HY. 52.
1....» .3

WWW-r

:F
mi

.fi, .m;
. .3wflrfiea. 2.2,.
s u... fiafimfifi

M
u
19

e
i-

It}...

I“

{3| .1: 741.12.:
15.55,»):
I. 1 . an .Idl

. 4.12;). .l‘ .313;
5!?!

T1 HES IS

This is to certify that the

dissertation entitled -

ANTHROPOMETRY, PHYSIQUE, AND PHYSICAL FITNESS
OF 6 TO 11 YEAR OLD CHILDREN FROM A RURAL
AND AN URBAN COMMUNITY IN OAXACA, SOUTHERN MEXICO

‘
.5.

presented by

Swee Kheng Tan

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in Kinesio logy

 

 

mam

 

. Major professor 0

Date {/41};

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN REI'URN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

J“ 22 mi

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DateDue.p65-p.15

ANTHROPOMETRY, PHYSIQUE, AND PHYSICAL FITNESS OF 6 TO 11
YEAR OLD CHILDREN FROM A RURAL AND AN URBAN COMMUNITY
IN OAXACA, SOUTHERN MEXICO

By

Swee Kheng Tan

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Kinesiology

2002

ABSTRACT
ANT HROPOMETRY, PHYSIQUE, AND PHYSICAL FITNESS OF 6 TO 11
YEAR OLD CHILDREN FROM A RURAL AND AN URBAN COMMUNITY IN
OAXACA, SOUTHERN MEXICO

By

Swee Kheng Tan

The present study investigated the physical fitness of rural and urban school
children in Oaxaca, southern Mexico, in the context of anthropometric correlates,
somatotype, and nutritional status. First, the contributions of body size and specific
anthropometric dimensions to variation in indicators of health- and performance-related
fitness were estimated. Second, relationships between somatotype and indicators of
fitness were considered. Third, the growth status, physical fitness and relative fat
distribution of stunted (height z-score below -2.0) and non-stunted children (height 2-
score equal to or greater than -2.0) were compared.

The sample comprised 329 rural (154 boys, 175 girls) and 318 urban (161 boys,
157 girls) school children 6.00-1 1.99 years of age. Anthropometric dimensions included
weight, height, sitting height, four skeletal breadths, two limb circumferences, four
skinfolds, and several derived variables. Somatotype was estimated with the Heath-Carter
anthropometric protocol. Fitness items included right and left grip (strength), 35 yard
dash (32.3 m, speed), standing long jump (explosive power), sit-and-reach (flexibility),
timed sit-ups (30 seconds, abdominal strength and endurance), and a distance run (8
minutes in grades 1-3, 12 minutes in grades 4-6, cardiovascular endurance). Partial

correlation, multiple regression, analysis of covariance, multiple analysis of covariance,

and principal components analysis were used in the comparisons of rural and urban and
of stunted and non-stunted children.

Age, height, and weight explained most of the variance in the performances of
rural and urban children. The addition of other anthropometric dimensions to the
regression accounted for only slightly more of the variance and altered the estimated
contributions of age, height and weight. The increase in the amount of variance explained
in fitness was significant in the sum of right and left grip strength while the significance
in the variance increased for the other fitness items were variable. There were no urban-
rural contrasts in estimates.

Rural and urban boys, but not girls, differed in somatotype. Endomorphy
discriminated urban and rural boys. Endomorphy was negatively correlated to fitness,
whereas mesomorphy and ectomorphy were variably correlated with fitness. Age and
somatotype accounted for a major portion of the variance in fitness, but estimated
contributions varied with the tests.

Non-stunted children were larger than stunted children in height (as expected) and
also in all other dimensions and derived variables except for the sitting height to standing
height ratio, which indicated proportionally shorter legs in stunted children. Non-stunted
children performed better than stunted children on only one item, absolute strength.
Strength per unit estimated arm muscle was also greater in non-stunted children. Other
fitness items did not significantly differ between stunted and non-stunted children.
Relative subcutaneous fat distribution (trunk-extremity skinfold ratio and principal

components) did not differ between stunted and non-stunted children.

Copyright by
TAGQEDWTHEBUHEBKE

2002

ACKNOWLEDGMENTS

The present dissertation was supervised by Dr. Robert M. Malina of the Department of
Kinesiology, Michigan State University. I am sincerely grateful to Dr. Malina for his
scholarly advise, encouragement, and support throughout my dissertation. Dr Malina’s
critical reading and constructive comments of this dissertation has greatly improved the

final version of this dissertation.

To the other members of my doctoral committee, Dr. Gail Dummer, Dr. Deborah Kashy,
Dr. Nigel Paneth, and Dr. Christopher Womack, I am appreciative for the generosity of

their time and guidance.

Special thanks to Dr. Maria E. Pefia Reyes and colleagues in the physical anthropology
department at Escuela Nacional de Antropologia e Historia (ENAH), Mexico, for their

invaluable contribution during data collection.

Table of Contents

Page
CHAPTER 1
INTRODUCTION ............................................................................... l
SIGNIFICANCE OF STUDY .................................................................... 4
RESEARCH QUESTIONS .................................................................... 6
HYPOTHESES ................................................................................... 7
DELIMITATIONS .............................................................................. 9
LIMITATIONS ................................................................................. 10
BIBLIOGRAPHY .............................................................................. 11
CHAPTER 2 - GENERAL REVIEW OF LITERATURE
NATURE OF MALNUTRITION ........................................................... l6
UNDERNUTRITION AND GROWTH .................................................... 19
UNDERNUTRITION AND PERFORMANCE ........................................... 23
STUNTING AND FAT DISTRIBUTION ................................................. 26
RURAL-URBAN DIFFERENCES IN GROWTH ....................................... 28
RURAL-URBAN DIFFERENCES IN PERFORMANCE .............................. 30
ANTHROPOMETRIC CORRELATES OF PERFORMANCE ........................ 31
Body Dimensions .......................................................................... 31
Physique ..................................................................................... 35
Body Composition ........................................................................ 39
BIBLIOGRAPHY ............................................................................. 43
CHAPTER 3 — METHODS AND PROCEDURES
COMMUNITIES STUDIED .................................................................. 57
PARTICIPANTS .............................................................................. 59
ANTHROPOMETRIC VARIABLES ...................................................... 6O
DERIVED VARIABLES ..................................................................... 61
PHYSICAL FITNESS ........................................................................ 63
MEASUREMENT VARIABILITY AND RELIABILITY ............................... 65
SELECTION AND TRAINING OF ASSISTANTS/EXAMINERS ................... 65
STATISTICAL ANALYSES ................................................................ 66
BIBLIOGRAPHY .............................................................................. 71

vi

CHAPTER 4 - CORRELATES OF PHYSICAL FITNESS IN RURAL AND
URBAN CHILDREN IN SOUTHERN MEXICO

INTRODUCTION ............................................................................. 77
METHODS AND PROCEDURES .......................................................... 79
RESULTS ....................................................................................... 80
DISCUSSION .................................................................................. 83
BIBLIOGRAPHY .............................................................................. 89

CHAPTER 5 — SOMATOTYPES OF RURAL AND URBAN CHILDREN IN

SOUTHERN MEXICO
INTRODUCTION ........................................................................... 106
METHODS AND PROCEDURES ........................................................ 108
RESULTS ..................................................................................... 110
DISCUSSION ................................................................................. l 1 1
BIBLIOGRAPHY ............................................................................ 114

CHAPTER 6 — RELATIONSHIP BETWEEN SOMATOTYPE AND PHYSICAL
FITNESS IN RURAL AND URBAN MEXICAN CHILDREN

INTRODUCTION ........................................................................... 126
METHODS AND PROCEDURES ........................................................ 128
RESULTS ..................................................................................... 129
DISCUSSION ................................................................................. 130
BIBLIOGRAPHY ............................................................................ 1 34

CHAPTER 7 — GROWTH AND PHYSICAL FITNESS OF STUNTED AND NON-
STUNTED IN A RURAL AND AN URBAN COMMUNITY IN

OAXACA, MEXICO
INTRODUCTION ........................................................................... 140
METHODS AND PROCEDURES ........................................................ 142
RESULTS ..................................................................................... 143
DISCUSSION ................................................................................. 145
BIBLIOGRAPHY ............................................................................ 147

vii

CHAPTER 8 - SUBCUTANEOUS FAT DISTRIBUTION OF STUNTED AND
NON-STUNTED C I ILDREZN FROM tAX.—KCA, MEXICO

INTRODUCTION ........................................................................... 1 6 1
METHODS AND PROCEDURES ......................................................... 163
RESULTS ..................................................................................... 1 64
DISCUSSION ................................................................................. 165
BIBLIOGRAPHY ............................................................................ 168

CHAPTER 9 - GENERAL SUMMARY, DISCUSSION AND CONCLUSION

SETTING AND METHODS ............................................................... 174

DISCUSSION ................................................................................. 175
Question 1 ................................................................................. 175
Question 2 ................................................................................. 178
Question 3 ................................................................................. 179
Question 4 ................................................................................ 181

CONCLUSION ............................................................................... 184

RECOMMENDATIONS .................................................................... 1 85

BIBLIOGRAPHY ............................................................................ 1 87

APPENDICES

Appendix
A. Letter of UCRIHS Approval ...................................................... 191
B. Protocol for Anthropometry and Fitness Testing ............................... 193

viii

LIST OF TABLES

CHAPTER 2
Table 2.1. Summary of studies that used second-order partial correlations between age,

height, weight, and physical fitness tasks in children 6-15 years of age ...... 55

Table 2.2. Mean somatotypes for children 6 to 11 years of age in several Latin American

countries ................................................................................ 56

CHAPTER 3 _
Table 3.1. Indicators of social, economic, educational and health conditions in the rural
and urban communities with comparative data for the city of Oaxaca and the

state of Oaxaca based on the national census for 2000 ........................... 75

Table 3.2. Numbers of stunted and non-stunted children 6-1 1 years of age by sex within

each community ........................................................................ 76

CHAPTER 4

Table 4.1. Age-adjusted means and standard errors of anthropometric and derived
variables in rural and urban boys 6-11 years of age, and F ratios for the
ANCOVAs with age as the covariate .............................................. 93

Table 4.2. Age-adjusted means and standard errors of anthropometric and derived
variables in rural and urban girls 6-11 years of age, and F ratios for the
ANCOVAs with age as the covariate .............................................. 94

Table 4.3. Age-adjusted means and standard errors of physical fitness and derived
variables in rural and urban boys 6-11 years of age, and F ratios for the
ANCOVAs with age as the covariate .............................................. 95

Table 4.4. Age-adjusted means and standard errors of physical fitness and derived
variables in rural and urban girls 6-11 years of age, and F ratios for the

ANCOVAs with age as the covariate .............................................. 96

Table 4.5. Second-order partial correlations between age and body size, and physical

fitness items in rural and urban boys 6-11 years of age .......................... 97

Table 4.6. Second-order partial correlations between age and body size, and physical

fitness items in rural and urban girls 6-11 years of age .......................... 98

Table 4.7. Multiple regression of age, height, and weight on physical fitness items in

rural and urban boys 6-11 years of age ............................................. 99

Table 4.8. Multiple regression of age, height, and weight on physical fitness items in
rural and urban girls 6-11 years of age ........................................... 100

Table 4.9. Beta coefficients from multiple regressions of age, height, weight and other
anthropometric characteristics on physical fitness items in rural boys 6-11

years of age ........................................................................... 101

Table 4.10. Beta coefficients from multiple regressions of age, height, weight, and
additional anthropometric characteristics on physical fitness items in urban

boys 6-11 years of age ............................................................ 102

Table 4.11. Beta coefficients from multiple regressions of age, height, weight, and
additional anthropometric characteristics on physical fitness items in rural

girls 6-11 years of age ............................................................ 103

Table 4.12. Beta coefficients from multiple regressions of age, height, weight, and
additional anthropometric characteristics on physical fitness items in urban

girls 6-11 years of age ............................................................. 104

Table 4.13. Summary of studies that used second-order partial correlations between age,

height, weight, and physical fitness tasks in children 6-15 years of age 105

CHAPTER 5
Table 5.1. Reproducibility of anthropometric estimates of somatotype (n = 79) ........ 117
Table 5.2. Means and standard deviations of somatotype in rural boys by age group .. 118

Table 5.3.

Table 5.4.

Table 5.5.

Table 5.6.

Table 5.7.

Table 5.8.

Table 5.9.

Means and standard deviations of somatotype in urban boys age group 1 19

Means and standard deviations of somatotype in rural girls by age group 120

Means and standard deviations of somatotype in urban girls by age group

.......................................................................................... 121

Age-adjusted means and standard errors, and rural-urban differences in

somatotype of rural and urban children 6-11 years of age ..................... 122

Age-adjusted means and standard errors, and sex differences in somatotype of

rural and urban children 6-11 years of age ....................................... 123

Summary of forward stepwise discriminant function analyses for rural-urban

and sex differences in somatoype ................................................. 124

Mean somatotypes of children 6-11 years of age in the present study and

samples of Latin American children .............................................. 125

CHAPTER 6
Table 6.1. Third-order partial correlations of age and somatotype, and physical fitness in

Table 6.2.

rural and urban boys 6-11 years of age .......................................... 136

Third-order partial correlations of age and somatotype, and physical fitness in
rural and urban girls 6-11 years of age ........................................... 137

xi

Table 6.3. Multiple regression of physical fitness on age and somatotype in rural and

urban boys 6-11 years of age ....................................................... 138

Table 6.4. Multiple regression of physical fitness on age and somatotype in rural and

urban girls 6-11 years of age ....................................................... 139

CHAPTER 7
Table 7.1. Prevalence of stunting in 6-11 year old children in the rural and urban

communities .......................................................................... 1 50

Table 7.2. Age-adjusted means and standard errors of anthropometric and derived

variables of stunted and non-stunted rural boys 6-11 years of age ............ 151

Table 7.3. Age—adjusted means and standard errors of anthropometric and derived
variables of stunted and non-stunted rural girls 6-11 years of age ............ 152

Table 7.4. Age-adjusted means and standard errors of anthropometric and derived

variables of stunted and non-stunted urban boys 6-11 years of age .......... 153

Table 7.5. Age-adjusted means and standard errors of anthropometric and derived

variables of stunted and non-stunted urban girls 6-11 years of age ........... 154

Table 7.6. Age-adjusted means and standard errors of somatotype in stunted and non-

stunted rural and urban boys 6-11 years of age ................................. 155

Table 7.7. Age-adjusted means and standard errors for somatotypes of stunted and non-

stunted rural and urban girls 6-11 years of age .................................. 156

Table 7.8. Age-adjusted means and standard errors of physical fitness and derived

variables of stunted and non-stunted rural boys 6-11 years of age ............ 157

Table 7.9. Age-adjusted means and standard errors of physical fitness and derived

variables of stunted and non-stunted rural girls 6-11 years of age ............ 158

xii

Table 7.10. Age-adjusted means and standard errors of physical fitness and derived

variables of stunted and non-stunted urban boys 6-11 years of age 159

Table 7.11. Age-adjusted means and standard errors or physical fitness and derived

variables of stunted and non-stunted urban girls 6-11 years of age ......... 160

CHAPTER 8
Table 8.1. Age-adjusted means and standard errors for indicators of body size,
subcutaneous fatness and relative subcutaneous fat distribution of stunted and

non-stunted boys, 6-11 years of age, within each community ................. 170

Table 8.2. Age-adjusted means and standard errors for indicators of body size,
subcutaneous fatness and relative subcutaneous fat distribution of stunted and

non-stunted girls, 6-11 years of age, within each community ................. 171

Table 8.3. Summary of the principal components analysis: Loadings on the four

skinfolds on each principal component ........................................... 172

Table 8.4. Age-adjusted means and standard errors of scores on the two principal

components for stunted and non-stunted rural and urban children by sex .173

xiii

LIST OF FIGURES

CHAPTER 3
Figure 3.1. Map of the Valley of Oaxaca, indicating the location of the communities
studied ................................................................................. 74

xiv

CHAPTER 1

Introduction

The present study is an extension of a broader study of secular change in the
growth status and physical fitness of school children and living conditions of two
communities in the Valley of Oaxaca, Southern Mexico, which were initially studied in
1968 and 1972, and more recently in 1999-2000 (Malina et al., 1972, 1980; Malina and
Selby, 1982; Malina, 1999, 2002; Perla Reyes, 2002). Both communities have a history of
chronic undernutrition and associated marginal living conditions. The current study
examines in more detail issues related to the physical fitness of children living in the rural
and urban communities.

Body size, proportions, physique, and body composition are factors that influence
physical fitness. Stature (standing height) and body weight have been extensively used
with chronological age and sex in the study of fitness and motor performance in the
general population and in unique populations, such as the chronically undernourished.
Although skinfold thicknesses are included as an independent component of health-
related physical fitness, they can also exert an independent effect on other components of
fitness (American Alliance for Health, Physical Education, Recreation and Dance, 1984).

A reasonably extensive research has evaluated the relationships of age, height,
and weight with performances in a variety of fitness and motor tasks since the 19505.
Most of the studies were limited to correlational analyses, i.e., the strength of the linear
association between chronological age, size, and the performance variables (Seils, 1951;

Rarick and Oyster, 1964; Montoye et al., 1972; Malina and Buschang, 1985; Rocha

Ferreira et al., 1991). Although the concept of biological maturity, e.g., skeletal age, is
not within the scope of the present study, several studies examined the contribution of
biological maturity, in addition to age, height and weight, to physical fitness (Seils, 1951;
Rarick and Oyster, 1964; Beunen et al., 1983; Katzmarzyk et al., 1997). Studies that
considered the contributions of anthropometric dimensions other than body size to
variation in performance are limited (Beunen etal., 1983; Malina and Buschang, 1985;
Béne’fice and Malina, 1996; see also Malina, 1994).

Marginal living conditions and chronic undernutrition can alter body size and
composition, which in turn may influence physical fitness and performance. Changes in
body dimensions and composition can influence performance negatively or positively
(Malina, 1975, 1994; Malina and Bouchard, 1991). Positive changes in body size over
time are ordinarily viewed as an indicator of improved public health, diet, and general
living conditions (Eveleth and Tanner, 1990). Physical performance in some tasks, e.g.,
strength and power tasks, have also improved over time and reflect to a large exgtent the
increase in fimction of body size (Malina, 1978).

The growth status of children is commonly used as an index of the health and
nutritional conditions in a community (World Health Organization [WHO], 1997).
Improved living circumstances are reflected in greater growth in body size, whereas
marginal living conditions and chronic undernutrition often result in stunted growth
which, in turn, can influence physical fitness. Although it is recognized that genotype is
intrinsic to body dimensions and physique, marginal living conditions and chronic

undernutrition can compromise or alter genetic potential (Bouchard et al., 1997).

The evaluation of the growth. and nutritional status of children in a community is
based largely on the use of anthropometric indicators: weight-for-age, height—for-age, and
weight-for-height (Waterlow et al., 1977: WHO, 1997). Growth stunting and wasting are
estimated from height and weight relatr . .- to reference data for tell-nourished children.
The growth charts for United States children in the 19705 (Hamill et al., 1977) are used
most often as the reference. Criteria for stunting and wasting are z-scores for height-for-
age and weight-for—height, respectively, that are more than two standard deviations below
the reference, i.e., z-scores greater than —2.0. Although the growth charts for the US.
population are universally accepted as the reference (Yip and Scanlon, 1994; WHO,
1997), the applicability of this reference may have limitations, especially in developing
countries (Walker and Walker, 1997).

Data on the height and weight of children from different parts of the world living
in rural and urban communities are reasonably extensive (Meredith 1979, 1982). In
general, children living in urban areas are heavier and taller age-for—age than their rural
counterparts (Meredith, 1979, 1982; Eveleth and Tanner, 1990). Such rural-urban
comparisons have been recorded since the early 18703, primarily among children of
European ancestry (Meredith, 1979, 1982). Similar data for Latin American populations,
including Mexico, Central and South America, and the Caribbean, are also available, but
are less extensive (Malina, 1990). Mexico, specifically the Valley of Oaxaca in southern
Mexico, is the focus of this study.

Disparity in living conditions is evident in many developing countries where
resources are distributed inequitably among areas. In Mexico, for example, there was

economic progress and rapid growth of urban centers that was accompanied by poverty,

underdevelopment, and undernutrition (Malina, 1990). The accelerated growth of
children in urban centers is often associated with improved health and nutritional
conditions associated with rural to urban migration (Turner, 1976). Such migratory
movements from rural areas in Latin America often result in the formation of irregular,

squatter settlements on the edges of cities, which are called colonias populates in Mexico

 

(Selby and Murphy, 1979; Murphy and Stepick, 1991). Although the settlements are
‘urban’ or ‘suburban’, facilities available for health care and sanitation are far from the
expectations of urbanized centers. Efforts to improve the living conditions of these
communities are an ongoing process.

A primary concern with many developing countries, including Mexico, is the
marginal living conditions in which children are reared and the presence of chronic mild-
to-moderate undernutrition among children. Regional differences in the nutritional status
of children are often prevalent within a country, mainly between rural and urban regions
(Oumarou et al., 1993). In 1996, the estimated prevalence of mild-to-moderate
undernutrition in Oaxaca, one of the poorest states in Mexico, was 17.1%, compared to
6.9% in Mexico City (Secretaria de Salud, 1998). Chronic mild-to-moderate
undernutrition has implications for associated morbidity and mortality of children, and
long term debilitating consequences on growth and functional development, which in

turn, can have economic repercussions in a community.

Significance of the Study
In addition to height and weight, information on the relationship between physical

fitness and other anthropometric dimensions is rather limited (Malina, 1975, 1994). The

association between other anthropometric dimensions (segment lengths, skeletal breadths,
limb circumferences, and skinfold thickness) and physical fitness tasks should be
explored in a variety of populations as variation in body dimensions and proportions may
influence outcomes differently. Reported correlations obtained in most studies indicate
considerable overlap by sex, ethnicity, living conditions, and socioeconomic status (SES).
There is a need to consider the effects of the living conditions and perhaps chronic
undernutrition as sources of variation in the physical fitness of children in developing
areas of the world.

Differences in living conditions between rural and urban communities are often
prevalent in developing countries and countries in economic transition (e.g., countries in
Eastern Europe). Given the differences in living circumstances, attained body size and
composition may differ, and may influence physical fitness. As such, comparisons ‘
between children living in rural and urban conditions may provide insights into the
contributions of specific anthropometric dimensions and estimates of body composition
to explaining the variance in the fitness of children reared under marginal circumstances.
Although the adverse consequences of chronic undernutrition are reflected in reduced
body size and altered body composition, which in turn affect performance, the physical
fitness of children from communities with a history of chronic undernutrition has not
been extensively examined.

The populations examined in most studies of chronic undernutrition usually
involve preschool children, i.e., those under 5 years of age. Although children below the
age of 5 years are more sensitive to unfavorable living conditions and often show high

morbidity and mortality, children of school age represent a population that. has survived

the vigorous selection processes associated with environments characterized by a high
prevalence of infectious and parasitic diseases and marginal nutrition. Chronic
undernutrition during the preschool years is associated with growth stunting, i.e.,
interference with linear growth. The consequences of early growth stunting for later
performance need further study. Specifically, what are the consequences of stunting on
physical fitness at school age? Satyanarayana et al. (1979) suggested that growth stunting
by five years of age affected the absolute work capacity of rural Indian boys during
adolescence. Power output (PWC 170) per unit body weight was generally similar in
well-nourished and malnourished boys, but was considerably reduced in boys were who
extremely stunted in growth. Some evidence for African children (Be'néfice et al., 2001a,
2001b) and Guatemalan adults (Schroeder et al., 1999) also suggests that stunting during
the preschool years is associated with a proportionally greater accumulation of

subcutaneous fat on the trunk during adolescence and adulthood.

Research Questions

There are two aspects to the questions in the present study. The first involves a
comparison of children from the rural and urban communities, and the second involves a
comparison of stunted and non-stunted children within the respective communities. Using
the same population sample and with the inclusion of slightly older children (6 to 13
years of age), Pefia Reyes (2002) reported significant rural-urban differences in body
size, but inconsistent differences in the skeletal breadths, circumferences and skinfolds.

For example, skinfolds did not significantly differ, except between 10-13 year old rural

and urban boys. There were also significant rural-urban differences, in favor of the urban

children, for most measures of physical fitness.

In the context of previous studies in the Valley of Oaxaca in southern Mexico, the

present study investigated the following questions:

1.

What are the anthropometric correlates of physical fitness in rural and urban school
children, respectively? Specifically, what are the contributions of anthropometric
dimensions to variation in the health- and performance-related physical fitness of
rural and urban children?

What is the relationship between somatotype and the health— and performance-related
physical fitness of rural and urban children? The Heath-Carter anthropometric
somatotype protocol (Carter and Heath, 1990), with few exceptions, has not been
used in samples of children with a history of marginal living and nutritional
conditions; hence, this aspect of the study is exploratory and unique. In this context,
the somatotypes of rural and urban children are initially described and compared.
How do growth stunted and non-stunted children within each community compare in
body proportions, relative muscularity, subcutaneous fatness, relative subcutaneous

fat distribution, and physical fitness?

Hypotheses

1.

The following hypotheses were tested:
In addition to age, height, and weight, the anthropometric dimension(s) that add

significantly to explaining the remaining variation in the performance of:

b)

Static strength (sum of right and left grip strength) is estimated arm muscle
circumference;

Running speed (35 yard / 32 meters dash) is estimated calf muscle circumference
and estimated leg length;

Explosive power (standing long jump) is estini :d am .rd calf muscle
circumference;

Lower back flexibility (sit and reach) is estimated leg length;

Muscular endurance (sit-ups) is estimated arm and calf muscle circumference;

Cardiovascular endurance (distance run) is sum of four skinfolds.

. Rural and urban children 6-11 years of age do not significantly differ in Heath-Carter

anthropometric somatotypes.

. After controlling for age and the other two somatotype components, in urban and

rural children 6-11 years of age:

a)

b)

C)

Endomorphy has a negative relationship to cardiovascular endurance (distance
full);

Mesomorphy has a positive relationship to measures of strength (sum of right and
left grip strength), running speed (35 yard / 32 meters dash), explosive power
(standing long jump), and muscular endurance (sit-ups); and

Ectomorphy has a positive relationship to lower back flexibility (sit and reach).

4. a) Stunted and non-stunted children 6-11 years of age do not differ in relative

muscularity and subcutaneous fatness, but do differ in relative body proportions
related to sitting height and leg length.
b) Non-stunted children 6-11 years of age perform better in tests of health- and

performance-related physical fitness comoart .1 to stunted c‘2ildren.

Stunted children 6-1 1 years of age have a proportionally greater accumulation of
subcutaneous fat on the trunk than on the extremities compared to non-stunted

children.

Delimitations

1.

The present study was developed within the context of a broader study designed to
evaluate secular changes in growth status, physical fitness, and living conditions
(Malina, 1999, 2002; Pefia Reyes, 2002).

The sample is limited to children of primary school age range (grades 1-6 in Mexico)
who were attending primary school in each community the time of data collection.
The sample is limited to children 6-11 years. The numbers of children 12 years and
older enrolled in the primary schools in each community were small and to avoid
potentially confounding effects associated with the onset of the adolescent growth

spurt, they were not included in the present study.

Limitations

1. Although the school children were apparently healthy and free from overt disease, it
is possible that some may have had conditions that could influence their performance
on the fitness tests.

2. All participants were assumed to be cooperative and motivated throughout the test
sessions.

3. The performance scores obtained for each of the physical fitness task items were
assumed to be a true reflection of the participants’ best efforts.

4. The nutritional status of the samples was based on their growth status (height and

weight) and not on records of food consumption.

Bibliography

American Alliance for Health, Physical Education, Recreation and Dance [AAHPERD].
1984. Technical Manual, Health Related Physical Fitness. Reston, VA: American
Alliance for Health, Physical Education, Recreation and Dance.

Be'ne'fice E, Garnier D, Simondon KB, Malina RM. 2001a. Relationship between stunting
in infancy and growth and fat distribution during adolescence in Senegalese girls.
Eur J Clin Nutr 55:50-58.

Be’ne’fice E, Garnier D, Simondon KB, Malina RM. 2001b. Differences in growth and
body composition during puberty between Senegalese adolescent girls stunted or
not stunted in infancy. Biom Hum et Anthropol 19:55-61.

Bénéfice E, Malina RM. 1996. Body size, body composition and motor performances of
mild-to-moderately undernourished Senegalese children. Ann Hum Biol 23:307-
321.

Beunen GP, Malina RM, Ostyn M, Renson R, Simons J, Van Gerven D. 1983. Fatness,
grth and motor fitness of Belgian boys 12 through 20 years of age. Hum Biol
55:599-613.

Bouchard C, Malina RM, Perusse L. 1997. Genetics of Fitness and Physical Performance.
Champaign, IL: Human Kinetics.

Carter JEL, Heath BH. 1990. Somatotyping - Development and Application. Cambridge:
Cambridge University Press.

Eveleth PB, Tanner JM. 1990. Worldwide Variation in Human Growth, 2"d edition.

Cambridge: Cambridge University Press.

11

Hamill PVV, Johnson CL, Reed RB, Roche AF. 1977. NCHS growth curves for children
birth-18 years. DHEW Pub] No. 78-1650, Vital and Health Statistics, Series 11,
No. 165. Washington: US Government Printing Office.

Katzrnarzyk PT, Malina RM, Beunen GP. 1997. The contribution of biological
maturation to the strength and motor fitness of children. Ann Hum Biol 24:493-
505.

Malina RM. 1975. Anthropometric correlates of strength and motor performance. Exerc
Sport Sci Rev 3:249-274.

Malina RM. 1978. Secular changes in growth, maturation, and physical performance.
Exerc Sport Sci Rev 6:203-255.

Malina RM. 1990. Growth of Latin American children: socioeconomic, urban-rural and
secular comparisons. Revista Brasileira de Ciencia e Movimiento 4:46-75.

Malina RM. 1994. Anthropometry, strength and motor fitness. In: Ulijaszek SJ, Mascie-
Taylor CGN, editors. Anthropometry: The Individual and the Population.
Cambridge: Cambridge University Press. P 160-177.

Malina RM. 1999. Secular change in size, strength and motor fitness in the Valley of
Oaxaca, Mexico. Proposal submitted to the National Science Foundation,
Proposal and Grant No. BCS-9816400.

Malina RM. 2002. Secular change in size and physical fitness in the Valley of Oaxaca,
Mexico. Final report submitted to the National Science Foundation, Proposal and
Grant No. BCS-9816400.

Malina RM, Bouchard C. 1991. Growth, Maturation, and Physical Activity. Champaign,

IL: Human Kinetics.

12

Malina RM, Buschang PH. 1985. Growth, strength and motor performance of Zapotec
children, Oaxaca, Mexico. Hum Biol 57:163-181.

Malina RM, Selby HA. 1982. Zapotec restudy: a ten year follow-up of Zapotec children.
Report of activities and results of the research supported by National Science
Foundation grant BSN 78-10642, submitted to the National Science Foundation,
July 1, 1982.

Malina RM, Selby HA, Buschang PH, Aronson WL. 1980. Growth status of school
children in a rural Zapotec community in the Valley of Oaxaca, Mexico, in 1968
and 1978. Ann Hum Biol 7:367-374.

Malina RM, Selby HA, Swartz LJ. 1972. Estatura, peso, circunferencia del brazo en una
muestra transversal de nifio zapotecos de 6 a 14 afios. Anales de Anthropologia
9: 143-155.

Meredith HV. 1979. Comparative findings on body size of children and youths living at
urban centers and in rural areas. Growth 43:95-104.

Meredith HV. 1982. Research between 1950 and 1980 on urban-rural differences in body
size and growth rate of children and youths. Adv Child Develop Behav 17:83-
138.

Montoye HJ, Frantz ME, Kozar AJ. 1972. The value of age, height and weight in
establishing standards of fitness for children. J Sports Med Phys Fitness 12: 1 74-
179.

Murphy AD, Stepiek A. 1991. Social Inequality in Oaxaca: A History of Resistance and

Change. Philadelphia: Temple University Press.

13

Oumarou M, Nestel P, Rutstein SO. 1993. Nutrition and health status of infant and young
children in Niger: Findings from the 1992 Niger demographic and health survey.
Columbia, MD: Macro International.

Per'ia Reyes ME. (2002). Growth status and physical fitness of primary school children in
an urban and a rural community in Oaxaca, Southern Mexico. Doctoral
Dissertation, Michigan State University.

Rarick GL, Oyster N. 1964. Physical maturity, muscular strength, and motor performance
of young school-age boys. Res Quart 35:523-531.

Rocha Ferreira MB, Malina RM, Rocha LL. 1991. Anthropometric, functional and
psychological characteristics of eight-year-old Brazilian children from low
socioeconomic status. In: Shephard RJ, Parizkova J, editors. Human Growth,
Physical Fitness and Nutrition. Basel: S. Karger. p 109-118.

Satyanarayana K, Nadamuni Naidu A, Narasinga Rao BS. 1979. Nutritional deprivation
in childhood and the body size, activity, and the physical work capacity of young
boys. Am J Clin Nutr 32:1769-1775.

Schroeder DG, Martorell R, Flores R. 1999. Infant and child growth and fatness and fat
distribution in Guatemalan adults. Am J Epidemiol 149:177-185.

Secretaria de Salud. 1998. Aspectos relevantes sobre la estadistica de deficiencias de la
nutricion. Salud Publica de Mexico 40: 206-214.

Seils LG. 1951. The relationship between measures of physical growth and gross motor

performance of primary-grade school children. Res Quart 22:244-260.

Selby HA, Murphy AD. 1979. The City of Oaxaca. Office of Urban Development,
Technical Assistance Bureau, Agency for International Development.
Washington, DC.

Turner FC. 1976. The rush to the cities of Latin American. Science 192:955-962.

Walker ARP, Walker BF. 1997. Moderate to mild malnutrition in African children of 10-
12 years: roles of dietary and non-dietary factors. Int J Food Sci Nutr 48:95-101.

Waterlow JC, Buzina R, Keller W, Lane JM, Nichaman MZ, Tanner JM. 1977. The
presentation and use of height and weight data for comparing the nutritional status
of groups of children under the age of 10 years. Bull World Health Organ 55:489-
498.

World Health Organization (WHO). 1997. Physical Status: The Use and Interpretation of
Anthropometry. Geneva: World Health Organization, Technical Report Series,
No. 854.

Yip R, Scanlon K. 1994. The burden of malnutrition: A population perspective. J Nutr

124:2043 S-2046S.

15

CHAPTER 2

General Review of Literature

The review of literature is organized in a manner that first discusses general
aspects of growth status and then the influence of growth status on the physical
fitness of children with mild-to-moderate undernutrition and those living in rural
and/or urban areas. The contributions of body size, physique, and composition to
variation in physical fitness are then considered. Although the factors are addressed

separately, it must be recognized that they are not mutually exclusive.

Nature of Malnutrition

Nutrition is a process that involves the relationship between food consumption
and the functional ability of an individual (Malina and Bouchard, 1991). Basic
nutritional requirements need to be met for normal growth and firnctional
development. Any disturbances or imbalances in health and/or nutrition can
ultimately affect child growth (de Onis et al., 1993; de Onis and Habicht, 1996).
Protein-energy malnutrition (PEM) occurs when the protein and/or energy
requirements to ensure optimal growth and function are not satisfied through the diet
(Torun and Chew, 1994; WHO 2001). Currently, it is estimated that one in every four
children in the world is affected with PEM, and 10.9 million children die from
malnutrition per year (WHO, 2001 ). The primary cause of PEM is inadequate food
intake, while the secondary cause is diseases “. . .that lead to low food ingestion,
inadequate nutrient absorption or utilization, increase nutritional requirements, and/or
increased nutrient losses” (Tortin and Chew. 1994, p 950). The poorly nourished are

more vulnerable to infections and it is estimated that an undernourished child suffers

160 days of illness per year (WHO, 2001). Insufficient nutritional intake and/or
repeated digestive infections are two circumstances often prevalent in developing
countries where nutritional and living conditions are marginal (Keller, 1988; de Onis
et al., 1993).

The severity of PEM lies on a continuous spectrum so that its classification is
somewhat arbitrary. The severe forms of PEM are kwashiorkor, marasmus, marasmic
kwashiorkor, and undifferentiated depending on the signs, symptoms, and pathology
of the conditions (Scrimshaw and Béhar, 1961; Torun and Chew, 1994). Mild to
moderate forms of PEM are usually assessed with the use of anthropometry and age:
weight-for-age, height-for—age, and weight-for—height (Waterlow et al., 1977; WHO,
1997). Although weight-for—age is the most valid indicator for children under 1 year
of age, it does not discriminate between acute and chronic malnutrition in older
children, especially those above 5 years of age (Waterlow et al., 1977). Weight-for-
height is an indicator of current nutritional status, whereas height-for-age is an
indicator of nutritional history (Waterlow, 1972).

Deficit in weight-for-age is defined as underweight, while deficits in height-
for-age and weight-for—height are defined as stunted and wasted, respectively
(Waterlow, 1972, 1973; WHO, 1997). Growth stunting and wasting, and underweight
are estimated from height and weight relative to the US. reference data (Hamill et al.,
1977). The criteria for classifying children as underweight, stunted, or wasted are
based on z-scores that are > 2 standard deviations below age- and sex-specific
reference values for United States children (WHO, 1997). Although the US.
population is universally accepted as the reference for comparison between
populations (Yip and Scanlon, 1994; de Onis and Habicht, 1996; WHO, 1997), it is

recognized that the applicability of this reference has limitations. One limitation is

applying the growth status of children from a developed country as the reference for
those in developing countries (Walker and Walker, 1997). A second is the assumption
that“. . .all child populations throughout the world have the same genetic potential for
growth in size” (Waterlow et al., 1977, p 490; Waterlow, 1973), implying that there is
no genetic variation in the growth of children. However, variation in growth in
length/height among well-off children under 7 years of age in different populations is
small compared to that between children from the extremes of SES within a
population (Habicht et al., 1974).

At present, there is a global database that provides a standardized compilation
of child growth and nutrition data for children under 5 years of age. The data were
obtained from nutritional surveys conducted around the world since 1960, and are
maintained and routinely updated by the World Health Organization (WHO, 1997).
Available statistics showed that approximately 70% of the children with PEM live in
Asia, 26% in Africa, and 4% in Latin America and the Caribbean (WHO, 2001).
Geographically, Latin America and the Caribbean include countries within the
Caribbean, Mexico, Central America, and South America (de Onis et al., 2000;
WHO, 2001).

Given the context of the present study, attention is focused on growth stunting
in Mexico and Central America. In 2000, the projected prevalence of stunting in
preschool children was, on average, 32.5% (range 28.0-37.0) in developing countries
of Africa, Asia, and Latin America and the Caribbean. Central America including
Mexico had an estimated prevalence of stunting of 24%, which is about 3.9 million
(range, 1.6-6.2 million) children (de Onis et al., 2000). Data for Mexico in 1996
indicated an overall prevalence of stunting of 33.9% in children under 5 years of age

living in rural areas (de Onis et al., 2000). lnfonnation on the prevalence of stunting

l8

for older children in Mexico is limited. Data for 1994 indicated that 43.4% of first
grade children (i.e., at school entry) in the state of Oaxaca were stunted. Only the
state of Chiapas had a higher prevalence, 44.1% (Secretaria de Salud, 1998).

Recent data also indicate increasing rates of overweight and to a lesser extent
obesity in developing countries that are already struggling with the overwhelming
presence of undernutrition (de Onis and Blossner, 2000). Overweight is increasing in
deve10ping countries where segments of the population are undergoing rapid
demographic change associated with economic development and rural-to-urban
migration, and nutritional changes due largely to western influences (Popkin et a1,
1996, 2001; Schroeder et al, 1999). The shift from undernutrition to overweight and
perhaps obesity is mainly due to a change towards higher fat and lower carbohydrate
diets, and reduced levels of physical activity (Popkin et al., 2001). In 1995, the
prevalence of overweight among children under the age of 5 years in Latin America
and the Caribbean was 4.4% (approximately 2.4 million) and 3.5% in Central
America including Mexico (de Onis and Blossner, 2000). An increase in the
prevalence of “diet-related” diseases in developing countries has increased the costs
of public health care to comparable magnitudes as the costs for undernutrition

(Popkin et al., 2001).

Undernutrition and Growth

Studies on undernutrition are often carried out in developing countries where
nutritional and living conditions are marginal. Chronic undernutrition and/or episodes
of infections lead to weight loss and impaired linear growth. Growth stunting most

often occurs during the first two or three years of life and tends to persist into

adolescence and adulthood. Stunted individuals are by definition shorter and are also
lighter with reduced muscle mass (Spurr, 1984; Hoffman et al., 2000a).

Two questions are commonly addressed by studies examining the long-term
consequences of undernutrition on growth and development. One is whether the early
effects of nutritional deprivation result in permanent growth stunting, and the other is
whether improvements in the nutritional environment enhance “catch-up” growth
(Golden, 1996). The available data are somewhat inconclusive as to the “catch-up”
growth hypothesis.

School children 6-12 years of age in the Valley of Oaxaca in the 1970s were
shorter and lighter than better-off children of the same age and sex in the United
States and Mexico (Malina, 1983). The Oaxaca data included children from rural and
urban communities, and among these children, those from rural indigenous (Zapotec-
speaking) communities were the shortest and lightest (Malina et al., 1981). The
growth status of the children of Oaxaca suggests a profile of chronic, mild-to-
moderate undernutrition.

Studies on African children, 4.0 to 6.5 years of age, reported that well-
nourished children had greater body dimensions compared to two groups of
malnourished children (Bénéfice et al., 1996). One of the malnourished groups
included children who were exposed to chronic mild-to-moderate undernutrition,
while the other group included children who were hospitalized for severe
undernutrition during infancy and were subsequently nutritionally rehabilitated
(Bénéfice et al., 1996; Béne’fice et al., 1999). Upon diagnosis, the nutritionally
rehabilitated children were hospitalized for 6-8 weeks where they received medical
and nutritional care and only left the center when the weight-for-height z-score

reached —1.0 standard deviation. There were no differences among the three

20

nutritional groups in weight-for—height and the body mass index (BMI) at the time of
the study (Bénéfice et al., 1999), indicating that the children were proportionate in
weight for height. The severely malnourished children had larger arm circumferences
and suprailiac skinfolds than the mild-to-moderately undernourished children
(Bénéfice et al., 1996). Though speculative, the greater accumulation of fat in the
severely undernourished children could be due to hormonal changes during the course
of starvation (Sawaya et al., 1998), or the body’s tendency to store fat as a potential
reserve for periods of energy inadequacy (Bénéfice et al., 1996).

Walker and Walker (1997) cautioned against mild-to-moderate undernutrition
being viewed as totally underlying the grth and functional disadvantages observed
in underweight school children. This caution was based on small differences in the
health, well-being, learning, and physical abilities of African children 10-12 years of
age who were below and above the 5th percentile of US. reference values for weight.
Although the caution is warranted, using weight-for-age as an indictor of mild-to-
moderate undernutrition may not be the most appropriate criterion to assess
undernutrition in children 10-12 years of age. As noted earlier, weight-for—age is not
sufficiently sensitive to differentiate the nature (acute or chronic) and severity (mild,
moderate or severe) of undernutrition in children 5 years and older (Waterlow et al.,
1977)

Several studies have examined the physical development of adolescents and
adults with a history of undernutrition. ln Senegal, childhood stunting did not affect
the BMI of girls during puberty; rather, body mass and subcutaneous fat mass were
comparable to those of adolescent girls who were not stunted (Bénéfice et al., 2001 a,
2001 b). Body height, lengths, and diameters of the “stunted” girls were, however,

consistently shorter and/or smaller than the “non-stunted” girl.c indicating a failure of

21

“catch-up” growth after the original insult. The previously stunted adolescent girls
also had greater accumulation of subcutaneous fat on the trunk compared to their non-
stunted counterparts (Béne’fice eta1., 2001a, 2001b), possibly indicating a greater
stimulation of hormonal changes in fatness due to puberty (Béne’fice et al., 2001a).
However, the hormonal changes in fat could also be a consequence of malnutrition as
suggested by Sawaya et a1. (1998). Similarly in rural Guatemala, stunting in early
childhood was associated with greater abdominal fatness and percent body fat
(Schroeder etal., 1999) and reduced fat-free mass (Martorell et al., 1992) in
adulthood. Stunting in samples in Colombia and the Ecuadorian Amazon, however,
was not associated with reduced fat-free mass (Orr et al., 2001). These studies,
though suggestive, illustrate difficulties with field assessments of body composition.
The index of abdominal fatness in the Guatemalan study, for example, was the waist-
hip ratio, which is not necessarily an accurate indicator of relative fat distribution.
Moreover, the waist-hip ratio may not be an appropriate indicator for describing
subcutaneous fat distribution in children and adolescents due to potentially
confounding effects of proportional changes in muscular and skeletal growth (Malina,
1974; Johnston, 1992; Sarria, 1992).

Wilson and colleagues (1999) reported a negative relationship between
gastrointestinal parasitic infection and weight and stature of 6 to 16 year old
Colombian boys. Infected boys were 1.61 times more likely to be stunted than their
non-infected counterparts. The infected boys also had proportionally more trunk
subcutaneous fat than their non-infected peers.

Nutritionally stunted children in developing countries tend to have a higher
risk of becoming obese than non-stunted children. After controlling for the effects of

income, Popkin et al. (1996) reported risk ratios of overweight for being stunted in 3

22

to 9 year old children as 7.8, 3.5, 2.6, and 1.7 in Russia, China, South Africa, and
Brazil, respectively. Chronically undernourished stunted children appear to fluctuate
considerably in body weight from underweight to overweight/obese when food
sources are available. Hoffman and colleagues (2000b) suggested that stunted
children have an impaired ability to oxidize fat, which may underlie the increased

prevalence of overweight among stunted individuals.

Undernutrition and Performance

Relatively few researchers have considered chron ' PE as a factor in the
negative consequences for physical performance in school age children and
adolescents (Malina, 1984). A good deal of the work has focused on infancy and
early childhood. The few studies of school age children and adolescents have
generally compared the performances of undernourished and well-nourished samples
to evaluate the association between nutritional status and performance.

Chavez and Martinez ( 1984), for example, studied two groups of infants 2-24
months from rural Mexico. One group of infants received nutritional supplement, i.e.,
mashed food and bottled milk, while the other did not. Children who received the
nutritional supplement were more physically active than the unsupplemented group.
The differences in activity level between the groups increased with age. These
findings are consistent with previous studies of 7-18 month old Indian infants where
the well-nourished infants had significantly greater activity scores and time in play
than undernourished infants (Graves, 1976, 1978).

The strength and motor performance of rural children, 6-13 years of age, from
a Zapotec-speaking community in Oaxaca and living under conditions of chronic

mild-to-moderate undernutrition were compared to better nourished American

23

children of the same age (Malina and Buschang, 1985). The absolute levels of
performance in strength and motor tasks were significantly poorer in the rural
Zapotec children, but when performances were expressed relative to body size,
performance levels were commensurate with reduced body size. Malina and
Buschang (1985) concluded that absolute body size and nutritional status influenced
performance. A subsequent analysis investigated the strength and motor performance
of the children from rural Oaxaca, Mexico, with a sample from coastal Papua New
Guinea (Pere village on Manus Island), and a well-nourished American sample.
Although the children from Oaxaca and Pere village were weaker, the motor
performances of children from Pere compared favorably to the American children
(Malina et al., 1987).

Well-nourished Senegalese children, 4-6 years of age, performed better in the
standing long jump, ball throw for distance, shuttle run, and static strength than two
malnourished groups of children of the same age (Bénéfice et al., 1996). Mild-to-
moderately malnourished children performed poorer in the power and coordination
tasks than severely malnourished children. The severely malnourished group included
children who were severely undernourished during infancy and were subsequently
nutritionally rehabilitated for about 6 to 8 weeks with medical and nutritional care at
the time of study. A principal components analysis of the anthropometric dimensions
of these samples resulted in two factors, general body size and body composition,
which were subsequently used in an analysis of variance in performance of the three
samples of children. General body size had high, positive loadings that accounted for
41% of variance in performance, whereas body composition had moderately high,
positive loadings that accounted for 25% of the variance in performance. Overall,

general body size had a positive influence on the endurance run, jump, throw, agility,

24

and hand-grip, whereas body composition had a negative influence on all the task
performance.

A subsequent analysis of the Senegalese children, 4-6 years of age, examined
the effects of undernutrition on motor coordination and performance (Béne’fice et al.,
1999). Well-nourished children performed better than the undernourished children on
all coordination and motor fitness items. Children who were severely malnourished
during infancy and rehabilitated performed significantly poorer on grip strength
compared to mild-to-moderately undernourished children. Body dimensions
explained 7% to 50% of variation in motor performances, with stature as the main
predictor. Performances were significantly influenced by body size; hence, reduced
body size as a consequence of chronic undernutrition is clearly a disadvantage for
performance. Growth stunting and reduced body size generally had a negative
association with performance among the Senegalese children. Ten year old Afiican
boys who were undernourished (low BMI) had significantly greater grip strength per
unit body mass compared to overweight boys, although they had less absolute grip
strength (Naidoo, 1999). However, the performances and fitness of daily activities do
not include adjustments for body size. Hence, being ‘bigger is better’ for many fitness
and performance tasks (Spurr, 1984).

Mildly undernourished Colombian boys, 6-16 years of age, had reduced
absolute oxygen uptakes, but similar maximal uptakes per unit body weight compared
to well-nourished boys (Spurr et al., 1983; Spurr, 1983). Spurr and colleagues (1983)
suggested that the differences observed between the absolute and relative VOzmax
values in undernourished Colombian boys were due to smaller body size and possibly

differences in body composition. In a subsequent analysis of the Colombian boys,

25

small body size associated with parasitic infections also contributed to lower VOgmax
(Wilson et al., 1999).

Small body size, regardless of its origin or cause, tends to adversely affect
performances. Guatemalan children who were stunted at 3 years of age or earlier had
reduced grip strength as adults (18 years or older) compared to those who were not
stunted during childhood (Martorell et al., 1992). In communities where nutritional
and living conditions are marginal, the economy depends largely on intensive manual
labor, reduced strength and work capacity can have significant repercussions on the
productivity and livelihood of the population (Malina, 1986).

The timing and intensity of nutritional stress is central to the resulting effects
on physical fitness (Malina and Bouchard, 1991 ). It is assumed that the earlier onset
and greater intensity of undernutrition, the poorer would be performances. Under the
unfavorable conditions of undernutrition, males are generally more sensitive and are
more afi’ected by these environmental stresses compared to females (McCance, 1966;

Malina et al., 1985; Stinson, 1985).

Stunting and Fat Distribution

Significant changes in the relative distribution of subcutaneous adipose tissue
occur during puberty and the adolescent grth spurt (Malina and Bouchard, 1991;
Malina, 1996). During adolescence, boys tend to develop thicker skinfolds on the
trunk relative to the extremities, while girls tend to accumulate similar amounts of
subcutaneous fat on the trunk as on the extremities (Baumgartner et al., 1986;
Baumgartner and Roche, 1988; Malina 1996). Studies have also shown ethnic
variation in the distribution of subcutaneous fat (Mueller, 1988; Malina etal., 1995;

Malina 1996; Naidoo, 1999), although results vary among ethnic groups. American

26

adolescent girls of Asian and Mexican ancestry have proportionally more
subcutaneous fat on the trunk than those of European and African ancestry (Malina et
al., 1995), whereas some data also indicate proportionally more subcutaneous fat on
the trunk among Americans of African ancestry (Mueller, 1988; Malina, 1996).
Among 10 year old African boys, those of East Indian ancestry had a greater
accumulation of adipose tissue on the trunk relative to the extremities compared with
boys of African and European ancestry (Naidoo, 1999).

Most studies of subcutaneous fat distribution have been conducted on
children, adolescents and adults from industrialized countries (Gam, 1955;
Baumgartner et al., 1986; Baumgartner and Roche, 1988; Malina et al., 1995; Malina,
1996; Van Lenthe et al., 1996). Few studies have evaluated the distribution of
subcutaneous fat in populations of developing countries where undernutrition is
chronic (Schroeder et al., 1999; Be’néfice et al., 2001a, 2001 b).

Studies that have examined the long-term consequences of early stunting on
the distribution of subcutaneous fat are limited to two follow-up studies, one in
Guatemala (Schroeder et al., 1999) and the other in Senegal (Bénéfice et al., 2001a,
2001b). These follow-up studies evaluated the growth status of adolescent and adult
samples who were classified as stunted or non-stunted at 6-18 months (Bénéfice et
al., 2001a, 2001 b) and at 36 months (Schroeder et al., 1999) of age, respectively. The
evidence suggested proportionally greater accumulation of subcutaneous fat on the
upper body in stunted adolescents and adults. The study of adolescents used
skinfolds, whereas the study of adults used the waist-hip ratio. Be'néfice et al. (2001a)
suggested that the tendency to accumulate subcutaneous fatness on the trunk “. . .may
represent transitory changes in fatness under the influence of hormonal stimulation

during puberty” (p. 57). Based on an elevated prevalence of overweight in mildly

stunted 7-11 year old girls from a shanty town (favela) in sao Paulo, Brazil, Sawaya
et a1. (1998) proposed that the hormonal changes associated with malnutrition may
render individuals more susceptible to weight (and presumably fat) gain in the

presence of high fat diets.

Rural-Urban Differences in Growth

Differences in the living conditions under which children are reared have an
effect on growth. Under apparently ‘better’ living circumstances, growth potential is
more likely to be achieved, whereas under marginal conditions growth can be
compromised. Earlier data from Europe consistently show that children living in
urban areas are larger and mature earlier than those living in rural areas (Meredith,
1979, 1982; Eveleth and Tanner, 1990). The positive effects of urbanization on
growth are presumably associated with improved living conditions related to health
care and diet. These include, for example, a constant food supply, clean access to
public water and sanitation, health services, available quality medical services,
educational institutions, and recreational and welfare facilities within a community
(Eveleth and Tanner, 1990). Although urban children are, in general, taller and
heavier than their rural peers (Meredith, 1979, 1982), socioeconomic status (SES) can
influence the benefits associated with urbanization. More recent European data do not
show marked rural-urban differences expect for several countries in Eastern Europe.
Data from the United States indicate negligible rural-urban differences in growth
status (Hamill et al., 1972).

Growth studies of children from developing countries, specifically Latin
America, indicate rural-urban differences, but nutritional status and access to food

and health resources are mediating factors. At present, Latin America is faced with

28

increased rural to urban migration that has resulted in a sector of the lower class that
does not have access to resources although it has the “advantage” of living in a city.
In developing countries, rapid urban growth due to expansions of urban slums and
marginal areas are not necessarily associated with improved growth status in children,
but relatively few studies have compared the growth status of Latin American
children living in rural and urban communities (Graham et al., 1980; Malina et al.,
1981; Perla Reyes, 2002).

Malina et a1. (1981) compared the growth status of children 6-14 years from
rural and urban communities in southern Mexico. The children were from three I
different types of communities: two cg_lo_ni_a_s_ it: the city of Oaxaca (urban), two rural
Ladino communities, and two rural Zapotec-speaking communities in the Valley of
Oaxaca. The Ladino communities were more westemized in lifestyle and agricultural
practices compared to the rural Zapotec-speaking communities, which practiced
subsistence agriculture using traditional techniques. The rural children from the
Zapotec-speaking communities were shorter and lighter, and had reduced muscle
mass compared to children from the Ladino communities and urban colonias.
Children from the rural Ladino communities were slightly larger in height and weight
than children from the colonias, implying that children living in rural communities
with westemized lifestyles are better off than those living in the urban slums.

Similar findings were reported on Peruvian children from four relatively
prosperous rural agricultural villages and a poor urban community in Lima (Graham
et al., 1980). Slight height differences favored the urban children, particularly among
boys. There was no difference in stature between rural and urban girls. The rural
children also had lower weights and thus less weight-for-height compared to the

urban children.

29

Cameron and colleagues (1992) compared two groups of urban children from
average to high SES and two groups of rural children (from presumably lower SES),
5-19 years of age, in South Africa. Anthropometric dimensions included weight,
height, sitting height, biacromial and bi-iliac breaths, head and arm circumferences,
and four skinfolds. The higher SES urban children were consistently, though not
significantly, taller and heavier than all other groups of children. However, ‘average’
SES urban children were consistently and at times significantly, smaller and lighter
than their rural counterparts. Among 9 year old Nigerian boys, urban boys were, in
general, larger in body size and skeletal robustness than rural boys. Arm and
abdominal skinfold thicknesses were also greater for the urban than rural boys
(Spurgeon etal., 1984). Among Japanese children, 9-17 years of age, the amount of
subcutaneous fat in urban children was significantly greater compared to rural

children, although height and weight were comparable (Matsui and Tamura, 1975).

Rural-Urban Differences in Performance

In contrast to comparisons of the growth status of rural and urban children,
corresponding comparisons of performance are relatively limited. A comparison of
the weight, height and physical fitness (dash, agility, ball throw, vertical jump, and
squat-thrust) of 10 year old rural and urban Polish children indicated rural-urban
differences in fitness that favored the urban children (Pilicz and Sadowska, 1973). In
this sample of Polish children, rural and urban boys did not differ in height and
weight, whereas urban girls were significantly taller, but not heavier, than rural girls.
Children from four rural and two urban districts of Japan were compared for aerobic
capacity (Matsui and Tamura, 1975). Rural children were superior in their endurance

ability to urban children.

30

Similarly, Corlett and Mokgwathi (1987) compared the endurance running
ability of rural and urban Tswana children and noted superior endurance performance
in the rural children. Rural and urban children, aged 7-12 years, living in Bostwana
were also compared in height, weight, upper arm circumference and hand grip
strength (Corlett, 1988). The urban children were stronger than their rural peers. Even
when body size was statistically controlled, the superior performance in grip strength
persisted with increasing age, but to a lesser extent.

Cultural and social conditions of a community also influence the development
of motor proficiency. Munetaka et al. (1971) evaluated the contribution of community
differences to the growth and motor development of 4-5 year old children living in
three different communities in Japan: an island, a housing development, and an urban
area. Four body dimensions and 13 motor ability items were measured. The average
family size was larger in the island sample, while the urban sample had a higher
percentage of parents who were either senior high school or college graduates. Body
dimensions between the groups were not significantly different. The island sample
was superior in coordination and flexibility test items, while the urban sample was

superior in all other motor ability items.

Anthropometric Correlates of Fitness

Body Dimensions

Children who are shorter and lighter generally tend to perform poorer on
fitness tasks than taller and heavier peers (Malina, 1975, 1994; Malina et al., 1987).
Body weight is negatively correlated with performances on jumping and running
tasks, and positively correlated with performances on throwing tasks (Seils, 1951;

Rarick and Oyster, 1964; Malina, 1975, 1994). These trends imply that lighter

31

children perform better in activities involving movement or projection of the body
through space (jumping and running) compared to heavier peers, who perform better
in object projection tasks (throwing). Absolute body mass is also a major contributor
to static muscular strength (Malina, 1975, 1994).

Stature and weight correlate better with strength than with other fitness
variables, particularly during middle childhood, indicating that a child who is bigger
tends to be also stronger (Malina, 1975; Malina and Buschang, 1985; Malina and
Bouchard, 1991; Bénéfice and Malina, 1996). However, height and weight have only
low to moderate correlations with other motor fitness variables such as the grip
strength, jump, dash, shuttle run, and distance throw (Seils, 1951; Espenschade, 1963;
Rarick and Oyster, 1964; Montoye et al., 1972; Malina, 1975; Rocha Ferreira et al.,
1991; Be'néfice and Malina, 1996), and the correlations overlap by sex,
socioeconomic status and ethnicity (Malina, 1994).

Since age, stature, and weight are related to performance, studies have used
partial correlation analyses to evaluate relationships among age, body size, and
performances on a variety of tasks (Table 2.1). The partial correlations are variable
and overlap by sex, SES, ethnicity, and nutritional status (Seils, 1951; Rarick and
Oyster, 1964; Malina and Buschang, 1985; Rocha Ferreira et al., 1991; Malina, 1975,
1994; Bénéfice and Malina, 1996). When stature and weight are statistically
controlled, age is positively related with performances on several motor tasks, which
is consistent with the notion that neuromuscular maturation and/or experience
associated with age is positively related with some fitness (Malina and Bouchard,
1991; Malina, 1994). Similarly, correlations between stature and performances are
mainly positive after age and body weight are controlled. Body weight, on the other

hand, is often negatively related to performances, afier controlling for age and stature,

32

specifically those performances involving projection or movement of the body
through space.

Body mass is a predictor of static strength in children and accounts for a major
part of the variance in strength tests (Malina and Buschang, 1985; Rocha Ferreira et
al., 1991; Bénéfice and Malina, 1996; Katzmarzyk et al., 1997). As noted earlier,
Malina and Buschang (1985) compared the grip strength and motor performances of
6-13 year old children from a rural Zapotec-speaking community in the Valley of
Oaxaca, Mexico, to better nourished American children. The absolute size, strength
and motor performances of undernourished Zapotec-speaking children were less-than
those of the American children. However, when body size was controlled, the
differences were reduced. In fact, the Zapotec children had better throwing
performances per unit body size and '- 2nilar g: .p strengths compa :d to the American
children. Hence, variation in performance is influenced by absolute body size. The
residuals of other anthropometric dimensions (a skinfold, estimated muscle
circumference, segment lengths, and skeletal breadths), after controlling for the
combined effects of age, height and weight, were regressed on grip strength and
motor fitness in the sample of Zapotec children. Few of the variables added
significantly to explaining the remaining variation in the performances of Zapotec
children.

Bénéfice and Malina (1996) examined the contribution of body dimensions
and composition to the variation in the fitness of Senegalese children. Stature and
weight explained about 30% to 50% of the variation in the performances of fitness
tasks in children younger than 10 years of age. Among children older than 10 years,
weight was the best predictor of performances of the fitness tasks, but the proportion

of variance in motor performance explained by weight alone was only 10 to 25%.

33

Stature, on the other hand, was the main predictor of performance in the fitness tasks
for children younger than 6 years of age.

Studies evaluating the relationships between specific segment lengths and
circumferences and performance are limited. Malina (1975) examined the relative
relationship of leg length (sitting height subtracted from stature) to the performances
of selected fitness tasks for children 6-12 years of age. The relationships between leg
length and running, jumping, and throwing were low and perhaps reached the
moderate range (correlation coefficients: 0.00 to +0.34). Previous studies also
indicated low relationships between leg length and various performance tasks
(Clarke, 1957; Clarke and Degutis, 1964; Baacke, 1964). Correlations between leg
length and performance were, in general, of the same magnitude or slightly less than
those between stature and performance. The comparable association of stature and leg
length with performance is expected, as leg length is a major component of stature.

Limb circumferences are positively related to performance, particularly
strength tasks (Malina, 1975, 1994; Béne’frce and Malina, 1996). The strength of a
muscle is proportional to its cross-sectional area (Malina, 1984). Limb
circumferences are related to body mass, so that a question of interest is the
relationship between limb circumference and performance after statistically
controlling for body mass. Small body size and/or a reduction in muscle mass is
associated with less absolute strength. Children with smaller arm circumferences have
lower static strength measurements compared to those with larger arm circumferences

(Malina et al., 1987; Malina and Buschang, 1985).

34

Physique

Physique refers to the general shape or form of the body as a whole (Malina,
1975, 1978, 1992; Malina and Bouchard, 1991). There are several methods for
assessing physique (Sheldon et al., 1940, 1954; Parnell, 1958), but the most common
method and that which will be used in the present study is the Heath-Carter
anthropometric somatotype protocol (Carter and Heath, 1990; Malina, 1995). The
Heath-Carter method uses several anthropometric dimensions to estimate the three
components of somatotype: endomorphy, mesomorphy, and ectomorphy (Carter and
Heath, 1990; Malina, 1995).

Endomorphy refers to relative fatness, and softness and roundness of contours
throughout the body. It is estimated from the sum of three skinfolds adjusted for
height in the Heath-Carter anthropometric protocol.

Mesomorphy refers to the robustness of musculoskeletal development, i.e., a
predominance of muscle, bone, and connective tissues. It is estimated from two
extremity skeletal breadths (biepicondylar breadth of the humerus and bicondylar
breadth of the femur), two limb circumferences corrected for skinfold thickness
(flexed arm circumference and the triceps skinfold, and calf circumference and the
medial calf skinfold), and height.

Ectomorphy refers to relative linearity or slendemess, and the predominance
of surface area over body mass. It is estimated from the reciprocal ponderal index,
height divided by the cube root of weight.

The lowest possible rating with the Heath-Carter protocol is 0.1, and the upper
end is open. In practice, most scores range between 1 and 7, which is the range of the
fixed scale initially described by Sheldon and colleague (1940, 1954). A somatotype

is a composite that represents the contribution of each component. On average,

35

children in the age range of the present study (6-1 1 years), tend to have a balanced
somatotype, e.g., 3-4-3 or 3-3-3, where the first number refers to endomorphy, the
second to mesomorphy and the third to ectomorphy. A somatotype of 2-5-2 indicates
dominant mesomorphy, while somatotypes of 5-2-2 and 2-3-5 indicate dominant
endomorphy and ectomorphy, respectively.

Variability in the range of ratings of somatotype components is reasonably
large within and/or between populations. Among males, ratings range from 1.0 to
10.5 for endomorphy; 1.0 to 9.5 for mesomorphy, and 0.5 to 9.0 for ectomorphy.
Corresponding ranges among females are 1.5 to 10.0 for endomorphy, 0.5 to 6.0 for
mesomorphy, and 0.5 to 6.5 for ectomorphy (Carter and Heath, 1990). On average,
males tend to become more mesomorphic, while females become more endomorphic
with age during childhood (Carter and Heath, 1990; Malina and Bouchard, 1991).

The Heath-Carter anthropometric somatotype method was developed on
adults and the samples used to validate the method were largely adult males (Carter
and Heath, 1990). Hence, the applicability of the method “. . .to the growing and
maturing individual, to females, or to the clinically ill, may require adjustment”
(Malina, 1992, p. 94-95).

It is important to note that the three components of somatotype together
constitute an individual’s physique. However, the specific component ratings are
often analyzed individually without controlling for the effects of the other two
components, thus limiting the essence of the somatotype concept. Cressie et al.
(1986) recommended a multivariate analysis of variance (MANOVA) approach to
analyzing the Heath-Carter somatotype so as to maintain “. . .the essential quality of

component dominance together with the relationship of the three components in the

individual subject” (p.197).

36

Studies examining somatotype have largely described or compared the
physiques of individuals in different populations, and athletes in a variety of sports
(Carter and Heath, 1990). The Heath-Carter anthropometric somatotype protocol has
also been applied to samples of healthy children and adolescents primarily of
European ancestry (Carter and Heath, 1990; see also Malina and Bouchard, 1991),
but has been used to a lesser extent in Latin American children.

The protocol was used with school children of both sexes 7 through 18 years
of age in Guatemala (Alexander et al., 1993) and Venezuela (Alexander, 1992), and
with boys 6 to 15 years of age in Chile (Godoy et al., 1994). The studies in
Guatemala and Venezuela were done in the context of national physical fitness
surveys of school children and the results were presented as descriptive statistics by
age and sex. Relationships between somatotype components and physical fitness
were not addressed. The study in Chile was limited to boys attending a private school
(i.e., upper class) in Santiago and focused on age changes in somatotype components
in small samples followed longitudinally over one year.

In contrast, the Heath-Carter anthropometric somatotype method has not been
commonly applied to populations living under marginal health and nutritional
conditions. Murguia et al. (1990) applied the protocol to a large sample of children
and adolescents of both sexes 5-20 years of age from different areas of Yucatan,
Mexico. The samples were from a traditional Indian community (subsistence maize
agriculture) and from municipalities which were more “modern” areas specializing in
fishing, sisal production, and cattle raising. Although there were body size differences
among the samples, there was considerable overlap in mean somatotypes.

Endomorphy tended to be lower in boys and girls 6-11 years of age from the

37

subsistence agricultural community, but mesomorphy and ectomorphy were similar.
Unfortunately, the study was largely descriptive with little statistical analysis.

Godoy and Barcos (1995) used the Heath-Carter protocol with Chilean girls 4-
1 1 years to address somatotype variation associated with socioeconomic and
nutritional status, while Diaz Gamboa and Fuentes Heinrich (1996) compared the
somatotypes of Aymara boys and girls, 6-15 years of age, resident at high altitude
(3000-4500 meters) in northern Chile. Table 2.2 presents a summary of mean
somatotypes from these studies of Latin American children. A question of interest is
the applicability of the method on samples of children living under varied nutritional
and other environmental conditions.

Correlations between somatotype components and performance in running
and jumping activities are generally low to moderate in 7-12 year old boys and girls,
with endomorphy and ectomorphy more closely relatec’ to running and jumping than
mesomorphy (Slaughter et al., 1977, 1980). The most consistent relationship of
somatotype and performance is a negative association between endomorphy and
running, jumping and agility tasks (Malina, 1975, 1992; Slaughter etal., 1977). Some
data indicate a negative correlation between mesomorphy and a moderate distance
run, the 600 yard run (Slaughter et al., 1977, 1980). The authors suggested that the
negative relationship may be influenced to some extent by fatness as heavier children
tend to be higher in mesomorphy. However, the analysis did not control for variation
in the other two somatotype components (see above).

Subcutaneous fat is the major aspect of endomorphy in the Heath-Carter
protocol and represents ‘dead weight’ that has no mechanical advantage in
performances of activities. Hence, it often hinders the performance of activities that

involve the projection or movement of the body through space (Malina, 1992).

38

Conversely, endomorphy is positively related to performances of static strength,
emphasizing the contribution of overall body size to strength. However, relationships
between mesomorphy and ectomorphy and performance are not consistent across
studies. For example, endomorphy and mesomorphy have a moderately positive
correlation with muscular strength while ectomorphy correlates negatively to
muscular strength (Malina, 1975). This indicates that overall body size, particularly
fat-free mass, is important for strength, which implies that variation in strength is
accountable by the amount of muscle mass.

Most studies relating somatotype to performance are limited to correlations
between one component and the performance task. As noted earlier, somatotype is
defined by the three components together so that treating a component in isolation
from the other two may provide limited information. Since the three components are
interrelated, it is more appropriate to use second-order partial correlations or
multivariate techniques to analyze relationships between the somatotype and

performance.

Body Composition

Body composition refers to the partitioning and quantification of the primary
tissue components of body mass (Malina and Bouchard, 1991). The two-compartment
model, fat-free mass and fat mass, is most commonly used (Beunen et al., 1982;
Malina and Bouchard, 1991), although recent developments in methodology permit
the use of multi-component modes. Anthropometric dimensions provide an indirect
estimate of body composition. Skinfold thicknesses at specific sites are often used as
an indication of subcutaneous fat (Malina, 1996) and are good predictors of body fat

because the majority of fat is subcutaneous (N organ, 1991).

39

Correlations between body composition and performance are generally low to
moderate, but vary somewhat across studies. Fat-free mass and measures of
muscularity are positively correlated to motor performance (Malina, 1984; Be'néfice
and Malina, 1996). Fat mass is negatively associated with performance, especially
items that require displacement or projection of the body (Malina, 1975, 1994;
Beunen etal., 1983; Thomas and Thomas, 1988; Malina and Bouchard, 1991;
Bénéfice and Malina, 1996). Although the performance of sit-ups does not require
great displacement of the body, it is also negatively associated with fatness (Riendeau
et al., 1958; Cureton et al., 1975; Pate et al., 1989). Static strength, on the other hand,
is positively associated with fatness since fatter children and adolescents tend to be
taller and heavier (Beunen et al., 1983; Malina and Bouchard, 1991).

Compared to males, females, on average, have a greater prOportion of fat
mass. Although this is apparent in childhood, the sex difference becomes more
defined during adolescence when the fat mass in girls increases at an estimated rate
that is almost twice that of boys (Malina and Bouchard, 1991). The sex difference in
fatness is often invoked as underlying sex differences in performance.

The fitness of the fattest (obese) and leanest children in several age groups of
national samples of Belgian boys and girls was compared (Beunen et al., 1983;
Malina et al., 1995). Within each sex and age group, the leanest children performed
better, on average, on a variety of motor fitness tests. Static shoulder strength and
absolute power output (PWC170, measured in girls only) were exceptions. Absolute
values were higher in obese children as being heavier in absolute body mass,
regardless of the composition of fat mass and fat-free mass, often results in better
performances. However, when adjusted for body weight, the leanest children

performed better per unit body mass.

40

Naidoo (1999) reported that fatness, estimated as the sum of five skinfolds,
accounted for 31% of the variation in grip strength per unit mass of 10 year old South
African boys. Grip strength per unit of body weight was negatively related to fatness
indicating the non-functional contribution of subcutaneous fat to performance,
particularly for the fattest boys. F atness, on the other hand, explained only a small
proportion, between 1% to 5%, of the variation in the dash, distance run, push-up, sit
and reach, sit-ups, and standing long jump of the 10 year old boys.

In a chronically undernourished population, muscularity and subcutaneous fat
have a limited influence on the variance in several motor performances (Béne’fice and
Malina, 1996). Muscularity and subcutaneous fat predicted the performances only in
6-10 year old boys, but not in younger and older age groups of boys. However,
muscularity and fatness contributed more significantly to the variance in the
performances of girls. Among rural Zapotec children in Oaxaca, fatness as estimated
from the triceps skinfold was negatively related to the performances in boys but
positively related to the performances in girls 6-13 years of age (Malina and
Buschang, 1985). In a subsequent analysis of a subsample of the Zapotec boys 9-12
year of age, relative fatness predicted from four skinfolds (triceps, subscapular,
midaxillary, and suprailiac) was positively related to grip strength, but had no
relationship with running and jumping performances (Malina and Little, 1985). The
relatively close relationship between fatness and performance among extremely lean
children suggests that there might be a threshold below which fat mass does not exert
a negative influence on motor performance, but rather relates positively to
performance. Conversely, there may also be a threshold above which excess fat mass
exerts a negative influence on performance (Malina and Little, 1985; Bénéfice and

Malina, 1996).

41

Taken together, anthropometric correlates of physical fitness are often
analyzed through partial correlational analyses. Few studies have conducted
multivariate analyses to explain the variation in performance associated with body
segments, lengths, and proportions and body composition. The Heath-Carter
somatotype protocol has not been extensively used with Latin American children,
particularly those from marginal nutritional and living conditions. Furthermore, the
difficulty in keeping the integrity of the three scores that make up somatoype as a
composite has limited the exploration of the relationship between somatotype and
physical fitness. Physical growth, especially height and weight, of stunted children
have been examined to a great extent, while relatively few studies have examined the
consequences of stunting on the performance of physical fitness as well as the

distribution of subcutaneous fat in stunted and non-stunted children.

42

Bibliography

Alexander P (1992) Caracteristicas fisicas y morfologicas del Venezolano. In
Memorias IV Congreso Nacional de Educacion F isica, Deporte y Recreacion,
Ponencias Trabajos libres y Talleres. Caracas: Corporacion Venezolana de
Guayana Fundeporte, pp 170-223.

Alexander P, Mota D, Arevalo A (1993) Norrnas para la evaluacion de la aptitud
fisica y caracteristicas morfologicas del estudiante Guatemalteeo cle 7.5 a 18.4
anos. Ciencias de la Actividad F isica 7: 1-54.

Baacke LW. 1964. Relationship of selected anthropometric and physical perfoMance
measures to performance in the running hop, step, and jump. Res Quart
35:107-115.

Baumgartner RN, Roche AF. 1988. Tracking of fat pattern indices in childhood: the
Melbourne grth study. Hum Biol 60:549-567.

Baumgartner RN, Roche AF, Guo SM, Lohman T, Boileau RA, Slaughter MH. 1986.
Adipose tissue distribution: the stability of princip ..1 components by sex,
ethnicity and maturation stage. Hum Biol 58:719-735.

Be'ne'fice E, F ouéré T, Malina RM. 1999. Early nutritional history and motor
performance of Senegalese children, 4-6 years of age. Ann Hum Biol 26:443-
455.

Bénéfice E, Fouéré T, Malina RM, Beunen G. 1996. Anthropometric and motor
characteristics of Senegalese children with different nutritional histories.
Child Care Health Dev 22:151-165.

Bénéfice E, Garnier D, Simondon KB, Malina RM. 2001a. Relationship between
stunting in infancy and growth and fat distribution during adolescence in

Senegalese girls. Eur J Clin Nutr 55:50-58.

43

Bénéfice E, Garnier D, Simondon KB, Malina RM. 2001b. Differences in growth and
body composition during puberty between Senegalese adolescent girls stunted
or not stunted in infancy. Biom Hum et Anthropol 19:55-61.

Be'ne'fice E, Malina RM. 1996. Body size, body composition and motor performances
of mild-to-moderately undernourished Senegalese children. Ann Hum Biol
23:307-321.

Beunen GP, Malina RM, Ostyn M, Renson R, Simons J, Van Gerven D. 1982.
Fatness and skeletal maturity of Belgian boys 12 through 17 years of age. Am
J Phys Anthropol 59:387-392.

Beunen GP, Malina RM, Ostyn M, Renson R, Simons J, Van Gerven D. 1983.
Fatness, growth and motor fitness of Belgian boys 12 through 20 years of age.
Hum Biol 55:599-613.

Cameron N, Kgamphe J S, Leschner KF, Farrant PJ. 1992. Urban-rural differences in
the growth of South African black children. Ann Hum Biol 19:23-33.

Carter JEL, Heath BH. 1990. Somatotyping - Development and Application.
Cambridge: Cambridge University Press.

Chavez A, Martinez C. 1984. Behavorial measurements of activity in children and
their relation to food intake in a poor community. In: Pollitt E, Amante P,
editors. Energy Intake and Activity. New York: Alan R. Liss. p 303-321.

Clarke HH. 1957. Relationships of strength and anthropometric measures to physical
performances involving the trunk and legs. Res Quart 28:223-233.

Clarke HH, Degutis EW. 1964. Relationships between standing broad jump and
various maturational, anthropometric, and strength tests of 12-year-old boys.
Res Quart 35:258-264.

Corlett JT. 1988. Strength development of Tswana children. Hum Biol 60:569-577.

44

Corlett JT, Mokgwathi M. 1987. Running performances of Tswana children. Phys Ed
Rev 10:110-113.

Cressie NAC, Withers RT, Craig NP. 1986. The statistical analysis of somatotype
data. Yrbk Phys Anthropol 29:197-208.

Cureton K, Boileau R, Lohman T. 1975. Relationship between body composition
measures and AAHPER test performances in young boys. Res Quart 46:218-
229.

de Onis M, Blossner M. 2000. Prevalence and trends of overweight among preschool
children in developing countries. Am J Clin Nutr 72:1032-1039.

de Onis M, Frongillo EA, Bldssner M. 2000. Is malnutrition declining? An analysis
of changes in levels of child malnutrition since 1980. Bull World Health
Organ 78:1222-1233.

de Onis M, Habicht J-P. 1996. Anthropometric reference data for international use:
recommendations from a World health Organization expert committee. Am J
Clin Nutr. 64:650-658.

de Onis M, Monteiro C, Akre' J, Clugston G. 1993. The worldwide magnitude of
protein-energy malnutrition: an overview from the WHO global database on
child growth. Bull World Health Organ 71 :703-712.

Diaz Gamboa J, Fuentes Heinrich E (1996) Composicion corporal y somatotipo de
infantes Aymara que habitan en elaltiplano del norte de Chile. Archivos de la
Sociedad Chilena de Medicina del Deporte 41 :87-95.

Espenschade AS. 1963. Restudy of relationships between physical performances of
school children and age, height, and weight. Res Quart 34:144-153.

Eveleth PB, Tanner JM. 1990. Worldwide Variation in Human Growth, 2“‘1 edition.

Cambridge: Cambridge University Press.

45

Garn SM. 1955. Relative fat patterning: an individual characteristic. Hum Biol 27:75-
89.

Godoy JD, Bareos M. 1995. Potencia aerobica maxima, somatotipo y composicién
corporal de nifias escolares de 4 a 1 l afios de edad, de nivel socioeconomico
bajo. Archivos de la Sociedad Chilena de Medicina del Deporte 40:21-28.

Godoy JD, Ugarte J, Donoso V, Zupeuc F, Herrera A, Valle A. 1994. Somatotipos de
escolares varones Chilenos: cambios con el crecimiento (Seguimento de
cohortes y estudios transversales durante 5 aflos). Archivos de la Sociedad
Chilena de Medicina del Deporte 39:1 19-134.

Golden MHN. 1996. The effect of early nutrition on later growth. In: Henry CJ K,
Ulijaszek SJ, editors. Long-Term Consequences of Early Environment:
Growth, Development and the Lifespan Developmental Perspective.
Cambridge: Cambridge University Press. p 91-108.

Graham GG, MacLean WC, Kallman CM, Rabold J, Mellits ED. 1980. Urban-rural
differences in the growth of Peruvian children. Am J Clin Nutr 33:338-344.

Graves PL. 1976. Nutrition, infant behavior, and maternal characteristics: A pilot
study in West Bengal, India. Am J Clin Nutr 29:305-319.

Graves PL. 1978. Nutrition and infant behavior: A replication in the Katmandu
Valley, Nepal. Am J Clin Nutr 31:541-551.

Habicht J-P, Martorell R, Yarbrough C, Malina RM, Klein RE. 1974. Height and
weight standards for preschool children: How relevant are ethnic differences
in growth potential? Lancet 1:611-615.

Hamill PVV, Johnson CL, Reed RB, Roche AF. 1977. NCHS growth curves for
children birth-l8 years. DHEW Publ No. 78-1650, Vital and Health Statistics,

Series 11, No. 165. Washington: US Government Printing Office.

46

Hamill PVV, Johnston FE, Lemeshow S. 1972. Height and weight of children:
socioeconomic status. DHEW Publ No. 73-1606, Vital and Health Statistics,
Series 11, No. 119. Washington: US Government Printing Office.

Hoffman DJ, Sawaya AL, Coward WA, Wright A, Martins PA, de Naseimento C,
Tucker KL, Roberts SB. 20003. Energy expenditure of stunted and non-
stunted boys and girls living in the shantytowns of sao Paulo, Brazil. Am J
Clin Nutr. 72:1025-1031.

Hoffman DJ, Sawaya AL, Verreschi 1, Tucker KL, Roberts SB. 2000b. Why are
nutritionally stunted children at increased risk of obesity? Studies of metabolic
rate and fat oxidation in shantytown children from $50 Paulo, Brazil. Am J
Clin Nutr 72:702-707.

Johnston FE. 1992. Developmental aspects of fat patterning. In: Hernandez M,
Argente J, editors. Human growth: Basic and clinical aspects. Amsterdam,
Netherlands: Elsevier Science Publishers. p 217-226.

Katzmarzyk PT, Malina RM, Beunen GP. 1997. The contribution of biological
maturation to the strength and motor fitness of children. Ann Hum Biol
24:493-505.

Keller W. 1988. The epidemiology of stunting. In: Waterlow JC, editor. Linear
Growth Retardation in Less Developed Countries. New York: Raven Press. p
17-39.

Malina RM. 1974. Adolescent changes in size, build, composition and performance.
Hum Biol 46:117-131.

Malina RM. 1975. Anthropometric correlates of strength and motor performance.

Exerc Sport Sci Rev 3:249-274.

47

Malina RM. 1978. Secular changes in growth, maturation, and physical performance.
Exerc Sport Sci Rev. 6:203-255.

Malina RM, 1983. Growth and maturity profile of primary school children in the
Valley of Oaxaca, Mexico. Garcia de Orta Serie de Antropobiologia: Revista
Do Instituto de Investigacao Cientifica Tropical (Lisboa) 2 ( l & 2): 153-157.

Malina RM. 1984. Physical activity and motor development/perfonnance in
populations nutritionally at risk. In: Pollitt E, Amante P, editors. Energy
Intake and Activity. New York: Alan R. Liss. p 285-302.

Malina RM. 1986. Child growth and health studies in Oaxaca. Practical Anthropology
8:13-15.

Malina RM. 1992. Physique and body composition: effects on performance and
effects on training, semistarvation, and overtraining. 1n: Brownell KD, Rodin
J, Wilmore J H, editors. Eating, Body Weight and Performance in Athletes:
Disorders of Modern Society. Philadelphia: Lea & Febiger. p 94-111.

Malina RM. 1994. Anthropometry, strength and motor fitness. In: Ulijaszek SJ,
Mascie-Taylor CGN, editors. Anthrop :netry: '. rte Individual and the
Population. Cambridge: Cambridge University Press. P 160-177.

Malina RM. 1995. Anthropometry. In: Maud PJ, Foster C, editors. Physiological
Assessment of Human Fitness. Champaign, IL: Human Kinetics. p 205-219.

Malina RM. 1996. Regional body composition: age, sex, and ethnic variation. In:
Roche AF, Heymsfield SB, Lohman TG, editors. Human Body Composition.
Champaign, IL: Human Kinetics. p 217-255.

Malina RM, Bouchard C. 1991. Growth, Maturation, and Physical Activity.

Champaign, IL: Human Kinetics.

48

Malina RM, Buschang PH. 1985. Growth, strength and motor performance of
Zapotec children, Oaxaca, Mexico. Hum Biol 57:163-181.

Malina RM, Himes JH, Stepick CD, Gutierrez Lopez F, Buschang PH. 1981. Growth
of rural and urban children in the Valley of Oaxaca, Mexico. Am J Phys
Anthropol 55:269-280. .

Malina RM, Huang Y-C, Brown KH. 1995. Subcutaneous adipose tissue distribution
in adolescent girls of four ethnic groups. Int J Obes 19:793-797.

Malina RM, Little BB. 1985. Body composition, strength, and motor performance in
undernourished boys. In Binkhorst RA, Kemper HCG, Saris WHM, editors.
Children and Exercise XI. Champaign, IL: Human Kinetics. p 293-300.

Malina RM, Little BB, Buschang PH, DeMoss J, Selby HA. 1985. Socioeconomic
variation in the growth status of children in a subsistence agricultural
community. Am J Phys Anthropol 68:385-391.

Malina RM, Little BB, Shoup RF, Buschang PH. 1987. Adaptive significance of
small body size: strength and motor performance of school children in Mexico
and Papua New Guinea. Am J Phys Anthropol 73:489-499.

Martorell R, Rivera J, Kaplowitz H, Pollitt E. 1992. Long-tenn consequences of
growth retardation during early childhood. In: Hernandez M, Argente J,
editors. Human Growth: Basic and Clinical Aspects. Amsterdam,
Netherlands: Elsevier Science. p 143-149.

Matsui H, Tamura Y. 1975. Characteristics in physical fitness of urban and rural
people. In: Asahina K, Shigiya R, editors. Human Adaptability Volume 3:
Physiological Adaptability and Nutritional Status of the Japanese. Tokyo:

University Tokyo Press. p 41-68.

49

McCance RA. 1966. The effects of normal development and of undernutrition on the
grth of the calcified tissues. Tijd Soc Geneesk 44:569-573.

Meredith HV. 1979. Comparative findings on body size of children and youths living
at urban centers and in rural areas. Growth 43295-104.

Meredith HV. 1982. Research between 1950 and 1980 on urban-rural differences in
body size and grth rate of children and youths. Adv Child Develop Behav
17:83-138.

Montoye HJ, F rantz ME, Kozar AJ. 1972. The value of age, height and weight in
establishing standards of fitness for children. J Sports Med Phys Fitness
12:174-179.

Mueller WH. 1988. Ethnic differences in fat distribution during growth. In: Bouchard
C, Johnston FE, editors. Fat distribution during growth and later health
outcomes. New York: Plenum. p 127-145.

Munetaka H, Matsuura Y, Munetaka H. 1971. On the study if the differences in motor
ability in children between three different communities; island, housing
development and urban area. Res J Phys Ed 16:91-97.

Murguia R, Dickinson F, Cervera M, Ligia UC. 1990. Socio-economic activities,
ecology and somatic differences in Yucatan, Mexico. Stud Hum Ecol 9:11 1-
134.

Naidoo RB. 1999. The effects of socioeconomic status, ethnicity, and nutritional
status on the growth and physical fitness of 10 year old south African boys.
Doctoral Dissertation, Michigan State University.

Norgan NO. 1991. Anthropometric assessment of body fat and fatness. In J. H. Himes
editor. Anthropometry Assessment of Nutritional Status. New York: Wiley-

Liss. p 197-212.

50

Orr CM, Dufour DL, Patton JQ. 2001. A comparison of anthropometric indices of
nutritional status in Tukanoan and Achuar Amerindians. Am J Hum Biol
13:301-309.

Parnell RW. 1958. Behaviour and Physique: An Introduction to Practical and Applied
Somatometry. London: Edward Arnold.

Pate RR, Slentz CA, Katz DP. 1989. Relationship between skinfold thickness and
performance of health related fitness test items. Res Quart Ex Sport 60: 1 83-
189.

Pefia Reyes ME. (2002). Growth status and physical fitness of primary school
children in an urban and a rural community in Oaxaca, Southern Mexico.
Doctoral Dissertation, Michigan State University.

Pilicz S, Sadowska J. 1973. Z badan nad rozwojem I sprawnoscia fizyczna mlodziezy
szkél podstawowych. Wychowanie F izyczne i Sport 1:3-10.

Popkin BM, Horton S, Kim SW, Mahal A, J in SG. 2001. Trends in diet, nutritional
status, and diet-related noncommunicable diseases in China and India: the
economic costs of the nutrition transition. Nutr Rev 59:379-390.

Popkin BM, Richards MK, Montiero CA. 1996. Stunting is associated with
overweight in children of four nations that are undergoing the nutrition
transition. J Nutr 126:3009-3016.

Rarick GL, Oyster N. 1964. Physical maturity, muscular strength, and motor
performance of young school-age boys. Res Quart 35:523-531.

Riendeau R, Welch B, Crisp C, Crowley L, Griffin P, Brockett J. 1958. Relationships
of body fat to motor fitness test scores. Res Quart 29:200-203.

Rocha Ferreira MB, Malina RM, Rocha LL. 1991. Anthropometric, functional and

psychological characteristics of eight-year-old Brazilian children from low

51

socioeconomic status. In: Shephard RJ, Parizkova J, editors. Human Growth,
Physical Fitness and Nutrition. Basel: S. Karger. p 109-118.

Sarria A. 1992. Methods for assessing fat patterning in children. In: Hernandez M,
Argente J, editors. Human growth: Basic and clinical aspects. Amsterdam:
Elsevier Science. p 233-243.

Sawaya A, Grillo L, Verreschi L, Silva AD, Roberts S. 1998. Mild stunting is
associated with higher susceptibility to the effects of high fat diets: studies in
a shantytown population in $50 Paulo, Brazil. J Nutr 128:4158-420S.

Schroeder DG, Martorell R, Flores R. 1999. Infant and child growth and fatness and
fat distribution in Guatemalan adults. Am J Epidemiol 149: 177-185.

Scrimshaw NS, Béhar M. 1961. Protein malnutrition in young children. Science
133:2039-2047.

Secretaria de Salud. 1998. Aspectos relevantes sobre la estadistica de deficiencias de
la nutricion. Salud Publica de México 40: 206-214.

Seils LG. 1951. The relationship between measures of physical growth and gross
motor performance of primary-grade school children. Res Quart 22:244-260.

Sheldon WH, Dupertuis CW, McDermott E. 1954. Atlas of Men: A Guide for
Somatotyping the Adult Male at All Ages. New York: Harper and Brothers.

Sheldon WH, Stevens SS, Tucker WB. 1940. The Varieties of Human Physique. New
York: Harper and Brothers.

Slaughter MH, Lohman TG, Misner JE. 1977. Relationship of somatotype and body
composition to physical performance in 7-to-12-year-old boys. Res Quart

48: 159-168.

52

Slaughter MH, Lohman TG, Misner J E. 1980. Association of somatotype and body
composition to physical performance in 7-12 year-old-girls. J Sports Med
20:189-198.

Spurgeon JH, Meredith HV, Onuoha GBI, Giese WK. 1984. Somatic findings at age
9 years on three ethnic groups of Nigerian urban and rural boys. Growth
48:176-186.

Spurr GB. 1983. Nutritional status and physical work capacity. Yrbk Phys Anthropol
26:1-35.

Spurr GB. 1984. Physical activity, nutritional status, and physical work capacity in
relation to agricultural productivity. In: Pollitt E, Amante P, editors. Energy
Intake and Activity. New York: Alan R. Liss. p 207-261.

Spurr GB, Reina JC, Dahners HW, Barac-Nieto M. 1983. Marginal malnutrition in
school-aged Colombian boys: functional consequences in maximum exercise.
Am J Clin Nutr 37:834-847.

Stinson S. 1985. Sex differences in environmental sensitivity during grth and
development. Yrbk Phys Anthropol 28:123-147.

Thomas JR, Thomas KT. 1988. Development of gender differences in physical
activity. Quest 40:219-229.

Torun B, Chew F. Protein-Energy Malnutrition. 1994. In Shils ME, Olson JA, Shike
M, eds. Modern Nutrition in Health and Disease, Vol. 2, 8‘h edition.
Philadelphia: Lea & F ebiger. p 950-976.

Van Lenthe F, Kemper H, Mechelen WV, Post G, Twisk J, Welten D, Snel J. 1996.
Biological maturation and the distribution of subcutaneous fat from
adolescence into adulthood: the Amsterdam growth and health study. Int J

Obes Rel Metab Disord 20:121-129.

53

Walker ARP, Walker BF. 1997. Moderate to mild malnutrition in African children of
10-12 years: roles of dietary and non-dietary factors. Int J Food Sci Nutr
48:95-101.

Waterlow JC. 1972. Classification and definition of protein-calorie malnutrition. Brit
Med J 3:566-569.

Waterlow JC. 1973. Note on the assessment and classification of protein-energy
malnutrition in children. Lancet 2:87-89.

Waterlow JC, Buzina R, Keller W, Lane JM, Nichaman MZ, Tanner JM. 1977. The
presentation and use of height and weight data for comparing the nutritional
status of groups of children under the age of 10 years. Bull World Health
Organ 55:489-498.

Wilson WM, Dufour DL, Staten LK, Barac-Nieto M, Reina JC, Spurr GB. 1999.
Gastrointestinal parasitic infection, anthropometrics, nutritional status, and
physical work capacity in Colombian boys. Am J Hum Biol 11:763-771.

World Health Organization (WHO). 1997. Physical Status: The Use and
Interpretation of Anthropometry. Geneva: World Health Organization,
Technical Report Series, No. 854.

World Health Organization (WHO). 2001. http://www.who.int/nut/pem.htm.

Yip R, Scanlon K. 1994. The burden of malnutrition: A population perspective. J

Nutr 124:20438-20468.

54

and 853me .I. and ”385. 352me u >83... ”Eweobm atO 638m .1. AFC £83 weed @525me n 3m
828.5388 .8309 8:865 08: 832 m me 3:35 08>» S: bEwe was 5.8 05 mm :25 88:

BE: .85 mama on... .8585 08 beam 8888 05 E wow: 88: 8 8:85 28 863m 2: =m 8 .8888 3mm" 8.“ 28:28.80 3:0
.585 08 3:05.88 biozmzfim 08 83355, o>5 850 85 5:3 38: 30:5 2: £3, Emmy: 8 .336: .owa .8253 98:22.80...

 

 

385:. ED mad 8 mod ”:38? 8:82:88 .82 95
2.5 .75 e3 2 8.5 - His: 82 u 5 2.8 2 2:52 8525 see see
fine a 38 8.5 2 8.5 no? 92 n 5 28 .28» 2A 522% oaoawoeom 68: 2%: eon 888m
25 e8
2.8.15 25 - 2 :5 - mama? so u 5 250 mm 32 .28»
:28 2 cm 5.5 2 So 55o: as u 5 28 w 222% 322 2:85 :8: ._e a eaten £82
325 a8
25 .3m 35 2 a; - ”5225 can 2 N? u 5 82:8 Lo a=e> 2: see .22;
san— ear? m3 2 Be seen: 28 es... 25 2e 522% Boga 52on £3: weeeoaem one 2%:
32.5 28 m3 - 2 Rd - 2863 22.528 523% 2222
3m omd 2 25 ”:38: 8 8&3 28¢ £89» arm when
:25 Bevan mmd 9 Cd Home. 8v u 5 mxom 88:8 3me 2888 585:2 coco: 8530 can 8th
325 28 :5 2 a2 - ”5525 8225228 52
3m 85 2 8.5 - 5%: an” n 5 28 5 2852 238 see .22; me
ease 2&2. 2.5 2 So ”on... SA u 5 .28 222% 2.3 8255 e885. :3: azom
38% 28:39:00 .8 omega e xom 295m 83m

 

.owm .8 £8.» 2-5 56:5

E 8.22 mmofiu 165.38 use .8383 can 330: .oma c8253 macaw—oboe 3:8 8286883 com: :2: 883m .8 Egm _.~ 635,—.

55

.m 8820 E 82:88: 8m 83— dfiom 8:: 8:88 .8088 02882828: 880.583 85 £23 b88888 :8 £88882:
£88888 mass—:23 :8 88:89: 3.28% 05. 98% mm: some :_ mofim 2:88 8.: 88m: .8: 88:: 0582—809“ :8: 882839..

 

 

a: .. 3 .. Z. .. am 1 5 22: 2 comics: 2228 82 2 68: 8:8: 822...:
e: . 3. .. mm .. Ame n 5 m»om 82.5 .28» :2 82:8 28.2: e8 22:22: 88
85 32 $8: 282:
NN a. New a. 9m ... AN: H :5 2:0 88» :6 £18 82.8w Sam—EU 8:: .8800
2.5 8.: 2 2282
Z 3. .2 2% u 5 2o: .28» _ 7e .82: 82:2 28.28 2.8 c 8 so 28:
ad ad Em $3; n :5 £20 . .833 8:28: .88.»
Z 2. Z 8%.: u 5 8o: :2 .2228 228a 82228: :8: 8 so 2282...
E em 2 38.: n 5 22: 822 22222 .28»
e.~ 3. 2 3:: u 5 a2: _ :2 82:8 .88... 22.88., 38: 82282
Z .. 2 . em . 62 u 5 22:
2 . S. .. 2 _. Q: n 5 8o: :82 228; 2 8282282
e: .. 3. .. Wm .. 82 u 5 220 :22 822228 .265 222288:
S . 3. .. ed .. $2 .1. 5 a»om 2225 22: .28» :2 22:8 8282 88: 8 so 2822
398888”: 3:88:82 3:82:28”.— xom £95m beam

 

85:88 :88: 8:3 1:38 :2 own .8 88» Z 9 o 58:5 :8 8:59:88 :82 NN 032%

56

CHAPTER 3

Methods and Procedures

The present study was developed within the activities of the general project:
“Secular change in size, strength and motor fitness in the Valley of Oaxaca, Mexico”,
with Robert M. Malina as the principal investigator (Malina, 1999, 2002). The data were
collected in Fall 1999 and Spring 2000, and descriptive analyses of rural and urban
children were reported by Pefia Reyes (2002). This study extends the analyses to specific
issues related to the anthropometric correlates of physical fitness, somatotype, and

growth stunting.

Communities Studied

The communities under study, the participants surveyed, and anthropometric and
physical fitness data, are briefly described. More detailed information has been
previously reported by Pefia Reyes (2002).

The state of Oaxaca is in the southern part of Mexico, and the Valley of Oaxaca
lies in the center of the state (Figure 3.1). The altitude of the valley floor ranges from
1420 to 1740 m; the valley is surrounded by mountains that reach over 3000 m (Kirkby,
1973). The raised valley floor requires that the issue of altitude affecting the performance
of physical fitness tasks, especially cardiovascular endurance, be considered. The
potential influences of physiological changes associated with such a moderate altitude are
negligible. Altitude appears to affect the threshold level of performance and it seems to

be prevalent with training at altitudes above approximately 2,500 meters (Astrand, 1954).

57

Performance at high altitude is also affected by level of acclimatization, usually viewed
in field studies as duration of residence at a specified altitude. Given the rather moderate
altitude of the Valley of Oaxaca and the long term residence of the children at this
altitude, the influence of altitude on fitness or performance in the present study is not an
issue.

The city of Oaxaca de Juarez is located at the junction of the Y-shaped valley.
School children in two communities, one rural and the other urban, in the Valley of
Oaxaca comprise the basis for this study. Living conditions in the rural and urban
communities in Oaxaca are quite different.

Santo Tomas Malzaltepec is a rural community located about 23 km northwest of
the city of Oaxaca. The community is largely based on a tradition of kinship with small-
scale family subsistence farming, although a small but increasing number are engaged in
other economic activities. In addition, a significant number of adult males have migrated
to Mexico City and the United States, and regularly send funds back to the community.
There is a health center that is staffed by a public health nurse and a physician who goes
to the center daily during the week, except on weekends. Essential utilities such as
sewage systems, water availability within each household, and water treatment, though
improving in recent years, are generally lacking in the community. The number of paved
roads is also limited.

San Juan Chapultepec is an urban community located on the slopes of the hills
west of the city of Oaxaca. Most of its population works in the city of Oaxaca. San Juan
is an irregular settlement on the edge of the city. Such settlements are called m

populares in Mexico, although historically, San Juan Chapultepec was an independent

58

community (Graedon, 1976; Murphy and Stepick, 1991). Facilities for health care and
sanitation in San Juan are far from the expectations of an urbanized center, but they are in
advance of those available in Santo Tomas. San Juan, for example, is within several
kilometers of a hospital and related public health facilities.

Both communities were categorized as rural and urban, respectively, based on the
census on the national census for Mexico in 2000 (INEGI, 2002). Comparative data for
indicators of social, economic, educational and health conditions in Santo Tomas (rural),
San Juanito' (urban), the city of Oaxaca and the state of Oaxaca are summarized in Table

3.1.

Participants

An initial sample of 708 primary school children (355 boys, 353 girls), 6.01 years
to 15.74 years (9.3 i 2.0 years), was measured during surveys of the rural and urban
communities in the Fall 1999 and early Spring 2000. According to Mexican law, it is
mandatory for all children to attend school. Hence, the children represent the total student
body in each school at the time of the surveys, except for one child from the urban
community. All of the children were apparently healthy and showed no overt signs of
being in a diseased state and no obvious physical disabilities. It is possible that several
children might have had minor anomalies, but none were immediately apparent. It is
likely that children with disabilities or congenital anomalies died before school age.
Several cases of Down Syndrome were noted in the mortality records for the 19803 and

19905, and all deaths occurred before 5 years of age (Malina, field notes).

For the purpose of the present study, the sample was limited to children between
the ages of 6.00 to 11.99 years of age, which includes approximately 91% of the initial
sample. Numbers of children older than 12.0 years were small in specific age groups
within each sex and community. The rural sample thus comprised 329 children, 154 boys
and 175 girls with mean ages of 8.8 i 1.5 years and 9.1 i 1.7 years, respectively. The
urban sample included 318 children, 161 boys and 157 girls with mean ages of 8.8 i 1.7
years and 9.0 j; 1.7 years. Chronological ages were calculated from date of measurement
and birth dates taken from official school enrollment records.

The study was approved by the University Committee for Research Involving
Human Subjects at Michigan State University (Appendix A). Local authorities and school
officials of each community also approved the project. Parents provided informed
consent for their children to participate in the study. Self-assent was also obtained from
school children 10 years of age and older. The statement of participation was read to the
parents of the children and to the older children, and they were informed that their

children or they could participate and/or withdraw from the study at any time.

Anthropometric Variables

Measurements were taken following the protocol described in Lohman et a1.
(1988; also see Malina 1995). Bilateral dimensions were taken on the left side, which was
consistent with the measurements taken in earlier studies done in Oaxaca. Body weight
(kg) was measured using a portable scale accurate to 100 grams. The children were
measured without shoes, but wore light clothing with all accessories removed. Stature

(cm) was measured with a portable field anthropometer with the children standing with

60

arms hanging relaxed at the sides in an erect posture and the eyes in a horizontal plane,
with heels placed together without shoes and body weight evenly distributed between
both feet. Sitting height (cm) was measured with the portable field anthropometer with
the children sitting erect on a table with feet hanging freely and hands positioned on the
thighs with palms down. Biacromial and bicristal breadths were measured with the upper
end of the anthropometer used as a large sliding caliper, while bicondylar and
biepicondylar breadths were measured with a small sliding caliper. Relaxed arm, flexed
arm, and calf circumferences were measured with a flexible, non-stretchable tape. The
triceps, subscapular, suprailiac and medial calf skinfolds were measured with the Lange
skinfold caliper to the nearest 0.5 mm. The weighing scale and all anthropometric
equipment (anthropometer, calipers, and tapes) were checked and calibrated daily before
measurements were taken. Detailed descriptions of the anthropometric techniques are

described in Appendix B.

Derived Variables

The anthropometric dimensions were used to derive several additional variables:
(a) the body mass index (BMI, - kg/mz); (b) estimated leg length (subischial length) —
standing height minus sitting height; (c) the sitting height to standing height ratio — sitting
height divided by stature multiplied by 100; (d) the sum of four skinfolds (triceps,
subscapular, suprailiac and medial calf); and (e) the trunk-extremity ratio — sum of
subscapular and suprailiac skinfolds divided by sum of triceps and medial calf skinfolds.

Estimated arm and calf muscle circumferences were derived from relaxed arm

61

circumference and the triceps skinfold, and calf circumference and the medial calf
skinfold, respectively, using the following formula:

Muscle circumference = C — (T! S)
where C is relaxed arm or calf circumference (cm) and S is the triceps or medial calf
skinfold (cm). The derived values provide an estimate of the relative muscularity of the
upper arm and lower leg, respectively (Malina, 1995).

Somatotype was estimated with the Heath-Carter anthropometric protocol using
the algorithms provided by Carter and Heath (1990, p 374):

1. Endomorphy = -0.7182 + 0.1451(X) — 0.00068 (X2) + 0.0000014(X3)
where X is the sum of the triceps, subscapular, and suprailiac skinfolds and multiplied by
170.18/height in cm.

2. Mesomorphy = [(0.858 x biepicondylar breadth) + (0.601 x bicondylar

breadth) + (0.188 x corrected arm circumference) + (0.161 x corrected calf

circumference)] — (stature x 0.131) + 4.50
where the corrected arm circumference is flexed arm circumference (cm) minus the
triceps skinfold (cm) and the corrected calf circumference is calf circumference (cm)
minus the medial calf skinfold (cm).

3. Ectomorphy = HWR x 0.732 — 28.58
where HWR is height (cm) divided by the cube root of weight (kg). If HWR is less than
40.75 but more than 38.25, then Ectomorphy = HWR x 0.463 — 17.63, and if HWR is

equal or less than 38.25, a rating of 0.1 is given.

62

The somatotype of a given child is defined by the three components: endomorphy (first
component), mesomorphy (second component), and ectomorphy (third component),

respectively.

Physical Fitness
The physical fitness battery included a combination of performance- and health-
related tasks (Malina, 1991). Three of the tests were used in earlier studies in the rural

community — grip strength, dash, and standing long jump (Malina and Buschang, 1985).

The other three, all indicators of health-related fitness (American Alliance for Health,

Physical Education, Recreation and Dance, 1980), were selected for their suitability in

field conditions, specifically for ease of administration without extensive equipment. All

tests were administered during the school day. The tests were demonstrated and
explained to the children, and a warm-up, largely stretching, was provided. Detailed

descriptions of each fitness test are described in Appendix B.

1. Static strength - Static grip strength of the left and right hands was measured with a
Stoelting adjustable dynamometer to the nearest 0.5 kg. Three trials were given with
each hand and the best score was retained for each. The children were instructed to
exert maximum effort on each attemp‘

2. Speed - Running speed was measured as tne time clasped in a 35 yard dash (32.3
meters) from a stationary start. Two trials were given and the faster time to the
nearest 0.1 second was used. The children ran individually and were instructed to run
as fast as they could at the command “go”. A rest time between the trials was

allowed.

63

. Power - Explosive power was measured as the standing long jump. Three trials were
given and the furthest distance jumped (cm) was used in the analysis. The children

were estimated to jump as far as possible on each attempt.

. Flexibility - The sit and reach was used to estimate the flexibility of the lower back

and upper thigh. Three trials were given and the furthest distance reached (cm) was
used. The children were instructed to reach straight out without bouncing.

Muscular endurance — Endurance of the abdominal musculature was measured as
timed sit-ups, the number completed in 30 seconds. One trial was given. The children
were instructed to do as many sit-ups as they could when the cue counting time starts
and only stopped when they were told.

Cardiovascular endurance - The distance (meters) run and/or walked in 8 minutes for
children in grades 1 to 3 and 12 minutes for grades 4 to 6 was used as an indicator of
cardiovascular fitness. One trial was given. The children were instructed to run at
their own pace for the duration and not to stop moving until the end of the test.

Static grip strength was measured at the time of anthropometry. The dash and the

distance run were conducted on a concrete surface at a designated area within the school

compound for the urban children and at the central plaza of the community for the rural

children. Markers (plastic cones) were used to mark the boundaries and distance covered

for the distance run. The remaining items were also conducted outdoors but in a separate

area.

The sum of left and right grip strength was used as the estimate of overall static

strength. The distance run was converted to an average running speed expressed as

meters per minute (Pate et al., 1989).

64

Measurement Variability and Reliability

All anthropometric dimensions were collected by a single experienced and
qualified individual, while the physical fitness tests were collected with the assistance of
trained physical anthropology students enrolled in Escuela Nacional de Antropologia e
Historia (ENAH) in Mexico, D.F. Each assistant was assigned to conduct a specific
physical fitness test throughout the whole duration of the survey. Intra-observer technical
errors of measurement (TEM) for anthropometric dimensions were within acceptable
ranges and compares favorably to intra- and inter-examiner estimates in earlier studies in
Oaxaca (Buschang, 1980) and in several national health surveys in the United States
(Malina, 1995). Replicate measurements for all anthropometric dimensions taken about
one month apart in the urban community (n = 36) and about two months apart in the rural
community (n = 43). The TEM was calculated for each dimension based on the
measurements taken at both occasion were as follows: weight (0.52kg), height (0.32 cm),
sitting height (0.40 cm), skeletal breadths (0.09-0.35 cm), arm and calf circumferences
(0.21-0.29 cm), and skinfold thicknesses (0.63-0.83 m) (Pefia Reyes, 2002). Using
repeated measures analyses with age as the covariate, within day reliabilities for fitness
tasks that had more than one trial ranged from 0.85 to 0.95 (Pefia Reyes, 2002),

indicating high consistency in performances.

Selection and Training of Assistants/Examiners
There were three phases to the selection and training process of the
assistants/examiners. First, students enrolled in the physical anthropology department at

Escuela Nacional de Antropologia e Historia (ENAH) were identified and interviewed by

65

the field coordinator. Six assistants were selected. Second, a working session was
conducted to provide more detailed information about the project and specific training on
the field protocols. Third, standardizing sessions were conducted for anthropometry and
physical fitness testing. For anthropometry, the specific dimensions were initially
described and demonstrated, after which the assistants practiced taking the measurements
on each other and subsequently on other adults and children. The administration of the
static hand grip strength test was also taught at this time. The assistants were trained in
anthropometry in case they had to take the measurements for the children. However, only
one person was required to collect all the anthropometric measurements in the present
study. For physical fitness, the assistants first familiarized themselves with specific
protocols and equipment in the laboratory. The tests were then performed outdoors on a
basketball court at ENAH and subsequently administered to a group of children, 8-12
years of age. Before administering the anthropometry and physical fitness protocols on
the study sample, the protocols were administered on a group of urban school children in

Cuemavaca, Morelos.

Statistical Analyses
The respective analyses for each question were conducted using the SPSS

statistical software in the following sequence:
Question 1. Anthropometric correlates of physical fitness in rural and urban children.

a) Sex-specific descriptive statistics were calculated for anthropometric dimensions,

derived variables and physical fitness items for each community. Sex-specific

66

b)

d)

multivariate analyses of covariance (MANCOVA) with age as the covariate were
used to test urban-rural differences in performance.

Second order partial correlations were calculated for each of the physical fitness items
in relation to age, stature and body weight for boys and girls separately in each
community. Second order partial correlations were used to statistically control for the
other two variables so that the discrete effects of age, stature, or weight, on physical
fitness could be estimated. The partial correlations were done primarily for
comparative reasons since earlier studies, especially those of well-nourishedchildren,
have used this approach.

Multiple regression was employed first, to estimate the amount of variance in
physical fitness variables explained by age, stature, and weight, and second, to
estimate the contribution of age, height, weight and other anthropometric dimensions
to variation in each physical fitness item. The other anthropometric variables included
the BMI; sitting height and estimated leg length (segment lengths); biacromial,
bicristal, bicondylar, and biepicondylar breadths (skeletal robustness); estimated arm
and calf muscle circumferences (muscularity); and the sum of four skinfolds (fatness).
It should be noted that the correlational analyses were conducted for the purpose of
comparison with those of earlier studies, while the multivariate analyses conducted in
the present study are the primary focus; hence, family error rates in the analyses are

not an issue.

67

Question 2. Heath-Carter somatotypes of rural and urban children.

The analysis of somatotype followed the multivariate statistical procedure

recommended by Cressie et a1. (1986).

a)

b)

d)

Descriptive statistics were initially calculated by age and sex within each community.
Subsequently, an ANCOVA was conducted to test the age effects.

An overall sex-specific multivariate analysis of covariance (MANCOVA) was
conducted for each community with Wilks’s lambda as the test statistic and with age
as the covariate. The subsequent analyses were conducted only for MAN COVAs that
were significant.

Community- and sex-specific pairwise comparisons were made using Hotelling’s T2
with a Bonferroni adjusted alpha level to determine which components contributed to
the significant urban-rural and sex differences.

Forward stepwise discriminant analyses were conducted to determine which
somatotype component(s) best discriminated between rural and urban boys and
between rural and urban girls.

Forward stepwise discriminant analyses were also conducted to determine which
somatotype component(s) best discriminated between boys and girls within each

community.

Question 3. Relationship between the Heath-Carter somatotype components and physical

fitness of rural and urban children.

3)

Sex-specific third order partial correlations were used to statistically control for age

and the other two somatotype components to examine the association of each

68

somatotype components with each of the physical fitness tasks in the children from
each community, e.g., the correlation between endomorphy and standing long jump
when age, mesomorphy and ectomorphy were controlled; mesomorphy and standing
long jump when age, endomorphy and ectomorphy were controlled; ectomorphy and
standing long jump when age, endomorphy and mesomorphy were controlled.

b) Sex- and community-specific multiple regression analysis was used to estimate the
variation in each physical fitness task that was contributed by age and each of the

somatotype components.

Question 4. Comparison of physical fitness of stunted and non-stunted children in
their respective rural and urban communities.

The children were categorized as stunted and non-stunted on the basis of height.
Children whose height-for-age z-scores were more than 2.0 standard deviations below
age- and sex-specific medians for the United States reference were considered stunted.
This is the criterion of the World Health Organization (1997). All other children were
considered non-stunted. The most recent Centers for Disease Control and Prevention
(2000) growth charts were used to calculate age- and sex-specific z-scores for each child.
The prevalence of stunting in the study sample of 6-11 year old children was 28% for
both boys and girls in Santo Tomas, and 13% for girls and 16% for boys in San Juan
(Table 3.2). The subsequent analyses of stunted and non-stunted children were as
follows:

a) Descriptive statistics for the anthropometric and physical fitness variables were

calculated for stunted and non-stunted children by sex within each community.

69

b)

Sex-specific multivariate analysis of covariance (MANCOVA) was used to compare
the anthropometric and physical fitness characteristics of stunted and non-stunted

children within each community.

Question 5. Subcutaneous fat distribution of stunted and non-stunted children.

a)

b)

Descriptive statistics for the trunk-extremity ratio were calculated by sex and
community for stunted and non-stunted children.

Sex-specific analyses of covariance (AN COVA) were used to compare the trunk-
extremity ratio between stunted and non-stunted children within each community.

A principal components analysis (Baumgartner et al., 1986; Malina et al., 1995) was
also used to evaluate subcutaneous fat distribution in stunted and non-stunted
children. The four skinfolds were transformed to natural logarithms and the mean log
skinfold was used as the index of subcutaneous fat for the individual. Log values for
skinfold thicknesses were then separately regressed on the mean log skinfold
thickness for each individual, thus controlling for overall subcutaneous fatness. The
residuals were then subjected to principal component analysis. Components with
eigen values greater than 1.0 were retained for further analysis (Baumgartner et al.,
1986). Component scores of stunted and non-stunted boys and girls within each

community were compared with ANCOVA with age as the covariate.

70

Bibliography

Astrand P. 1954. The respiratory activity in man exposed to prolong hypoxia. Acta
Physiol Scand 30:343-368.

American Alliance for Health, Physical Education, Recreation and Dance. 1980. Health
Related Physical Fitness Test Manual. Reston, VA: American Alliance for Health,
Physical Education, Recreation and Dance.

Baumgartner RN, Roche AF, Guo SM. Lohmar: T, Boiluu RA, Slaughter MH. 1986.
Adipose tissue distribution: the stability of principal components by sex, ethnicity
and maturation stage. Hum Biol 58:719-735.

Buschang PH. 1980. Growth status and rate of school children 6 to 13 years of age in a
rural Zapotec-speaking community in the Valley of Oaxaca, Mexico. Doctoral
Dissertation, University of Texas at Austin.

Carter JEL, Heath BH. 1990. Somatotyping - Development and Application. Cambridge:
Cambridge University Press.

Centers for Disease Control and Prevention (CDC). 2000. National Center for Health
Statistics CDC grth charts: United States.

http://www.cdc.gov/gromhcharts.htm.

 

Cressie NAC, Withers RT, Craig NP. 1986. The statistical analysis of somatotype data.
Yrbk Phys Anthropol 29:197-208.
Graedon TLF. 1976. Health and nutritional status in an urban community of southern

Mexico. Doctoral Dissertation. University of Michigan.

71

Instituto Nacional de Estadistica, Geografia e Informatica (IN EGI). 2002. Sistema para la
Consulta de Informacion Censal. Oaxaca. XII Censo de Poblacion y Vivienda
2000. Aguascalientes: INEGI.

Kirkby AVT. 1973. The Use of Land and Water Resources in the Past and Present Valley
of Oaxaca, Mexico. University of Michigan, Museum of Anthropology, Memoirs,
No. 5, Ann Arbor, Michigan.

Lohman TG, Roche AF, Martorell R, eds. 1988. Anthropometric Standardization
Reference Manual. Champaign, IL: Human Kinetics.

Malina RM. 1990. Growth of Latin American children: socioeconomic, urban-rural and
secular comparisons. Revista Brasileira de Ciencia e Movimiento 4:46-75.

Malina RM. 1991. Fitness and performance: Adult health and the culture of youth. In
Park RJ, Eckert HM, editors. New Possibilities, New Paradigms? Champaign, IL:
Human Kinetics. p 30-38.

Malina RM. 1995. Anthropometry. In: Maud PJ, Foster C, editors. Physiological
Assessment of Human Fitness. Champaign, IL: Human Kinetics. p 205-219.

Malina RM. 1999. Secular change in size, strength and motor fitness in the Valley of
Oaxaca, Mexico. Proposal submitted to the National Science Foundation,
Proposal and Grant No. BCS-9816400.

Malina RM. 2002. Secular change in size and physical fitness in the Valley of Oaxaca,
Mexico. Final report submitted to the National Science Foundation, Proposal and
Grant No. BCS-9816400.

Malina RM, Buschang PH. 1985. Growth, strength and motor performance of Zapotec

children, Oaxaca, Mexico. Hum Biol 57:163-181.

72

 

Murphy AD, Stepick A. 1991. Social Inequality in Oaxaca: A History of Resistance and
Change. Philadelphia: Temple University Press.

Pate RR, Slentz CA, Katz DP. 1989. Relationship between skinfold thickness and
performance of health related fitness test items. Res Quart Ex Sport 60: 1 83-1 89.

Pefia Reyes ME. (2002). Growth status and physical fitness of primary school children in
an urban and a rural community in Oaxaca, Southern Mexico. Doctoral
Dissertation, Michigan State University.

World Health Organization (WHO). 1997. Physical Status: The Use and Interpretation of
Anthropometry. Geneva: World Health Organization, Technical Report Series,

No. 854.

73

  

 

At 88:82.0 £52 :8 21:88:82: 88$ 25m "33m

 

 

 

 

855653 2: .3 5:30. 2: 9.2355 .3580 .3 ao=a> 05 be .32 . 3.. 0.53".
a... one”) coo:
. \ooa 0. '0...’
.59qu ...... .. 2).... .a g ..
_o >U:D> 6...... .. .. m
x... . ~ . ....
m... fl... .../...
no u. no. ’
swung.- o.’ .a...../
.. ....-...L 1.. .2...
us: \5an W
. .4..." a... In a m .00: >3...) .0 some I...
.o\l J. c (on \ o \to. o '0 co
’ I on. a.
J u .1. A .. E .—
....ooolc -. g C n .
o ‘0 o o\~. “on o \c. 09% u no" all]
.. o. o
. ...- ...o. .... .... u ...
...\.. a . I}... - - . i . ... 7...... ..... .
Ill . ova-.00 \QM cocoo\ooHoOoV no” ‘o
00-40.00. o so no not u’s .o
....... 1... otoflom o:— ... .. .. .55: u. .... u. ._ XV
o! OOOOOOOO 0’... open“ I! 0.00.. 6|. 00.0.“.‘im 0.. 0090‘ch
oo 5...-.- uo/a “.50: n- .v no 000 oo'ooso 9\~\ a... sh.
. .\. : u 090x00” ..
.oo \\ m. one. 00\ co: 0.0 no 0 loo. 0 no 9“. no
vooJ .1. .I. a... U. o I co \
... p"; .. so J...\\+w
.- .- ..\\.oo¢ n‘. o‘ 0“

0
C
C
o'-’ 00'
0’ I O.
‘ ’ ‘ O
00' 0‘ O ”I. ’0'
c 0‘ r). 0’
on a o "I"

 

 

74

Table 3.1. Indicators of social, economic, educational and health conditions in the rural
and urban communities with comparative data for the city of Oaxaca and the state of

Oaxaca based on the national census for 2000.

 

 

Santo Tomas San Juan City of State of
(rural) (urban) Oaxaca Oaxaca

Total population 1,939 16,279 256,130 3.4 million
Household Characteristics:
Number of households 422 3,782 60,612 741,005
Sewage connected to street 1% 77% 78% 26%
Sewage to septic tank 14% 5% 6% _ 20%
No sewage installation 84% 18% 14% 54%
Radio 82% 90% 91% 71%
Television 71% 87% 89% 57%
Refi'igerator 29% 61% 72% 37%
Male head of household 85% 73% 70% 77%
Female head of household 15% 27% 30% 22%
Income (% economically active population):
< 1 minimum wage 15% 15% 12% 20%
l to 2 minimum wages 16% 29% 25% 24%
2 to 5 minimum wages 15% 34% 28% 18%
No wages 49% 6% 4% 28%
Health and Education:
Insured population 6% 39% 48% 23%
Uninsured population 93% 59% 50% 76%
Number children/woman 3.1 2.2 1.9 2.1
Literate, 15+ years 91% 91% 94% 78%
Illiterate, 15+ years 9% 9% 5% 20% _
Years of schooling 6 yrs 8 yrs 10 yrs 6 yrs
Did not complete primaria 22% 14% 1 1% 25%

 

Adapted from INEGI (2002) and Pena Reyes (2002).

75

Table 3.2. Numbers of stunted and non-stunted children 6-1 1 years of age by sex within
each community.

 

 

Rural Urban
Boys Girls Boys Girls
Stunted 49 43 25 21
Non-Stunted 126 l l l 132 140
Total 175 154 15 7 161

 

76

 

 

CHAPTER 4

Correlates of Physical Fitness in Rural and Urban Children in Southern Mexico

Introduction

Urbanization is assumed to have a positive influence on the growth status of
children, which in turn may lead to improvements in physical fitness. This assumption
follows fi'om the fact that children residing in urban areas are, in general, taller and
heavier than those living in rural areas (Eveleth and Tanner, 1990), and larger body
size is positively related to commonly used indicators of fitness in children and '
adolescents (Malina, 1975, 1994). These observations are derived largely from studies
in developed countries of Europe and Asia. Data for the United States indicate
negligible rural-urban differences in grth status (Hamill et al., 1972; Meredith,
1979), but urban-rural comparisons of physical fitness are lacking.

The positive effects of urbanization on grth are associated primarily with
improved living conditions related to health care and diet. These include, for example,
ready availability of a constant food supply, access to potable water and sanitation,
and access to health services and quality medical care, educational institutions, and
recreational and welfare facilities within the community (Eveleth and Tanner, 1990).

Urban living is not always associated with improved growth status and
physical fitness (Cameron et al., 1992; Malina et al., 1981; Meredith, 1979). Rapid
urban growth particularly in developing countries is often due to expansions of urban
slums and marginal living areas without access to or availability of the health and
service infrastructures (Murphy and Stepick, 1991).

Physical fitness and performance of children are related to body size and

composition (Malina, 1975, 1994; Malina and Bouchard, 1991; Norgan, 1994).

77

Although body dimensions and composition are influenced by genotype, reductions in
body size and muscle mass are associated with chronic undernutrition. Small body
size and reduced muscle mass tend to contribute to lower levels of fitness and
performance (Spurr, 1984; Malina and Buschang, 1985; Malina and Little, 1985;
Benefice and Malina, 1996).

Undemourished children tend to show reductions in performance or fitness
that commensurate with their reduced body size and specifically muscle mass. When
performances are expressed per unit height or weight, the performances of well-
nourished and undernourished children are reasonably similar, emphasizing the
importance of body size per se (Malina et al., 1987; Spurr, 1984). Such a relationship
has implications for the functional significance of reduced body dimensions in
children with a history of chronic undernutrition. However, the performance and
fitness demands of daily activities do not include adjustments for body size. Simply
stated, ‘bigger is better’ for many fitness and performance tasks (Spurr, 1984).

Attempts to explain individual differences in physical fitness and performance
have been approached in several ways. Previous studies focused primarily on age,
height, and weight (Montoye etal., 1972; Malina, 1975). Some studies include an
indicator of biological maturity, specifically skeletal age (Seils, 1951; Rarick and
Oyster, 1964; Beunen et al., 1997; Katzmarzyk et al., 1997). Relatively few studies
have considered other body dimensions, specifically after controlling for the effects of
age, height, and weight, as possible factor(s) contributing to the variation in physical
fitness and performance (Malina and Buschang, 1985; Benefice and Malina, 1996).

The present study examined the relationship between anthropometric
characteristics and physical fitness of -. hildren from a rural and an urban community

in Oaxaca in southern Mexico. Both communities have a history of marginal living

78

and nutritional conditions suggesting chronic undernutrition (Malina, 1983; Malina et
al., 1980, 1981). The estimated contributions of age and body size to variation in
physical fitness are first considered, and then the estimated contributions of age,

height, weight, and other anthropometric dimensions to fitness are considered.

Methods and Procedures
The communities, sample, anthropometric procedures and fitness tests were
described earlier (Chapter 3). The variables used in the analysis are first listed and the
analytical procedures are then described.
Anthropometric variables included the following:
1) Body size: Weight, height, body mass index (BMI);
2) Segment lengths: Sitting height, estimated leg length;
3) Skeletal breadths: Biacromial, bicristal, biepicondylar, bicondylar;
4) Estimated limb musculature: Estimated midarm muscle circumference (EAMC),
estimated calf muscle circumference (ECAC);
5) Subcutaneous fatness: Sum of 4 skinfolds - triceps, subscapular, suprailiac and
medial calf skinfolds.
Physical fitness variables included the following:
1) Performance-related fitness: 35 yard (32.3 m) dash -— running speed, standing long
jump (SLJ) — explosive power;
2) Health-related fitness: Sum of right and left grip strength (Sum RL) — static
strength, sit-and-reach (SAR) — lower back and upper thigh flexibility, timed sit-ups —
abdominal muscular strength and endurance, distance run - cardiovascular endurance.
The analyses proceeded in four stages. First, sex-specific descriptive statistics

were calculated for anthropometric dimensions, derived variables and physical fitness

79

items for each community. Second, sex-specific multivariate analyses of covariance
(MANCOVA) with age as the covariate were used to test urban-rural differences
determined in anthropometric dimensions and fitness. Third, second-order partial
correlations were calculated for each physical fitness item in relation to age, stature
and body weight for boys and girls separately in each community to allow for
comparison with previous studies. Fourth, multiple regression analyses were used to
estimate the amount of variance in physical fitness items explained by age, stature,
and weight, and then to estimate the contribution of age, height, weight, and other

anthropometric dimensions to variation in each fitness item.

Results

Sex-specific age-adjusted means and standard errors for the anthropometric,
derived, and physical fitness variables of the rural and urban children are presented in
Tables 4.1 to 4.4. Urban children were, on average, significantly taller, heavier, and
more muscular than rural children. Body segments of urban children were also
broader and longer than rural children. Subcutaneous fatness was significantly greater
among urban than rural boys, but did not differ between rural and urban girls. Urban
children had, on average, better performances than rural children with the exception
of grip strength per unit body weight, the dash and distance run, but statistical
significance varied among tasks.

Partial correlations between age, height, and weight and each physical fitness
task, after controlling for the other two variables, are presented in Tables 4.5 and 4.6.
Overall, the correlations were low to moderate, ranging from - 0.26 to 0.43. After
controlling for height and weight, age was positively correlated with performances in

both sexes with the exception of the sit and reach in urban girls. Similarly, height was

80

positively related with performances in both sexes except for the sit and reach in boys
after controlling for age and weight. In contrast, weight was negatively related with
performance except for grip strength in both sexes and the sit and reach in urban boys.

Among physical fitness items, age, height, and weight accounted for a major
portion of the variance in strength, 74% to 82% (Tables 4.7 and 4.8). Age and body
size accounted for 38% (boys) and 40% (girls) of the variance in the dash for rural
children, and 52% (boys) and 57% (girls) in urban children. Age explained most of
the variation in running speed. Slightly more of the variation in strength and speed
were explained by age and body size in females than in boys.

Approximately 31% to 40% of' the var. -=.nce in the standing long jump was
accounted for by age, height and weight. Age contributed most to the variance in
power, except in rural girls where weight was the main contributor. Age and body size
explained more of the variance in power in boys than in girls.

Age, height, and weight explained 26% and 23% of the variance in the
distance run for urban boys and girls, respectively, but only 9% and 6% of the
variance for rural boys and girls. Although height was negatively related to the
distance run, it explained the largest portion of the variance for rural children of both
sexes and urban boys. Age was the main contributor to the variance in cardiovascular
endurance in urban girls.

Age and body size explained more of the variance for sit-ups in boys (24% -
rural; 18% - urban) than in girls (9% - rural; 14% - urban). Height contributed most to
the variance in sit-ups for rural children, whereas age contributed most for the
variance in urban children. In both rural and urban children, weight was negatively
related to abdominal endurance. Variation in the sit and reach explained by age and

body size was very low, ranging from 1% to 6%.

81

When age, height, weight, and other anthropometric characteristics were
regressed on each physical fitness variable, the increase in explained variance was
rather small, ranging from 2% to 14% (Tables 4.9 to 4.12). The increase in the
amount of variance explained in fitness was significant in the sum of right and left
grip strength across all four samples, while the significance in the variance increased
for the other fitness items were variable. The sit and reach (14%) and sit-ups (12%) in
urban boys and the distance run (14%) in rural boys were the only fitness items for
which other anthropometric dimensions explained an additional 10% or more of the
explained variance.

The addition of other anthropometric dimensions to the regression analyses
altered the contributions of age, height, and weight to each physical fitness test. A
decreasing trend in beta coefficients was observed for several items. Weight was an
exception; there was an increasing trend in the coefficients. Beta coefficients indicate
the amount of variance contributed by age and specific anthropometric dimensions,
and the magnitude of change in performance (positive or negative) associated with an
increase or decrease of one standard deviation. Positive beta coefficients suggest
improvements, while negative coefficients suggest decrements in performance
associated with changes in particular anthropometric variables.

In general, the sum of skinfolds contributed negatively to physical fitness.
Across sex and community, a one standard deviation increase in the sum of four
skinfolds decreased performances by 0.65 to 0.01 units, except for the dash in urban
boys and rural girls, and the distance run in urban girls, where there was a marginal
increase. Height accounted for the most variance in strength for rural and urban girls,

while weight contributed the most to the variance in strength for rural boys and

82

estimated arm muscle circumference contributed most to the variance in strength for
urban boys.

Though negative, variation in the sit and reach was largely explained by leg
length in urban boys and rural girls. Biacromial breadth and height were the main
contributors to variation in the sit and reach in rural boys and urban girls, respectively.
Weight contributed negatively and accounted for most of the variance in sit-ups,
except in rural girls, among whom most of the variance was accounted for by
biepicondylar breadth.

In boys, age was the main contributor to variation in the standing longjump.
Estimated arm muscle circumference contributed a comparable amount of variance in
the standing long jump to age in urban boys. The main contributor to the standing
long jump in rural girls was weight and the sum of four skinfolds in urban girls. Both
weight and the sum of skinfolds contributed negatively to variation in thejump.

Similarly, age was the main contributor to variance in the dash, with the
exception of urban boys where weight was the primary factor. The most of the
variance in the distance run was explained by weight in urban boys and rural girls, but

by height in rural boys and urban girls.

Discussion

Living conditions have an effect on the growth status and physical fitness of
children. Results from the present study are consistent with the general perception that
“better” living conditions encourage attainment of growth and performance. Children
living in the urban community were taller, heavier, and more muscular compared to
those from the rural community. Nevertheless, the rural and urban children are, on

average, smaller and lighter than children in other urban centcrs of 'x-iexico (Pena

83

Reyes et al., 2002). Their heights and weights fall, on average, in the lower
percentiles of the new United States growth charts (Centers for Disease Control and
Prevention, 2000). Mean heights of urban boys and girls fall at or above the 10‘'1
percentiles but approach the 25th percentiles at several ages, whereas mean heights of
rural boys and girls fluctuate between the 5th and 10th percentiles. Mean weights of
urban children of both sexes tend to be above the 25‘h percentiles of the reference
values. In contrast, mean weights of rural boys and girls are at or below the 25th
percentiles. The trends for mean heights and weights suggest elevated weight-for-
height in both the urban and rural samples, although the prevalence of overweight is
low (Pena Reyes, 2002).

Previous studies on the grth status of children 6-14 years of age from rural
and urban communities in Oaxaca in the 19705 also reported that rural children were
shorter and lighter, and had reduced muscle mass compared to children resident in
urban minis (Malina et al., 1981). Similar to earlier studies in Africa (Spurgeon et
al., 1984; Cameron et al., 1992) and Japan (Matsui and Tamura, 1975), subcutaneous
fatness was greater in urban males compare to rural males.

In addition to being larger in body size and estimated musculature, urban
children also performed better in fitness tasks, although the differences were not
consistently significant. Taller and heavier children tend to perform better on fitness
tasks than their shorter and lighter peers (Malina, 1975, 1994; Malina et al., 1987).
Body weight is the major contributor to static strength, so a child who is bigger tends
to also be stronger. However, although urban children were absolutely stronger,
reflecting their larger body size, strength per unit body mass tended to be greater in
rural children, which was consistent with earlier studies in Oaxaca (Malina and

Buschang, 1985; Malina et al., 1987).

84

A question of interest is the fitness of the Oaxaca children compared to well-
nourished children. Comparable data are not available for Mexico. Data for a mixed-
longitudinal sample of Philadelphia children 6-13 years of age tested in the mid-19605
include the standing longjump, the 35-yard dash, and grip strength (Malina, 1968).
The jumping and running protocols were the same as in the present study, but a
Narragansett dynamometer was used to measure grip strength in the Philadelphia
study and an adjustable Stoelting dynomometer was used in the present study. Mean
performances of rural and urban Oaxaca children on the standing long jump and 35-
yard dash were consistently lower than corresponding age- and sex-specific values for
the mixed-longitudinal sample of Philadelphia children. Allowing for differences in
type of dynamometer, the grip strength of Oaxaca children was also consistently
lower than that of Philadelphia children.

The American Alliance for Health, Physical Education, Recreation and Dance
(1980) provide reference values for the distance run (yards) in 9 minutes. Expressing
the medians and means as yards per minute and meters per minute, respectively, thus
provides an indirect comparison. Allowing for the differences in measurement units,
the distance run performance of rural and urban Oaxaca children, estimated as meters
per minute, is reasonably similar to the reference. The aerobic performance of the
Oaxaca children is perhaps more impressive in the context of their smaller body size,
specifically leg length, which is related to stride length. This would suggest better
aerobic fitness in the Oaxaca children.

The F itnessgram (Cooper Institute for Aerobic Research, 1994) scores the sit-
and-reach on a pass-fail basis with passing scores of 20 cm in boys and 23 cm in girls
6-11 years of age. Mean values for the sit-and-reach of rural and urban Oaxaca

children exceed the passing levels of the Fitnessgram and approximate the medians of

85

the American Alliance for Health, Physical Education, Recreation and Dance (1980)
reference. The Fitnessgram also includes modified sit-ups (curl-ups), but it is not a
timed test. Children are instructed to perform the curl-ups at a slow and controlled
pace of about 20 per minute. Hence, the data are not comparable to values in the
present study.

Although the urban and rural children from Oaxaca are smaller and lighter,
their health-related physical fitness as reflected in the distance run and sit and reach
appears to compare favorably to reference values for taller and heavier American
children. In contrast, their performance-related fitness as reflected in grip strength, the
standing long jump and dash does not compare as well. These tests are related to body
size, and the smaller body size of the Oaxaca children is a disadvantage.

Results of the partial correlations in the present study were similar to those of
previous studies of well-nourished (Seils, 1951; Rarick and Oyster, 1964) and low
SES and/or undernourished samples (Malina and Buschang, 1985; Rocha Ferreira et
al., 1991; Bénéfice and Malina, 1996; see Table 4.13). Age is positively related with
performances in both sexes suggesting the notion that neuromuscular maturation
and/or experience is an important factor in the performances of children 6-11 years of
age (Malina and Bouchard, 1991; Malina, 1994). Height is also largely positively
related with performances after controlling for age and weight, while weight is
negatively related with performances except for grip strength after age and height
were controlled.

Age, height, and weight explained most of the variance in the performances of
rural and urban children. Consistent with previous studies, the results of the multiple
regression analyses showed body size explained most of the variation in static

strength, which was consistent with those of previous studies (Malina and Buschang,

86

1985; Rocha F erreira et al., 1991; Be’néfice and Malina, 1996; Katzmarzyk et al.,
1997). After static strength, body size also accounted for a major part of the variation
in the distance run and sit-ups. Similar to previous studies (Malina and Buschang,
1985), the variation in performance was affected and influenced by absolute body
size. However, in the present study, age explained most of the variation in the dash
and standing long jump.

Sex differences were also examined in the present study and the variations of
several performances were explained by age and body size. More of the variation in
strength and speed was explained by age and body size in girls than in boys, whereas
more of the variance in power was explained by age and body size in boys than in
girls.

With the addition of other anthropometric dimensions to age, height, and
weight in the multiple regression analyses, the variance explained in physical fitness
increased only marginally. The increase in the variance explained was consistently
significant for the sum of right and left grip strength across the four samples while the
significance was variable for the other five fitness tasks.

Few of the other anthropometric dimensions added significantly to the
variation in performance which was consistent with earlier studies of Mexican
(Malina and Buschang, 1985), Brazilian (Rocha Ferreira et al., 1991), and African
(Naidoo, 1999) children. These earlier studies, however, used a different analytical
approach so that the results may not be directly comparable. Nevertheless, the
addition of other anthropometric dimensions to the regression analyses in the present
study altered the contributions of age, height, and weight to performances on the
physical fitness tasks. This observation suggests overlap or interactions among

variables in their estimated contributions to performance.

87

Among the anthropometric variables considered, several contributed
significantly to the proportion of the variance in performance that was accounted for.
In general, fatness (sum of four skinfolds) had a consistently negative influence on
physical fitness. Of interest, the sum of skinfolds contributed negatively (significantly
so in three of the four samples, Tables 4.9-4.12) to static strength after age, height,
and weight were statistically controlled. Estimated arm muscle circumference
contributed most to the variance in strength, sit-ups, standing long jump, and dash in
urban boys, while leg length was the main contributor in the sit and reach. In contrast,
there was no consistent pattern in the contributions of other anthropometric variables
to the performances of rural boys. Biepicondylar breadth contributed significantly to
grip strength, sit-ups, and the standing long jump among rural girls, while there was
no consistent pattern in the estimated contributions of other anthropometric variables
in urban girls.

In summary, performances in physical fitness tests were influenced by body
size. Although variation in performance appeared to be largely influenced by age,
height, and weight, there was overlap in the contribution of other anthropometric
characteristics to performance. The selective inclusion or exclusion of anthropometric
dimensions in such analyses can over- or underestimate the contributions of age,

height, and weight to variation in performance.

88

Bibliography

American Alliance for Health, Physical Education, Recreation and Dance. 1980.
Health Related Physical Fitness Test Manual. Reston, VA: American Alliance
for Health, Physical Education, Recreation and Dance.

Be'néfice E, Malina RM. 1996. Body size, body composition and motor performances
of mild-to-moderately undernourished Senegalese children. Ann Hum Biol
23:307-321.

Beunen G, Malina RM, Lefevre J, Claessens AL, Renson R, Vanden Eynde B,
Vanreusel B, Simons J. 1997. Skeletal maturation, somatic growth and
physical fitness in girls 6-16 years of age. Int J Sports Med 18:413-419.

Cameron N, Kgamphe J S, Leschner KF, Farrant PJ. 1992. Urban-rural differences in
the growth of South African black children. Ann Hum Biol 19:23-33.

Centers for Disease Control and Prevention (2000) National Center for Health
Statistics CDC growth charts: United States.

http://www.ch.gov/growthcharts.htm.

 

Cooper Institute for Aerobics Research. 1994. The Prudential Fitnessgram: Test
Administration Manual. Dallas, TX: Cooper Institute for Aerobics Research.

Eveleth PB, Tanner JM. 1990. Worldwide Variation in Human Growth. NY:
Cambridge University Press.

Hamill PVV, Johnston FE, Lemeshow S. 1972. Height and weight of children:
socioeconomic status. DHEW Publ No. 73-1606, Vital and Health Statistics,
Series 1 1, No. 119. Washington: US Government Printing Office.

Katzmarzyk PT, Malina RM, Beunen GP. 1997. The contribution of biological
maturation to the strength and motor fitness of children. Ann Hum Biol

24:493-505.

89

Malina RM. 1968. Growth, maturation and performance of Philadelphia Negro and
White elementary school children. Doctoral Dissertation, University of
Pennsylvania.

Malina RM. 1975. Anthropometric correlates of strength and motor performance.
Exerc Sport Sci Rev 3:249-274.

Malina RM, 1983. Growth and maturity profile of primary school children in the
Valley of Oaxaca, Mexico. Garcia de Orta Serie de Antropobiologia: Revista
Do Instituto de Investigacao Cientifica Tropical (Lisboa) 2 (1 & 2): 153-157.

Malina RM. 1994. Anthropometry, strength and motor fitness. In: Ulijaszek SJ,
Mascie-Taylor CGN, editors. Anthropometry: the individual and the
population. NY: Cambridge University Press. p 160-177.

Malina RM, Bouchard C. 1991. Growth, maturation, and physical activity.
Champaign, IL: Human Kinetics.

Malina RM, Buschang PH. 1985. Growth, strength and motor performance of Zapotec
children, Oaxaca, Mexico. Hum Biol 57:163-181.

Malina RM, Himes JH, Stepick CD, Lopez FG, Buschang PH. 1981. Growth of rural
and urban children in the Valley of Oaxaca, Mexico. Am J Phys Anthropol
55:269-280.

Malina RM, Little BB. 1985. Body composition, strength, and motor performance in
undernourished boys. In Binkhorst RA, Kemper HCG, Saris WHM, editors.
Children and Exercise XI. Champaign, IL: Human Kinetics. p 293-300.

Malina RM, Little BB, Shoup RF, Buschang PH. 1987. Adaptive significance of
small body size: strength and motor performance of school children in Mexico

and Papua New Guinea. Am J Phys Anthrop 73:489-499.

90

Malina RM, Selby HA, Buschang PH, Aronson WL. 1980. Growth status of
schoolchildren in a rural Zapotec community in the Valley of Oaxaca, Mexico,
in 1968 and 1978. Ann Hum Biol 7:367-374.

Matsui H, Tamura Y. 1975. Characteristics in physical fitness of urban and rural
people. In: Asahina K, Shigiya R, editors. Human Adaptability Volume 3:
Physiological Adaptability and Nutritional Status of the Japanese. Tokyo:
University Tokyo Press. p 41-68.

Meredith HV. 1979. Comparative findings on body size of children and youths living
at urban centers and in rural areas. Growth 43:95-104.

Montoye HJ, Frantz ME, Kozar AJ. 1972. The value of age, height and weight in
establishing standards of fitness for children. J Sports Med Phys Fitness
12:174-179.

Murphy AD, Stepick A. 1991. Social Inequality in Oaxaca: A History of Resistance
and Change. Philadelphia, PA: Temple University Press.

Naidoo RB. 1999. The effects of socioeconomic status, ethnicity, and nutritional
status on the growth and physical fitness of 10 year old south African boys.
Doctoral Dissertation, Michigan State University.

Norgan NO. 1994. Anthropometry and physical performance. In: Ulijaszek SJ,
Mascie-Taylor CGN, editors. Anthropometry: the individual and the
population. NY: Cambridge University Press. p 141-159.

Pena Reyes ME. (2002). Growth status and physical fitness of primary school
children in an urban and a rural community in Oaxaca, Southern Mexico.

Doctoral Dissertation, Michigan State University.

91

Pena Reyes ME, Cardenas Barahona EE, Cahuich MB, Barragan A, Malina RM.
2002. Growth status of children 6-12 years of age from two different
geographic regions of Mexico. Ann Hum Biol 29:11-25.

Rarick GL, Oyster N. 1964. Physical maturity, muscular strength, and motor
performance of young school-age boys. Res Quart 352523-531.

Rocha Ferreira MB, Malina RM, Rocha LL. 1991. Anthropometric, fiinctional and
psychological characteristics of eight-year-old Brazilian children from low
socioeconomic status. In: Shephard RJ, Parizkové J, editors. Human Growth,
Physical Fitness and Nutrition. Basel: S. Karger. p 109-118.

Seils LG. 1951. The relationship between measures of physical growth and gross
motor performance of primary-grade school children. Res Quart 22:244-260.

Spurgeon JH, Meredith HV, Onuoha GBI, Giese WK. 1984. Somatic findings at age 9
years on three ethnic groups of Nigerian urban and rural boys. Growth 48:176-
186.

Spurr GB. 1984. Physical activity, nutritional status, and physical work capacity in
relation to agricultural productivity. In: Pollitt E, Amante P, editors. Energy

intake and activity. NY: Alan R. Liss. p 207-261.

92

92333 @3933 8a momofieoaa E come—0:0 £38: Z 56 v a .1. .36 v a ...

 

 

:3: 3 a 3 2; as am a 8%
1.2.8 3 SN 8 SN so .028
:32 3 <2 3 ma :5 62%
18.8 3 EN 2 is so .26 :8
its 3 S: S 3: a. .26 E...
a: ed 3 od 3 :5 55:83.5
:32 9o 3 3 E .5 5:285
:23 3 o. _ N 3 Sm :5 .3255
3.3% 3 SN 3 SN :5 .aéega
£52 2 new no New 8. .3 was
:3. a mo 26 mo ”.8 so .E em
:8.” do a: do 08 Ne} .95
:46: to $2 to 3.2 .8 .23.:
1.2.2 to EN so EN me. .3903
A: U 3 A? U 3 as» a?

a mm :82 mm :82 mozۤ

:2 ”5935

9.2 u 5 35m

 

665.56 05 we own 53>

m<>OUZ< 2: 8m 85E m can own no 38% :6 as 53.5 was 35.. 5 833.5»
eczema new 05088855 .3 Echo Begum new 8on Bufi36w< ._.v 2%...

93

228326 wavefim 03 835283 E 68226 $3632 56 v a 1.. .36 v a ..

 

94

 

:..o 3 Em 3 2m 22 .mm v 25m
:88 S 0.2 3 e. a :5 636.2
1.3% 3 NE 3 0.2 :8 62$
1.8.2 do on do 32 so .2626

.2... S 3: 3 2: ea .26 22
.22 ed 2. S. w... :5 558286
1.3.2 3 E 3 2 :5 2.68.6
33.24 S 4. a 3 3N :5 .3256
.anmm 3 ma 3 2a .5 422226
1.38 no 3% 2. Sm :8 .5 men
1.3.: S 22 3 MS .5 .2: 2m
.9... mo .5 .8 6.2 NEE .95
1.3.8 3 32 to 3.2 so .22.:
1%.: .3 fix to Now me. .2283
A: U ad 9. d 3 22% am...
2 mm :82 mm 822 2322.9
62 u 5 .22: a: u 5 =23.

 

33.830 05 we own .23
m<>OUZ< 05 28 mouse «— can .0?“ mo menu.» :6 me 535 was 3:: E 333:;
33.6.6 98 3.3683855 .8 macho @3986 new 282: noumzmneéw/x .Né 63m...

_qova..
chva .

 

.668 2 ode 3 2.5: 522 .52 8.226
6.5 35 as 3.5 o: 8m .58
1.6.8 Z «N: 3 $5 :5 .15
1.2.. :. 2 5.2 no 3 a .2525
.8... 3 EN to 3N :5 2%
o; :3 :3 :5 a; 229. 625.56 ._
.22 5.5 8.5 So 5.. was. .552 2.5
2; 4.5 new to 2.2 we. :2 22m

,2 mm 282 mm as: 2322>

:5 "5.22:

9.2 u 5 123.

3.2.8.60 2: we own 53,

m<>OUZ< 05 Ba 8:8 m Ba .03 .«o menu» :6 when 53.5 use .83 E 33.25;
336w can «18:5 3393a mo £86 @3986 BB 338 voamsmnmémxx .mé 2an

95

~o.o V Q .I.

 

 

33 v a ..

a: 2 :3: E 32 ESE as: 8:92:
23 :3 N: :3 N: 8m 38:
:38 E :3 2 3.3 :5 .15
1.2.2 3 3:: m3 3 a .255
$3. 3 :3 2 3.8 .5 32m
:3 :3 m3 :3 m3 :5} US$30 a
:3: : :3 23 :3 33 33 .55: Sam
23 :3 3m 3 3.3 3 .3 saw

.3 mm 982 mm :82 m..._e§>

:2 "5535

9.2 u 5 as:

.Bmtgoo 05 mm owm 53>

m<>OUZ< 05 How moss.— m was .0?“ m0 m3?» :6 me 535 98 3:: E why—nab;
32:6 98 $025 136.33. mo 28.8 9353 can 282: woamsmuméwxx .v.v £an

96

8:35.833 .533 £333 2:: 832 a 3 3:35 825 mcwfi a
:3 v 3 ...... .33 v 3 _.

:33 - 5:3 :83 5.55 .55 85.35
:3 - 333 2.333 8... .23:
N33 :3 1.333 .5 .33
:3 - :3 :23 a .8255
333 - 333 - :3 so .323

:83 1.333 23 mg .55 33 53m

:2 n 5 :35
.333 - :3 .23 55>: .55 8:55:
:3 - 33 , :23 03 £3:
83 - 33 :33 :5 .3m

:33 - :33 3:3 3 £5-36
333 - :3 - 33 :5 32m

:33 :23 :83 33. .55 .2 saw

a: u 5 .833

 

Emmom cam ow< Emma? 98 ow< Emma? 98 Ewmo: no.5 wcfiobcoo
53303 536: 0?

 

.0? .8 £3» :6 3.89 555 2.3 :22 E man:
305% 333.3.“ was .86 >95 .owa wcofim macaw—9:8 3th 838-28% .mé Bank

97

858.2822. .883 23:30 0:5 39.52 3 33 3:35 225 2&5 a
:3 v 3 .. .83 v 3 .

23 - 23 :83 5.55 .53 835.5
23 - .233 :23 8... .295
:3 - 23 :83 .8 .75
~33 - 23 23 a .2533
S3 :3 333 - so .333
.....mm.o imvd 12d 3 .320 AM Sam
:32 n 5 5.35

.. 3.3 - 33.3 2.3 55>: £3— 85.55
.23 - :83 :33 8. .295
:83 - .333 .23 .8 .33
.23 - .23 333 c .2523
.23 - 233 .23 so .33
:23 .23 :23 3.. .55 33 Sam

2: "av—35M

 

Emfim Ea ow< Ewfi? EB 09.. Emma? EB 2302 no.“ wcfobzou
£903 536: um...

.03 no 23% 2% me 535 98 .32: E 25:
335.2 33233. .35. .onm Egon .owa wcosa 28:22.28 Etna 82:06:83 .04. 033.

98

0023:2823 .523 2250 0:5 .532 3 mm 322:: 803 3me a
83 v a 1.333 v a .

 

 

.33 - 2.3 $3 :23: £3 _2 5.5:. .33. 8535

83 - 23 23 :32 m3 _2 8. .225

83 333 23 :82 32.3 _2 .5 .3m

23- :3 $3 .13.: 23 _2 582-5

23 - 23 - 23 $3 83 _2 :5 32m

£3 £3 23 :53: E3 .2 3 3:0 .2 cam
535

:3 - S3 23 :35. 333 N2 .258 332 8:35

83 - 23 3.3 :32: 23 32 8. .295

33 - 33 23 :32 23 32 :5 .3m

:3 - 53 R3 :22 £3 «2 a .253

333 S3 - 83 23 S3 2 .8 32m

23 $3 :3 .2332 23 2.2 m: 35 .2 83m
332

£333 236: ow< ,2 N3 =
3566000 Sum

 

.owm .«o 282 :6 when
933 3:3 :22 E 25: 3852 32935 so £303 28 .590: .09“ mo :23»:on 25232 .5. 03.5.

99

8.38.2.2. .323 23:33 3.2. .332 3 33 38.332 3.33 8.me 3
.33 v 3 .. .83 v 3 .

100

 

83 - 83 83 :8... 83 8. 5.5... .83 8.5.5

333 - 83 83 :88 83 8. 8. .295

83 - 83 83 :88 83 8. .5 .33

83 - 23 83 :83 E 3 8. .. 3.5-8

83 - 83 :3 - 3.... 83 8. .5 .333

83 3.3 .23 :23... 83 8. 33 3.5 ..3 :53
:85

83 - 23 83 .88 83 32 5.5... .....3 8.5.5

83 - 83 3.3 :88 3.3 .2 8... .235

83 - 83 83 : .58 83 8. .5 .33

83 - 3.3 23 :83 333 8. .. 332-.....

83 - 333 83 .28 83 8. .5 3%

83 83 83 :33.me 83 8. 3. 3.5 33 .53
3.5%

33.33 23.8. 33.. 3 3 ..

 

83205300 3.3m

 

.33 .«o 3.33% :6 22m
33...: 2.3 .32... 2 3832 3335... 323.35 .8 2233 2.3 .232. .ow3 mo 32.83.33. 22:22 .3... 333.3

33:38.23: .332. 2328 3:2. .332 3 33 32.22 3.33 3:me 3

 

 

.q3V3..
83 v 3 .
83 .33 83 3.3 83 .33 3 3 33.3
83 83 83 83 83 83 N3
:83 - 83 - 83 - 333 - :3 - 83 - .3... .33 .V .53
83 - 83 .33 - 83 - 83 83 - .5 .253
3.3 - 3.3 333 .83 :3 - 83 .5 .025
83 23 83 83 :3 333 .5 5.853.323
83 83 - :3 - 83 - :3 :3 .5 5.853...
83 83 83 83 - 3 . 3 - 83 - .5 5.5.3
83 - 83 :3 - 83 .83 :3 .5 5.5.5.:
83 83 83 - 83 83 - :3 .5 .5 35
83 83 - 83 83 - 83 - .83 3. 53.33
.33 - 333 :3 83 :3 - 8.3 .5 3.3.3:
83 :53 :83 83 83 .23 33...
53 85.35 25: 33 332-8 3.33 3.5 :3 .53 5.55., 3.5538...

 

.ow3 mo 3.33.» :6 3.3.. .35. 2 3:82 835... 8292: :o 32.35.33.323
2.8.3.8955 .23.: 3:3 £333 Jams... .33 .30 3:28.23. 22:2: :8... 32.20538 Sam 6.: 2:32.

10]

33:39:33.3: .3339 3.3333 3:... .326— 3 33 38.3.5. 3.33 m:w_m 3

 

 

5.0 v a .1.

83 v 3 .
33.3 33.3 .33 33.3 33.3 33.3 .3 . .82....3.m

83 83 8.3 38.3 3.3 .33 .3

38.3 - 2.3 38.3 - .33 - 83 - :83 - .3... .33 v .83
8.3 3.3 .33 - 8.3 - .83 - 83 - .8 .3233
3.3 :83 :83 :83 333 13.3 .8 .3323
:3 - .83 - :3 - .83 - 8.3 - 8.3 .8 8.838.323
83.3 - 8.3 8.3 :83 .83 8.3 .8 8.838...
8.3 - 8.3 8.3 - 2.3 3.3 8.3 .8 33.8.3
83 .83 .83 8.3 .83 8.3 .8 33.88.33
33.3 - .83 3.3 8.3 :83 - .33 .8 .3 8..
33.3 - 8.3 - 8.3 - :3... - 8.3 - 83 3. 3.3.33
83 8.3 - 83 - 8.3 8.3 8.3 .8 3.3.3:
.833 .833 :83 .83 :3 :3 .3...

33. 88...: .38: 33 83-33 3<m 3.5 d. .83 8.3.8., 8388.3...

 

.33 .3 3.33.» 2-0 3.8.. :3... 5 8:3: 835... 323.3: :3 33.38.383.35
2.3388859. 8:32.333 3:3 .2833 3.3.3.. .3w3 33 3:238:33. 2:23.: So... 35.35333 3.3m .o _ .v 333...

102

oocnscotoa 5on £960 2:: 332 a 3 v2.52: 20>» mama a

 

 

:2 v a :

85 v a ..
Ed 23 m3 3o 2:. 3o 4 m Hfiouemmm

NS 98 Rd 25 :5 23 Na

o3 - 85 £5 - Rd - m2 - :85 - es mi v 8.5
ate 3o 2:. ad - Ed - 3o - :5 .028
Ed - mad - :3 m2 - 2d - 5o .8 dz<m
NS 3o *5. 3.2. as :35 so $388305
23 - 8o 8... NE 8d 3d .8 53285
So - :.o - :5 2d - m3 - So so .3355
:3 - :3 2.9 - 3o - 85 Ed .5 42828.5
35 NS Rd 2:. .1. a; - mod .5 .5 MB
m2 - ad - a; - :.o - one 3... 9. 3%?
as - :.o MS 25 as; :3 :3 imam
”do and as Ed 25 1mg ow<

Se 8535 Ewan 3m 25% 5a 35 d new 833.5 Eugene?

 

.0? .«o E8» :6 me :5: E was: mmofic Romans :0 8558320
oEoEOQoEHSW 38:63 23 “Emma? .236: a? mo 8238on 032:8 an £56508 «:5 ._ 3» 28¢.

103

00580802— 0033 £03.00 08: 0032 a 8 00:02: 0.03 3me d

 

 

56 V a .1.

mod v d ..
2d 8d 3d 3d Rd 8d 4 u .sodewa

Rd Rdd Sd 2d ddd Rdd Na

:d ddd - :de - Rdd - mdd - :Rd - ea ma v 55
odd - t. :d ddd - Rd 8d - 8d .5 .0200
2d - 2d 2d ed Rd .3 so .025
8d - 8d ddd - mdd - 8d - 2d .5 5328320
2d - 8d - Rd «dd 2d 3d :5 5:885
2d - .de - 2d - ddd Rd - add .5 .3255
2d Rdd - odd - 2d ddd odd E .35235
Rd :d dd Rd Rmd - 2d - so .5 03
3d Rd - *de dmd - Rd - Rd 00. .253
odd dmd «dd 2d - Ed *Sd :3 .206:
1. 3d :Rmd :Rd 0 _ d :d - Ed 00¢.

5m 8:35 “can 3m 335 5% $6 .2 saw m2d§> 608083

 

.0? mo £00» :6 me 53.5 5 £20: 0005: .86.?3 :0 mozmtflofidnu
030880.550 3:25:00 93 .Ewm03 .290: .0? no acommmawe 033:8 82% 856508 80m .26 £an

104

.53 002805 n :3. ”38.5. 00555 n 39:5. mfiwcgm £00 0.8% u 3.5 ”95:. wcoq wEESm M 3m
005.50%3 00:3 $806.: 08: .032 a 8 00:02: 0.03 S: bzmwm v.8 fig 05 mm :03. was:

.008: 8.. m:w_m 05. .552? 05 >35 E03... 06 E 000.: 5:0: 8 5:86 28 86:6 0:. =0 2 58:80 300“ 5.. 28:20.08 3:0
.859? 0.0 00:05:00 £005.50 03 830:? 95 .050 05 :05? 0.80: 30.5... 05 53» :30? 8 Jaws: .0? 503.09 30:50:00...

 

 

mad 8 86 ”3203 8.2258800 3.5.. 0.5
385 .25 Rd 2 add - 2.0.0: 8... n 5 2.5 2 3...:2. 022.3. .52. 22.
.3m .55 Rd 2 ddd 0?. Add. u 5 :8 .28» :2 22.0.20 80.58% add: “2.5,. .23 8.02.00
2;. .25 2d - 2 Rd - 2.0.03 Add u 5 2.6 ma 32 .28»
.75 .53 ed 2 8d 220...... .3. u 5 whom d 22.2.5 .822 5:220 :8: d. .0 2.2.0“. 203.
265 .25 and 2 2d - ӣ0.03 add 2 Rm u 5 ago do 2:5, 2.. 22. .282
3m .58 Rd 2 8d 2.0.5. 2.6 .22 MW3m. 2-0 222.5 32.8 28.5.). .23: 022.050 22 .22.5,.
2d - 2 Rd - 2.0.03 2.5.228 522.3 2.6.2
385. omd 3 2d ”530: 8 00am: 80¢ .030» M75 .98..
.75 £me mmd 9 :6 .09.. Se n 5 0.3m BB8 008w 9.80m snows—5x $00 3 .020 can 03th
:.o 2 mmd - ”£303 m0E§EEoo .58
38.: Rd 2 odd - 2.0.2. 2R u 5 2.5 2 28:8 2.2... 22. .28. 3
.3m .55 9d 2 Ed New... RR n 5 5Om. 522.5 2020 $2.... 525:2 :3: 2.0m
:3. a... Rm. u 5 535 830 do 2:;
2.2-5 9... 2 Rd - ”20.03 c: u 51.5.5.6 2.. 2 b.5228 :85 5.
.23 .25 Rd 2 mod - 2.0.2. .8. n 5 535 .22 .22 a 220 22% :d
.75 Emma 93 8 ood - .03. AN? u 5 FEM “whom £06.30 .038 50.0.02 35% 2.085
38,—. mecca—0.80 mo 098m... xom 0.95m Exam

 

.0mm mo £00» 2-0 20.230

E 8.53 305$ €06.39 van .3303 .530: .03 502503 macaw—2.80 Etna 820-3503 “.00: 35 3630 ac b.2555 m _ .v 030...

105

CHAPTER 5

Somatotypes of Rural and Urban Children in Southern Mexico

Introduction

Physique refers to the overall shape, form, or configuration of the entire body
(Carter and Heath, 1990). It is most often quantified as a somatotype, which is a
composite of the contributions of three components: endomorphy (fatness, roundness),
mesomorphy (musculoskeletal development), and ectomorphy (linearity).

There are several methods for estimating somatotype (Sheldon et al., 1940, 1954;
Parnell, 1958; Carter and Heath, 1990). The most commonly used method at present is
the Heath-Carter anthropometric protocol (Carter and Heath, 1990; Malina, 1995), which
uses several anthropometric dimensions to estimate the three components of somatotype.
Scores for each component ranged from 1 (low) to 7 (high) in the original method of
Sheldon et al (1940, 1954) and also in Parnell’s (1958) anthropometric approach. In the
Heath-Carter anthropometric method, the lowest possible score is 0.1 and the high end of
the scale is open-ended (Carter and Heath, 1990).

During childhood, somatotype of boys and girls tend to be generally balanced, 3-
4-3 or 3-3-3, but boys tend to be, on average, slightly more mesomorphic and girls tend
to be more endomorphic. The sex difference in somatotype is not marked in means, but is
especially apparent in the distributions. Although most boys and girls have somatotypes
that are balanced, more boys have somatotypes in the mesomorphic pole whereas more

girls have somatotypes in the endomorphic pole of the distributions. Considerable

106

overlap characterizes the distributions of somatotypes among and between boys and girls
(Carter and Heath, 1990; Malina and Bouchard, 1991).

Studies examining somatotype have largely described and/or compared the
physiques of adults in different populations, and especially athletes in a variety of sports
(Carter and Heath, 1990). The Heath-Carter anthropometric protocol has also been
applied to samples of healthy children and adolescents primarily of European ancestry
(Carter and Heath, 1990; Malina and Bouchard, 1991), but has been used to a lesser
extent with Latin American children. The protocol was used with school children of both
sexes 7-18 years of age in Guatemala (Alexander et al., 1993) and Venezuela (Alexander,
1992), and with boys 6-15 years of age (Godoy et al., 1994) and girls 4-11 years of age
(Godoy and Barcos, 1995) in Chile. The studies in Guatemala and Venezuela were done
in the context of national physical fitness surveys of school children and the results were
presented as descriptive statistics by age and sex. The studies in Chile were limited to
boys of medium to high, and to girls of low socioeconomic status (SES) attending school
in Santiago. The study of boys focused on age changes in somatotype components in
small samples followed longitudinally over one year, while that of girls compared low ‘
SES girls relative to nutritional status.

Murguia et a1. (1990) applied the Heath-Carter anthropometric protocol to a large
sample of children and adolescents of both sexes 5-20 years of age from several different
areas of Yucatan, Mexico. The samples were from a traditional Indian community
(subsistence maize agriculture) and from three municipalities that were more “modern”,
specializing in fishing, sisal production, and cattle raising, respectively. Although there

were body size differences among the samples, there was considerable overlap in mean

107

somatotypes. Endomorphy tended to be lower in boys and girls 6-11 years of age from
the subsistence agricultural community. but mesomorphy and ectomorphy, were similar
in the children and adolescents from the four communities (Murguia et al., 1990).
Unfortunately, the study was largely descriptive with little statistical analysis.

Diaz Gamboa and Fuentes Heinrich (1996) described the somatotypes of Aymara
children, 6-15 years of age, resident at high altitude (3 000-4500 meters) in northern
Chile. Mean somatotypes indicated balanced endomorphy and mesomorphy and low
ectomorphy in girls, and slightly greater mesomorphy than endomorphy and low
ectomorphy in boys.

The present study compares Heath-Carter anthropometric somatotypes of rural
and urban children 6 to 11 years of age living under marginal health and nutritional

conditions in Oaxaca, southern Mexico.

Methods and Procedures
The communities, sample, and anthropometric procedures were described earlier
(Chapter 3). The dimensions required by the Heath-Carter anthropometric somatotype
protocol included weight (kg), height (cm), bicondylar and biepicondylar breadths (cm),
flexed arm and calf circumferences (cm), and the triceps, subscapular, suprailiac and
medial calf skinfolds (mm). The algorithms for the Heath-Carter anthropometric protocol
were used to derive the three components of somatotype (Carter and Heath, 1990, p 374):
1. Endomorphy = -0.7182 + 0.1451(X) — 0.00068 (X2) + 0.0000014(X3)
where X is the sum of the triceps, subscapular and suprailiac skinfolds, multiplied by

170.18/height in cm.

l08

 

2. Mesomorphy = [(0.858 x biepicondylar breadth) + (0.601 x bicondylar ‘

breadth) + (0.188 x corrected arm circumference) + (0.161 x corrected calf

circumference)] —- (stature x 0.131) + 4.50
where the corrected arm circumference is flexed arm circumference (cm) minus the
triceps skinfold (cm) and the corrected calf circumference is calf circumference (cm)
minus the medial calf skinfold (cm).

3. Ectomorphy = HWR x 0.732 - 28.58
where HWR is height (cm) divided by the cube root of weight (kg). If HWR is less than
40.75 but more than 38.25, then Ectomorphy = HWR x 0.463 - 17.63; and if HWR is
equal or less than 38.25, a rating of 0.1 is given.

Measurement variability for the anthropometry was also described in Chapter 3.
The reproducibility of anthropometric somatotypes was estimated in 79 children.
Intraclass reliability coefficients were between 0.97 and 0.98, technical errors of
measurement varied between 0.16 and 0.23 somatotype units, and coefficients of
variation ranged from 4% to 10% (Table 5.1). The differences between means were less
than 0.07 units for any somatotype component. Thus, the Heath-Carter anthropometric
somatotype was highly reproducible in the sample of rural and urban children. The
intraclass coefficients, technical errors of measurements, and coefficients of variation
were smaller than reported in the Quebec Family Study (Bouchard, 1985).

The analysis was done in several steps using the multivariate statistical procedure
recommended by Cressie et al. (1986). Age-, sex- and community-specific descriptive
statistics were initially calculated. Sex-specific multivariate analyses of variance

(MANCOVA), with age as the covariate, were calculated to compare differences

109

between rural and urban children. Community-specific multivariate analyses of variance
(MANCOVA), with age as the covariate, were calculated to compare sex differences in
the rural and urban samples. If a MANCOVA was significant, univariate F-tests were
performed to determine which components contributed to the significant difference(s).
Sex-specific forWard stepwise discriminant analyses were also conducted to determine
which somatotype component(s) best discriminated between rural and urban children

within each sex, and between the sexes within each community.

Results

Age- and sex-specific means and standard deviations for somatotypes of rural and
urban children are presented in Tables 5.2 to 5.5. An ANCOVA indicated no significant
age effects on somatotype among the 6-11 year old children except in urban boys (Table
5.3). It appears that the significant age effect was due to elevated endomorphy and
mesomorphy, and reduced ectomorphy in 10 year old boys.

Age-adjusted means and standard errors of somatotypes by sex and community
are shown in Table 5.6. Rural and urban boys differed significantly in somatotype (p=
0.021); the univariate F-tests indicated that urban boys were significantly more
endomorphic than rural boys (p=0.004). In contrast, the MANCOVA for girls was not
significant between the communities (p=0.226).

Age-adjusted means and standard errors of somatotypes by sex within each
community are shown in Table 5.7. Somatotype differed significantly between boys and
girls within each community (p<0.001). girls were more endomorphic (p<0.001) and

boys were more mesomorphic (p<0.001), while ectomorphy did not differ.

110

Results of the forward stepwise discriminant function analyses rural-urban
comparisons of somatotype within sex and comparisons of somatotype by sex within
each community are presented in Table 5.8. Endomorphy was the best discriminator
between rural and urban boys; none of the somatotype components discriminated
between rural and urban girls. Within the rural community, endomorphy was the best
discriminator between boys and girls followed by mesomorphy. All three somatotype
components discriminated between boys and girls in the urban community. Mesomorphy
was the best discriminator between urban boys and girls, followed by endomorphy and

then ectomorphy.

Discussion

Although there was no significant age-effect in the somatotypes of rural children,
mean somatotypes showed a tendency to increase or decrease with age in rural children.
In contrast, mean somatotypes fluctuated from age to age in urban children, particularly
boys among whom there was a significant age effect (Table 5.2). This appeared to be due
to increased endomorphy and mesomorphy, and decreased ectomorphy in the 10 year old
boys.

Rural-urban differences in somatotype were apparent among boys but not among
girls. Endomorphy discriminated rural and urban boys, while none of the components
discriminated between rural and urban girls. Urban boys were more endomorphic than
rural boys. The urban-rural differences in endomorphy reflect differences in

subcutaneous fatness (Table 4.1) and perhaps in habitual physical activity. Rural boys

11]

spend more time daily in moderate and at times vigorous household and agricultural
chores (Pefia Reyes, 2002).

The lack of an urban-rural difference in the somatotypes of may be related to the
lifestyles of girls in both communities. Urban and rural girls do not differ in
subcutaneous fatness (Table 4 .2), and girls in both communities spend a good deal of
their time in activities related to food preparation and do not differ markedly in physical
activity (Pena Reyes, 2002).

A previous study in Yucatan, Mexico, reported that children of both sexes from
better living conditions were more endomorphic than those living in traditional
subsistence agricultural communities (Murguia et al., 1990). However, the differences in
somatotype among the communities were not statistically compared in this study.

It has also been suggested in the nutritional and anthropological literature that the
females are more resistant to adverse environmental influences (McCance, 1966; Stinson,
1985). The lifestyle of the children in the rural and urban communities related to diet and
physical activity was mentioned earlier, and mral boys were more active than urban
boys. Urban boys may also have had chronically excessive energy intake, but data are not
available to document this. In contrast, urban and rural girls did not differ in fatness and
activity.

It is difficult to speak of environmental influences on the somatotype of children
due to lack of data on children and, perhaps due to limitations of the Heath-Carter
anthropometric protocol with children. However, studies of adults indicate that extreme
environmental circumstances, such as semi-starvation in males (Lasker, 1947),

disordered eating in females (Malina, 1992; Malina et al., 2002), or resistance training in

112

males (Tanner, 1952), are needed to significantly modify somatotype. Changes in
somatotype with starvation and disordered eating can be reversed with appreciate diet
(Lasker, 1947, Malina, 1992), whereas gains in mesomorphy with resistance training
return to pre-training values after the training program is stopped (Tanner, 1952).

Within each community, sex differences in somatotype were evident in
endomorphy and mesomorphy. Girls in both communities were more endomorphic while
boys were more mesomorphic. Endomorphy was the best distinguishing component of
somatotype between boys and girls in the rural community followed by mesomorphy. In
contrast, all three somatotype components discriminated between the sexes in the urban
community, beginning with mesomorphy, then endomorphy and finally ectomorphy.

The results of the present comparison of boys and girls are consistent with
somatotype data for European and North American children and adolescents (Malina and
Bouchard, 1991). Comparative data for samples of Latin American children 6-11 years of
age are summarized in Table 5.9. Data for Latin American children are also generally
consistent. There is considerable overlap in mean somatotypes among studies, including
the present samples of rural and urban children from Oaxaca.

It should be cautioned, perhaps, that the Heath-Carter anthropometric somatotype
method was developed on adults and the samples used to validate the method were
largely adult males (Carter and Heath, 1990). The applicability of the method to children,
especially in those living under marginal economic and nutritional conditions, needs

further study.

113

Bibliography

Alexander P (1992) Caracteristicas fisicas y morfologicas del Venezolano. In Memorias
IV Congreso Nacional de Educacion F isica, Deporte y Recreacion, Ponencias
Trabajos libres y Talleres. Caracas. Venezuela: Corporacion Venezolana de Guayana

Fundeporte, pp 170-223.

Alexander P, Mota D, Arevalo A (1993) Norrnas para la evaluacion de la aptitud fisica y
caracteristicas morfologicas del estudiante Guatemalteeo de 7.5 a 18.4 anos. Ciencias
de la Actividad F isica 7:1-54.

Bouchard C. 1985. Reproducibility of body composition and adipose-tissue
measurements in humans. In Roche AF, editor. Body-Composition Assessments in
Youth and Adults: Report of the Sixth Ross Conference on Medical Research.
Columbus, Ohio: Ross Laboratories. p 9-13.

Carter JEL, Heath BH. 1990. Somatotyping - Development and Application. Cambridge:
Cambridge University Press.

Cressie NAC, Withers RT, Craig NP. 1986. The statistical analysis of somatotype data.
Yrbk Phys Anthropol 29:197-208.

Diaz Gamboa J, Fuentes Heinrich E (1996) Composicion corporal y somatotipo de
infantes Aymara que habitan en elaltiplano del norte de Chile. Archivos de la
Sociedad Chilena de Medicina del Deporte 41 :87-95.

Godoy JD, Barcos M. 1995. Potencia aerobica maxima, somatotipo y composicion
corporal de nifias escolares de 4 a 11 afios de edad, de nivel socioeconomico bajo.

Archivos de la Sociedad Chilena de Medicina del Deporte 40221-28.

114

Godoy JD, Ugarte J, Donoso V, Zupeuc F, Herrera A, Valle A. 1994. Somatotipos de
escolares varones Chilenos: cambios con el crecimiento (Seguimento de cohortes y

estudios transversales durante 5 ar‘ios). Archivos de la Sociedad Chilena de Medicina

del Deporte 39:119-134.

Lasker GW. 1947. The effects of partial starvation on somatotype. Am J Phys Anthropol
5:323-341.

Malina RM. 1992. Physique and body composition: effects on performance and effects
on training, semistarvation, and overtraining. In: Brownell KD, Rodin J, Wilmore JH,
editors. Eating, Body Weight and Performance in Athletes: Disorders of Modern
Society. Philadelphia: Lea & Febiger. p 94-111.

Malina RM. 1995. Anthropometry. In: Maud PJ, Foster C, editors. Physiological
Assessment of Human Fitness. Champaign, IL: Human Kinetics. p-205-219.

Malina RM, Battista RA, Siegel SR. 2002. Anthropometry of adult athletes: Concepts,
methods and applications. In: Driskill JA, Wolinsky I, editors. Nutritional

Assessment of Athletes. Boca Raton, FL: CRC Press. p 135-175.

Malina RM, Bouchard C. 1991. Growth, Maturation, and Physical Activity. Champaign,

IL: Human Kinetics.

McCance RA. 1966. The effects of normal development and of undernutrition on the

growth of the calcified tissues. Tijd Soc Geneesk 442569-573.

Murguia R, Dickinson F, Cervera M, Ligia UC. 1990. Socio-economic activities, ecology
and somatic differences in Yucatan, Mexico. Stud Hum Ecol 9:111-134.
Parnell RW. 1958. Behaviour and Physique: An Introduction to Practical and Applied

Somatometry. London: Edward Arnold.

115

Pena Reyes ME. (2002). Growth status and physical fitness of primary school children in
an urban and a rural community in Oaxaca. Southern Mexico. Doctoral Dissertation,
Michigan State University.

Sheldon WH, Dupertuis CW, McDermott E. 1954. Atlas of Men: A Guide for
Somatotyping the Adult Male at All Ages. New York: Harper and Brothers.

Sheldon WH, Stevens SS, Tucker WB. 1940. The Varieties of Human Physique. New
York: Harper and Brothers. ‘

Stinson S. 1985. Sex differences in environmental sensitivity during growth and»
development. Yrbk Phys Anthropol 28:123-147.

Tanner JM. 1952. The effects of weight-training on physique. Am J Phys Anthropol

10:427-462.

116

.EaéEwfi go: 895 $8332 .«0 £82: 05 c8252. moocobta

o3 CA \ 8.20 232.2005 M cos—~93 mo Emma—beau - >0

ZN \ NASA 1 as w» n 85 32.22; - E.

 

 

S , mmd nod 56 :6 _m.~ 390895
a. Ed was :3 23 N3. £98.80:
a So 8o 86 S .o NE 29885
BED 830580 memo—2 cooéom
o\o .>U abofiEom HF 2228:: 8:20me mm 532 oEatm>

 

.35 n 3 0930888 no 338:8 63088855 mo bzéozvoaom fin 05m...

117

Hanoamawmm 8: n m2

 

 

 

m2 m2 m2 a
R; a; w: ow:
8:33; \o Mambgw.
3.: mod 2 owd w _ .v 2 mad cod 2 :
S6 and cm wed 3% cm mod m ~ .m cm 2
Nwd mmd _m 35 wmé Tm cad m_.m 2“ o
and 2N cm 36 mod om mud Ed on w
Rao mod cm and 86 cm cad 35 cm N.
2.6 an; em mud «56 em mad 2.1m om o
Gm 532 : Gm :32 : Gm 532 : ow<
390883— »:EoEOmo—z asfioEowcm

 

Agnew own 3 $39 3.5.. E 250388 mo 82332. @8938 ES memo—z .N.m 033.

118

8.3 a ..

 

 

_q¢va**

* #3.. .w Q

0:” a: 03” am“ 0

0050.5; \0 flabunw
Nod moN om wad omé om mod cod cm I
<2 mm; _m .0: mod 3 mod 3a 3 3
mod voN mm mmd wmé mm $6 omfi mm a
mad end cm mod 03» om amd mm.m cm w
Nod EN am and om;V om cod ~_.m om s
end we; om 2: mod cm 86 wm.m cm 0

Gm 502 : Gm 502 : Gm 502 : 0w<

 

Enrofiofim 3308802 EEoEoccm

 

98% 0mm 3 macs 53.5 E 0938088 .«0 223300 93:06 98 $802 .m.m 030,—.

119

8005838 8: u m Z

 

 

 

 

m2 m2 m2 0
$0 3.0 $0 80 0
0.02.58; \0 flubuew
cod mmd 0m 36 8.0 «m 3.6 3.0 0m 2
m: cmd mm 0: 5.0 mm we; voé mm 3
a: mm.~ mm m: 2.0 0N :4 3; em a
mo; wmd hm «md amé 0m mad gem mm w
:3 2mm mm 36 voé mm id 3mm mm b
and o0; on and 3.0 an mad mm.m am e
Cm 8002 : Gm :002 : Cm 8002 : 0w<
390880m 8908802 FEoEovam

 

.968 0w0 .3 me .08: 8 0953088 .8 8200300 80808 .80 8002 in 0309

120

8005:me 8: u mZ

 

 

 

 

mz mz m2 0
00o :: 8.0 000 8
08.8.8; \0 8050200
00; Sum mm 006 cod mm mud RM mm 2
m0._ mw.m mm nod ww.m mm two SM mm o—
ood MON mm 56 5.0 mm 36 Ram mm 0
Ed and mm mud 3.0 mm end mod mm w
E: 2N mm id 3.0 mm Rd $.m mm 5
3d 00; 0m Rao $0 mm 006 NBA 0m 0
Gm :002 : Gm :002 : Gm :002 : 0w<
30.8885 8908802 £80828m

 

95% 0w0 E 28w :38 :0 0938088 no 8200300 0:00:08 0:0 8:002 .m.m 0509

121

 

 

 

 

00.0 00.0 00.0 00.0 009082000

00.0 0:. 00.0 0:. 0090508:

00.0 00.0 00.0 00.0 009050000

000.0 03 00.0 02 0: 00009088
2.00

000.0 3.. 00.0 :0 00.0 00.0 00088200

020 00.0 00.0 00.0 00.0 00.0 0090508:

00.0 00.0 00.0 00.0 00.0 0:0 0008808:

00.0 00.0 00.0 S. 02 000.2088
000m

0 008... 0083 .303 8 :82 a 8 :82 0 8

8:20.005 :0003-_08m :0003 08%

.0w0mo 800%

_ To :000000 .8 0938088 8 008000.030 :008408: 0:0 .880 0:00:08 0:0 0:008 008300-090 .0.m 030,—.

122

 

 

 

 

000.0 0: 00.0 00.0 00. 00.0 000 _0_ 00808900
500 v 00.0_ 00.0 03 00. 00.0 $0 6. 000850002
0000 00.0 00.0 00.0 02 00.0 00.0 § 008060000
300 v 00.00 00.0 02 § 000020.000

0005
000.0 3.0 00.0 00.0 00_ 00.0 00.0 02 00905200
000.0 00.0 00.0 0: 0: 00.0 03 02 00035802
0000 v 0000 00.0 00.0 00. 00.0 000 02 009050000
.000 v 00.00 00.0 00_ 02 0008.250

10.5w—

: 0:0.— m 0080A .0053 mm :032 : mm :002 : xom
8:80.005 xom :302 35%

 

.uw0 0o 0:090 :6

5:00.? :090: 0:0 ~05: .00 0938088 5 08:0:ob_0 080 0:0 .885 0:00:50 0:0 0:02: 080:_.00-ow< .06 030p.

123

 

:000 v 00.0 $0.00 00806200 60.000 008050000: 000.000 009050002 0005

:000 v 00.0 00.000 009020002 @000: 009080000 .83:
0005:0005 xom

000.0 00.0 20.8 3908005 000m
0005:0me 58:40:3—

 

0 0083 .0003 0020 0 020 _ 020

 

0900:0800 E 0005:0030 :80 0:0 50:07—05: :0.“ 0000—050 830:3 60:05:80 0005080 0:03:00 00 b08806 .m.m 030.0

124

.m 880:0 8 000::0000 0:0 800: .500: 0:0 8:00: 8030:: 0508000380 8:00-500: 0:: :23 3908200 0:0 n39080008
“03808005 :0 0000—00—00 0:: :0.: 0802890 050000 0:: 030% 0w0 :000 8 000:0 0:0—800 :0: 8:00.000 :0:.:0 :008 00:08::0m .._

 

 

0.: _.. 0.0 .. 0.0 .. , :00 n 5 00:0 0:: 00000000: 0:03:00 00:0 :0 600:: 00:50: 00:50:
0.: ... 0.0 .. 0.0 ... :00 u 5 0000 00E: 0:80 :-0 022:0 08:20 0:0 80800 00:0
00:0 30: :000: 08:00
0.0 .. 0.0 __. 0.0 .. :00: u 5 0:00 0:80 ::-0 0:00 :8000 08:20 0:0 00:00
00:0 00:0 0: 80:02::
EN m0 0.0” :80 n 5 000m 00:00» 2-0 .0000: 80:00 :0220 $000: :0 :0 000006
0.0 0.0 0.0 000.: n 5 0:00 02:00 000000 .0000
0.0 0.0 :.0 800.: n 5 0000: ::-0 022:0 02:00 5:080:00: :000: :0 :0 5000002
0.0 0.0 0.0 000.: u 5 0:00 02:00 0000000 0:80
00 0.0 :.0 :00 :.: u 5 0000 ::-0 022:5 02:00 0:808; 000:: 5000002
:.0 _.. 0.0 .. 0.0 _, :00: n 5 0:00
0.0 ... 0.0 _. 0.0 .. 0:: u 5 0000: H000:0 00:80.0 0: 80:00:25.:
0.: ... 3. .. 0.0 0. :00: n 5 00:0 000 0:008:08 00:00: 0:00:00:
0..— 0 0.0 _. ON .0 300: n 5 000m 80:05 80:: 0:00» 2-0 820—30 :000002 800: ._0 :0 08902
0.0 0.0 0.0 :00: u 5 0:00
NN 0.0 Wm 20— H :5 000m 800:3 0000000 .:0 >0=0> 05 0: b80888
0d .010 ad Am: n :5 0:5 :008 :0 0:0 :08: 0 80:: 0:00»
ON mé NM 902 n :5 000m ”—08% 2-0 820—30 30:00 :00802 000::m 8000::
0330885 39080002 02908005: x0m 0:080m N003m

 

80:0:20 000::08< 8:04 .:0 003800 0:0 00:0 8000:: 0:: 0: 0w0 .:0 0:00» I 0: 0 :0:0_E0 .:0 0090:8088 :002 .06 030B

125

CHAPTER 6
Relationship between Somatotype and Physical Fitness in Rural and Urban

Mexican Children

Introduction

Physical fitness is influenced by variety of factors including, size, physique, and
body composition, among others (Malina and Rarick, 1973; Malina, 1975, 1992).
Physique as a factor influencing physical fitness, in particular performance-related
fitness, has received considerable attention given the reasonably unique distributions of
physiques among athletes in specific sports or in different events within a sport (Carter
and Heath, 1990; Malina et al., 2002). In contrast, the relationship between physique and
performance in the general population, especially children and adolescents, has received
relatively less attention (Malina and Rarick, 1973; Malina, 1975).

Physique refers to the total configuration of the body, and is most often quantified
as a somatotype (Sheldon et al., 1940, 1954; Parnell, 1958; Carter and Heath, 1990).
Somatotype is a composite based on the contributions of three components -
endomorphy, mesomorphy and ectomorphy. At present, the Heath-Carter anthropometric
protocol (Carter and Heath, 1990; Malina, 1995) is the most commonly used method to
estimate somatotype.

Studies relating physique to performance generally focus on individual
components of somatotype. This is a limitation because the three components together
define an individual’s somatotype. It is necessary to control for variation in the other two

components when considering relationships between a component and performance.

126

Allowing for this limitation of earlier studies, correlations between somatotype
components and strength and motor performance tend to be low to moderate in
magnitude (Malina, 1975, 1992; Malina and Rarick, 1973). They rarely exceed 0.5;
hence, explained variances are low and have limited predictive utility.

Nevertheless, the literature suggests several trends. Correlations between
endomorphy and performances requiring projection or movement of the body through
space are consistently negative in children and adolescents of both sexes. These include
tasks such as the standing long jump, vertical jump, dashes, pull-ups, and distance runs.
The excess fatness associated with endomorphy represents dead weight, which must be
projected or moved, and thus presents a mechanical disadvantage. In contrast,
endomorphy is positively related to measures of static strength, which emphasizes,
perhaps, a significant contribution of muscularity to ratings of endomorphy.
Mesomorphy contributes variably to strength and motor performance. Correlations tend
to be positive and low in both sexes in childhood and adolescence. Ectomorphy is not
consistently related to motor performance during childhood, but tends to be negatively
related to strength and power tasks in adolescent males. The low negative correlations
for strength tasks emphasize the lack of muscular development associated with extreme
ectomorphy or linearity of build (Malina 1975, 1992; Malina and Rarick, 1973).

These trends are derived from primarily from samples of well-nourished
European and North American children. The situation may be somewhat different for
children living under marginal health and nutritional conditions. Chronic undernutrition, .
for example, is associated with reduced body size and muscularity and altered

proportions, specifically proportionally shorter lower extremities.

127

This study considers the relationship between somatotype and performances on
several physical fitness tests in rural and urban children in Oaxaca, southern Mexico. The
study used two approaches. First, the relationship between somatotype components and
each physical fitness item was correlated while statistically controlling for age and the
other two components. Second, the contribution of somatotype to variation in physical

fitness was estimated with multivariate techniques.

Methods and Procedures

The communities, sample, anthropometric procedures and fitness tests were
described earlier (Chapter 3). Procedures for estimating somatotype with the Heath-
Carter protocol and age, sex and rural-urban variation were described earlier (Chapter 5).
The physical fitness battery included a combination of performance- and health-related
tests:
1. Static strength — Static grip strength of the left and right hands,
2. Running speed — 35 yard (32.3 meters) dash,
3. Power — Standing long jump (SLJ),
4. Flexibility - Sit and reach (SAR),
5. Muscular endurance - Timed sit-ups, number in 30 seconds,
6. Cardiovascular endurance — Distance (meters) run and/or walked in 8 minutes for

children in grades 1 to 3 and 12 minutes for grades 4 to 8.

The specific procedures for each test were described in Chapter 3. The sum of left and

right grip strength (Sum of RL) was used as the estimate of overall static strength. The

128

distance run was converted to an average running speed expressed as meters per minute
(Pate et al., 1989).

Two analyses were done. Sex- and community-specific third order partial
correlations between a somatotype component and a fitness task, controlling for age and
the other two somatotype components. were initially calculated. Then, sex- and
community—specific multiple regression analyses were conducted to estimate the
variation in each physical fitness task that was contributed by age and each somatotype

component.

Results

Third-order partial correlations between somatotype components and physical
fitness variables, controlling for age and the other two components, in rural and urban
boys and girls 6-11 years are presented in Tables 6.1 and 6.2. Correlations were generally
low and but several were moderate, —0.39 to +0.28. The highest negative correlation, —
0.39, was between endomorphy and the distance run in rural boys, whereas the highest
positive correlation, 0.28, was between ectomorphy and sit-ups in rural boys.
Endomorphy correlated negatively with all fitness items except strength in rural children
and the dash in rural girls; the two positive correlations, however, approximated zero
(0.03 to 0.10). The three somatotype components were negatively correlated with the
distance run except in rural girls. Overall, the partial correlations indicated that each
somatotype component was significantly correlated, at most, with only two physical

fitness items, but the correlations were moderate at best.

129

Results of the multiple regression analyses indicated that age and somatotype
accounted for 1% to 71% of the variance in the six fitness tasks (Tables 6.3 and 6.4), and
there were no consistent urban-rural contrasts by sex. The independent variables
contributed proportionally more to the variance in static strength (58% to 71%) compared
to the other fitness tests. Contributions of age and somatotype to the variance in other
fitness tests were more variable. Age and somatotype accounted for moderate amounts of
the variance in the dash, 37% to 56%, and standing long jump, 31% to 41%.
Corresponding estimates for the distance run were 26% and 25% of the variance in urban
boys and girls, 21% of the variance in rural boys, but only 4% of the variance in rural
girls. Age and somatotype accounted for somewhat lesser proportions of the variance in
timed sit-ups in urban and rural girls, 14% and 10%, respectively, compared to urban and
rural boys, 19% to 28%, respectively. In contrast to the other fitness items, the variance
in the sit and reach accounted for by age and somatotype was low in the four groups, 1%

to 9%.

Discussion

Partial correlations between somatotype and physical fitness were generally low,
with several of moderate magnitude. Endomorphy was consistently and negatively
correlated, whereas as mesomorphy and ectomorphy were variably correlated with the
fitness items included in this study. The results are reasonably consistent with earlier
correlational studies (Malina, 1975, 1992; Malina and Rarick, 1973; Slaughter et al.,
1977, 1980). However, the low to moderate correlations between somatotype and static

grip strength in the present study (Tables 6.1 and 6.2), especially for endomorphy and

130

mesomorphy, were at variance with the trends suggested in the literature. The highest
correlations in the present study were for mesomorphy and strength among urban
children, 0.34 in boys and 0.17 in girls, whereas all other correlations were low, -0.02 to
0.10 for endomorphy and -0.02 to 0.15 for ectomorphy.

Results of the multivariate analysis indicated that age and somatotype accounted
for 1% to 71% of the variance in the six fitness tasks. Age appeared to be the main
contributor to five of the performance items in contrast to Heath-Carter anthropometric
somatotype components. Except for a negative influence of endomorphy, there did not
appear to be any consistent trends in the contributions of somatotype components to
variation in performance in these samples of Mexican children living under marginal
nutritional and health conditions. There also were no consistent rural-urban and sex
differences in the estimated contributions of age and somatotype to variation in
performance. Age and somatotype contributed proportionally most of the variation in
static strength (58% to 71%), followed by the dash (3 7% to 56%) and the standing long
jump (31% to 41%). The estimated contributions of age and somatotype to variation in
performance were lower and more variable for the distance run (4% to 26%) and sit-ups
(10% to 28%), and lowest for the sit and reach (1% to 9%).

Previous studies of somatotype and fitness of children did not include multiple
regression analyses; hence, comparisons are not possible. Results of the present study
indicated a primary role for age and a lesser role for somatotype in explaining the
variance in the health- and performance-related fitness of children 6-11 years of age

living under marginal health and nutritional circumstances. The important role for age

l31

per se probably reflects the significant contributions of neuromuscular maturation and
experience to fitness tasks at these ages.

Data based on questionnaires for fourth, fifth, and sixth grade children suggest a
limited scope of activities among both urban and rural children. Urban children appeared
to have less regular physical activity in the context of sports and household chores, but a
greater variety of activities than rural children. Rural children, in contrast were more
regularly involved in household-related activities of moderate intensity and in some
instance activities of moderate-to-vigorous intensity, but a reduced scope of activities
(Pena Reyes, 2002).

The results may also reflect sampling variation per se, the relatively narrow range
of somatotypes in the sample, and/or the small body size of the children compared to I
other studies based on samples of European and North American children. The small size
of the current samples was indicated earlier. The range of variation in somatotypes was
relatively reduced compared to studies of well-nourished children. None of the standard
deviations for rural boys exceeded 1.0. Age-specific standard deviations for endomorphy
ranged from 0.6 to 1.1, and only two of the values exceeded 1.0. Corresponding values
for mesomorphy ranged from 0.6 to 1.1, and only four exceeded 1.0. Standard deviations
for ectomorphy ranged from 0.7 to 1., and seven exceeded 1.0. Values >1.0 occurred
most often among 9-10 year old girls and boys (see Tables 5.2 to 5.5). In contrast,
standard deviations for mean components approximate 1.0 to 1.5 somatotype units in
samples of North American and European children and adolescents (Carter and Heath,

1990; Malina et al., in press).

132

The reduced variation in somatotypes, or more specifically, the lack of extreme

 

somatotypes in the present sample of Mexican children, may influence the contributions
of physique to performance. It is at the extremes of somatotype that performances on
fitness tests may be affected. Endomorphy, mesomorphy and ectomorphy are,
respectively, indicators of fatness, musculoskeletal development and linearity. Extreme
ectomorphs tend to be deficient in muscle mass and strength, whereas extreme
mesomorphs are characterized by muscular development and strength. Extreme
endomorphs, on the other hand, have excess fat but also reasonably well development

mesomorphy (Malina, 1975, 1992; Malina and Rarick, 1973).

133

Bibliography

Carter JEL, Heath BH. 1990. Somatotyping - Development and Application. Cambridge:
Cambridge University Press.

Malina RM. 1975. Anthropometric correlates of strength and motor performance. Exerc
Sport Sci Rev 3:249-274.

Malina RM. 1992. Physique and body composition: effects on performance and effects
on training, semistarvation, and overtraining. In: Brownell KD, Rodin J, Wilmore JH,
editors. Eating, Body Weight and Performance in Athletes: Disorders of Modern
Society. Philadelphia: Lea & Febiger. p 94-111.

Malina RM. 1995. Anthropometry. In: Maud PJ, Foster C, editors. Physiological
Assessment of Human Fitness. Champaign, IL: Human Kinetics. p 205-219.

Malina RM, Bouchard C, Bar Or 0. in press. Growth, Maturation, and Physical Activity,
2"d edition. Champaign, IL: Human Kinetics.

Malina RM, Battista RA, Siegel SR. 2002. Anthropometry of adult athletes: Concepts,
methods and applications. In: Driskill JA, Wolinsky I, editors. Nutritional
Assessment of Athletes. Boca Raton, FL: CRC Press. p 135-175.

Malina RM, Rarick GL. 1973. Growth, physique, and motor performance. In: Rarick GL,
editor. Physical Activity: Human Growth and Development. New York: Academic
Press. p 124-153.

Parnell RW. 1958. Behaviour and Physique: An Introduction to Practical and Applied
Somatometry. London: Edward Arnold.

Pate RR, Slentz CA, Katz DP. 1989. Relationship between skinfold thickness and

performance of health related fitness test items. Res Quart Ex Sport 60: 1 83-1 89.

134

 

Sheldon WH, Dupertuis CW, McDermott E. 1954. Atlas of Men: A Guide for
Somatotyping the Adult Male at All Ages. New York: Harper and Brothers.

Sheldon WH, Stevens SS, Tucker WB. 1940. The Varieties of Human Physique. New
York: Harper and Brothers.

Slaughter MH, Lohman TG, Misner J E. 1977. Relationship of somatotype and body
composition to physical performance in 7-to-12-year-old boys. Res Quart 48:159-

168.

Slaughter MH, Lohman TG, Misner JE. 1980. Association of somatotype and body
composition to physical performance in 7-12 year-old-girls. J Sports Med 20: 1 89-

198.

135

 

ooze—Epics Eton £950 0:5 832 a ma Etna: 0.53 mcwmm a

85 v a Bsaficam _.
:5 v a oogéawa :

 

 

mod - o_.o - std - 58>: .Si 3565
cod - mod :6 - com .2me
86 2d N2. - =5 .15
mod ood m mo - c .mqDém
Ed - mod - ivmd .. Eu rm<m
2d :42 So - 3 duo i saw

3&3
...:.o - mod - iomd - 58>: and 8535
mod cod mod - oom .2“:me
wed - mod - Ed - :5 .3m
.33 a; :3 - a .aaaa
mod - mod mod - :8 Jam
93 o; o; 3 d5 5. Sam

:23—

802 6.62m dw< 20m .095 dw< Bum .032 .ow< com wax—05:00

zanuoEBom 3958082 Efiofiovcm

 

.03 no 88» :6 when :35 can .83
5 30:5 3063a can .oqbofiEOm can own he macaw—oboe 3th 820-22; 4.0 033.

136

wanna—88¢ .8809 25:3 0:5 832 a mm 3:35 203 mcwmm a

mod v a 8936mm _.
:3 v a 8§£§m 1.

 

mod - 3:. .. mod - 58>: and 8ng
mod 86 86 .. com .asmmD
Ed - mod - ivmd - Eu .3m
36 - cod Ed - a Essa
Ed 36 mod - Eu rm<m
25 Ed 8d - w; .95 .2 8%

5&3
cod vod mod - SEE and 853me
:6 No.0 86 com Mamma—
..:..N~.o Ema vod - Eu .3m
*Ed :6 8.0 .. : .82-:m
2d - 3d - fiend - :8 Jam
3o - 23 3o 3 .36 .E cam

REM

082 695 dw< Sam 6ow dw< Sum .802 .ow< com wizobcou

Eacofiouom 3908082 3908095

 

.0? mo menu» :6 me SES use :22
E 3025 Bowman—n 96 .330888 wan own Mo 30:22.80 3th BEoéhfi. .N6 038.

I37

ooggotoa 533 £960 08: .832 a me 3:02: 203 8me a
85 v a uosucawa ..
5d v S ougocawfi 1..

138

 

 

:d - dmd - *vmd - .1..de 1.3.6.— cmd $— 5.58 .53 8:8me
Odd - 2d 2d - .1..ch 1.5.2. de G~ com .anmaa
odd mmd Ed - fixed .3de 3d 3: So .75
vod 3d w—d - 32d .1. _ md 2d $~ : .maDfim
mmd - :d - immd - vod End wdd $~ :8 Jam
86 1. 3o 3o - 1.1.; 13.8 mg § 3 .35 .2 Sam
:35
*mmd - Ed - 1..Omd - Immd .Iddd Rd NE 58>: .55 3535
vod add vod - iced 1.8.5 nmd mm ~ com .230
2d - 8d - mod - 13d inflo— de v2 :5 .3w
*Lmd ...dmd mod .. Ihmd 1.3.: and cm ~ a .33ém
2d - Nod :d - Nod 3d 8d em _ Eu Jam
2d 2d odd 39d :vndm wmd Vm _ 3 £ch Ad Esm
33m
39889on 2308802 Enrofiowcm ow< m «M =
3565000 Sum

 

.03 do was?» :6 939 53.5 28 3.5.. 5 0930388 can own so mmofic Bummba mo cofimawe 032:2 .md 2an

8:85:85: 5qu 29.60 25. .532 a mu 3:35 825 39m a

mod v a 8505:me ..
:3 v a oofiomama 1.

 

 

m _d - Lmd - mod - Idvd ..:..w_.m_ mmd mm 5 55>: E3— 3:850
:d nod mdd - {.mnd 13.: wmd 02 08 “.5me
Ed - odd - *mmd - Imod Ivndm nmd 02 :8 .3m
mod .. vod odd - iwmd 1.36 3d mm _ : .335m
vod 2d 2d - 2d omd mod em 5 Eu Jam
26 Rao . cod 1&2 1.8.3 8d :2 3 $5 5. new
:35
3d dod m _d - mod om; vod d2 55>: 55% 8:9me
Gd Nod vdd Inmd Iwmfim dvd :5 com .555me
idvd omd nod - iced 1.3.3 Rd m: 50 .Bm
de dmd mod - Lmd 1.5.21V d—d m: : £335
omd - Ed - ibmd - Ed Imdd odd m2 Eu Jam
3d 2d cod .1..de ..:..3d3 Ed m5 3 5:0 AM 53m
33¢
13905835 5905802 590503m 09.. m NM :
$550580 Sum

 

.0? mo 83% :6 55m :35 :5 3:: 5 0&8358 55 own :0 32.5 35993 m0 533%»: 25:32 .vd 2an

139

CHAPTER 7

 

Growth and Physical Fitness of Stunted and Non-Stunted Children in a Rural and

an Urban Community in Oaxaca, Mexico

Introduction

Stunting is a concept introduced in the early 19703 by Waterlow (1972, 1973) to
indicate retardation of linear growth. The term describes deficits in attained length or
height for chronological age compared to an international reference (Waterlow, 1972;
World Health Organization [WHO], 2000). The World Health Organization (2000)
describes stunting as “. . .a process of failure to reach linear growth potential as a result of
suboptimal health and/or nutritional conditions.” Chronic undernutrition and illness early
in life, usually the first three years, are primary factors in the etiology of stunting. The
prevalence of stunting among children under 5 years of age in developing countries
varies from 5% to 65% (de Onis et al., 1993). Stunting is defined as height-for-age that is
more than two standard deviations below age- and sex-specific reference values for
healthy, well-nourished children (WHO, 2000).

Chronic undernutrition is common in most developing countries where poverty,
inequality, and/or economic underdevelopment are prevalent. The growth status of
children is very responsive to an environment where the nutritional conditions are
marginal, and stunted growth in length/height is a common result. Most studies on
stunted populations have considered the retardation of growth per se and/or catch-up
growth (Bénéfice et al., 2001a, 2001b; Neumann and Harrison, 1994; Walker et al.,

1996). The behavioral, biological, intellectual, and social consequences of chronic mild

I40

to moderate undernutrition during childhood have also been investigated on a more or
less regular basis (Post and Victora, 2001; Walker and Walker, 1997; Neumann and
Harrison, 1994; Martorell et al., 1992), but the consequences of stunting for physical
fitness have been studied to a lesser extent (Satyanarayana et al., 1979; Bénéfice et al.,
1996)

Stunting is a continuous variable that is a reflection of compromised growth in
stature that occurred in early childhood. Although short stature has a generally negative
influence on many standard performance tasks (Malina and Buschang, 1985; Malina et
al., 1987; Malina and Bouchard, 1991), the effects of stunting per se on physical fitness
of school age children has not been systematically evaluated. An exception is the work
of Satyanarayana et al. (1979), who considered the effects of stunting at 5 years of age on
the physical working capacity of Indian boys during adolescence. The most severely
stunted boys were most compromised in working capacity. In contrast, most research on
nutritionally at risk populations has been conducted on preschool children, and children
who were nutritionally compromised during the first two years of life tend to show
deficits in several measures of physical performance (Benefice et al., 1996). Although
children below the age of 5 years are more sensitive to unfavorable living and nutritional
conditions, children of school age represent a population that has survived the
compromised environmental conditions that are generally associated with high morbidity
and mortality during the preschool years.

The purpose of this study is to compare the anthropometric and physical fitness
characteristics of stunted and non-stunted school children 6-11 vears of age resident in an

urban and in a rural community in southern Mexico.

141

 

Methods and Procedures

The communities, sample, anthropometric procedures and physical fitness tests
were described earlier (Chapter 3). The children were classified as stunted and non-
stunted on the basis of height. A child whose height-for-age z-score was > 2 standard
deviations below the age- and sex-specific median of the United States reference was
considered stunted. This is the criterion of the World Health Organization (1997). Other
children were considered non-stunted. The new Centers for Disease Control and
Prevention (2000) growth charts were used to calculate age- and sex-specific z-scores for
each child. The prevalence of stunting in the sample of 6-11 year old children was
approximately 28% for both boys and girls in the mral community, and 15% for girls and
boys in the urban community (Table 7.1).

Anthropometric variables included the following:
1) Body size: Weight, height, body mass index (BMI);
2) Segment lengths: Sitting height, estimated leg length;
3) Skeletal breadths: Biacromial, bicristal, biepicondylar, bicondylar;
4) Limb circumferences: Relaxed and flexed arm circumference, calf circumference,
estimated midarm muscle circumference (EAMC), estimated calf muscle circumference
(ECAC);
5) Subcutaneous fatness: Sum of 4 skinfolds - triceps, subscapular, suprailiac and medial
calf skinfolds;
6) Proportions: Sitting height/standing height ratio.
Heath-Carter anthropometric somatotypes were also calculated for each child (Carter and

Heath, 1990; see Chapter 5).

I42

Physical fitness variables included the following:
1) 35 yard (32.3 m) dash — running speed,
2) Standing long jump (SLJ) — explosive power,
3) Sum of right and left grip strength (Sum RL) — static strength,
4) Sit-and-reach (SAR) — lower back and upper thigh flexibility,
5) Timed sit-ups -— abdominal muscular strength and endurance,
6) Distance run (Dist Run) — cardiovascular endurance.

The analyses proceeded in several steps. Descriptive statistics for all variables
were initially calculated for stunted and non-stunted children by sex within each
community. Sex-specific multivariate analyses of covariance (MAN COVA), with age as
the covariate, were used to compare tr» anthropometric and physical fitness
characteristics of stunted and non-stunted children within each community. Sex-specific
multivariate analyses of covariance (MANCOVA) were used to compare the

somatotypes of stunted and non-stunted children within each community.

Results

Anthropometry. Age-adjusted means and standard errors of anthropometric and
derived variables in stunted and non-stunted children by sex in each community are
presented in Tables 7.2 to 7.5. All anthropometric characteristics differed significantly
between stunted and non-stunted children; stunted children were smaller in all
anthropometric dimensions and derived variables. The sitting height/standing height

ratio, however, was higher in stunted children (significantly so except urban girls),

4/

143

 

indicating that the shortness of stunted children was due in large part to proportionally
shorter legs.

Somatotype. Within each community and sex, stunted and non-stunted children
differed significantly in somatotype (Tables 7.6 and 7.7). However, univariate F -tests for
specific components did not differ significantly between stunted and non-stunted children
except for endomorphy and ectomorphy in rural boys. Stunted rural boys were
significantly less endomorphic and ectomorphic than non-stunted boys. Overall, stunted
children of both sexes were slightly higher in mesomorphy, and lower in ectomorphy and
endomorphy compared to non-stunted children.

Physical Fitness. Community- and sex-specific age-adjusted means and standard
errors for the physical fitness characteristics of stunted and non-stunted children are
summarized in Tables 7.8 to 7.11. Only absolute strength (sum of right and left grip) and
left grip strength per unit estimated arm muscle circumference differed significantly
between stunted and non-stunted children of both sexes within each community. The
only other significant differences between stunted and non-stunted children were the
dash in rural males and the standing long jump in rural girls; non-stunted children
performed better. Although the differences were not significant, stunted rural children
had somewhat better average performances on the distance run expressed as meters per
minute compared to non-stunted rural children, but corresponding means for stunted and
non-stunted urban children were virtually identical. Otherwise, mean performances of
stunted children on the sit and reach and timed sit-ups were comparable to those of non-

stunted children.

144

 

Discussion

Anthropometric characteristics were consistently larger in the non-stunted
compared to stunted children. The sitting height to standing height ratio, however, was
higher in stunted children indicating that growth of the lower extremities (leg length) was
specifically compromised. Bénéfice et al. (2001a, 2001b) reported similar observations in
stunted (malnourished) Senegalese girls. Undemourished children and adults, who were
not necessarily stunted, also were shorter and lighter, and had reduced muscle mass
compared to well-nourished age and sex peers (Spurr, 1984; Be’néfice et al., 1996, 1999;
Hoffman et al., 2000).

Absolute body size and nutritional status are related to physical performance
(Spurr, 1984; Malina and Buschang, 1985). However, in the present study, non-stunted
children performed better than stunted children only in absolute strength (sum of right
and left grip) and strength per unit estimated midann muscle circumference. The former
reflects the influence of overall small body size, while the latter suggests changes in
muscle mass associated with growth stunting. The reduced strength per unit arm muscle
in stunted children may, perhaps, reflect qualitative changes in muscle tissue associated
with undernutrition early in life. Interestingly, strength per unit body mass did not differ
between stunted and unstunted children, which further implicates subtle changes in
muscle tissue. Performances in the dash of rural males and the standing long jump of
rural females also favored the non-stunted children, and may also possibly reflect
changes in muscle mass per se and qualitative changes in muscle tissue associated with

early undernutrition.

145

 

Otherwise, the performances of stunted and non-stunted children on the other
fitness tests did not consistently differ. There were negligible differences in the sit and
reach (flexibility) and timed sit-ups (abdominal muscular strength and endurance). On
the other hand, stunted children in the rural community were, on average, better in the
distance run (meters per minute) than non-stunted children, whereas stunted and non-
stunted children in the urban community did not differ in the distance run.

Heath-Carter anthropometric somatotypes differed between stunted and non-
stunted children. Stunted children of both sexes, urban and rural, had somatotypes that
were, on average, slightly more mesomorphic and less endomorphic and ectomorphic
compared to non-stunted children. Although stunted children were, on average smaller in
the body dimensions used to estimate the Heath-Carter somatotype, they were obviously
shorter (as per the criterion of stunting) but had disproportionally shorter in estimated leg
length compared to non-stunted children. Although the shortness relates to the definition
of stunting, the short stature may influence the estimated somatotype. Each component of
the Heath-Carter anthropometric protocol involves an adjustment for height.
Unfortunately, comparative somatotype data for other samples of stunted and non-

stunted children are not available to verify this suggestion.

146

Bibilography

Bénéfice E, Fouére’ T, Malina RM. 1999. Early nutritional history and motor
performance of Senegalese children, 4-6 years of age. Ann Hum Biol 26:443-455.

Be'néfice E, Fouéré T, Malina RM, Beunen G. 1996. Anthropometric and motor
characteristics of Senegalese children with different nutritional histories. Child: Care,
Health Develop 22: 1 5 1-165.

Béne’fice E, Garnier D, Simondon KB, Malina RM. 2001a. Relationship between stunting
in infancy and growth and fat distribution during adolescence in Senegalese girls. Eur
J Clin Nutr 55:50-58.

Béne'fice E, Garnier D, Simondon KB, Malina RM. 2001b. Differences in growth and
body composition during puberty between Senegalese adolescent girls stunted or not
stunted in infancy. Biom Hum et Anthropol 19:55-61.

Carter JEL, Heath BH. 1990. Somatotyping - Development and Application. Cambridge:
Cambridge University Press.

Centers for Disease Control and Prevention (CDC). 2000. National Center for Health

Statistics CDC growth charts: United States. http://www.cdc.gtrv/gromhchartshtm.

 

de Onis M, Monteiro C, Akré J, Clugston G. 1993. The worldwide magnitude of protein-
energy malnutrition: an overview from the WHO global database on child growth.
Bull World Health Organ 71:703-712. v

Hoffman DJ, Sawaya AL, Coward WA, Wright A, MartinsPA, de Nascimento C,
Tucker KL, Roberts SB. 2000. Energy expenditure of stunted and non-stunted boys
and girls living in the shantytowns of 850 Paulo, Brazil. Am J Clin Nutr 7221025-

1031.

I47

Malina RM, Bouchard C. 1991. Growth, Maturation, and Physical Activity. Champaign,
IL: Human Kinetics.

Malina RM, Buschang PH. 1985. Growth, strength and motor performance of Zapotec
children, Oaxaca, Mexico. Hum Biol 57:163-181.

Post CLA, Victora CG. 2001. The low prevalence of weight-for-height deficits in
Brazilian children is related to body proportions. J. Nutr. 131:1290-1296.

Satyanarayana K, Nadamuni Naidu A, Narasinga Rao BS. 1979. Nutritional deprivation
in childhood and the body size, activity, and the physical work capacity of young
boys. Am J Clin Nutr 32: 1 769-1 775.

Spurr GB. 1984. Physical activity, nutritional status, and physical work capacity in
relation to agricultural productivity. In: Pollitt E, Amante P, editors. Energy Intake

and Activity. New York: Alan R. I iss. p 207-261.

Walker ARP, Walker BF. 1997. Moderate to mild malnutrition in African children of 10-
12 years: roles of dietary and non-dietary factors. Int J Food Sci Nutr 48:95-10].

Walker SP, Grantham-McGregor SM, Himes JH, Powell CA, Chang SM. 1996. Early
childhood supplementation does not benefit the long term growth of stunted children
in Jamaica. J Nutr 126:3017-3024.

Waterlow JC. 1972. Classification and definition of protein-calorie malnutrition. Brit
Med J 32566-569.

Waterlow JC. 1973. Note on the assessment and classification of protein-energy

malnutrition in children. Lancet 1:87-89.

148

World Health Organization (WHO). 1997. Physical Status: The Use and Interpretation
of Anthropometry. Geneva: World Health Organization, Technical Report Series, No.
854.

World Health Organization (WHO). 2000. http://www.who.int/nut/pem.htm.

149

Table 7.1. Prevalence of stunting in 6-11 year old children in the rural and urban

 

 

communities.
Rural Urban
Males Females Males Females
Stunted 49 43 25 21
Non-Stunted 126 1 1 1 132 140
Total 175 154 157 161

 

150

 

 

:3 v no 3... S ”.8. ea .mm a. saw
:3 v 3 SN 2 o. a :5 628
:5 v S ”.2 3 a: so 62%.
:3 v S ”3 No SN .5 .26 :6
:3 v 3 2: mo _.: so .26 :5
:3 v 3 9m 3. 3. :5 5:28.305
:3 v 3 as S i so 5:285
:3 v 3 2m 3 .3. so .355
:5 v 3 3m 2 SN :5 428885
:3 v S N: do 3m .x. .5: \ a: em
:3 v do 9% to 2m .5 .5 m3
:5 v 2 Sb 3 3e .8 .E 5
mod v No ”.2 No 92 NEE ._2m
:3 v 3 was no 3: so .290:
:5 v 3 3m 2 SN 3 .Ewaa

a mm :82 mm 53: §£E>

2 _ T 5 8555-82

Amv u 5 3:55

 

own do memo.» :6 when :22 3256-5: 28 3:53 «o

$32.53 Egret 98 05088855 do Echo duodenum was 2:88 vodmzmdaomax .ms 033.

151

 

 

mod v ad can «A hem ES .mm v Ezm

 

5.5 v No SN No 2m :5 .028
:3 v 3 5.2 do 3; so .025
:3 v mo in no «.8 so .2526
:3 v No 3: we 5.: . so .26 :5
:3 v od 3 3 5 :5 5328505
:3 v Qo 5 3 3 so 53285
:3 v 3 ”.8 No we so .3655
:3 v 3 35 mo EN :5 .35285
:3 v 3 3m 2. 3m .x. .5: \ 5 5m
:3 v No 9% to 3% so .5 we;
:3 v No is no 35 so .5 5m
85 v S m2 2 we NE} .95
:3 v 3 $2 2 E: .8 .250:
5.5 v to 98 9o 92 3 5%?

5 mm :82 mm :82 $325»

$5 u 5 Beameoz av u 5 Beam

 

.0? mo man?» :6 me Eh: 32:528.. can @855 we
.83th» 33:6 was oEoEomeficm go 323 35:85 98 «:88 Ban—mumbmxx .ms 2an

152

 

 

5.5 v so . 3m 2 man as am v saw
:3 v 3 2m to v. a so .028
:3 v S «.8 I. of :5 .usZm
85 v No 0.2 .3 3N .5 .2628
:3 v we 3: to n: so .25 :5
:3 v 3 3 3 2. :5 5328505
:3 v 3. 3w 3 3 :5 £32505
5.5 v S N. a no ”.2 .8 .3555
:3 v 3 mom 3 3N :5 55.285
:3 v S New 2 3m .x. .5: \ 5 5m
5.5 v do 2% we 1% so .5 m3
:3 v we 25 3 5.8 so .5 5m
85 v 3 1: no as .53 .EE
:3 v 3 5.5 we 3: so 59%
:3 v to 2m 2 2m 5. .5503

5 mm :82 mm :82 835$

153

a: u 5 Baumeoz :N u 5 855m

 

.0? mo mac» :6 when :33 “5853-8: was 3:53 .«o
8.85? co>tov 25 05082—853 me 383 @8386 98 £808 Rum: 93-095. .vs 2an

 

 

:3 v 3 3m 3 3m 22 .5 v 2.5
:3 v 3 3m 3 IN .8 .255
:3 v 3 3: 3 :2 so .0325
:3 v 3 3m 3 3N :5 .2556
:3 v 3 3. 3 5.: so .25 :5.
:3 v 3 :3 3 3. .5 538285
:3 v 3 3 3 3 so .5385
:3 v 3 o. 5 3 3N :5 .3555
:3 v 3 3m 3 EN :5 52225
mz 3 3m 3 3m 2.. .5 \ 5 .5
:3 v 3 3m 3 3m .8 .5 5.5
:3 v 3 35 3 35 .5 .5 5
33 v 3 n: 3 3: N2} .55
:3 v 3 35 3 3: so .5565
:3 v 3 3m 3 3m 3 .2553
a mm :82 mm :82 25.5.5

154

2.2 n 5 coEBméoZ Gm H 5 BEBm

 

.53 .3 559m :6 55% SEE 62:35-50: 95 3:55 we
335:; 3355 Ea 05583853 mo 58.5 uawgm 95 £508 353.355-0532 .2. 2an

 

 

23 5.. 333 :3 33 53 33.. 5 523225

155

 

 

 

353 33 333 $3 3: 53 33 5 5223802

53 m3 :3 33 3: 53 33 5 523285

:33 33 333 252325
:35

533 $3 333 23 : _ N3 33 Q 523225

333 I: :3 $3 : _ N3 33 3 5232202

533 :3 333 33 _ _ _ m _ 3 33 3. 523285

coed oo. _ ~ Nwd ombemfiom
:53—

3 one 5 8:3 533 mm :82 3 mm :82 a 553
8:255 3% 3525 3253.82 Beam

.53 .«o 55?»

_ To when :35 95 556 3:35-50: 35 58:53 E 250588 mo 58.5 23255 was. 252: 3533-36. 66 033.

 

 

036 CNN Ed Sam I: mNd 3N mm EEO—:98“

156

 

 

 

2% 3c So m3 5 :5 a? 8 2908802
33 SN :3 m2 .2 m; own 2 298.85
Sad 36 3d onboEEom
:BHD
wood 3% 3o 3% § Ed m _ .N 3 Eeoesom
mm 3 SN 85 :4. «2 m3 m? S 39380:
2&6 m: mod a: «N. m; N: 3 £98.85
:56 Q? So abosaom
3.5%
a 038 m 8qu 353 mm :82 : mm :82 : ow£=>
oocobba £58m wag—3m 3:55-202 338m
. .09» .«o m8?“

2% me :35 can 3:: cofiaméoc 28 @853 5 2.30388 mo 38.5 @8283 can 282: BE: €9ow< SK 033.

 

€35:me 8: u m Z

 

 

m2 5 3w: 3 22 58E. .5. 85:5
3o v 35 o: 8.9 as am .53
m2 2 ma 2 So .8 .75
m2 to 3 9o 3 a .25-:m
mz to 3m 8 SN .5 42m
:3 v :3 as 85 $6 8&3 62556 4
m2 So 84 8d 3o 33 £24m 8%
:3 v to 2m 9o 0. a 3 .E .55

a mm :82 mm :82 853:;

as u 5 Seawaoz

5 n 5 36%

 

.0?“ mo E8» :6 983 BE 3286-5: 98 323% Mo
83%.? 33:6 Ba $055 :8?an .«o Echo 23:8,.“ 28 838 vofiswvméw/x .ws 2an

157

 

Educfiwfi 8: n m2

 

 

m2 3 :3: 33 c .3 5.55 .55 85:5
m2 :3 E. :3 S 3 3m .58
33 v f 35 3 3.: :5 .Bm
m2 3 :3 3 5.. a £5-55
m2 3 3.8 3 2N :5 .53
:3 v :3 who 83 $3 523 US$55 4
m2 :3 33 83 33 33 .35: 5.5
:3 v 3 new 3 EN 5. .3 :am

3 mm :82 mm :32 3355/

as u 5 35532

5. u 5 Beam

 

.03 50 £3: 2-0 mEm :28 3255-5: 28 @855 Mo
83559 83.8: 98 $025 3063: no £88 235% was 2:38 teamsqnaéwxx in 033.

158

3:35:me 8: n m2

 

 

m2 2 332 3 33: 5.55 .53 8535
mz :3 3: :3 32 ca 38:
32 2 33: 33 M32 50 .33
32 v3 3: 3: 3.: a 333.3
m2 3 3.3 3: SN .5 .33
:3 v :3 N33 33 $3 559. 632835 3
32 :3 33 :3 33 353 323 5.3
:3 v 3 3.3 N: m. a m: .3 5:3

3 mm :82 mm :82 $22.;

:33. n 5 35.33.82 5 n 5 Beam

 

.33 .«o m3?» :6 93.: 53:: 35:33-8: :53 @855 mo
833:3 :35: :3 385m 3293: .«o 28.5 33:83 :5 332: woamsavm-uwxx .o _ .5 033:.

159

 

 

:85:me :o: n m Z

 

 

m2 2 32 3“ 22 5.55 .55 855:
m2 83 we: :3 3.» 8m 33:
m2 2 :3 I 3.3 :5 .Em
m2 3 S: 3 :3 a 52.3
m2 3 SN 3 SN .8 :73
:3 v :3 £3 :3 33 so} 625:5 :
m2 :3 83 :3 :3 5:9. .55: Sam
:3 v 3 3m .3 3N 9. :2 saw

3 mm :8: mm :82 mgfifi>

as n 5 355.82 A? u 5 Beam

 

.03 («o m3?» 2-: me :35 :2:3m-:o: ::m :25; .«o
3355» :35: :5 $05.: .853: mo 82.5 :3::Sm :5 3:2: :Bmsagbw/w A _.n 2an

160

CHAPTER 8
Subcutaneous Fat Distribution in Stunted and Non-Stunted Children from Oaxaca,

Mexico

Introduction

It is common to discuss obesity and its comorbidities in developed or
industrialized countries. However, the prevalence of overweight and obesity is increasing
in developing countries where previously undernourished populations are undergoing
rapid economic development, rural-to-urban migration, and nutritional changes
associated with development and western influences (Popkin et al., 1996, 2001;
Schroeder et al., 1999). There appears to be a paradox in developing countries. The
presence of chronic undernutrition, especially among preschool children, is apparently
continuing in the face of an increasing prevalence of overweight and obesity, specifically
in adolescents and adults. A rapid increase in the prevalence of “diet-related” diseases in
developing countries, in turn, has implications for costs of public health care that are
comparable in magnitude to the costs associated with chronic undernutrition (Popkin et
al., 2001).

Changes in the distribution of subcutaneous adipose tissue occur during puberty
and the adolescent growth spurt (Malina and Bouchard, 1988, 1991; Malina, 1996). Boys
tend to accumulate more subcutaneous fat on the trunk relative to the extremities during
puberty, which is due in part to a decrease in the thickness of extremity skinfolds at this
time. In contrast, girls gain proportionally similar amounts in both trunk and extremity

skinfold thicknesses during puberty. There is also ethnic variation in the relative

161

 

distribution of subcutaneous fat (Mueller, 1988; Malina et al., 1995; Malina 1996).
Compared to individuals of European ancestry, those of African, Asian, and Mexican
ancestry tend to have proportionally more subcutaneous fat on the trunk than on the
extremities.

It has been suggested that growth stunting during infancy and early childhood
may influence subcutaneous fat distribution in adolescence and adulthood. Data are
presently limited to adolescent Senegalese girls (Béne’fice et al., 2001a, 2001b) and
Guatemalan adults (Schroeder et al., 1999) who were stunted in early childhood.
Senegalese adolescent girls who were classified as stunted between 6 and 18 months of
age had proportionally thicker trunk skinfolds than those who were not stunted (Béne'fice
et al., 2001a, 2001b). Guatemalan young adults who were stunted by 36 months of age
had a proportionally greater accumulation of subcutaneous fat on the upper (higher waist-
hip ratios) body compared to those who were not stunted (Schroeder et al., 1999).

Bénéfice et al. (2001a) suggested that the trend to accumulate proportionally
more subcutaneous fat on the trunk “. . .may represent transitory changes in fatness under
the influence of hormonal stimulation during puberty” (p. 57). Sawaya et al. (1998)
proposed that hormonal changes as a consequence of early malnutrition may be a major
contributory factor to greater than expected weight gain in mildly stunted girls, 7-11
years of age, from a shanty town @fl) in 850 Paulo, Brazil. The authors suggested a
specific role for high-fat diets.

Corresponding data on the relative fat distribution of stunted children during

childhood are apparently not available. The present study thus compares the

162

subcutaneous fatness and relative subcutaneous fat distribution of stunted and non-

stunted rural and urban Mexican children, 6-11 years of age.

Methods and Procedures
The communities, sample, and anthropometric procedures were described earlier
(Chapter 3). A child whose height-for-age z-scOre was > 2 standard deviations below
age- and sex-specific medians of the United States reference (Centers for Disease Control
and Prevention, 2000) was considered stunted. Other children were considered non-
stunted. The prevalence of stunting in the sample of 6-11 year old children was
approximately 28% for both boys and girls in the rural community, and 15% for girls and
boys in the urban community (see Table 7.1).
Anthropometric variables included the present analysis were as follows:
1) Body size: Weight, height, body mass index (BMI);
2) Subcutaneous fatness: Triceps, subscapular, suprailiac and medial calf skinfolds;
sum of the four skinfolds; provided an estimate of overall subcutaneous fatness.
3) Relative subcutaneous fat distrii ution: Tmnk-thremity Ratio (TER) - ratio of the
sum of two trunk skinfolds (subscapular and suprailiac) to the sum of two
extremity skinfolds (triceps and medial calf).
The analyses proceeded in three steps. Descriptive statistics were initially
calculated for all variables. Sex-specific multivariate analyses of covariance
(MANCOVA), with age as the covariate, were used to compare the size and fatness

characteristics of stunted and non-stunted children within each community.

163

 

 

Principal components analysis (PCA), as described by Baumgartner et al. (1986)
and Malina et al. (1995), was also used to evaluate relative subcutaneous fat distribution.
The PCA was used to reduce the number of variables and possibly attain a smaller set of
variable components that might suggest a trend in the evaluation of relative fat
distribution (Rencher, 1995). The four skinfolds were transformed to natural logarithms
and the mean 10g skinfold was used as the index of subcutaneous fat for the individual.
Log values for skinfold thicknesses were then separately regressed on the mean log
skinfold thickness for each individual, thus controlling for overall subcutaneous fatness.
The residuals were subjected to the principal components analysis. The resulting
components provided an indication of relative subcutaneous fat distribution. Components
with eigen values greater than 1.0 were retained for further analysis. Component scores
of stunted and non-stunted boys and girls within each community were compared with

analyses of covariance (ANCOVA), with age as the covariate.

Results

Age-adjusted means and standard errors of the body size and fatness
characteristics of stunted and non-stunted children by sex within each community are
presented in Tables 8.1 and 8.2. Within each community, non-stunted children were
significantly larger in body size and skinfold thicknesses than stunted children with the
exception of the triceps and medial calf skinfolds in rural girls. On the other hand, the
TER did not significantly differ between stunted and non-stunted children except in rural
girls. Mean ratios were quite similar in stunted and non-stunted boys in each community,

but were higher in non-stunted than in stunted girls in each community. Note, however,

164

that the standard errors for the TER in stunted children of both sexes were considerably
larger than corresponding errors in non-stunted children.

Results of the principal components analysis are summarized in Table 8.3. Two
components with eigen values greater than 1.0 resulted and accounted for approximately
60% of the variance. The first principal component contrasted the two trunk skinfolds
(negative loadings) with the two extremity skinfolds (positive loadings), suggesting a
trunk-extremity component, which accounted for 32.2% of the variation in fat
distribution in the sample. The first component highlights high extremity fatness and low
trunk fatness. The second principal component had positive loadings for all four skinfold
and accounted for about 25% of the variance in fat distribution of the sample. The two
trunk skinfolds had moderately high positive and similar loadings, and the two extremity
skinfolds had low, positive loadings. It also suggested a trunk-extremity contrast with an
emphasis on trunk compared to extremity fatness.

Sex-specific age-adjusted means and standard errors for scores on the two
principal components in rural and urban, stunted and non-stunted children, respectively,
are presented in Table 8.4. Mean component scores did not differ significantly between

stunted and non-stunted children by sex within each community.

Discussion

Stunted children, both rural and urban, were shorter (as expected) and lighter, and
had a lower BMI and thinner skinfold thicknesses than non-stunted children in both
sexes. The TER, however, did not differ significantly between stunted and non-stunted

children, except among rural girls. The TER was, on average, higher in non-stunted boys

165

and girls, both urban and rural, which suggests proportionally more subcutaneous fat on
the trunk in non-stunted children. This contrasts observations based on skinfolds in
Senegalese adolescent girls (Be’néfice et al., 2001a, 2001b) and on the waist-hip ratio in
Guatemalan young adults (Schroeder et al., 1999) who were stunted in growth before
three years of age.

The variability in results among the three studies of samples who were stunted
may be due to several factors. The present study was based on a largely prepubertal
sample 6-11 years of age. Only two 11 year old girls, one urban and one rural, were post-
menarcheal. In well-nourished samples, individual and in particular sex differences in
relative fat distribution become established during adolescence (Malina, 1996). It is also
possible that the proportionally greater accumulation of subcutaneous fat on the trunk in
stunted children occurs during adolescence as suggested by Be'néfice et al. (2001a).

Another factor may be the indicators of relative subcutaneous fat distribution
used in the three studies. The present study was based on four skinfold thicknesses, two
on the trunk and two on the extremities, a trunk-extremity skinfold ratio, and principal
components analysis. The study of Senegalese adolescents also used skinfolds, but the
same skinfolds were not measured at all ages in the mixed-longitudinal sample (Be’néfice
et al., 2001a, 2001b), and a different analytical strategy was used. The study of
Guatemalan young adults was based on the waist-hip ratio (Schroeder et al., 1999). The
validity of the waist-hip ratio as an indicator of relative subcutaneous fat distribution has
not been established, particularly in adolescents. It may be more of an indicator of
primarily abdominal fatness in adults in contrast to fatness on the upper (subscapular)

and lateral (suprailiac) aspects of the trunk. Further, the waist-hip ratio is not the

166

appropriate indicator of relative subcutaneous fat distribution in children and adolescents
(Johnston, 1992; Sarria, 1992), and some of the subjects in the Guatemalan study were in
fact late adolescents.

Results of the present study and corresponding studies of Senegalese adolescents
and Guatemalan young adults emphasize the need for further study of the potential
effects of early growth stunting on relative subcutaneous fat distribution. Studies utilizing
more clinically relevant indicators of fat distribution, e.g., magnetic resonance imaging

and computerized axial tomography, in stunted and non-stunted children are warranted.

167

Bibliography

Baumgartner RN, Roche AF, Guo SM, Lohman T, Boileau RA, Slaughter MH. 1986.
Adipose tissue distribution: the stability of principal components by sex, ethnicity
and maturation stage. Hum Biol 58:719-735.

Bénéfice E, Garnier D, Simondon KB, Malina RM. 20013. Differences in growth and
body composition during puberty between Senegalese adolescent girls stunted or
not stunted in infancy. Biom Hum et Anthropol 19:55-61.

Bénéfice E, Garnier D, Simondon KB, Malina RM. 2001b. Relationship between
stunting in infancy and growth and fat distribution during adolesence in
Senegalese girls. Eur J Clin Nutr 55:50-58.

Centers for Disease Control and Prevention. 2000. National Center for Health Statistics

CDC growth charts: United States. http://www.cdc.gov/gowthchartshtm.

 

Johnston FE. 1992. Developmental aspects of fat patterning. In: Hernandez M, Argente J,
editors. Human growth: Basic and clinical aspects. Amsterdam: Elsevier Science.
p 217—226.

Malina RM. 1996. Regional body composition: age, sex, and ethnic variation. In: Roche
A, Heymsfield S, Lohman T, ed‘tors. Human body composition. Champaign, IL:
Human Kinetics. p 217-255.

Malina RM, Bouchard C. 1988. Subcutaneous fat distribution during growth. In:
Bouchard C, Johnston FE, editors. Fat distribution during growth and later health
outcomes. New York: Plenum. p 63-84.

Malina RM, Bouchard C. 1991. Growth, Maturation, and Physical Activity. Champaign,

IL: Human Kinetics.

168

Malina RM, Huang Y-C, Brown KH. 1995. Subcutaneous adipose tissue distribution in
adolescent girls of four ethnic groups. Int J Obes 191793-797.

Mueller WH. 1988. Ethnic differences in fat distribution during growth. In: Bouchard C,
Johnston FE, editors. Fat distribution during growth and later health outcomes.
New York: Plenum. p 127-145.

Popkin BM, Richards MK, Montiero CA. 1996. Stunting is associated with overweight in
children of four nations that are undergoing the nutrition transition. J Nutr
126:3009-3016.

Popkin BM, Horton S, Kim SW, Mahal A, Jin SG. 2001. Trends in diet, nurtitional
status, and diet-related noncommunicable diseases in China and India: the
economic costs of the nutrition transition. Nutr Rev 59:379-390.

Rencher AC. 1995. Principal component analysis. In: Methods of Multivariate Analysis.
NY: John Wiley and Sons, Inc. p 415-444.

Sarria A. 1992. Methods for assessing fit patterning in children. In: Hernandez M,
Argente J, editors. Human growth: Basic and clinical aspects. Amsterdam:
Elsevier Science Publishers. p 233-243.

Sawaya A, Grillo L, Verreschi L, Silva AD, Roberts S. 1998. Mild stunting is associated
with higher susceptibility to the effects of high fat diets: studies in a shantytown
population in S50 Paulo, Brazil. J Nutr 128:4158-420S.

Schroeder DG, Martorell R, Flores R. 1999. Infant and child growth and fatness and fat

distribution in Guatemalan adults. Am J Epidemiol 149:177-185.

169

Table 8.1. Age-adjusted means and standard errors of body size and indicators of fatness

and relative subcutaneous fat distribution of stunted and non-stunted boys, 6-11

years of age, within each community.

 

 

Stunted Non-Stunted
Variables Mean SE Mean SE p
Rural
Weight, kg 21.9 0.5 26.6 0.3 < 0.01
Height, cm 116.6 0.5 125.5 0.3 < 0.01
BMI, kg/mz 16.0 0.2 16.8 0.2 < 0.05
Triceps, mm 6.9 0.3 8.1 0.2 < 0.01
Subscapular, mm 5.7 0.3 7.0 0.2 < 0.01
Suprailiac, mm 7.2 0.5 8.9 0.3 < 0.01
Medial Calf, mm 7.0 0.3 8.3 0.2 < 0.01
Sum 4 SF, mm 26.8 1.2 32.2 0.8 < 0.01
TER, % 93.0 3.1 97.0 1.9 NS
Urban ‘

Weight, kg 22.5 1.1 28.5 0.4 < 0.01
Height, cm 117.4 0.8 127.0 0.3 < 0.01
BMI, kg/m2 16.3 0.5 17.4 0.2 < 0.05
Triceps, mm 7 .2 0.4 8.5 0.2 < 0.01
Subscapular, mm 6.1 0.6 7.4 0.2 < 0.05
Suprailiac, mm 8.5 0.6 10.0 0.2 < 0.05
Medial Calf, mm 7.5 0.4 8.7 0.2 < 0.05
Sum 4 SF, mm 29.2 1.7 34.6 0.7 < 0.01
TER, % 99.7 4.6 100.5 1.8 NS

 

Based on MAN COVA with age as the covariate. Sum 4 SF = Sum of 4 Skinfolds
(Triceps, Subscapular, Suprailiac, Medical Cali), TER = Trunk-Extremity Ratio
NS = Not Significant

170

Table 8.2. Age-adjusted means and standard errors of body size and indicators of fatness
and relative subcutaneous fat distribution of stunted and non-stunted girls, 6-11 years of

age, within each community.

 

 

 

Stunted Non-Stunted
Variables Mean SE Mean SE p
Rural
Weight, kg 22.5 0.6 27.6 0.4 < 0.01
Height, cm 117.7 0.5 126.9 0.3 < 0.01
BMI, kg/mz 16.1 0.3 16.9 0.2 < 0.05
Triceps, mm 8.7 0.3 9.3 0.2 NS
Subscapular, mm 7.5 0.4 8.8 0.3 < 0.05
Suprailiac, mm 9.4 0.4 10.7 0.3 < 0.05
Medial Calf, mm 9.1 0.4 9.8 0.2 NS
Sum 4 SF, mm 34.7 1.4 38.6 0.9 < 0.05
TER, % 93.3 2.7 101.6 1.7 < 0.05
Urban
Weight, kg 23.2 1.0 29.2 0.4 < 0.01
Height, cm 118.7 0.9 129.1 0.4 < 0.01
BMI, kg/m2 16.3 0.4 17.3 0.2 < 0.05
Triceps, mm 8.2 0.3 8.8 0.1 < 0.05
Subscapular, mm 7.0 0.5 8.6 0.2 < 0.01
Suprailiac, mm 8.9 0.5 10.4 0.2 < 0.01
Medial Calf, mm 8.5 0.4 9.6 0.2 < 0.01
Sum 4 SF, mm 32.5 1.4 37.5 0.6 < 0.01
TER, % 94.8 3.9 102.8 1.7 NS

 

Based on MANCOVA with age as the covariate. Sum 4 SF = Sum of 4 Skinfolds
(Triceps, Subscapular, Suprailiac, Medical Calf), TER = Trunk-Extremity Ratio
NS = Not Significant

l7l

Table 8.3. Summary of the principal components analysis: Loadings on the four skinfolds

on each principal component.

 

Principal Components

 

Skinfold PCl PC2

Triceps 0.748 0.281
Subscapular - 0.278 0.684
Suprailiac - 0.216 0.701
Medial Calf 0.777 0.169
Eigen Value 1.287 1.066
Variance, % 32.2 26.7

 

172

Table 8.4. Age—adjusted means and standard errors of scores on the two principal

components for stunted and non-stunted rural and urban children by sex.

 

 

PC] PC2
Stunted Non-Stunted Stunted Non-Stunted
M SE M SE M SE M SE
Males
Rural - 0.0030 0.15 0.0012 0.09 - 0.0009 0.16 0.0004 0.10
Urban - 0.0892 0.21 0.0133 0.08 - 0.0388 0.22 0.0058 0.09
Females
Rural 0.0252 0.12 - 0.0099 0.08 0.0088 0.13 - 0.0035 0.87
Urban - 0.0160 0.20 0.0031 0.09 - 0.0090 0.20 0.0017 0.09

 

173

CHAPTER 9

General Summary, Discussion and Conclusions

The present study examined the anthropometric and physique characteristics of
rural and urban children as correlates of health- and performance-related physical fitness.
It also considered the anthropometric, physique, fatness, and fitness characteristics of
stunted and non-stunted children in the rural and urban communities. In this chapter, the

results are discussed in light of the questions/hypotheses and findings of previous studies.

Setting and Methods

The setting of the study was two communities, one rural and the other urban,
in the Valley of Oaxaca in southern Mexico. Both communities were categorized as
rural and urban, respectively, based on the census on the national census for Mexico
in 2000.

The rural community is located about 23 km northwest of the city of Oaxaca.
It is a subsistence agriculture community in which a major segment of the population
speaks an indigenous language, Zapotec. The community has a health center that is
regularly staffed by a public health nurse and a physician. Essential utilities such as
sewage systems, water availability within each household, and water treatment,
though improved in recent years, are generally limited.

The urban community is located on the slopes of the hills west of the city of

Oaxaca. It is an irregular settlement on the edge of the city (colonias populares),

 

although historically it was an independent community (Murphy and Stepick, 1991).

174

Most of its population works in the city of Oaxaca. Facilities for health care and
sanitation are far from expectations of an urbanized center, but are in advance of
those available in the rural community.

The sample included children 6.00 to 11.99 years of age. Chronological ages
were calculated from date of measurement and birth dates taken from official school
enrollment records. The rural sample comprised 329 children, 154 boys and 175 girls,
and the urban sample included 318 children, 161 boys and 157 girls.

Anthropometric dimensions included weight, height, sitting height, four
skeletal breadths, three limb circumferences and four skinfold thicknesses. Derived
variables included the body mass index, estimated leg length, the sitting height to
standing height ratio, the sum of four skinfolds, the ratio of trunk to extremity
skinfold thicknesses, and estimated arm and calf muscle circumferences. Heath-
Carter anthropometric somatotypes were also derived.

The physical fitness battery included a combination of performance-related
tests, 35 yard (32.3 meter) dash, standing long jump, and health-related tests: sum of

right and left grip strength, sit and reach, timed sit-ups, distance run.

Discussion

The subsequent discussion is set in the context of the specific questions and

hypotheses of the study.

Question 1. What is the contribution of anthropometric dimensions to variation in the

performance of rural and urban children on physical fitness tasks?

175

This question examined the contribution of age, height, and weight, and of other
anthropometric indicators to variation in health- and performance-related physical fitness.
Based on multiple regression analyses, the variance explained by age, height and weight
ranged from 74% to 82% for strength, 38% to 57% for the dash, 31% to 40% for the
jump, 6% to 26% for the distance run, 9% to 24% for timed sit-ups and 1% to 6% of the
sit and reach. The addition of other anthropometric variables to the regression analyses
increased the explained variance in fitness tests only marginally, 2% to 14%. The
addition of other anthropometric dimensions to the analyses also altered the contributions
of age, height and weight to the explained variance in the fitness tests, which suggests
overlap or interactions among variables in their contributions to performance. The results
are generally consistent with previous studies of children in Mexico (Malina and
Buschang, 1985), Brazil (Rocha Ferreira et al., 1991), and Africa (Naidoo, 1999)
children, although different analytical procedures were used.

Results of the analyses did not support the hypotheses:

a) It was hypothesized that estimated arm muscle circumference would explain a
significant portion of the variation in static strength. The anthropometric dimensions that
significantly contributed to static strength were variable across the four samples of
children. The main significant contributor was the sum of skinfolds except in rural boys.
Estimated arm muscle circumference contributed significantly only to the variation in the
static strength of urban boys.

b) It was hypothesized that estimated calf muscle circumference and estimated leg
length would explain a significant portion of the variance in the 35 yard dash. Age was

the main significant contributor of variation in running speed across all the four samples

176

of children, and calf muscle contributed significantly only to the variance in the dash in
urban girls.

c) It was hypothesized that arm and calf muscle circumferences would explain a
significant portion of the variance in the standing long jump. Estimated arm and calf
muscle circumferences contributed significantly to the variation of explosive power only
in urban boys. Age was the main contributor to the variance in the jump except in rural
girls.

d) It was hypothesized that estimated leg length would explain a significant
portion of the variation in the sit and reach. Estimated leg length did contribute
significantly to the variation in the sit and reach in urban boys and rural girls, but none of
the anthropometric dimensions consistently contributed to the explained variance in
flexibility.

c) It was hypothesized that the two estimates of muscularity, arm and calf muscle
circumferences, would explain a significant portion of the variance in timed sit-ups.
Estimated arm muscle circumference contributed significantly to the variance in the
abdominal muscular endurance of boys, whereas estimated calf muscle circumference did
not contributed significantly to the explained variance in the four samples.

f) It was hypothesized that the sum of four skinfolds would explain a significant
proportion of the variance in the distance run. The sum of skinfolds contributed to the
variation in the distance run only in rural boys. The primary contributor to variation in the
distance run in urban boys and girls was age.

There was substantial overlap in the estimated contributions of anthropometric

characteristics to the six fitness tests. The hypotheses were thus not supported.

177

 

Question 2. Do the somatotypes of rural and urban children 6 to 11 years of age living
under marginal health and nutritional conditions in Oaxaca, Mexico, differ?

It was hypothesized that rural and urban children 6-11 years of age do not
significantly differ Heath-Carter anthropometric somatotypes. Results of the analysis
generally supported the hypothesis. Urban boys were significantly more endomorphic
than rural boys, whereas urban and rural girls did not differ in somatotype. This may
reflect differences in lifestyle. Rural boys were more active than urban boys. Urban boys
may also have had chronically excessive energy intake, but data are not available to
document this. In contrast, urban and rural girls did not differ in fatness and activity
(Pena Reyes, 2002). A previous study in Yucatan, Mexico, reported that children of both
sexes from better living conditions were more endomorphic than those living in
traditional, indigenous, subsistence agricultural communities (Murguia et al., 1990).

It is difficult to speak of environmental influences on the somatotype of children
due to lack of data on children and, perhaps due to limitations of the Heath-Carter
anthropometric protocol with children. Studies of adults indicate that extreme
environmental circumstances, such as semi-starvation in males (Lasker, 1947), disordered
eating in females (Malina, 1992; Malina et al., 2002), or resistance training in males
(Tanner, 1952), are needed to significantly modify somatotype. Changes in somatotype
associated with starvation and disordered eating can be reversed with an appropriate diet
(Lasker, 1947, Malina, 1992), whereas gains in mesomorphy with resistance training
return to pre-training values after the training program is stopped (Tanner, 1952).

A possible explanation may be the suggestion in the anthropological literature that

females are generally more resistant to environmental influences than males (McCance,

178

1966; Stinson, 1985). In the present case, environmental influences related to the
accumulation of subcutaneous fat may be relevant, i.e., chronically excessive energy
intake and/or chronically low levels of physical activity. Most activities of girls in both
the rural and urban communities revolve about food preparation and are of low to
moderate intensity. On the other hand, rural boys engage in regular moderate to vigorous
activity associated with household and agricultural chores, whereas urban boys do not

have such responsibilities (Pefia Reyes, 2002).

Question 3. What is the relationship between somatotype and the health- and
performance-related physical fitness of rural and urban children?

It was hypothesized that after controlling for age and the other two somatotype
components, endomorphy has a negative relationship to cardiovascular endurance
(distance run); mesomorphy has a positive relationship to measures of strength, speed,
power, and muscular endurance; and ectomorphy has a positive relationship to flexibility.
The hypothesis was generally supported in the partial correlation analysis. Endomorphy
was consistently and negatively correlated, whereas as mesomorphy and ectomorphy
were variably correlated with the fitness items included in the study. The results are
reasonably consistent with earlier correlational studies (Malina, 1975, 1992; Malina and
Rarick, 1973; Slaughter etal., 1977, 1980). There also were no consistent rural-urban
contrasts in the correlations.

. Age and somatotype contributed proportionally more to the variation in static
strength, dash, and standing long jump, with smaller contributions to three of the health-

related fitness items, sit and reach, timed sit-ups, and distance run. Endomorphy generally

179

 

had a negative contribution to fitness, while the contributions of mesomorphy and
ectomorphy were variable. Previous studies of somatotype and fitness of children did not
use multiple regression analyses; hence, comparisons are not possible. Results of the
present study, however, suggest a primary role for age and a lesser role for somatotype in
explaining the variance in the health— and performance—related fitness of children 6-11
years of age. The important role for age per se probably reflects the significant
contributions of neuromuscular maturation and experience to age-related improvement in
fitness tasks and perhaps the increase in body size. On the other hand, there are relatively
small changes in somatotype during childhood (Carter and Heath, 1990; Malina and
Bouchard, 1991), and in the present study there was no age effect in three of the four
samples.

The results may also reflect sampling variation per se, the small body size of the
children compared reference data for well-nourished children, and/or the relatively
narrow range of somatotypes in the sample compared to other studies based on samples
of European and North American children. The current samples were, on average,
consistently below the 25th percentiles of the United States reference data and were
consistently smaller than samples of urban school children in other areas of Mexico. The
range of variation in somatotypes was relatively reduced compared to studies of well-
nourished children. The reduced variation in somatotypes in the sample of rural and
urban children, more specifically, the lack of extreme somatotypes, may influence the
correlation and regression analyses, and in turn estimates of the contribution of
somatotype to performance. In studies of well-nourished children, the influence of

somatotype on performance is more apparent at the extremes of the distributions, e.g.,

180

 

extreme ectomorphs are deficient in muscle mass and strength (Malina and Rarick, 1973;
Malina, 1975). The issue needs further consideration in other samples from developing

countries.

Question 4. How do growth stunted and non-stunted children within each community
compare in body proportions, relative muscularity, subcutaneous fatness, relative
subcutaneous fat distribution, and physical fitness?

Growth stunting is defined as a height more than two standard deviations below
the age- and sex-specific reference for well-nourished children. A z-score was calculated
for each child relative to the Centers for Disease Control and Prevention (2000) growth
charts for the United States. Children with height-for-age z-scores below -2.0 were
classified as stunted. About 28% of the rural school children (49 boys, 43 girls) and 15%
of the urban school children (25 boys, 21 girls), 6-11 years of age, were stunted. All other
children were classified as non-stunted, 126 and 111 rural boys and girls, respectively,
and 132 and 140 urban boys and girls, respectively.

Three hypotheses were generated from the literature on stunting:

a) Stunted and non-stunted children 6-11 years of age do not in differ relative
muscularity and subcutaneous fatness, but do differ in relative body proportions related to
sitting height and leg length. This hypothesis was partially supported. Stunted children
had proportionally shorter legs than non-stunted children, i.e., the sitting height to
standing height ratio was higher. However, stunted children, both rural and urban, were
significantly smaller in all anthropometric dimensions and derived indicators compared to

non-stunted children. In the context of the hypothesis, stunted children were less

181

muscular and had less subcutaneous fat than non-stunted children. Bénéfice et al. (2001a,
2001b) reported similar observations in adolescent Senegalese girls who were stunted in
early childhood. Undemourished adults and children, who were not necessarily stunted,
are, on average, shorter and lighter, and have reduced muscle mass compared to well-
nourished age and sex peers (Spurr, 1984; Be’néfice et al., 1996, 1999; Hoffman etal.,
2000).

b) Non-stunted children 6-11 years of age perform better in tests of health- and
performance-related physical fitness compared to stunted children. This hypothesis was
also partially supported. Non-stunted children performed significantly better than stunted
children only in absolute grip strength and in grip strength per unit estimated arm muscle
circumference. These observations reflect the small body size and reduced muscle mass
of stunted children also suggest changes in muscle mass associated with stunting. The
reduced strength per unit arm muscle in stunted children may reflect, perhaps, qualitative
changes in muscle tissue associated with undernutrition early in life. Otherwise, non-
stunted children of both sexes did not perform significantly better than stunted children
on the other fitness tests. Results were variable for the dash and standing long jump. Of
interest, three indicators of health-related physical fitness, the sit and reach (flexibility),
timed sit-ups (abdominal muscular strength and endurance), and distance run
(cardiovascular endurance) did not differ significantly between stunted and non-stunted
children in all of the comparisons.

c) Stunted children 6-11 years of age have a proportionally greater accumulation
of subcutaneous fat on the trunk than on the extremities compared to non-stunted

children. Subcutaneous fat distribution was estimated as the ratio of the sum of the two

182

trunk skinfold thicknesses (subscapular + suprailiac) to the sum of the two extremity
skinfold thicknesses (triceps + medial calf) (TER), and scores on two factors based on
principal components analysis of the four skinfolds. Two components that contrasted
trunk and extremity skinfolds were identified, i.e., had eigen values >1.0.

Although stunted children, both rural and urban, had a lower BMI and thinner
skinfold thicknesses than non-stunted children, the TER was not significantly different,
except among rural girls. Scores on the two principal components also did not differ
significantly between stunted and non-stunted children. Hence, relative subcutaneous fat
distribution did not differ between stunted and non-stunted children, and the hypothesis
was rejected. The results contrast observations based on skinfolds in Senegalese
adolescent girls (Béne’fice eta1., 2001a, 2001b) and on the waist-hip ratio in Guatemalan
young adults (Schroeder et al., 1999), who were stunted in growth during early
childhood. The variability in results among studies may be related to several factors. The
present study was based on a sample of largely prepubertal children 6-11 years of age.
Only two 11 year old girls, one urban and one rural, were post-menarcheal. The
proportionally greater accumulation of subcutaneous fat on the trunk in stunted children
may occur during adolescence and into adulthood (Bénéfice et al., 2001a). Another factor
may be the anthropometric indicators of relative subcutaneous fat distribution used in the
three studies. The present study was based on four skinfold thicknesses, two on the trunk
and two on the extremities. The study of Senegalese adolescents also used skinfolds, but
the same skinfolds were not measured at all ages in the mixed-longitudinal sample
(Be'ne'fice et al., 2001a, 2001b), and a different analytical strategy was used. The study of

Guatemalan young adults was based on the waist-hip ratio (Schroeder et al., 1999). The

183

validity of the waist-hip ratio as an indicator of relative subcutaneous fat distribution has
not been established. It may be more of an indicator of primarily abdominal fatness in
contrast to fatness on the upper (subscapular) and lateral (suprailiac) aspects of the trunk.
Further, the waist-hip ratio is not the appropriate indicator of relative subcutaneous fat
distribution in children and adolescents (Johnston, 1992; Sarria, 1992), and some of the

subjects in the Guatemalan study were in fact late adolescents.

Conclusions

Based on the results of the present study, several conclusions are warranted. First,
the anthropometric correlates of physical fitness do not differ between urban and rural
children of both sexes 6-11 years of age. Age, height and weight account for a large
preportion of the variance in the six fitness tasks. The addition of other dimensions to the
regression analysis adds relatively little to the explained variance, but alters the estimated
contributions of age, height and weight.

Second, rural and urban girls 6-11 years of age do not differ in somatotype, but
urban boys are more endomorphic than rural boys, perhaps reflecting lifestyle differences
associated with habitual physical activity and diet.

Third, the contribution of somatotype to the explained variance in the six fitness
tasks does not differ between rural and urban children.

Fourth, stunted children 6-11 years of age are smaller in anthropometric
dimensions, but have proportionally shorter legs than stunted children of the same age.

Fifth, non-stunted children are absolutely stronger and have greater strength per

unit estimated arm muscle than stunted children.

184

Sixth, stunted and non-stunted children do not differ in tests of performance- and
health-related fitness, with the exception of static strength.
Seventh, stunted and non-stunted children 6-11 years of age do not differ in

relative subcutaneous fat distribution.

Recommendations

The following are recommended for further study:

First, expansion of the performance protocol to include the movement patterns
associated with each task. For example, field observations suggest that many children
appeared to perform the standing long jump with a seemingly immature movement
pattern. Such an approach may afford insights into how marginal health and nutritional
conditions influence the development of specific movement patterns. This suggestion
also implies that future study should be extended into the preschool ages.

Second, validation of the Heath-Carter anthropometric somatotype protocol to
other samples of children living under marginal health and nutritional conditions in
different cultural groups, i.e., cross-culturally.

Third, results of studies of children in different cultural contexts suggest that age,
height and weight are primary factors affecting performances on a variety of tasks among
children 6-11 years of age. There is a need for further study of the contributions of age,
height and weight, and other anthropometric characteristics to variation in performance,
such as models incorporating interaction terms, or preliminary factor analysis to reduce

the number of specific anthropometric.

185

Fourth, given the difference between stunted and non-stunted children in absolute

 

strength and strength per unit estimated arm muscle mass, there is a need to consider
changes in muscle tissue associated with growth stunting early in life. The use of non-
invasive irnaging techniques may provide some insights.

Fifth, changes in relative fat distribution associated with growth stunting need
further study. If factors that regulate fat distribution are altered by chronic undernutrition
and specifically growth stunting in early childhood, when do the changes in fat
distribution occur? The use of non-invasive imaging techniques should provide valuable

insights into the study of relative fat distribution associated with growth stunting.

186

Bibliography

Bénéfice E, Fouéré T, Malina RM. 1999. Early nutritional history and motor performance
of Senegalese children, 4-6 years of age. Ann Hum Biol 26:443—455.

Bénéfice E, F ouéré T, Malina RM, Beunen G. 1996. Anthropometric and motor
characteristics of Senegalese children with different nutritional histories. Child:
Care, Health Develop 22:151-165.

Bénéfice E, Garnier D, Simondon KB, Malina RM. 2001a. Relationship between stunting
in infancy and growth and fat distribution during adolescence in Senegalese girls.
Eur J Clin Nutr 55:50-58.

Bénéfice E, Garnier D, Simondon KB, Malina RM. 2001b. Differences in growth and
body composition during puberty between Senegalese adolescent girls stunted or
not stunted in infancy. Biom Hum et Anthropol 19:55-61.

Carter JEL, Heath BH. 1990. Somatotyping - Development and Application. Cambridge:
Cambridge University Press.
Centers for Disease Control and Prevention. 2000. National Center for Health Statistics
CDC growth charts: United States. http://www.cdc.gov/growthcharts.htm.
Hoffman DJ, Sawaya AL, Coward WA, Wright A, Martins PA, de Nascimento C, Tucker
KL, Roberts SB. 2000. Energy expenditure of stunted and non-stunted boys and
girls living in the shantytowns of S50 Paulo, Brazil. Am J Clin Nutr 72:1025-
1031.

Johnston FE. 1992. Developmental aspects of fat patterning. In: Hernandez M, Argente J,
editors. Human growth: Basic and clinical aspects. Amsterdam: Elsevier Science.

p 217-226.

187

 

 

Lasker GW. 1947. The effects of partial starvation on somatotype. Am J Phys Anthropol
5:323-341.

Malina RM. 1975. Anthropometric correlates of strength and motor performance. Exerc
Sport Sci Rev 3:249-274.

Malina RM. 1992. Physique and body composition: effects on performance and effects
on training, semistarvation. and overtraining. In: Brownell KD, Rodin J, Wilmore
JH, editors. Eating, Body Weight and Performance in Athletes: Disorders of
Modern Society. Philadelphia: Lea & Febiger. P 94-111.

Malina RM, Battista RA, Siegel SR. 2002. Anthropometry of adult athletes: Concepts,

methods and applications. In: Driskill JA, Wolinsky I, editors. Nutritional

Assessment of Athletes. Boca Raton, FL: CRC Press. p 135-175.

Malina RM, Bouchard C. 1991. Growth, Maturation, and Physical Activity. Champaign,
IL: Human Kinetics.
Malina RM, Buschang PH. 1985. Growth, strength and motor performance of Zapotec

children, Oaxaca, Mexico. Hum Biol 57:163-181.

Malina RM, Rarick GL. 1973. Growth, physique, and motor performance. In: Rarick GL,
editor. Physical Activity: Human Growth and Development. New York:

Academic Press. p 124-153.

McCance RA. 1966. The effects of normal development and of undernutrition on the
growth of the calcified tissues. Tijd Soc Geneesk 44:569-5 73.
Murguia R, Dickinson F, Cervera M, Ligia UC. 1990. Socio-economic activities, ecology

and somatic differences in Yucatan, Mexico. Stud Hum Ecol 9:111-134.

188

Murphy AD, Stepick A. 1991. Social Inequality in Oaxaca: A History of Resistance and
Change. Philadelphia: Temple University Press.

Naidoo RB. 1999. The effects of socioeconomic status, ethnicity, and nutritional status on
the growth and physical fitness of 10 year old south African boys. Doctoral

Dissertation, Michigan State University.

Pefia Reyes ME. (2002). Growth status and physical fitness of primary school children in
an urban and a rural community in Oaxaca, Southern Mexico. Doctoral Dissertation,
Michigan State University.

Rocha Ferreira MB, Malina RM, Rocha LL. 1991. Anthropometric, functional and
psychological characteristics of eight-year-old Brazilian children from low
socioeconomic status. In: Shephard RJ, Parizkova J, editors. Human Growth,
Physical Fitness and Nutrition. Basel: S. Karger. p 109-118.

Sarria A. 1992. Methods for assessing fat patterning in children. In: Hernandez M,
Argente J, editors. Human growth: Basic and clinical aspects. Amsterdam:
Elsevier Science. p 233-243.

Schroeder DG, Martorell R, Flores R. 1999. Infant and child growth and fatness and fat
distribution in Guatemalan adults. Am J Epidemiol 149:177-185.

Slaughter MH, Lohman TG, Misner JE. 1977. Relationship of somatotype and body
composition to physical performance in 7-to-12-year-old boys. Res Quart 48:159-

168.

Slaughter MH, Lohman TG, Misner JE. 1980. Association of somatotype and body
composition to physical performance in 7-12 year—old-girls. J Sports Med 20: l 89-

198.

189

Spurr GB. 1984. Physical activity, nutritional status, and physical work capacity in
relation to agricultural productivity. In: Pollitt E, Amante P, editors. Energy

Intake and Activity. New York: Alan R. Liss. p 207-261.

Stinson S. 1985. Sex differences in environmental sensitivity during growth and

development. Yrbk Phys Anthropol 28: 1 23- 147.

Tanner JM. 1952. The effects of weight-training 0n physique. Am J Phys Anthropol

102427-462.

190

 

APPENDIX A

191

MICHIGAN STATE

UNIVERSITY
March29,2002

TO: Robert MALINA
128 IM Sports Circle

RE: IRB # 98-118 CATEGORYz- 2-8 EXPEDITED
RENEWAL APPROVAL DATE: March 29, 2002

TITLE: SECULAR CHANGE IN SIZE. STRENGTH AND MOTOR FITNESS IN THE VALLEY OF
OAXACA, MEXICO

The University Committee on Research Involving Human Subjects' (UCRIHS) review of this project
is complete and I am pleased to advise that the rights and welfare of the human subjects appear to
be adequately protected and methods to obtain informed consent are appropriate. Therefore. the
UCRIHS APPROVED THIS PROJECTS RENEWAL.

RENEWALS: UCRIHS approval is valid for one calendar year, beginning with the approval date
shown above. Projects continuing beyond one year must be renewed with the green renewal form. A
maximum of four such expedited renewal are possible. Investigators wishing to continue a project
beyond that time need to submit it again for complete review.

REVlSlONS: UCRIHS must review any changes in procedures involving human subjects. prior to
initiation of the change. If this is done at the time of renewal, please use the green renewal form. To
revise an approved protocol at any other time during the year, send your written request to the
UCRIHS Chair, requesting revised approval and referencing the project's IRB# and title. include in
your request a description of the change and any revised instruments, consent forms or
advertisements that are applicable.

PROBLEMS/CHANGES: Should either of the following arise during the course of the work. notify
UCRIHS promptly: 1) problems (unexpected side effects, complaints. etc.) involving human subjects
or 2) changes in the research environment or new information indicating greater risk to the human
subjects than existed when the protocol was previously reviewed and approved.

if we can be of further assistance, please contact us at 517 355-2180 or via email:
UCRIHS@pilot.msu.edu.

 

orncs or
32812113011 Sincerely.
ETHICS AND
STANDARDS M
Ilmulty Committee on

m 'mm' Ashir Kumar, MD.

Mm Subjects
UCRIHS Chair
w Sb University -
202 Old: Hall
m using. Ml
48824 AK: bd

517355-218)
Fm 517/4324503 cc: Swee Kheng Tan
museum/1m 128 N Sports Circle
EM W

“Wham
mam
mm '92
mum
WNW.

APPENDIX B

193

Protocols for Anthropometry and Fitness Testing

(Adapted from Pefia Reyes, 2002)

Anthropometry

1)

2)

3)

4)

Body mass (weight) was measured in kilograms using a portable scale
and recorded to the nearest 100 grams. Weight was measured without
shoes, but the children were light clothing with all accessories
removed (sweaters, sweatshirts, jackets, etc.).
Stature (standing height) was measured with the child standing erect -
posture, without shoes and with body weight evenly distributed
between both feet. The heels were placed together and the arms were
hanging relaxed at the sides. Height is the distance from the floor to
the top of the head positioned in the Frankfort horizontal plane.
Measurements were recorded to the nearest 0.1 centimeter (cm).
Sitting height corresponds to the distance from the table (sitting
surface) to the highest point at the top of the head (vertex). The
individual sat erect on the table with the knees hanging freely and the
hands positioned on the thighs. Sitting height was measured to the
nearest 0.1 cm.
Skeletal Breadths: Four skeletal breadths were measured to the nearest
0.1 cm.

a) Biacromial breadth was measured as the distance between the

left and right acromial processes of the scapulae with

194

 

b)

d)

application of firm pressure. The measurement was taken
from the rear.

Bicristal breadth was measured as the distance between the
most lateral points of iliac crests with application of firm
pressure. The measurement was taken from the rear.
Bicondylar breadth was measured as the distance across the
most medial and most lateral points of the femoral condyles.
The individual was seated with the knee flexed at 90°.
Biepicondylar breadth was measured as the distance between

the epicondyles of the humerus with the elbow flexed at 90°.

5) Limb Circumferences: Three limb circumferences were measured and

recorded to the nearest 0.1 cm.

a)

b)

Relaxed arm circumference was measured at the level midway
between the olecranon and acromial processes with the arm
handing loosely at the side. The tape was placed in the
horizontal position in contact with the skin but without
compression of the underlying soft tissues.

F lexed arm circumference was taken at the same level as
relaxed arm circumference, but the arm was flexed to a right
angle at the elbow.

Calf circumference was measured as the maximum
circumference of the calf with the subject in a standing

position and body weight evenly distributed between both

195

legs. The measuring tape was placed in a horizontal plane in
contact with the skin but without compressing the soft tissues.
6) Skinfolds: Four skinfolds were measured to the nearest 0.5 m using
a Lange skinfold caliper (as was used in the earlier studies in Oaxaca).
A double fold of skin and underlying soft tissue was raised with the
thumb and index finger of the left hand about one cm above the
specific site for each fold and the caliper was applied to the site.
a) Triceps skinfold was measured on the back of the arm over the
triceps muscle at the same level as relaxed arm circumference.
b) Subscapular skinfold was measured on the back immediately
beneath the inferior angle of the scapula following the natural
cleavage line of the skin.
c) Suprailiac skinfold was measured immediately above the iliac
crest in the midaxillary line.
(1) Calf skinfold was measured on the medial aspects of the calf at

the same level as calf circumference.

Physical Fitness

Physical fitness items included a combination of motor-and-health-related
tasks. Four tasks were used in an earlier study in Santo Tomas — right and left
grip strength, 35- yard dash, and standing long jump. Three additional health-

related fitness tests were added.

196

1)

2)

3)

4)

Static grip strength of the right and left hand was measured with a
Stoelting adjustable dynamometer. The subject held the
dynamometer in the line with the forearm at the level of the thigh.
Children were instructed to squeeze the dynamometer as
vigorously as possible so as to exert maximum force. Three trials
were administerd for each hand, the protocol alternating trials
between hands. The best trial with each hand was retained for the
analysis (Malina and Buschang, 1985).

The 35-yard dash (32.3 meters), an indicator of running speed,
was measured as the time elapsed from the starting signal to
crossing the finish line. The children ran on a flat concrete surface
on the respective school grounds. The elapsed time was recorded
to the 0.1 second. Two trials were administered, and the better of
the two was retained for the analysis. A sufficient rest period
between trials was included in the protocol (Malina and Buschang,
1985)

The standing long jump, a measure of power, was measured as the
distance from the take-off line to the point where the heels
touched the ground. The distance was measured to the nearest cm.
Three trials were administered and the best of the three was
retained for the analysis (Malina and Buschang, 1985).

The sit-and-reach, an indicator of lower back and upper thigh

flexibility, was measured as follows. Subjects were permitted to

197

5)

warm-up before the test by performing slow stretching
movements. The subject removed his/her shoes and was seated at
the test apparatus with the legs fully extended. The heels were
approximately shoulder width apart and the feet were flat against
the end board of the apparatus. The subjects were instructed to
extend the arms forward as far as possible with the hands placed
on top of each other. The subject leaned forward extending the
finger tips, with palms facing downward, as far alongthe ruler as
possible without jerking or bouncing. This effort pushed a sliding
marker along the scale. The child stretched forward along the top
of the box four times. On the fourth attempt, he/she held the
stretch for at least one second. The score was the farthest point
reached on the last stretch, which was recorded to the nearest
centimeter (Safrit, 1995). Three trials were recorded and the best
of the three was retained for the analysis.

Sit-ups, a measure of abdominal strength and endurance, were
measured as follows. The subject was in a supine position with the
knees bent at right angles and the feet approximately shoulder
width apart. The hands were placed at the side of the head with the
fingers over the ears and with the elbows pointed toward knees.
The hands and elbows had to be maintained in this position for the
entire duration of the test. The participant's ankles were held

throughout the test by the appraiser to ensure that heels were in

198

6)

constant contact with the mat. The subject was required to sit up,
touch the knees with the elbows, and return to the starting
positions (shoulder touching the mat). Each subject performed as
many sit-ups as possible within 30 seconds. Subjects were
permitted to pause for rest whenever necessary. Clear instructions
were given. Bouncing was not allowed and the subject’s buttocks
had to remain in contact with the mat. In curling down after the sit
up, the lower back had to be in contact with the mat before the
upper back and shoulders touched. One trial was given.

The 12 minute run/walk was used as a measure of cardiovascular
fitness. The subjects ran or walked for 12 minutes in an area at the
urban school or the central plaza of the rural community, which
was previously laid out with posts marking the limits and distance
covered. The number of laps ran was counted by the research
assistants. A warm-up and cool-down was provided done before
and after the test, respectively. Participants ran in groups of about
five. The distance ran was recorded to the nearest meter by the
research assistants. After testing several first grade children, it
was noted that younger children had a difficult time trying to
complete the 12 minutes walk/run. Therefore, the test was
modified for children in the younger ages, corresponding to
children at the 1St to 3'dgrades. The walk/run time was reduced

from 12 to 8 minutes. Only one trial was administered and the

199

children were verbally encouraged throughout the test to complete

the task.

200

   

111111 [11111101111111