
This work was supported by the UR 44 of
the Institut Français de Recherche Scientifique pour le Développement en Coopération (ORSTOM). I thank my research assistants Daouda Ndiaye, Modou m o w , and
Omar Sall, and all the parents and children
of Ndiokhane, Ndondol, Dodel, Thialaga,
and Diamandou for their participation in
the surveys. I a m especially indebted to the
editor for his assistance and suggestions on
the manuscript.
LITERATURE CITED
Amstrong N, Bray S (1991) Physical activity patterns
defined by continuous heart rate monitoring. Arch.
Dis. Child. 66:245-247.
Anjos LA, Boileau RA,Misner JE (1992) Maximal mechanical aerobic and anaerobic power output of lowincome Brazilian schoolchildren as a function of
growth. Am. J. Hum. Biol. 4:647-656.
Astrand PO, Rodahl K (1977) Textbook of Work Physiology. New York McGraw-Hill.
Bar-Or O (1983) Pediatrics Sports Medicine for the
Practitioner. Berlin, Heidelberg: Springer Verlag.
Bar-Or O, Malina RM (1995) Activity, fitness, and
health of children and adolescents. In Cheung LWY,
Richmond J B (eds): Child Health, Nutrition and
Physical Activity. Champaign; H i Human Kinetics,
pp. 79-123.
Bar-Or O, Baranowski T (1994) Physical activity, adiposity, and obesity among adolescents. Pediatr. Exerc. Sci. 6:348-360.
Bénéfice E (1992) Physical activity and anthropometric
and functional characteristics of mildly malnourished
Senegalese children. AM. Trop. Paediatr. 12:55-66.
Bénéfice E (1993) Physical activity, cardiorespiratory
PHYSICAL FITNESS OF SENEGALESE ADOLESCENTS
fitness, motor performance, and growth of Senegalese
pre-adolescents. Am. J. Hum. Biol. 5653467.
Beunen G, Malina RM, Renson R, Simons J, Ostyn M,
Lefevre J (1992) Physical activity and growth, maturation and performance: a longitudinal study. Med.
Sci. Sport Exer. 24:57&585.
Chevassus-Agnès S, Ndiaye AM (1981) Enquêtes de
consommation alimentaire de YORANA de 1977 à
1979: méthodologie et résultats. In Centre de Recherches pour le Développement International (ed):
Etat nutritionnel de la Population Rurale du Sahel.
Ottawa: C.R.D.I., pp. 57-66.
Diaham B, Wade S,Guiro A, Leite M, Sal1M, Coward A,
Prentice A, Ritz P, Lemonnier D (1992) Activité physique de dépense énergétique en période de travaux
agricoles dans les familles rurales du Senegal. 5em
Journées Internationales du Groupe d’etudes et recherches sur les Malnutritions (GERM). Balaruc.
Sept. 22-28, 1992.
Faleairette G. G a v m O. Bernard T. Hebbelinck (1996)
Eialuationof habi&al physical activity from a week’s
heart rate monitoring in French school children. Eur.
J. Appl. Physiol. 74:153-161.
Gazzaniga JM, Burns TL (1993) Relationship between
diet composition and body fatness, with adjustment
for resting energy and physical activity, in preadolescent children. Am. J. Clin. Nutr. 58:21-28.
Ghesquière J, Eeckels R (1984) Health, physical development of primary school children in Kinshasa. In
Ilmarinen J, Valimaki I (eds): Children and Sport.
Berlin: Springer Verlag, pp. 18-30.
Gilliam T, Freedson PS, Greenen DL, Shahraray B
(1981) Physical activity patterns determined by heart
rate monitoring in 6-7 year-old children. Med. Sci.
Sport Exerc. 1316547.
Gurney J, Jelliffe DB (1973) Arm anthropometry in nutritional assessment Nomogram for rapid calculation
of muscle circumference and cross-sectional muscle
and fat areas. Am. J. Clin. Nutr. 261912-915.
Kalkwarf H, Haas J, Belko AM, Roach RC, Roe DA
(1989)Accuracy of heart rate monitoring and activity
diaries for estimating energy expenditure. Am. J.
Clin. Nutr. 49:37-43.
Kemper HCG, Spekreijse M, Slooten J , Post GB, Welten
DC, Coudert J (1996) Physical activity in prebuscent
children: relationship with residential altitude and
socioeconomic status. Pediatr. Exerc. Sci. 857-68.
Janz KF, Golden JC, Hansen JR, Mahoney LT (1992)
Heart rate monitoring of physical activity in children
and adolescents: the Muscatine study. Pediatrics
89:256-261.
Li R, Deurenberg P, Hautvast J (1993) A critical evaluation of heart rate to assess energy expenditure in
individuals. Am. J. Clin. Nutr. 58:602-607.
Livingstone MBE, Coward WA, Prentice AM, Davies
PSW, Strain J J , McKenna PG, Mahoney CA, White
JA, Stewart CM, Kerr MJ (1992) Daily energy expenditure in free-living children: comparison of heartrate monitoring with t h e doubly labeled water
(2&180)
method. Am. J. Clin. Nutr. 56:343-352.
Livingstone MBE (1994) Energy expenditure and physical activity in relation to fitness in children. Proc.
Nutr. Soc. 531207-221.
Malina RM (1989) Growth and maturation: normal
variation and effect of training. In Gisolfi CV and
Lamb DR (eds): Perspectives in Exercise Science and
Sports Medicine, Volume 2: Youth, Exercise, and
Sport. Indianapolis: Benchmark Press, pp. 223-265.
Malina RM (1994) Anthropometry, strength, and motor
fitness. In Ulijaszek SJ, Mascie-taylor CGN (eds.):
Anthropometry: The Individual and the Population.
395
Cambridge: Cambridge University Press, pp. 160177.
Malina RM (1996) Regional body composition: age, sex,
and ethnic variation. In Roche AF, Heymsfield SB,
Lohman TG (eds.): Human Body Composition. Champaign, I L Human Kinetics, pp. 217-255.
Malina RM, Beunen GP, Claessens AL, Lefevre J,
Eynde BV, Renson R, Vanreusel B, Simons J (1995)
Fatness and physical fitness of girls 7-17 years. Obes.
Res. 3:221-231.
Malina RM, Buschang PH (1985) Growth, strength, and
motor performance of Zapotec children, Oaxaca,
Mexico. Hum. Biol. 57:163-181.
Malina RM, Little BB (1985) Boyd composition,
strength, and motor performance in undernourished
boys. In Binkhorst RA, Kemper HCG, Saris WHM,
(eds): Children and Exercise XI. Champaign, IL:Human Kinetics, pp. 293-300.
Malina RM, Little BB, Buschang PH (1991) Estimated
body composition and strength of chronically undernourished rural boys in Southern Mexico. In Shephard RJ, Parizkova J (eds): Human Growth, Physical
Fitness and Nutrition. Basel: Karger, pp. 119-132.
Matsudo VKR (1993) Physical fitness in developing
countries. In Claessens AL, Lefevre J , Van den Eynde
B (eds): World-Wide Variation in Physical Fitness.
Leuven: Institute of Physical education, pp. 111-125.
Morrow JR,Freedson PS (1994) Relationship between
habitual physical activity and aerobic fitness in adolescents. Pediatr. Exerc. Sci. 6r315-329.
Panter-Brick C, Todd A, Baker R, Worthman C (1996)
Comparative study of flex heart rate in three samples
of Nepali children. Am. J. Hum. Biol. 8:653-660.
Parizkova J (1996) Nutrition, Physical Activity, and
Health in Early Life. New York CRC Press, pp. 169188.
Rocha-Ferreira MB, Malina RM, Rocha LL (1991) Anthropometric, functional, and psychological characteristics of eight-year-old Brazilian school children
from low socioeconomic status. I n Shephard RJ,
Parizkova J (eds): Human Growth, Physical Fitness,
and Nutrition. Karger: Basel, pp. 109-118.
Rolland-Cachera F (1995) Prediction of adult body composition from infants and child measurements. In
Davies PSW, Cole TJ (eds): Body Composition Techniaues in Health and Disease. Cambridge:
- Cambridge
Uiiversity Press, pp. 100-145.
Saris WHM (1986) Habitual Dhvsicd activitv in children: methodolo& and findiGgsin health anä disease.
Med. Sci. Sport Exer. 18:253-263.
Simon I, Simondon K, Simondon F (1995) Enquête nutritionnelle d‘adolescents d’une région rurale au Senegal: relations avec leur état nutritionnel à Yâge préscolaire. Rencontres Francophones d‘Auxologie, Lyon,
NOV.13-14, 1995.
Slooten J, Kemper HCG, Post GB, Lujan C, Coudert J
(1994) Habitual physical activity in 10- to 12-year-old
Bolivian boys. The relation between altitude and socioeconomic status. Int. J. Sports Med. 15.-S106slll.
Spurr GB, Prentice AM, Murgatroyd PR, Goldberg GR,
Reina JC, Christman BS (1988) Energy expenditure
from minute-by-minute heart rate recording: comDarison with indirect calorimetrv. Am. J. Clin. Nutr.
28.5524359.
Spurr GB, Reina JC (1988) Influence of dietary intervention on artificially increased activity in marginally
undernourished ColÕmbian boys. Eur: J. Clin: Nutr.
42:835-846.
Spurr GB, Reina JC (1990) Daily pattern of %VOZand heart rates in normal and undernourished school
children. Med. Sci. Sport Exer. 22.-643-652.
396
E. BÉNÉFICE
Spurr GB, Reina JC (1989) Maximum oxygen consumption in marginally malnourished Colombian boys and
girls 6-16 years of age. Am. J.Hum. Biol. k11-19.
Strong WB,StanitskiCL, Smith RE, Wilmore J
H (1990)
Cardiovascular responses to exercise in childhood.
Am. J. Dis. Child. 114:1255-1260.
Sunnegardh J, Bratteby LE (1987) Maximal oxygen uptake, anthropometry, and physical activity in a randomly selected sample of 8-13 year old children in
Sweden. Eur. J.Appl. Physiol. 56:266-272.
Taylor W, Baranowski T (1991) Physical activity, cardiovascular fitness, and adiposity in children. Res.
Quart. Exer. Sport. 62:157-163.
Verschuur R, Kemper HCG (1985) Habitual physical
activity. Med. Sports Sci. 20:56-65.
Weiner JS, h u n e JA (1981) Practical Human Biology.
London: Academic Press.
World Health Organization (1983) Measuring Change
in Nutritional Status. Geneva: World Health Organization.