Citation
Growth patterns of rural and urban youth in northern Tanzania

Material Information

Title:
Growth patterns of rural and urban youth in northern Tanzania
Creator:
Wulff, Raymond Jerry
Publication Date:
Language:
English
Physical Description:
xiii, 100 leaves : ; 28 cm

Subjects

Subjects / Keywords:
Youth -- Growth -- Tanzania ( lcsh )
Maasai (African people) ( lcsh )
Rural youth -- Growth -- Tanzania ( lcsh )
Urban youth -- Growth -- Tanzania ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 96-100).
General Note:
Department of Anthropology
Statement of Responsibility:
Raymond Jerry Wulff.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
268661312 ( OCLC )
ocn268661312
Classification:
LD1193.L43 2008m W84 ( lcc )

Full Text
GROWTH PATTERNS OF RURAL AND URBAN YOUTH
IN NORTHERN TANZANIA
by
Raymond Jerry Wulff
B.S., University of New Mexico, 2004
A thesis submitted to the
University of Colorado Denver
in partial fulfillment
of the requirements for the degree of
Masters of Arts
Anthropology
2008


This thesis for the Master of Arts degree
by
Raymond Jerry Wulff
has been approved by
OB'/oz/zooS
Date


Wulff, Raymond J. (M.A., Anthropology)
Growth Patterns of Rural and Urban Youth in Northern Tanzania
Thesis directed by Associate Professor David P. Tracer
ABSTRACT
Numerous studies have documented differences in growth trajectories
between rural and urban populations. Most identify differential access to health care
and nutritional resources as contributors to these differences. It is unclear whether
environmental and cultural practices may mediate these contributors and impact
growth across the ages. The objective of this project is to examine growth patterns in
rural Maasai pastoralist youth ages 7 through 16 who inhabit the Northern
Ngorongoro highlands of Tanzania, asking whether rural environmental and cultural
practices may contribute to variation in growth trajectories.
Anthropometric measurements (standing height, weight and BMI) of 717 rural
youth were contrasted with measurements from 327 urban dwelling youth residing in
the city of Mwanza in summer, 2006, to ascertain whether differences exist between
rural and urban dwelling youth. In addition, the mean growth statistics of the rural
Maasai youth were compared with published data sets from two East African
nomadic pastoralist groups to evaluate uniformity in growth among the populations.
Results indicated that rural and urban growth trajectories differed significantly,


but only at specific ages. For both male and female youth, height, weight and BMI
curves were largely indistinguishable from one another between the ages of 5 and 10
years, followed by growth faltering in the rural but not urban males and females, from
age 11 through age 15. This growth faltering in the rural sample coincides with
stressful local rites of passage that begin as youths reach their 11th year.
We suggest that the observed growth faltering in part could be attributable to
the bio-cultural influence of liminality ambiguous and stressful transitional social
states on individuals development and physical well-being. The growth patterns
discovered in this study appear to be consistent with the original premise that growth
of rural pastoralist youth is influenced by their unique lifestyle behaviors and
environment.
This study provides a foundation for establishing growth norms for Tanzania,
and for more in depth examination of lifestyle behaviors and cultural practices that
may impact child health and well-being in rural and urban environments.
This abstract accurately represents the content of the candidates thesis. I recommend
its publication.
OH
Signed
David P. Tracer


DEDICATION
To Ginny Laadt, who encouraged and supported me throughout this process. I also
dedicate this thesis to the youth of Tanzania and the school personnel who so
graciously allowed me to obtain the data for this study.


ACKNOWLEDGMENT
My thanks and gratitude to my advisor David P. Tracer, to Charles M. Musiba and
Robert S. Walker for their unwavering guidance and accessibility thought out this
process. Special thanks is given to Dr. Cassian C. Magori, Dean of Anatomy and
Histology of the Weill-Bungando University College of Health Sciences Center in
Mwanza, Tanzania for his ongoing help and support. I would also like to thank the
schools with which I interacted: Endulen, Misigio, Ndiyan, Mlimani, and Mongela
primary schools. Each Tanzanian school headmistress, Teddy, Elizabeth, Margaret
and Rose was invaluable. In addition, I thank my fellow field school students from
the University of Colorado Denver and the University of Calgary for their help in
acquiring data for this thesis. Finally, to Connie Turner of the Anthropology
Department, thank you most sincerely for your encouragement.


TABLE OF CONTENTS
Figures..........................................................x
Tables...........................................................xii
CHAPTER
1. INTRODUCTION...................................................1
2. BACKGROUND.....................................................6
Comparisons among Rural Pastoralists and Urban Dwellers.....6
Comparisons among Non-settled vs. Settled Pastoralists.....12
Comparisons among East African Pastoralists................14
3. STUDY POPULATION..............................................17
Nomadic Pastoralists.......................................17
Rural Nomadic Pastoralist Study Population..........22
Endulen Village Area..........................22
Endulen Primary School........................23
Misigio and Nydian Primary Schools............23
Urban Study Population.....................................23
Mwanza Area.........................................23
Mlimani Primary Schools.......................24
East African Pastoralists..................................25
Eyasi Datoga
.25


Ngisonvoka Turkana
27
4. HYPOTHESES AND RESEARCH DESIGN.................................29
Hypotheses and Predictions..................................29
Research Design.............................................32
5. MATERIALS AND METHODS..........................................34
Endulen Area Rural Population Sample........................35
Mwanza Urban Population Sample..............................36
Samples from Published Sources..............................37
Data Collection.............................................38
Measurement Variables.......................................40
Statistical Analysis........................................41
Data Reliability............................................43
6. RESULTS........................................................45
Hypothesis 1 Results........................................45
Hypothesis 2 Results........................................56
Hypothesis 3 Results........................................63
7. DISCUSSION.....................................................73
Growth and Environmental Differences between Rural Pastoralists vs.
Urban Dwellers and Non-settled vs. Settled Rural
Pastoralists................................................73
Preadolescent Dip or Nadir...........................76
viii


LiminalityLifestyle Transition from Youth to Adult
| Status..................................................77
| Growth Trends in Rural Non-settled vs. Settled Pastoralists....82
|
! School Lunches..........................................85
Physical Activity.......................................85
| Genetic Variation.......................................87
| Comparisons among East African Pastoralists....................87
Conclusion.....................................................93
BIBLIOGRAPHY.................................................................96
1
IX


LIST OF FIGURES
Figure
3.1 Map of Tanzania showing area of study..................................21
3.2 Map of Lake Eyasi Region...............................................26
3.4 Map of South Turkana region, Kenya.....................................28
4.1 Research Design........................................................32
6.1 Male Rural and Urban Standing Height for Age Z-Scores..................47
6.2 Male Rural and Urban Weight for Age Z-Scores...........................47
6.3 Male Rural and Urban BMI for Age Z-Scores..............................48
6.4 Female Rural and Urban Standing Height for Age Z-Scores................52
6.5 Female Rural and Urban Weight for Age Z-Scores.........................52
6.6 Female Rural and Urban BMI for Age Z-Scores............................53
6.7 Male Non-settled and Settled Standing Height for Age Z-Scores..........57
6.8 Male Non-settled and Settled Weight for Age Z-Scores...................57
6.9 Male Non-settled and Settled BMI for Age Z-Scores......................58
6.10 Female Non-settled and Settled Standing Height for Age Z-Scores........60
6.11 Female Non-settled and Settled Weight for Age Z-Scores.................61
6.12 Female Non-settled and Settled BMI for Age Z-Scores....................61
6.13 Male East African Pastoralists Mean Standing Height (cm)...............65
6.14 Male East African Pastoralists Mean Weight (kg)........................65
- x -


6.15 Male East African Pastoralists Mean BMI...........................................66
6.16 Female East African Pastoralists Mean Standing Height (cm)..................69
6.17 Female East African Pastoralists Mean Weight (kg)...........................69
6.18 Female East African Pastoralists BMI..........................................70
- xi -


LIST OF TABLES
Table
3.1 Urban parental occupation....................................................25
5.1 Rural primary school cross-sectional anthropometric data.....................36
5.2 Summary rural primary school cross-sectional data............................36
5.3 Urban primary school cross-sectional anthropometric data.....................37
5.4 Urban cross-sectional data summary...........................................37
6.1 Male rural pastoralist and urban dwelling descriptive data...................46
6.2 Male rural pastoralist and urban dwelling statistical analysis...............50
6.3 Female rural pastoralist and urban dwelling descriptive data..............51
6.4 Female rural pastoralist and urban dwelling statistical analysis..........54
6.5 Male rural non-settled and settled descriptive data..........................56
6.6 Male rural non-settled and settled statistical analysis......................59
6.7 Female rural non-settled and settled descriptive data.....................59
6.8 Female rural non-settled and settled statistical analysis.................62
6.9 Male and female East African pastoralists descriptive data...................64
6.10 Male East African pastoralist one-way ANOVA..................................66
6.11 Parametric/non-parametric test summary-Male East African
comparison...................................................................67
6.12 Male East African pastoralist Post Hoc test results........................68


6.13 Female East African pastoralist one-way ANOVA..........................70
6.14 Parametric/non-parametric test summary-Female East African
comparison..............................................................71
6.15 Female East African pastoralist Post Hoc test results................71
7.1 Comparison of latitude, elevation and low-high annual rainfall........92
- xiii -


CHAPTER 1
INTRODUCTION
Human growth and development reflect ongoing interactions among the
biology of our species, the physical environment in which we live, and the socio-
cultural environment. The basic pattern of modem growth shared by all humans is
founded in our evolutionary history. As such, human growth and development
reflects the bio-cultural nature and evolutionary history of our species (Bogin, 1999).
Adult human morphology, physiology and behavior are plastic; modifiable to
different degrees in response to changes in the environment, particularly when these
changes are stressful to the organism (Bogin, 1999). When human biology and
behavior are considered together (a bio-cultural perspective) it appears that human
beings are perhaps the most plastic or adaptive of all species. In part, this is related to
the human lifespan being long relative to most other animals, especially during
formative years where a long juvenile period allows for greater environmental impact.
The human adult phenotype is realized after many years during which a variety of
factors combine to influence the final outcome (Bogin, 1999).
The long process of human growth, driven by the interplay between genetics
and the environment, determines the specific path by which individual morphology is
expressed. Differences between populations result from these genetic and
environmental differences, and in the interactions between the two (Eveleth & Tanner,


1990). Because humans have ongoing contact with and impact upon climate, land,
plant and animal species, and these environmental elements have a reciprocal impact
on humans; it is proposed that ecological differences contribute to human phenotypic
variation.
Numerous studies have examined growth and development trends between
rural pastoralist and urban dwellers in sub-Saharan Africa and reflect their differing
lifestyles (Aspray et al., 2000; Carlin et al., 2001; L.R. Pawloski, 2002; Sellen, 1999;
D. W. Sellen, 2000; Shell-Duncan & Obiero, 2000). Studies of child and adolescent
growth and development reflect morphological expressions that are reported to be the
result of differing nutrition and physical activity (L.R. Pawloski, 2002; Sellen, 1999;
Shell-Duncan & Obiero, 2000). Accordly, malnutrition, diminished growth, and
subsequent growth stunting reflect the marginal living conditions of rural agricultural
and pastoral youth. In contrast, urban youth tend to follow a more accelerated growth
curve trajectory as a result of their greater and more consistent access to nutritional
intake and a more sedentary lifestyle (L.R. Pawloski, 2002; Shell-Duncan & Obiero,
2000).
Child and adolescent growth have long been a focus of anthropological
research because of their unique sensitivity to environmental conditions. This
research examines growth patterns of rural nomadic Maasai pastoralist youth
inhabiting the Ngorongoro highlands in the eastern part of the Serengeti Plains,
2


within the Endulen Ward, and compares it with growth among urban dwelling youth
inhabiting the city of Mwanza on the southern shores of Lake Victoria in Tanzania,
East Africa. In addition, growth differences in height, weight, and BMI are examined
within the rural Maasai (between the rural non-settled and rural settled Maasai
pastoralist youth). This rural data sample provides the opportunity to examine if the
provision of school lunches affords any growth advantage or any number of other
factors that may differ between the two samples.
The specific aims of this study are:
1) to examine if growth patterns differ between the rural nomadic Maasai
pastoralist inhabiting Endulen Village and urban dwelling youth living in
Mwanza, Tanzania; and to explore what bio-cultural factors are intrinsic to
rural and urban habitation, and thus might contribute to any differences found.
2) to examine if growth patterns differ between two rural sub-groups, the rural
non-settled nomadic Maasai pastoralists and rural settled Maasai pastoralists
youth inhabiting Endulen Village, Tanzania; and to explore what bio-cultural
factors, resulting from rural non-settled / rural settled habitation, might
contribute to any differences found.
3) to examine if growth patterns differ between the rural Maasai pastoralists of
Endulen Ward, and other previously reported studies on African pastoral
3


populations such as the Datoga of Lake Eyasi in Tanzania and the Turkana of
Lake Turkana in Kenya.
Many prior studies of rural sub-Saharan African growth trends compare their
growth results to available data sets such as World Health Organization (WHO),
National Center for Health Statistics (NCHS) and/or National Health and Nutrition
Examination Survey (NHANES) (M. A. Little & Gray, 1990; M. A. Little & Johnson,
1987; L.R. Pawloski, 2002; D. W. Sellen, 2000; Shell-Duncan & Obiero, 2000).
These large data sets of published normal growth trajectories portray standard growth
curves between various populations. This Maasai study found the occurrence of a
unique growth faltering in the rural Maasai population (male and female) that is not
reflected in growth data of other East African pastoralist or other published research.
This growth faltering, a leveling or negative direction appears in male Maasai
pastoralists between ages 10 through 13 and in female Maasai pastoralists between
ages 9 through 12. Further, I explore what factors might contribute to these unique
growth trajectories (especially diet, nutrition level, physical activity, lifestyle and
cultural behavior). Of special interest are the effects of liminality, particularly the
Maasai pastoralist rite of passage, culminating in the ceremony of genital cutting.
Employing a bio-cultural perspective to examine the forces that drive decisions that
affect diet, nutrition, and physical activity levels provides deeper inquiry than the
traditional focus employed in todays research. This bio-cultural prospective
4


provides for the recurring interaction between the biology of human development and
the socio-cultural environment (Bogin, 1999). Accordingly, a more in depth
examination of the human growth process, indeed all of human biology must be from
a bio-cultural perspective in order to fully understand the interactions between
humans and the environment.
Human growth and development as a bio-cultural process demands an
integrated bio-cultural analysis of the intrinsic and extrinsic factors that regulate
individual development. The interrelatedness of biology and culture and the
difficulties in determining which, if either, has a greater influence over the other are
impossible to separate, but this is something we dont want to do, therefore the best
study of human growth and development comes from a distinctly bio-cultural
pursuit.
5


CHAPTER 2
BACKGROUND
Comparisons among Rural Pastoralists
and Urban Dwellers
Theoretical ecological anthropology focuses upon the complex relationships
between individuals and their environments. Phenotypic variation in humans is in
part attributable to the organisms adaptive response to environmental or ecological
conditions (Molnar & Molnar, 2000). Darwin proposed that this adaptive relationship
between an organisms physiology and behavior within their environment was the
determining factor regarding which individuals would survive, mature, reproduce and
pass their adaptive traits onto the next generation (McGee & Warms, 2000). A bio-
cultural perspective is utilized in this comparative study of physical growth variation
in male and female school age youth inhabiting rural and urban Tanzania. Of specific
interest are the growth differences between school age children of a traditional rural
Maasai population and an urban Mwanza population.
Normal childhood growth follows a similar course across time periods and
geographic boundaries emphasizing the commonality of human growth and the
predictability of the human growth pattern. Even though this pattern of growth is
predictable, it is influenced by heredity, nutrition, illness, socioeconomic status and
psychological well-being (Bogin, 1999). A fundamental impetus in studying human
6


growth is to understand the basic biological processes in relationship to the
environment of the organism. Studies of growth patterns from childhood to
adulthood, have demonstrated that diet and physical activity affect body morphology
(Norris et al., 2004). This variation in size, rates of growth and timing of
maturational stages and events has been studied at many levels: across populations,
across groups within populations, within populations over time, across individuals
within populations, and within individuals (Himes, 2004).
Differences in growth have been found to exist between people living in rural
versus urban environments. Recent studies of sub-Saharan Africa rural to urban
migratory populations have shown that urban dwellers on average are taller and
heavier when compared to recent rural migrants (Carlin et al., 2001; Mbanya,
Ngogang, Salah, Minkoulou, & Balkau, 1997; L.R. Pawloski, 2002; Sellen, 1999; D.
W. Sellen, 2000; Shell-Duncan & Obiero, 2000; Sobngwi et al., 2004). Studies of
more traditional rural populations undergoing the process of culture change to an
urban lifestyle have documented similar morphological changes in height and weight
(Aspray et al., 2000; Hoffmeister, Lyaruu, & Krawinkel, 2005; Maletnlema, 2002;
D.W. Sellen, 2000; Sobngwi et al., 2004). In addition, significant variation in
childhood growth patterns between rural and urban populations have been found in
sub-Saharan Africa and other developing countries (L.R. Pawloski, 2002; Rogol,
Clark, & Roemmich, 2000; D. W. Sellen, 2000; Shell-Duncan & Obiero, 2000).
7


Further support for lack of height and weight increase in the absence of exposure to
urban lifestyle behavior comes from studies of rural nomadic pastoralists living in
eastern Tanzania (D. W. Sellen, 2000; D.W. Sellen, 2000). These studies have
determined that the phenotypic variation found is largely related to differences
between rural and urban environments (D.W. Sellen, 2000; Sobngwi et al., 2004),
including differences related to diet (caloric density, quality, quantity, accessibility
and availability of food resources) and physical activity (strenuousness of exercise
and degree of sedentism) (Aspray et al., 2000; Carlin et al., 2001).
Rural pastoralist lifestyle is characterized by a more traditional protein-based,
physically-active nomadic existence. Pastoralist diets consist of milk, meat, and
blood obtained from their animals and cereals either grown or obtained from trading
their animals. Milk and milk products account for 60 to 65 percent of the dietary
energy of the Maasai and Turkana pastoralists (Fratkin, 2001). Previous studies in
support of the declination of nutritional status show that rural African children, when
compared to WHO and NHANES norms, are experiencing malnutrition, starvation
and wasting (L.R. Pawloski, 2002; L. R. Pawloski, 2003; Shell-Duncan & Obiero,
2000). The nutritional status of the local Maasai pastoralist children inhabiting the
Ngorongoro area, the rural population examined here, is declining to the point of
malnutrition (McCabe, Perkin, & Schofield, 1992).
8


In contrast, when compared to the rural lifestyle, urbanization is associated
with a drastic decrease in physical activity, change in dietary habit from proteins to
carbohydrates, and increased psychological stress (Sobngwi et al., 2004). The urban
lifestyle characteristically includes more plentiful carbohydrate-based foods and more
sedentary way of life.
The relationship between physical size differences and health is not as clear.
Aspray (2000) and Carlin (2001) report in their study of rural to urban migration and
increasing metabolic disease, that differences in diet and physical exercise express
themselves in growth variation between populations. Carlin (2001) stated that ...
childrens growth in urban environments is likely to exceed that of their rural
counterparts (pg 5). Aspray et al. (2000) explored the differences in the prevalence
of diabetes between a rural village in Kilimanjaro district and an urban district in Dar
es Salaam, Tanzania. The studys aim was to determine the contribution of physical
activity and indices of general and abdominal obesity to the prevalence of diabetes.
The authors found in adults, an association among obesity, levels of physical activity,
and the prevalence of diabetes in urban environments when compared to rural
environments (Aspray et al., 2000).
Additional studies of rural-urban immigrants and their children have further
documented the effect of migration on growth and development. Urban children in
industrialized countries have an advantage in height and weight over their rural peers,
9


leading investigators to assume that this advantage is indicative of better health (L.R.
Pawloski, 2002; Sellen, 1999; Shell-Duncan & Obiero, 2000). However, this is not
uniformly the case (Carlin et al., 2001; Shell-Duncan & Obiero, 2000). Recent
research into child nutrition in both children and adult rural-urban migrants suggests
that urban life is in fact less conducive to health than rural life (Carlin et al., 2001;
L.R. Pawloski, 2002; Sellen, 1999; Shell-Duncan & Obiero, 2000). Even though it is
widely assumed that settlement results in nutritional improvements; when controlling
for individual (age, gender, and birth order) and household (economic status, family
size, family type, parental education, household head) variation, Shell-Duncan and
Obiero (2000), found the effect of community (market integration, infrastructure,
ecological conditions, subsistence strategy) and its associated changes in lifestyle and
subsistence, resulted in no evidence of nutritional changes or diminished nutritional
status in children in transition from a nomadic pastoralist lifestyle to a variety of
settled lifestyles. The authors used weight-for-height z-score (WHZ) as a dependent
variable indicative of childrens nutritional status, to measure the effects of
community on the childs nutritional status in comparison to that of the individuals
family.
Pawloskis (2002) article originally hypothesized that rural Segou girls would
show indicators of poorer nutritional status when compared with urban Segou girls.
However the original hypothesis was unable to be corroborated even though urban
10


girls were found to be taller and heavier than rural girls. The author concluded that
the greater workloads required of the rural lifestyle resulted in the rural girls body
composition having less body fat and greater arm muscle mass when compared to the
urban girls (L.R. Pawloski, 2002).
Additional evidence of phenotypic variation in children and adolescents
comes from studies of other worldwide populations living a traditional lifestyle,
including the Ache hunter-gatherers of Paraguay (Hill & Hurtado, 1996) and the Pima
Indians of Mexico (Knowler, Pettitt, Saad, & al., 1991; Knowler, Pettitt, Saad, &
Bennett, 1990; Ravussin, Valencia, Esparza, & al., 1994). These studies of rural
populations have documented the absence of height and weight increases typically
associated with exposure to western lifestyle (Hill & Hurtado, 1996; Knowler, Pettitt,
Saad, & al., 1991; Knowler, Pettitt, Saad, & Bennett, 1990; Ravussin, Valencia,
Esparza, & al., 1994). However, conflicting data were found by Davies (1974) when
rural Bantu children were compared to urban Bantu children attending primary school
in Dar es Salaam. No significant differences were found in height and weight
between the two genetically identical populations living in different environments
(Davies, 1974; Davies & Mbelwa, 1974).
Thus, there is no question that different ecological environments affect human
growth and development differently. The recurring interaction between the biology
of human development and the socio-cultural environment provides a bio-cultural
11


perspective to human growth (pg 387) (Bogin, 1999). This bio-cultural perspective
allows us to question, at a deeper level, the possible contributors to the nutrition
variations in lifestyle behavior that influence human growth and development.
Previous studies have identified the association of growth variation between rural
and urban populations being partially attributable to bio-cultural lifestyles reflective
of their environmental setting (D. W. Sellen, 2000; D.W. Sellen, 2000; Shell-Duncan
& Obiero, 2000).
Comparisons among Non-Settled vs. Settled Pastoralists
Present day Maasai are undergoing tremendous changes in their lifestyle that
reflect the transition from rural to urban environment that is underway in other
populations worldwide. According to Dr. Charles Musiba (personal communication,
November 5, 2005), within the rural nomadic Maasai pastoralist populations of the
Ngorongoro district, the demographic transition from non-settled (nomadic pastoralist
with minor agriculture dependency) to a settled (semi-permanent housing while
remaining pastoralists with increasing dependency on agriculture) lifestyle is
underway. The effects of this transition process on the growth differences between
school age children of two rural Maasai populations are examined here.
Studies of rural non-settled and settled Datoga pastoralists have documented
poor childhood growth with little catch-up growth occurring during childhood and
adolescence (Sellen, 1999; D. W. Sellen, 2000). Sellen (2000) detailed the effects of
12


differential levels of childrens energy expenditure and social and ecological
conditions related to growth variation found among the Datoga. Other studies of
physical activity level between non-settled and settled nomadic school children have
found that the non-settled nomadic children had about 25% higher energy expenditure
than their settled counterparts, largely attributable to the daily subsistence duties of
boys herding animals and girls fetching firewood, water and tending animals upon
returning home (M. A. Little, Galvin, & Leslie, 1988).
Similar growth pattern differences have been found in Turkana children (M. A.
Little & Gray, 1990). Little and Gray (1990) found that settled children were heavier
in body weight than the nomads at all ages with most differences between both boys
and girls being significantly different. However, the differences found were not as
great for height as they were for weight. Even though Little and Gray (1990) found
that settled children were generally larger in both height and weight than non-settled
children, they found that the non-settled children were relatively better off because
the nomadic lifestyle of self-sufficiency results in less nutritional variation compared
to the settled population, who are more affected by resource seasonality.
Brainards (1990) study of the Nakwamoru in Turkana District, Kenya
supports Little and Grays (1990) findings that children of a nomadic lifestyle are
better off. When compared to Littles (1983) study sample, younger children in the
nomadic Nakwamoru sample were found to be bigger in most dimensions than the
13


younger children of the settled Turkana sample. This suggests that the pastoral diet
of mostly milk and animal products, which are consistent in quantity and availability,
may be more satisfactory than the unpredictable and seasonal variability of the settled
diet of higher carbohydrates grain products and less animal protein in early life
(Brainard, 1990).
A comparative study (Corbett, Gray, Campbell, & Leslie, 2003) of adult non-
settled and settled Turkana pastoralists of Kenya revealed differences in body
composition as a result of diet, disease and activity. The author found the settled diet,
based primarily on grains, was higher in calories while the non-settled diet was lower
in calories and primarily high in protein. The differences in body composition (fat
stores and muscle tissue) were found to be attributed to differences in physical
activities (Corbett, Gray, Campbell, & Leslie, 2003).
Comparisons among East African Pastoralists
Hypothesis 3 examines the degree to which the research data of this study is
comparable to other rural pastoralists from the same region of East Africa.
Anthropologists long have been interested in how environments influence human
physical growth in sub-Saharan Africa, but associations with subsistence practices
and behaviors remain poorly investigated (Sellen, 1999). While there have been
recent studies of differences between rural and urban populations in sub-Saharan
Africa (Carlin et al., 2001; Galvin, 1992; Mbanya, Ngogang, Salah, Minkoulou, &
14


Balkau, 1997; Sellen, 1999; Sobngwi et al., 2004; Unwin et al., 1999), relatively few
published studies have examined patterns of growth without comparison to western
population growth standards. Growth pattern comparisons between populations can
become problematic due to the effects of genetics and varying environmental
conditions (Cameron, 1991).
African pastoralists are unique in that their lifestyle is reflective of their
herding practices on the savannah or rangeland. Studies reflective of pastoralist
populations occupying similar environments and herding lifestyles provide a more
credible comparison indicative of the unique pastoralist environment of Africa
(Eveleth & Tanner, 1990; Michael A. Little, 1989; Sellen, 1999). For example,
Sellen (1999) found the growth of Datoga children, to be poor in contrast to Western
reference data, but typical of African pastoralists. However, children in the Eyasi
Datoga pastoralist population appear to fall towards the lower end of the observed
range of anthropometric variation among other African pastoral populations (Sellen,
1999).
Previous studies of pastoralist human growth compare study populations in
order to determine if variability exists among different pastoralists and what forces
might be effecting these differences (Cameron, 1991; Gray, Wiebusch, & Akol, 2004;
Michael A. Little, 1989; Sellen, 1999). These comparisons provide a basis to better
understand successful pastoralist behavior and their biological ways of coping with a
15


harsh environment or to provide information needed to plan for almost inevitable
change in their lifestyle (Michael A. Little, 1989). The overall goal of this research is
to identify and understand the growth patterns of rural Maasai pastoralist youth from
a bio-cultural perspective.
16


CHAPTER 3
STUDY POPULATION
Tanzania, an East African country, is home to diverse populations including
hunter-gatherers, agriculturalists, pastoralists and urban dwelling individuals. This
study primarily focuses on two populations: 1) rural nomadic Maasai pastoralists
living in the western zone of the Ngorongoro Conservation Area in the eastern part of
the Serengeti Plains (Endulen village and its vicinity in northern Tanzania, in East
Africa, and 2) urban inhabitants residing within Mwanza, a city of 474,679 (Mbogoro,
2002) located on the southern shores of Lake Victoria. Additionally, comparison is
made using data from studies of two other nomadic East African pastoralist
populations, the Eyasi Dagota inhabiting the area around Lake Eyasi in Tanzania and
the Ngisonyoka Turkana of northwestern Kenya.
Nomadic Pastoralists
Nomadic refers to movement while pastoralism is a type of subsistence
strategy associated with animal herd management. Thus, nomadic pastoralism
describes a lifestyle specializing in animal herding requiring periodic movement in
search of viable grazing land. Pastoralism is associated with inhabiting and
exploiting marginal arid grasslands of tropical environments found in East Africa.
Arid grasslands, generally poor food sources for humans, promote the adaptive
strategy of converting low quality plant resources into portable, high quality animal
17


foods. Low population density and high mobility are characteristic among pastoral
populations as a result of the low level energy availability of the environment. East
African pastoralist lifestyle ranges from almost total dependence on animal herding,
to some horticulture practice generally associated with raising and maintaining cattle,
to that of sedentary agriculture. The type of subsistence strategy in use is governed
by the amount of seasonal rainfall and type and availability of soil nutrients. Long
periods of rainfall support nomadic pastoralism, since rains stimulate vegetation
growth. With more vegetation come increased milk production from livestock, and
the pastoralists respond to this with increases in food intake and rapid increases in fat
and lean tissue as a result. The opposite applies to the settled, agricultural dependent
pastoralist. Crops can not be harvested until after the rainy season is over, thus the
longer the rainy season, the greater the nutritional stress (Corbett, Gray, Campbell, &
Leslie, 2003).
Pastoralism, a subsistence lifestyle requiring large areas of land to support
relatively small populations, is threatened today because of population growth,
environmental degradation and changing patterns of land use (Crawford & Leonard,
2002). Land availability pressure has contributed to demands for more productive
subsistence practices, such as ranching and cultivation. Population pressure has
contributed to loss of grazing lands. Population increase and the resulting pressure on
the land have led to intensive subsistence practice changes. Crop cultivation has
18


taken over lands formerly used by pastoralists for herding, with governments seeking
to settle nomadic pastoralists in order to tax and control them (M. A. Little, Dyson-
Hudson, Leslie, & Dyson-Hudson, 1999; Thorton, Galvin, & Boone, 2003). In
addition, governments have produce changes in pastoral lands through the
development of hydroelectric and other projects that contribute to the transformation
and degradation of the environment (Fratkin, 2001). Todays rapidly changing
political borders further disrupt traditional migratory routes in ways that impact the
survival of nomadic pastoralism. Herders of the East Africa region view cattle,
donkeys and camels as their most important livestock depending on geographic
location. In addition, most groups maintain sheep and goats for subsistence and
donkeys for transport. The herd animals provide blood, milk, and meat for
subsistence. Any horticulture activity is practiced by the women while men are
traditionally the herders (Crawford & Leonard, 2002). The diet of the traditional
nomadic Maasai is predominantly high in animal proteins. Dr. Charles Musiba
(personal communication November 19, 2005) said that seventy five percent of their
diet consists of high protein meat and dairy products, while high carbohydrate com
and beans make up the remaining twenty five percent.
Traditional nomadic pastoralist families base their activities at semi-
permanent family homesteads or bomas. Each semi-permanent homestead consists of
clusters of timber, earth and dung huts as well as sheds for goats and calves, maize
19


stores, and kraals for larger livestock, all enclosed within a thorn brush stockade for
protection against predators and human raiders. All household members participate
in herding tasks. The main herders are resident young men and younger children.
Older men may participate in herding, especially when unmarried or when their
children or brothers are still young, but they will do less herding when children or
younger men are available. Children of both sexes begin herding sheep, young calves
and goats for short periods in the vicinity of the homestead at about four years of age,
and then are given charge of more and larger livestock for longer periods of the day
as they grow older and their abilities improve. After about 12 years of age, boys
begin spending many of their days away with the herd, and therefore miss midday
meals. At this age they learn advanced herding skills, archery and spear-throwing in
defense of the herds. Teenage boys have been observed to work extremely hard as
measured by time allocation (D.W. Sellen, 2000). In the dry season, they leave with
the cattle at dawn and herd them as far as 12 kilometers to water, returning exhausted
at nightfall.
For the girls, between five and nine years of age, much time is spent helping
their mothers in household chores, milking and fetching water and firewood.
Teenage girls spend increasing amounts of time in light household activities, social
visiting, and in preparation for marriage.
20


21


Rural Nomadic Pastoralist Study Population
Endulen Village Area
The Endulen Village Area is located in the western part of the Ngorongoro
Conservation Area Authority (NCA) containing a population of more than 55,000
(according to National Census of 2002) (Nicholas, 2004). Endulen Village and the
surrounding area are home to both non-settled and settled rural nomadic Maasai
pastoralists. Non-settled nomadic pastoralists are found at low density throughout the
NCA, moving often in search of food for their herds, subsisting primarily on products
from their livestock such as milk, blood and meat, supplemented by maize meal
during periods of low milk production. Agricultural products are traded for and
constitute a minor portion of the family diet.
Settled pastoralists live in semi-permanent housing and are more dependent on
cultivation for subsistence, while maintaining a measurable livestock. Endulen
Village pastoralists fall into this category, as they are currently in the process of
transitioning from a traditional nomadic pastoralist lifestyle to a more settled
pastoralist lifestyle.
There are three primary schools in the Endulen Village Area. The Endulen
primary school serves students who are categorized as settled pastoralists. For the
most part, the Misigio and Ndiyan primary schools serve students that are categorized
as non-settled pastoralists, except some Ndiyan students come from nearby
22


permanent bomas. The rural pastoralist population for this study is inclusive of a
cross-sectional convenience sample of 717 rural school age youth ranging in ages
from seven through sixteen. This group is inclusive of 435 males, 282 females from
the three rural schools as discussed below and (see table 5.2).
Endulen Primary School
Endulen primary school is attended predominantly by Maasai pastoral youth
from the more settled Endulen Village. In addition, approximately 3% of the student
population lives on-site in the dormitory facilities due to long distances between the
family home and school. These students return home only during school breaks. All
boarding students are provided three meals per day. Lunch also is provided to all
local village students who attend, unless they choose to return home at lunch time.
Misigio and Nydian Primary Schools
The Misigio and Nydian primary schools are predominantly attended by non-
settled Maasai pastoralist youth. Students walk long distances daily in order to attend
school. Lunch is not provided for the students at school, and students who do not
return home at lunch time.
Urban Study Population
Mwanza Area
Mwanza, located on the southern shore of Lake Victoria, is Tanzanias second
largest city with a population of approximately 2.8 million people. The city consists
23


of a profusion of industries and a busy lake port that handles much of the cotton, tea
and coffee grown in the region.
Families in urban Mwanza are generally dependent economically on a variety
of manual labor jobs as their main source of income, and as such the people within
these communities are of low socioeconomic status (see table 3.1).
Access to food quantity and variety is more dependable and consistent then in
the rural environment. Diets are predominantly high in unrefined staple foods such as
com with low levels of animal proteins. Typically, urban diets consist of seventy five
percent high carbohydrate com and beans, while meat and dairy products make up the
remaining twenty five percent.
Mlimani Primary School
The Mlimani Primary school is attended by urban youth inhabiting local
neighborhood areas. The urban population for this study is inclusive of a cross-
sectional convenience sample of 327 rural school age youth ranging in ages from
seven through sixteen. This group includes 143 males and 184 females (see table 5.3).
Parental occupational data obtained from students indicate parents hold medium to
low income producing jobs (see table 3.1). Some students choose to go home for
lunch since lunch is not provided to the students.
24


Table 3.1 Urban parental occupation
Occupation Frequency Percent of Total
Accountant 1 0.31
Business 98 29.97
Carpenter 8 2.45
Cook 1 0.31
Doctor 13 3.98
Driver 26 7.95
Farmer 31 9.48
Fisherman 4 1.22
Housewife 1 0.31
Mechanic 18 5.50
Nurse 16 4.89
Painter 15 4.59
Police 12 3.67
Postman 2 0.61
Printer 2 0.61
Sailor 21 6.42
Shopkeeper 6 1.83
Station Master 1 0.31
Tailor 7 2.14
Teacher 24 7.34
Watchman 20 6.12
Total 327 100.00
East African Pastoral ists
The two comparison East African groups include the Eyasi Datoga (Sellen,
1999) and the Ngisonyoka Turkana (M. A. Little, Galvin, & Mugambi, 1983).
Eyasi Datoga
The Datoga inhabits the Eyasi basin and Yadea depression on the Eastern rim
of the East African Rift Valley due East of Lake Eyasi and South of the Ngorongoro
crater. Traditionally they are nomadic pastoralists who depend on keeping cattle,
25


sheep and goats for milk, meat, and trade for grain (Sellen, 1999). Cultivation is
difficult (except for onions) in the Eyasi environments which are comprised of
seasonal marshlands, salt flats and dunes, savanna grasslands and baobab or acacia
thicket (D. W. Sellen, 2000).
During April to June of 1989 and January to March of 1991, D. W. Sellen
(1999) collected cross-sectional anthropometric measurements (height and weight) of
417 Datoga children and youth, age 1 through 18 years. Females (N=236) are better
represented than males (N= 181) in Sellens (1999) sample. For comparison to this
samples Maasai data, only Datoga data on male (N=72) and female (N =102) age 7
through 16 years are incorporated.
TANZANIA
ARUSHA
O Basoti;
StNGIDA
^ main areas of impermanent
settlement included in the study
0
50 100 nm
KOSDOA
Figure 3.2 Map of Lake Eyasi region
26


Ngisonyoka Turkana
The Ngisonyoka Turkana pastoralists live in the area bounded by the Turkwel
and Kerio Rivers and the southern border of the Turkana District (M. A. Little,
Galvin, & Mugambi, 1983). The environment consists of dry bushland or scrub
savanna with a very short growing season due to marked seasonality of rainfall (M. A.
Little & Johnson, 1987). Similar to the Endulen area pastoralists, the Turkana
maintain five species of livestock (camels, cattle, goats, sheep and donkeys) and
subsist almost entirely on the products of their animals (milk, blood and meat) with
some trading for maize meal. They exploit this region by nomadic movements on
average of every three to four weeks depending on seasonality, availability of forage,
and location of water.
During the wet season of 1981 (March and April) and the late dry season of
1981-1982 (December through March, M. A. Little (1983) collected anthropometric
measurements (height and weight) of 543 Turkana adults and children, age 1 through
54 years. Herding responsibilities requiring older boys to be away from home for
days in satellite camps made it difficult to acquire male measurements, making
females (N=303) better represented than males (N=240) in Littles (1983) sample.
For comparison to this samples Maasai data, only Turkana data on male (N=64) and
female (N =142) age 7 through 16 years are utilized.
27


Figure 3.3 Map of South Turkana region, Kenya
28


CHAPTER 4
HYPOTHESES AND RESEARCH DESIGN
The primary aim of this study is to examine physical growth patterns of rural
Ngorongoro Maasai pastoralists and urban Mwanza youth living in Northern
Tanzania. Questions are addressed through three hypotheses and six predictions.
Hypothesis #1 addresses, identifies and explains variability between rural nomadic
Maasai pastoralist and urban Mwanza youth, age seven through sixteen, using
anthropometric data such as height, weight, and body mass index (BMI). Hypothesis
#2 addresses, identifies and explains variability within the rural nomadic pastoralist
environment; rural non-settled versus rural settled Maasai pastoralist youth age seven
through 11, using anthropometric data such as height, weight, and body mass index
(BMI). Hypothesis #3 proposes to contrast growth patterns of the rural nomadic
pastoralist Maasai youth of Endulen Village area with other existing published data
sets from East African nomadic pastoralist populations of comparable ages.
Hypotheses and Predictions
This study will examine growth patterns in relationship to current
environmental and lifestyle shifts underway between rural and urban populations in
sub-Saharan Africa. In addition, this study will examine if the data collected from the
rural Maasai pastoralist population is unique or is typical of other East African
pastoralist populations.
29


Naturally occurring variations between rural and urban environments, within
non-settled and settled lifestyles, and between similar populations provide
opportunities to test predictions generated from hypotheses based on the influence of
lifestyle (Davies, 1974; Gray, Wiebusch, & Akol, 2004; M. A. Little & Gray, 1990;
L.R. Pawloski, 2002; D. W. Sellen, 2000; Shell-Duncan & Obiero, 2000). To
examine these issues, predictions generated from the following hypotheses will be
tested:
Hypothesis 1: Growth differences will be present between two samples of Tanzanian
youth: rural Maasai pastoralists inhabiting the Northern Ngorongoro volcanic
highlands and urban dwellers in the city of Mwanza on the shores of Lake Victoria.
Predictions (Hypothesis 1):
1. Cross-sectional anthropometric measurements of standing height (cm),
weight (kg), and BMI will be significantly lower in rural pastoralist males when
compared to urban dwelling males age seven through sixteen.
2. Cross-sectional anthropometric measurements of standing height (cm),
weight (kg), and BMI will be significantly lower in rural pastoralist females when
compared to urban dwelling females age seven through sixteen.
Hypothesis 2: Growth differences will be present between rural non-settled Maasai
pastoralists inhabiting the Northern Ngorongoro volcanic highlands and rural settled
Maasai pastoralists inhabiting Endulen Village and the surrounding area.
30


Predictions (Hypothesis 2):
1. Cross-sectional anthropometric measurements of standing height (cm),
weight (kg), and BMI will be significantly lower in rural non-settled pastoralist males
when compared to rural settled pastoralist males age seven through eleven.
2. Cross-sectional anthropometric measurements of standing height (cm),
weight (kg), and BMI will be significantly lower in rural non-settled pastoralist
females when compared to rural pastoralist settled females age seven through eleven.
Hypothesis 3: Growth trajectories will not differ between rural nomadic Maasai
pastoralist youths inhabiting the Endulen Village area and data sets from other East
African rural nomadic pastoralist youths inhabiting similar environments.
Predictions (Hypothesis 3):
1. Cross-sectional anthropometric measurement growth trajectories of mean
standing height (cm), weight (kg), and BMI will not differ between rural nomadic
Endulen Village area Maasai pastoralist males age seven through sixteen and data sets
from other East African nomadic pastoralist males age seven through sixteen.
2. Cross-sectional anthropometric measurement growth trajectories of mean
standing height (cm), weight (kg), and BMI will not differ between rural nomadic
Endulen Village area Maasai pastoralist females age seven through sixteen and data
sets from other East African nomadic pastoralist females age seven through sixteen.
31


Research Design
The overall research design is presented in Figure 4.1 as a summary schematic
of this studys relational hypotheses of possible growth differences within and
between East African rural and urban primary school youth.
r ~ \
Growth Differences among East African Youth Populations
^ --------------------------------------------- ---------------'
<0
o
> I
i Rural Youth Population -
Settled (Endulen School)
Hvd.2
V
Non-settled (Misigio,
N Ndiyan Schools)
Figure 4.1 Research Design
In summary, the goal of this research is to understand possible physical
growth differences between and within East African youth inhabiting difference
environments (rural vs. urban and nori-settled vs. settled), as well as examining if the
32


growth data of the rural Maasai pastoralist population conforms to other regional East
African pastoralist youth populations.
33


CHAPTER 5
MATERIALS AND METHODS
A cross-sectional approach was used to sample the growth of two groups of
East African youth ages seven through 16 years. One group includes rural Maasai
pastoralist youth who reside in the western zone of the Ngorongoro Conservation
Area in the eastern part of the Serengeti Plain within the Endulen Ward in northern
Tanzania, East Africa at latitude S3 13.158 and longitude E35 16.145, and 1833m
above sea level. The second group includes urban children and adolescents living in
the city of Mwanza, Tanzania located on the southern shore of Lake Victoria, at
latitude S2 31.863 and longitude E32 54.450 and 1574m above sea level. Standing
height and weight were measured in 1093 rural and urban students attending their
respective schools on the days the measurements were obtained in May and June,
2006 during the University of Colorado Denver, Department of Anthropology annual
Tanzanian Field School.
In addition, in order to compare the youth in this study with other similar
sample populations for hypothesis 3, data were obtained from published sources of
two East African pastoralist populations. Cross-sectional anthropometric
measurements from youth of the same age consist of standing height and weight.
Each of the sample groups of youth is presented in detail below.
34


Endulen Area Rural Population Sample
The rural sample consists of 717 Maasai pastoralist youth, age seven through
sixteen, inclusive of 435 males and 282 females (see table 5.2) from three schools
located within an 11 kilometer radius of Endulen Village (see figure 3.1).
Endulen Primary School is located within Endulen Village. At the time of this
study, enrollment for the 2005-2006 calendar year consisted of 648 male and 370
female students. Anthropometric measurements were obtained from 99 male and 70
female students age seven through twelve (see table 5.1).
Ndiyan Primary School is located within a 3 kilometer radius of Endulen
Village and had 333 male and 149 female students enrolled for the 2005-2006
calendar year. Anthropometric measurements were obtained from 22 male and 81
female Ndiyan students age eight through thirteen and one fifteen year old (see table
5.1).
Misigio Primary School is located within a 10.5 kilometer radius of Endulen
Village and had 323 male and 238 female students enrolled for the 2005-2006
calendar year. Anthropometric measurements were obtained from 190 male and 255
female Misigio students age seven through sixteen (see table 5.1).
35


Table 5.1 Rural primary school cross-sectional anthropometric data
Endulen Primary School Endulen Primary Students N-height measured N-weight measured
Males 99 99 99
Females 70 70 70
Total Endulen School 169 169 169

Ndivan Primary School Ndiyan Primary Students N-height measured N-weight measured
Males 22 22 22
Females 81 81 81
Total Ndiyan School 103 103 103

Misigio Primary School Misigio Primary Students N-height measured N-weight measured
Males 190 190 190
Females 255* 254* 255*
Total Misigio School 445* 444* 445*
* Rural female student left after weighing anc did not complete other anthropometric
measurements.
The total number of children both male and female, seen from the rural
pastoralist society comprised 34.79% of the total school population enrolled in the
three rural primary schools visited in and around Endulen Village (see table 5.2).
Table 5.2 Summary rural primary school cross-sectional data
Endulen Primary Schools (3) Rural Students Enrolled N-height measured % height measured N-weight measured % weight measured
Males 1304 311 23.85% 311 23.85%
Females 757 405* 53.50%* 406* 53.63%*
Total Rural 2061 716* 34.74* 717* 34.79%*
* Rural female student left after weighing and did not complete other anthropometric
measurements.
Mwanza Urban Population Sample
The urban data sample consists of 327 urban youth, age seven through sixteen,
inclusive of 143 males and 184 females (see table 5.3) from two schools in the city
36


Mwanza within the Nyamagana District in Tanzania (see figure 3.1). Both Mongela
and Mlimani primary schools are located on the same campus.
Table 5.3 Urban primary school cross-sectional anthropometric data
Mlimani Primary School Mlimani Primary Students N-height measured N-weight measured
Males 143 143 143
Females 184 184 184
Total Mlimani School 327 327 327
The total number of children both male and female, seen from the urban
population comprised 38.07% of the total school population enrolled in the both
urban primary schools.
Table 5.4 Urban cross-sectional data summary
Mwanza Primary Schools (2) Urban Students Enrolled N-height measured % height measured N-weight measured % weight measured
Males 421 143 33.97% 143 33.97%
Females 438 184 42.01% 184 42.01%
Total Urban 859 327 38.07% 327 38.07%
Samples from Published Sources
Comparative samples of East African populations used in hypothesis 3 come
from two sources of published data. D. W. Sellen carried out two cross-sectional
surveys between April-June 1989 and January-March 199Ion 444 Datoga children
living in the Lake Eyasi basin in Tanzania (Sellen, 1999; D. W. Sellen, 2000). M. A.
Little carried out a cross-section survey of the Kenya Ngisonyoka Turkana
pastoralists between March-April 1981 and December-March 1982 (M. A. Little,
37


Galvin, & Mugambi, 1983). Both populations are similar to the Maasai in that they
are nomadic pastoralists and inhabit similar environments.
Data Collection
Cross-sectional convenience data collection of all rural and urban students in
the study was accomplished through collaboration between researchers from the
University of Colorado Denver and the University of Calgary, Canada. Prior to travel
to Tanzania, arrangements for this study were discussed between Dr. Charles Musiba
and each of the heads of the schools where anthropometric measurements were to be
taken. Upon arrival, Dr. Musiba met with school personnel to coordinate the data
collection process and procedures. On the days anthropometric measurements were
taken, a demonstration was presented to the students in Kiswahili language by Dr.
Musiba. Each of the research assistants doing the measurements and weighing was
taught Kiswahili terminology in order to interact with the students in a positive
manner. In addition, each research assistant reviewed proper anthropometric
procedures prior to data collection to insure measurement consistency across student
subjects. Every research student was paired with a teacher or teaching assistant for
any additional assistance with communication necessary to insure the comfort level of
the students.
Students were lined up outside the classroom where the measurements were
taken. Each student in turn was interviewed and then moved through the different
38


work stations. A general atmosphere of light heartedness, fun and curiosity
prevailed with much laughing and talking among the students, teachers and
researchers. At the first station, demographic information on the students age and
school class level (standard) were obtained through interviews conducted by student
researchers with assistance from resident school personnel at the three rural Endulen
area schools. School teachers conducted the interviews of all urban Mwanza students
with assistance from researchers. Reliance on school officials knowledge of each
student was required since official birth records were not present at the time of pre-
measurement interviews. Because of this, the possibility of age error exists.
Secondary sex characteristics were not used for age verification because of the lack of
trained field staff and the conditions under which measurements were made.
Each student was given their data sheet to proceed through the ensuing work
stations. Data was recorded at each station. The data sheets were collected at the end,
reviewed for completeness and students thanked and dismissed to return to their
classmates. Anthropometric measurement procedures followed standard procedural
guidelines (Cameron, 2004; Frisancho, 1990; NHANES, 1988; Roche & Sun, 2003).
All standing height measurements were taken and recorded by the author, thus
avoiding inter-rater error. Standing height was measured to the nearest millimeter
using an anthropometer from Swan TM, Basel, Switzerland. Subject each stood
barefoot on a level concrete floor and was encouraged to stand erect with head in
39


level upright position. The author used the chin-support method to guide the subject
to the correct measuring height.
All weight measurements were taken by two different individuals on a single
weight scale, placed on an existing level classroom concrete floor. Weight was
measured using a Nikai Model NBS387E electronic LED scale from Nikai TM, Kobe,
Japan, having programmed readings to the nearest .5 kilogram. Shoes and hand
carried book bags were removed, but student clothing and jewelry were worn during
measuring. All subjects, both rural and urban, wore school uniforms consisting of
light weight kaki shorts and colored shirts for males, and light weight colored dresses,
or skirts and blouses for females. All weight measurements incorporated these
similar clothing items, but since the digital scale utilized .5 kg increments, the
clothing was deemed to have a negligible effect.
Body Mass Index (BM1) was derived by calculating (weight (kg) / height
(m)2).
Measurement Variables
Major dependent variables utilized in this study are height measured in
centimeters, weight measured in kilograms, and Body Mass Index (BMI) calculated
as (weight (kg) / height (m)2). Rationale for the use of each of these measurements is
as follows:
40


Height increases over time and as such, it is an indicator of the long term total
nutrition and health history of the child (Bogin, 1995,, 1999).
Weight can both increase and decrease over time and, therefore, relates more
to recent nutrition and health status (Bogin, 1995,, 1999). Additionally,
weight is relatively more labile and sensitive to short-term effects of nutrition
and infections than length or stature (Himes, 2004).
Body Mass Index is often used as a proxy for current nutritional status (Bogin,
1995).
The independent variables of age and gender have a functional relationship
with all three dependent variables and are used to categorize outcome results. Other
independent variables, acknowledged as having a potential effect on the dependent
variables but not controlled for or utilized in this study, are genetics, disease, access
to health care, ecology, culture, birth order, individual peer status, household
economic status, family size, and head of household.
Statistical Analysis
All graphic plotting and statistical analyses were performed with SPSS for
Windows version 15 graduate pack (SPSS, 2006). Scatter plots with Lowess
smoothing (50%) curves were used to produce height and weight growth curve graph
comparisons between rural and urban school age children. Height in centimeters,
41


weight in kilometers and BMI were graphically plotted in one year increments, based
on chronological ages as reported by the children and verified by school staff. The
potential for age-error in this sample set is acknowledged (see below Data Reliability
Pg 44).
Not all anthropometric dimensions are normally distributed (Frisancho, 1990).
A Shapiro-Wilk test was done to determine normality especially since some data age
sets contain low sample quantities. Test results did indicate a non-normal data
distribution of the growth curve data. In order to maximize the diagnostic
effectiveness of the anthropometric information a Z-score transformation was
performed on all data to normalize the variable distribution.
For hypothesis 1, an independent sample t-test was used to determine if height,
weight and BMI differences existed between rural and urban male and female youth.
An independent sample t-test was used for hypothesis 2 to test if significant
differences existed between rural settled and non-settled male and female youth
pastoralist in height, weight and BMI. For both hypotheses, a p-value of less than
0.05 is indicative of a significant difference of means and is noted in all tables with an
asterisk.
One-way ANOVA was used in hypothesis 3 to determine if significant
differences exist among mean height, weight, and BMI of rural pastoralist Maasai
population and the other East African pastoralist populations including the Eyasi
42


Datoga and Ngisonyoka Turkana. Because the data were found originally to be non-
normally distributed, a Levenes and Brown & Forsythes test of homogeneity of
variance was performed. A non-parametric Kruskal-Wallis test was also done to
verify the validity of results of the initial parametric One-way ANOVA test. To
determine where any differences between populations might exist, a Bonferroni post
hoc analysis was performed. A /rvalue of less than 0.05 is indicative of a
significance difference of means and is noted in all tables with an asterisk.
Data Reliability
Studying growth in a rural pastoralist environment in Tanzania poses many
challenges. In an effort to investigate the possibility of sample bias and reliability, an
examination of the central limit tendency of the data was performed. This study,
typical of many published studies of East African pastoralists (M. A. Little, Galvin, &
Mugambi, 1983; Sellen, 1999), is subject to possible sampling bias. Ages were
difficult to ascertain reliably because of the imprecision of determination where in
general, date of birth is not officially recorded. Attendance at each school determined
the number and gender of school children measured. To verify central limit tendency
in both male and female rural and urban samples for hypotheses 2 and 3, and in order
to facilitate comparison with other East African pastoral populations in hypothesis 1,
Sellens (1999) technique of examining the population for sample variance within 2
standard deviations (greater or less than) of the mean was employed. Of this studys
43


total sample population of 1044 individuals, 43 cases or 4.1% of height-for-age (HAZ)
Z-scores were found to be greater or less than 2 standard deviations from the mean.
Within rural and urban populations, the rural sample (n = 717) included 28
individuals or 4% that fell outside of 2 SD, and for the urban population (n = 327) 15
individuals or 5% fell outside the 2 SD criteria. This low percentage of sample
variation suggests that most age estimates are reasonably accurate and the data are not
systematically biased for either sex or any particular age grouping or developmental
stage. For purposes of this study, this analysis demonstrates that valid inferences may
be drawn about the growth status of youth within the populations and that this Maasai
sample is validated as representative of a typical East African pastoralist population.
Age-error probability remains in spite of the above analysis. Lack of official
birth records prevents comparison of this study group with a documented control
group. Even though differences are not evident, it is possible that statistics would be
significantly different if sample sizes were larger.
44


CHAPTER 6
RESULTS
Growth trajectory differences are examined using standing height for age
(HAZ), weight for age (WAZ) and BMI for age (BMIZ) Z-scores for hypotheses 1
and 2. For hypothesis 3, standing height, weight and BMI mean differences are
examined among three East African pastoralist populations. Maasai data include
anthropometric measurements obtained in conjunction with this study. Data for the
Turkana (M. A. Little, Galvin, & Mugambi, 1983) and the Datoga (Sellen, 1999) are
from the two published sources.
Hypothesis 1 Results
Predictions under hypothesis 1 propose that significant differences in growth
will be present between two samples of Tanzanian male and female youth populations
ages 7 through 16 years, rural Maasai pastoralists and urban dwellers of the city of
Mwanza. Data for males are presented below followed by data for females.
Male data presented in Table 6.1 includes descriptive information according
to age including summary data of standing height (cm), weight (kg), and BMI for
rural pastoralist and urban dwelling youth.
45


Table 6.1 Male rural pastoralist and urban dwelling descriptive data
Std.
Male ___________i_____i____________Height_________________________Weight__________________________BMI
Age (yrs) Rural / Urban N Mean (cm) SD Median Mean (cm) SD Median Mean SD Median
7 Rural 81 120.100 7.223 120.900 19.574 2.808 19.500 13.515 0.980 13.393
Urban 9 120.222 5.677 120.500 20.444 2.338 20.000 14.120 0.920 13.924
8 Rural 48 123.654 7.565 123.450 20.896 3.321 20.250 13.612 1.287 13.823
Urban 13 123.031 5.679 121.900 22.538 2.681 22.500 14.866 1.225 14.559
9 Rural 38 126.550 6.596 127.250 22.026 2.959 22.000 13.732 1.284 13.808
Urban 7 130.386 5.721 131.100 25.429 2.864 26.500 14.909 0.628 15.203
10 Rural 25 134.212 8.948 134.900 27.120 4.857 27.000 14.982 1.811 14.533
Urban 24 134.267 5.461 134.100 27.625 3.582 27.500 15.274 1.227 15.168
11 Rural 17 134.547 8.081 136.800 26.412 3.838 27.500 14.523 1.153 14.267
Urban 16 135.013 5.842 135.950 28.531 3.603 28.000 15.593 1.064 15.605
12 Rural 110 136.090 10.685 133.500 26.982 5.431 26.000 14.412 1.053 14.434
Urban 7 141.457 3.754 142.200 30.929 3.155 30.000 15.445 1.293 15.474
13 Rural 57 139.981 11.903 138.900 28.421 7.141 27.000 14.277 1.197 14.084
Urban 10 145.190 7.596 146.250 33.600 4.795 33.750 15.884 1.282 16.376
14 Rural 34 144.809 7.405 144.650 30.544 4.677 30.000 14.485 1.208 14.357
Urban 17 153.429 8.168 153.400 39.471 7.199 39.500 16.637 1.843 16.843
15 Rural 22 151.768 6.883 152.500 35.159 4.365 35.000 15.204 0.757 15.140
Urban 19 159.305 5.912 157.700 46.132 5.676 47.000 18.141 1.674 17.773
16 Rural 12 160.350 7.829 161.900 39.167 5.726 39.500 15.160 1.208 15.091
Urban 7 165.971 5.279 164.400 53.500 6.589 56.500 19.390 1.943 18.870
Figures 6.1, 6.2 and 6.3 display contrasting rural pastoralist and urban
dwelling male standing height, weight, and BMI in scatter plots with Lowess
smoothing using Z-score converted data. Discussion follows these results.
46


sex: Male
Rural Urban
1 Rural
Urban
' Rural
Urban
Figure 6.1 Male Rural and Urban Standing Height for Age Z-Scores
sex: Male
Rural Urban
§
Rural
Urban
Rural
Urban
Figure 6.2 Male Rural and Urban Weight for Age Z-Scores
47


sex: Male
Rural Urban
8
Rural
Urban
Rural
'" Urban
Figure 6.3 Male Rural and Urban BMI for Age Z-Scores
As reflected in figures 6.1. 6.2, and 6.3 for the males, growth trajectories of
rural and urban male Z-score standing height, weight and BMI curves indicate an
increased divergence between rural and urban male youth beginning at age 10 and
continuing until age 16. In contrast to the standing height curve, no growth recovery
is seen in the weight and BMI curves. Of special interest is the flat period of growth
faltering appearing at approximately the age of 12 in standing height, weight and BMI.
This interesting finding regarding the period of no-growth is discussed further in
chapter 7.
48


Independent t-test results on male standing height, weight and BMI Z-scores
are summarized in table 6.2. Significance found to be equal to or less than alpha 0.05
are indicated by asterisk (*) and in bold font.
No differences in standing height and weight are present from age seven
through age 13, except for weight at age 9 (p = .003). Standing height between rural
and urban male youth diverges significantly at ages 14 (p = .000) and 15 (p = .001).
After the age of 15, the rural male youth height growth curve returns to closer
proximity.
Weight diverges significantly between rural and urban male youth beginning
at age 13 {p = .031) and continues through age 16 at significance of (p = .000) for all
years. No recovery of weight difference is apparent.
Early differences in BMI appear at ages 8 (p = .003) and 9 {p = .023).
Consistent BMI differences are found to begin at age 11 (p = .009), age 12 ip = .0014)
and continue through age 16 at (p = .000).
49


Table 6.2 Male rural pastoralist and urban dwelling statistical analysis
HAZ WAZ BMIZ
Age (yrs) Rural / Urban N Mean SD Sig. (2- tailed) Mean SD Sig. (2- tailed) Mean SD Sig. <2- tailed)
7 Rural 81 -0.995 0.490 0.961 0.868 0.295 0.373 0.651 0.475 0.081
Urban 9 -0.987 0.385 0.777 0.246 0.358 0.446
8 Rural 48 -0.754 0.513 0.783 0.729 0.349 0.106 0.604 0.623 0.003*
Urban 13 -0.797 0.385 0.556 0.282 0.003 0.593
9 Rural 38 -0.558 0.447 0.157 0.610 0.311 0.007* 0.546 0.622 0.023*
Urban 7 -0.298 0.388 0.253 0.301 0.024 0.304
10 Rural 25 -0.039 0.607 0.980 0.075 0.511 0.682 0.059 0.877 0.514
Urban 24 -0.035 0.370 0.022 0.377 0.201 0.594
11 Rural 17 0.016 0.548 0.852 0.149 0.403 0.113 0.163 0.558 0.009*
Urban 16 0.015 0.396 0.074 0.379 0.355 0.515
12 Rural 110 0.088 0.724 0.190 0.089 0.571 0.060 0.217 0.510 0.014*
Urban 7 0.452 0.254 0.326 0.332 0.283 0.626
13 Rural 57 0.352 0.807 0.187 0.062 0.751 0.031* 0.282 0.580 0.000*
Urban 10 0.705 0.515 0.607 0.504 0.496 0.621
14 Rural 34 0.679 0.502 0.000* 0.285 0.492 0.000* 0.181 0.585 0.000*
Urban 17 1.264 0.554 1.224 0.757 0.861 0.892
15 Rural 22 1.151 0.467 0.001* 0.771 0.459 0.000* 0.167 0.366 0.000*
Urban 19 1.662 0.401 1.924 0.597 1.589 0.811
16 Rural 12 1.733 0.531 0.111 1.192 0.602 0.000* 0.146 0.585 0.000*
Urban 7 2.114 0.358 2.699 0.693 2.194 0.941
Female data presented in Table 6.3 includes descriptive information according
to age including summary data of standing height (cm), weight (kg), and BMI for
rural pastoralist and urban dwelling youth.
50


Table 6.3 Female rural pastoralist and urban dwelling descriptive data
Female Std. Height Weight BMI
Age Rural / (cm) (kg)
(yrs) Urban N Mean SD Median Mean SD Median Mean SD Median
7 Rural 54 118.782 7.611 118.100 18.833 3.067 18.500 13.292 1.274 13.185
Urban 8 115.813 5.982 116.150 18.813 2.267 19.000 13.973 0.510 13.873
8 Rural 33 126.173 6.661 125.300 21.773 3.385 21.000 13.606 1.132 13.437
Urban 10 128.810 5.677 128.900 25.900 4.267 24.750 15.530 1.702 15.721
9 Rural 27 127.785 7.333 128.100 23.648 3.107 23.500 14.453 1.092 14.188
Urban 13 128.623 5.598 127.900 24.462 2.780 24.000 14.745 0.737 14.709
10 Rural 16 130.206 4.097 130.350 25.563 2.971 25.000 15.067 1.536 14.796
Urban 21 133.419 5.413 133.700 27.214 2.840 28.000 15.277 1.193 15.515
11 Rural 20 125.780 7.215 124.600 22.475 3.607 22.000 14.125 1.020 13.878
Urban 22 139.268 6.534 138.850 31.091 4.081 30.250 15.967 1.073 16.002
12 Rural 63* 131.318 8.346 131.200 24.286 4.390 24.000 13.985 1.125 13.929
Urban 35 144.263 6.771 144.700 34.371 5.144 33.500 16.448 1.660 16.440
13 Rural 36 136.792 5.965 137.700 26.153 3.437 26.000 13.949 1.298 14.093
Urban 25 147.328 5.729 147.200 38.380 7.058 36.500 17.573 2.404 17.030
14 Rural 20 144.435 6.107 144.600 30.025 4.342 30.000 14.319 1.148 14.376
Urban 21 156.105 5.840 155.200 44.810 5.849 44.500 18.352 1.829 18.335
15 Rural 5 153.540 6.485 156.300 42.500 10.137 42.500 17.820 2.954 17.683
Urban 16 157.231 6.626 157.450 50.500 7.568 48.250 20.459 3.120 19.252
16 Rural 1 165.300 165.300 49.000 49.000 17.933 17.933
Urban 9 158.478 6.530 156.900 48.500 5.772 49.000 19.356 2.543 19.728
* Age 12 female weighted but left before other measurements could be taken
Figures 6.4, 6.5 and 6.6 display contrasting rural pastoralist and urban
dwelling female standing height, weight, and BMI in scatter plots with Lowess
smoothing using Z-score converted data. Discussion follows these results.
51


Zscore:
sex: Female
Rural Urban
Rural
Urban
Rural
Urban
9
Figure 6.4 Female Rural and Urban Standing Height for Age Z-Scores
sex: Female
Age (yrs)
Rural Urban
$
Rural
Urban
Rural
Urban
Figure 6.5 Female Rural and Urban Weight for Age Z-Scores
52


sex: Female
8.00-
Rural Urban
6 00-
O
- Rural
Uiban
4.00-
O
O
o
o
o
8 8 8 O
o
-4.00-
10 11 12 13 14 15 16
Age (yrs)
Figure 6.6 Female Rural and Urban BMI for Age Z-Scores
Similar to male youth, no differences initially are seen for females from age 7
to 10. Standing height growth divergence between rural pastoralist and urban
dwelling female youth begins at age 11 and continues until age 15. After age 15, the
rural growth curve returns to closer proximity to that of urban females. In contrast to
the male growth curve comparisons (Figures 6.1, 6.2, and 6.3) where recovery in
male standing height is seen, there is no recovery of female standing height, weight
and BMI. The same period of growth faltering in standing height, weight and BMI
found in rural males beginning at age 12 also appears in females beginning earlier at
approximately age 10. This period of no-growth is discussed further in chapter 7.
53


Independent t-test results on female standing height, weight and BMI Z-scores
are summarized in table 6.4. Significance found to be equal to or less than alpha 0.05
are indicated by asterisk (*) and in bold font.
No differences in standing height, weight and BMI are present from age seven
through age 10, except for weight at age 8 ip = .003) and BMI (p =.000). Standing
height between rural and urban female youth diverges significantly at age 11 through
age 14 (p = .000). The rural female youth growth curve returns to closer proximity at
age 15.
Weight diverges significantly between rural and urban female youth
beginning at age 11 and continuing through age 14 (p = .000). Recovery of weight
differences begins at age 15 and catches up to the urban female population at age 16.
A difference in BMI appears at age 8 (p = .000), but returns to insignificance
at age 9 and 10. BMI differences are found to begin consistently at age 11 and
continue through age 14 (p = .000). Beginning at age 15 and through age 16, partial
recovery appears to occur.
Table 6.4 Female rural pastoralist and urban dwelling statistical analysis
Female HAZ___________________WAZ___________________BMIZ
Age (yrs) Rural / Urban N Mean SD Sig. (2- tailed) Mean SO Sig. (2- tailed) Mean SD Sig. (2- tailed)
7 Rural 54 -1.085 0.516 0.296 0.946 0.322 0.985 0.759 0.617 0.143
Urban 8 -1.286 0.405 0.948 0.238 0.430 0.247
8 Rural 33 -0.584 0.451 0.264 0.637 0.356 0.003* 0.607 0.548 0.000*
54


Table 6.4 (Cont.) Female rural pastoralist and urban dwelling statistical analysis
HAZ WAZ BMIZ
Age (VS) Rural / Urban N Mean SD Sig. (2- tailed) Mean SD Sig. (2- tailed) Mean SD Sig. (2- tailed)
Urban 10 -0.405 0.385 0.203 0.449 0.325 0.825
9 Rural 27 -0.474 0.497 0.718 0.440 0.327 0.428 0.197 0.529 0.391
Urban 13 -0.418 0.379 0.354 0.292 0.056 0.357
10 Rural 16 -0.310 0.278 0.056 0.238 0.312 0.095 0.101 0.744 0.643
Urban 21 -0.093 0.367 0.065 0.299 0.202 0.578
11 Rural 20 -0.610 0.489 0.000 0.563 0.379 0.000 0.356 0.494 0.000
Urban 22 0.304 0.443 0.343 0.429 0.536 0.520
12 Rural 63* -0.235 0.566 0.000 0.373 0.462 0.000 0.424 0.545 0.000
Urban 35 0.642 0.459 0.688 0.541 0.770 0.804
13 Rural 36 0.136 0.404 0.000 0.176 0.361 0.000 0.441 0.629 0.000
Urban 25 0.850 0.388 1.109 0.742 1.314 1.165
14 Rural 20 0.654 0.414 0.000 0.231 0.457 0.000 0.262 0.556 0.000
Urban 21 1.445 0.396 1.785 0.615 1.691 0.886
15 Rural 5 1.271 0.440 0.288 1.542 1.066 0.071 1.434 1.431 0.112
Urban 16 1.521 0.449 2.384 0.796 2.712 1.511
16 Rural 1 2.068 0.351 2.226 0.937 1.489 0.610
Urban 9 1.606 0.443 2.173 0.607 2.178 1.231
* Age 12 female weighted but left before other measurements could be taken
Based on the results presented, significant differences exist between both male
and female rural pastoralist and urban dwelling youth ages 7 through 16. Thus,
hypothesis 1 is upheld, but with special considerations. Significant differences do not
exist at all ages and among all indices as expected. Differences exist only at certain
ages and indices. Possible explanations for the unique growth patterns in rural males
and females are discussed in chapter 7.
55


Hypothesis 2 Results
Predictions under hypothesis 2 propose that significant differences in growth
will be present between two samples of rural Endulen Village area male and female
youth ages 7 through 11, non-settled and settled Maasai pastoralists. Data for males
are presented below followed by data for females.
Male data presented in Table 6.5 includes descriptive information according
to age including summary data of standing height (cm), weight (kg), and BMI for
rural non-settled (attending Misigio and Ndiyan primary schools) and settled
(attending Endulen primary school) Maasai pastoralists youth.
Table 6.5 Male rural non-settled and settled descriptive data
Standing
Height Weight
Male ________________i_____i____________(cm)_________________________(kg)_______________________BMI
Age (yrs) Non-settied Settled N Mean SD Median Mean SD Median Mean SD Median
7 Misigio/Ndiyan 21 115.338 5.732 116.600 18.476 2.101 18.000 13.868 0.966 13.895
Endulen 60 121.767 6.979 122.150 19.958 2.935 20.000 13.391 0.963 13.325
8 Misigio/Ndiyan 38 122.774 7.358 122.850 20.408 3.200 20.000 13.494 1.341 13.823
Endulen 10 127.000 7.783 125.500 22.750 3.268 21.500 14.059 0.986 14.084
9 Misigio/Ndiyan 24 124.629 6.216 125.400 20.854 2.564 21.000 13.436 1.392 13.610
Endulen 14 129.843 6.081 128.650 24.036 2.530 24.000 14.238 0.909 14.143
10 Misigio/Ndiyan 13 137.846 9.630 137.600 28.115 5.005 29.000 14.687 1.328 14.337
Endulen 12 130.275 6.405 130.400 26.042 4.659 24.500 15.302 2.239 65.200
11 Misigio/Ndiyan 13 134.285 9.249 137.900 25.808 3.761 27.500 14.236 0.710 14.198
Endulen 4 135.400 2.191 135.200 28.375 3.903 29.500 15.457 1.886 16.161
Figures 6.7, 6.8, and 6.9 display contrasting non-settled and settled male
standing height, weight, and BMI in scatter plots with Lowess smoothing using Z-
score converted data. Discussion follows these results.
56


sex: Male
Non-settled -
Settled
$
Non-settled
Settled
Non-settled
Settled
Figure 6.7 Male Non-settled and Settled Standing Height for Age Z-Scores
sex: Male
Non-settled -
Settled
8
Non-settled
Settled
** Non-settled
-----Settled
Figure 6.8 Male Non-settled and Settled Weight for Age Z-Scores
57


sex: Male
Non-settled -
Settled
Non-settled
Settled
Non-settled
Settled
§
Figure 6.9 Male Non-settled and Settled BMI for Age Z-Scores
Growth curves of standing height, weight and BMI Z-scores are indicative of
minimal divergence differences between rural non-settled and settled male youth.
Independent t-test results on male standing height, weight and BMI Z-scores
are summarized in table 6.6. Significance found to be equal to or less than alpha 0.05
are indicated by asterisk (*) and in bold font.
No differences in standing height are present from age seven through age 11.
Weight diverges significantly only at age 9 (p = .007). Differences in BMI appear at
ages 8 (p = .003), at age 9 (p = .023) and at age 11 (p = 009).
58


Table 6.6 Male rural non-settled and settled statistical analysis
Male__________________________________HAZ_____________________ WAZ BMIZ
Age (yrs) Non-settled Settled N Mean SD Sig. (2- tailed) Mean SD Sig. <2- tailed) Mean SD Sig. <2- tailed)
7 Misigio/Ndiyan 81 -0.995 0.490 0.961 0.868 0.295 0.373 0.651 0.475 0.081
Endulen 9 -0.987 0.385 0.777 0.246 0.358 0.446
8 Misigio/Ndiyan 48 -0.754 0.513 0.783 0.729 0.349 0.106 0.604 0.623 0.003*
Endulen 13 -0.797 0.385 0.556 0.282 0.003 0593
9 Misigio/Ndiyan 38 -0.558 0.447 0.157 0.610 0.311 0.007* 0.546 0.622 0.023*
Endulen 7 -0.298 0.388 0.253 0.301 0.024 0.304
10 Misigio/Ndiyan 25 -0.039 0.607 0.980 0.075 0.511 0.682 0.059 0.877 0.514
Endulen 24 -0.035 0.370 0.022 0.377 0.201 0.594
11 Misigio/Ndiyan 17 0.016 0.548 0.852 0.149 0.403 0.113 0.163 0.558 0.009*
Endulen 16 0.015 0.396 0.074 0.379 0.355 0.515
Female data presented in table 6.7 includes descriptive information according
to age including summary data of standing height (cm), weight (kg), and BMI for
rural non-settled (attending Misigio and Ndiyan primary schools) and settled
(attending Endulen primary school) Maasai pastoralists youth.
Table 6.7 Female rural non-settled and settled descriptive data
Standing
Female__________________________________Height______________________Weight______________________BMI
Age (yrs) Non-settled Settled N Mean (cm) SD Median Mean (k9) SD Median Mean SD Median
7 Misigio/Ndiyan 29 117.324 7.081 117.100 18.741 2.871 18.500 13.546 1.106 13.546
Endulen 25 120.472 7.993 119.600 18.940 3.336 18.000 12.998 1.410 12.862
8 Misigio/Ndiyan 20 126.590 5.642 125.350 21.774 2.798 21.250 13.535 0.823 13.4.001
Endulen 13 125.531 8.196 124.200 21.769 4.260 21.000 13.715 1.524 13.527
9 Misigio/Ndiyan 11 124.355 6.631 124.700 22.045 2.018 21.500 14.260 0.822 14.188
Endulen 16 130.144 7.024 129.750 24.750 3.291 24.000 14.586 1.253 14.405
10 Misigio/Ndiyan 2 131.700 4.101 131.700 27.500 3.536 27.500 15.814 1.053 15.814
Endulen 14 129.993 4.204 130.350 25.286 2.927 24.750 14.961 1.594 14.566
11 Misigio/Ndiyan 19 125.479 7.282 124.500 22.132 3.353 22.000 13.985 0.829 13.832
Endulen 1 131.500 131.500 29.000 29.000 16.771 16.771
59


Figures 6.10, 6.11 and 6.12 includes three figures that display contrasting non-
settled and settled female standing height, weight, and BMI in scatter plots with
Lowess smoothing using Z-score converted data. Discussion follows these results.
sex: Female
Non-sett!ed -
Settled
8
Non-settled
Settled
Non-sett led
Settled
Figure 6.10 Female Non-settled and Settled Standing Height for Age Z-Scores
60


sex: Female
Noil-settled -
Settled
.> Non-settled
5 Settled
*** Non-settled
Settled
Figure 6.11 Female Non-settled and Settled Weight for Age Z-Scores
sex: Female
Non-settled -
Settled
£
Non-settled
SEttled
~ Non-settled
SEttled
Figure 6.12 Female Non-settled and Settled BMI for Age Z-Scores
61


Growth curves of standing height, weight and BMI Z-scores are indicative of
minimal divergence differences between rural non-settled and settled female youth.
Statistical data and independent t-test results on the female standing height,
weight and BMI Z-scores are summarized in table 6.8. Significance found to be
equal to or less than alpha 0.05 are indicated by asterisk (*) and in bold font.
No differences in standing height are present from age seven through age 11.
Weight diverges significantly only at age 9 (p = .007). Differences in BMI appear at
ages 8 (p = .003), at age 9 (p = .023) and at age 11 (p = 009).
Table 6.8 Female rural non-settled and settled statistical analysis
Female HAZ WAZ BMIZ
Age Non-settled Sig. Sig. Sig.
(2- (2- (2-
(yrs) Settled N Mean SD tailed) Mean SD tailed) Mean SD tailed)
7 Misigio/Ndiyan 54 -1.085 0.516 0.296 0.946 0.322 0.985 0.759 0.617 0.143
Endulen 8 -1.286 0.405 0.948 0.238 0.430 0.247
8 Misigio/Ndiyan 33 -0.584 0.451 0.264 0.637 0.356 0.003* 0.607 0.548 0.000*
Endulen 10 -0.405 0.385 0.203 0.449 0.325 0.825
9 Misigio/Ndiyan 27 -0.474 0.497 0.718 0.440 0.327 0.428 0.197 0.529 0.391
Endulen 13 -0.418 0.379 0.354 0.292 0.056 0.357
10 Misigio/Ndiyan 16 -0.310 0.278 0.056 0.238 0.312 0.095 0.101 0.744 0.643
Endulen 21 -0.093 0.367 0.065 0.299 0.202 0.578
11 Misigio/Ndiyan 20 -0.610 0.489 0.000* 0.563 0.379 0.000* 0.356 0.494 0.000*
Endulen 22 0.304 0.443 0.343 0.429 0.536 0.520
Growth curves of standing height, weight and BMI Z-scores are indicative of
minimal divergence differences between rural non-settled and settled male youth.
62


Statistical data and independent t-test results on the female standing height,
weight and BM1 Z-scores are summarized in table 6.8. Significance found to be
equal to or less than alpha 0.05 are indicated by asterisk (*) and in bold font.
No differences in standing height are present from age seven through age 11.
Weight diverges significantly only at age 8 (p = .003). Differences in standing height,
weight and BMI only appear at age 11 (p = .000).
Based on the results presented, contrary to hypothesis 2 predictions,
significant differences do not exist at all ages and among all indices between both
non-settled and settled male and female youth ages 7 through 11. Hypothesis 2 is
rejected. However, it must be noted that differences do exist at certain ages and
indices and at age 11, significant differences begins to appear especially in the
females. Due to lack of data in ages 12 through 16, it is impossible to determine if
growth trajectory deviations would appear. Possible explanations for growth patterns
found in non-settled and settled males and females are discussed further in chapter 7.
Hypothesis 3 Results
Predictions under hypothesis 3 propose that no significant differences in
growth will be present between three samples of East African rural pastoralist male
and female youth populations ages 7 through 16. Data for males are presented below
followed by data for females.
63


Male data presented in Table 6.9 includes descriptive information according
to age including summary mean data of standing height (cm), weight (kg), and BM1
for three male youth East African pastoralist populations (Maasai, Datoga, Turkana)
Table 6.9 Male and Female East African pastoralists descriptive data
MALES
FEMALES
Maasai Datoga Turkana Maasai Datoga Turkana
Measure Age N Mean N Mean N Mean N Mean N Mean N Mean
Std. Height 7 81 120.1 9 118.2 14 113.9 54 118.8 14 120.2 18 117.3
Std. Height 8 48 123.7 5 124.2 13 119.9 33 126.2 6 125.7 20 119.6
Std. Height 9 38 126.6 12 126.3 6 122.9 27 127.8 16 125.1 20 125.7
Std. Height 10 25 134.2 9 132.2 5 127.8 16 130.2 12 131.9 16 128.9
Std. Height 11 17 134.6 10 136.1 2 131.6 20 125.8 12 135.1 9 134.0
Std. Height 12 110 136.1 9 142.6 3 139.4 63 131.3 12 137.4 9 140.1
Std. Height 13 57 140.0 5 139.0 10 143.8 36 136.8 7 149.2 19 144.2
Std. Height 14 34 144.8 8 141.4 6 139.6 20 144.4 10 149.1 16 150.1
Std. Height 15 22 151.8 3 145.4 2 155.1 5 153.5 9 152.4 10 152.4
Std. Height 16 12 160.4 2 159.1 3 148.1 1 165.3 4 156.7 5 152.4
Weight 7 81 19.6 9 20.2 14 17.6 54 18.8 14 19.9 18 18.3
Weight 8 48 20.9 5 21.6 13 19.3 33 21.8 6 22.1 20 18.4
Weight 9 38 22.0 12 23.2 6 20.2 27 23.7 16 21.5 20 21.7
Weight 10 25 27.1 9 24.5 5 23.2 16 25.6 12 23.7 16 23.2
Weight 11 17 26.4 10 27.6 2 24.0 20 22.5 12 27.0 9 24.4
Weight 12 110 27.0 9 29.1 3 26.9 63 24.3 12 28.6 9 29.3
Weight 13 57 28.4 5 29.7 10 29.9 36 26.2 7 34.2 19 29.7
Weight 14 34 30.5 8 30.1 6 28.1 20 30.0 10 35.5 16 32.8
Weight 15 22 35.2 3 35.1 2 35.6 5 42.5 9 38.0 10 34.8
Weight 16 12 39.2 2 36.5 3 31.7 1 49.0 4 38.3 5 34.0
BMI 7 81 13.6 9 14.5 14 13.6 54 13.4 14 13.8 18 13.3
BMI 8 48 13.7 5 14.0 13 13.4 33 13.7 6 14.0 20 12.9
BMI 9 38 13.8 12 14.5 6 13.4 27 14.5 16 13.7 20 13.7
BMI 10 25 15.1 9 14.0 5 14.2 16 15.1 12 13.6 16 14.0
BMI 11 17 14.6 10 14.9 2 13.9 20 14.2 12 14.8 9 13.6
BMI 12 110 14.6 9 14.3 3 13.8 63 14.1 12 15.2 9 14.9
BMI 13 57 14.5 5 15.4 10 14.5 36 14.0 7 15.4 19 14.3
BMI 14 34 14.6 8 15.1 6 14.4 20 14.4 10 16.0 16 14.6
BMI 15 22 15.3 3 16.6 2 14.8 5 18.0 9 16.4 10 15.0
BMI 16 12 15.2 2 14.4 3 14.5 1 17.9 4 15.6 5 14.6
64


Line graphs of standing height, weight, and BMI means contrasting male East
African pastoralists are illustrated below in figures 6.13, 6.14 and 6.15.
Male East African Pastoralist
-0Maasai --B--Datoga Turkana
Figure 6.13 Male East African Pastoralists Mean Standing Height (cm)
Figure 6.14 Male East African Pastoralists Mean Weight (kg)
65


Male East African Pastoralists
I------------------------------------------------
| Q Maasai -B Datoga - - Turkana
Figure 6.15 Male East African Pastoralists Mean BM1
Line graph trajectories of standing height, weight and BMI means reflect no
noticeable differences among male East African pastoralist youth populations.
Table 6.10 Male East African pastoralist one-way ANOVA
Male East African Pastoralist Comparisons
Sum of Squares df Mean Square F Sig.
Std. Hgt. Between Groups 48.553 2 24.277 0.154 0.858
Within Groups 4256.954 27 157.665
Total 4305.507 29
Wgt. Between Groups 27.966 2 13.983 0.412 0.666
Within Groups 915.349 27 33.902
Total 943.314 29
BMI Between Groups 2.687 2 1.344 3.232 0.055
Within Groups 11.225 27 0.416
Total 13.912 29
Parametric One-way ANOVA analysis indicates no significant differences are
evident in male standing height and weight among Maasai, Datoga, and Turkana
66


pastoralist youth age seven through 16. However, BM1 results ofp= .055 approach
significance. A Post Hoc test (Bonferroni) was performed to determine which of
these three pastoralists groups might be influencing the BMI outcome. Because data
sample sizes are small and not normally distributed, a non-parametric test (Kruskal-
Wallis) was performed producing similar results between parametric and non-
parametric tests (see table 6.11).
Table 6.11 Parametric/non-parametric test summary-Male East African comparison
PARAMETRIC / NON-PARAMETRIC TEST SUMMARY Male East African Comparison
Levene Browne- Kruskal-
Dependent Variable Statistic (Sig.) ANOVA Forsythe Wallis
Standing Height 0.858 0.858 0.858 0.887
Weight. 0.954 0.666 0.666 0.665
BMI 0.689 0.055 0.057 0.051
The results of the Post Hoc test (Bonferonni) to determine which of these
three pastoralist groups might be influencing the BMI outcome ofp = .055
(approaching alpha ofp = .05) as seen in table 6.12, are presented in table 6.4.
Bonferroni results show no differences in male standing height and weight among the
three East African Pastoralist populations. BMI results show no significant
differences between the Maasai and the Datoga (p=.965) and the Turkana (p=.423).
Borderline significant differences are present between the Datoga and the Turkana (p
- .053). This finding is interesting because of the differing geographic proximity and
ecological differences between the groups. This relationship will be examined further
in chapter 7.
67


Table 6.12 Male East African pastoralist Post Hoc test results
POST HOC TESTS Male East African Comparison
Dependent Variable Group Mean Difference Std. Error Sig.
Std. Hgt Bonferroni Maasai Datoga 0.756 5.615 1.000
Turkana 2.996 5.615 1.000
Datoga Maasai -0.756 5.615 1.000
Turkana 2.240 5.615 1.000
Turkana Maasai -2.996 5.615 1.000
Datoga -2.240 5.615 1.000
Wgt. Bonferroni Maasai Datoga -0.130 2.604 1.000
Turkana 1.980 2.604 1.000
Datoga Maasai 0.130 2.604 1.000
Turkana 2.110 2.604 1.000
Turkana Maasai -1.980 2.604 1.000
Datoga -2.110 2.604 1.000
BMI Bonferroni Maasai Datoga -0.291 0.288 0.965
Turkana 0.437 0.288 0.423
Datoga Maasai 0.291 0.288 0.965
Turkana 0.728 0.288 0.053
Turkana Maasai -0.437 0.288 0.423
Datoga -0.728 0.288 0.053
Female data presented in Table 6.9 includes descriptive information according
to age including summary mean data of standing height (cm), weight (kg), and BMI
for three female youth East African pastoralist populations (Maasai, Datoga, Turkana).
Line graphs of standing height, weight, and BMI means contrasting female
East African pastoralists are illustrated below in figures 6.16, 6.17 and 6.18.
68


Female East African Pastoralists
-O-
Maasai --B--Datoga
Turkana
Figure 6.16 Female East African Pastoralists Mean Standing Height (cm)
Female East African Pastoralists
O Maasai --B--Datoga ---A--- Turkana
Figure 6.17 Female East African Pastoralists Mean Weight (kg)
69


Female East African Pastoralists
O Maasai --B--Datoga ---A--- Turkana
Figure 6.18 Female East African Pastoralists BMI
Line graph trajectories of standing height, weight and BMI means reflect
minimal differences among female East African pastoralist youth populations.
Table 6.13 Female East African pastoralist one-way ANOVA
Female East African Pastoralist Comparisons
Sum of Squares df Mean Square F Sig.
Std. Hgt. Between Groups 28.784 2 14.392 0.079 0.925
Within Groups 4941.548 27 183.02
Total 4970.332 29
Wgt. Between Groups 27.507 2 13.754 0.224 0.801
Within Groups 1658.358 27 61.421
Total 1685.866 29
BMI Between Groups 4.268 2 2.119 1.466 0.249
Within Groups 39.022 27 1.445
Total 43.26 29
70


Parametric One-way ANOVA analysis indicates no significant differences are
evident in female standing height and weight among Maasai, Datoga, and Turkana
East African pastoralist age seven through 16. Because data sample sizes are small
and not normally distributed, a non-parametric test (Kruskal-Wallis) was performed
producing similar results between parametric and non-parametric tests (see table
6.14).
Table 6.14 Parametric/non-parametric test summary-Female East African
comparison
PARAMETRIC / NON-PARAMETRIC TEST SUMMARY Female East African Comparison
Dependent Variable Levene Statistic (Sig.) ANOVA Browne- Forsythe Kruskal- Wallis
Standing Height 0.985 0.925 0.925 0.932
Weight. 0.624 0.801 0.801 0.790
BMI 0.109 0.249 0.257 0.244
A Post Hoc test (Bonferonni) was performed and the results are presented in
table 6.15.
Table 6.15 Female East African pastoralist Post Hoc test results
POST HOC TESTS Female East African Comparison
Dependent Variable Group Mean Difference Std. Error Sig.
Std. Hgt Bonferroni Maasai Datoga -2.269 6.050 1.000
Turkana -0.459 6.050 1.000
Datoga Maasai 2.269 6.050 1.000
Turkana 1.810 6.050 1.000
Turkana Maasai 0.459 6.050 1.000
Datoga -1.810 6.050 1.000
Wgt. Bonferroni Maasai Datoga -0.454 3.505 1.000
Turkana 1.766 3.505 1.000
Datoga Maasai 0.454 3.505 1.000
71


Table 6.15(Cont.) Female East African pastoralist Post Hoc test results
POST HOC TESTS Female East African Comparison
Dependent Variable Group Mean Difference Std. Error Sig.
Turkana 2.220 3.505 1.000
Turkana Maasai -1.766 3.505 1.000
Datoga -2.220 3.505 1.000
BMI Bonferroni Maasai Datoga 0.085 0.538 1.000
Turkana 0.836 0.538 0.394
Datoga Maasai -0.085 0.538 1.000
Turkana 0.751 0.538 0.521
Turkana Maasai -0.836 0.538 0.394
Datoga -0.751 0.538 0.521
Bonferroni results show no differences in female standing height and weight
between the three East African Pastoralist populations. BMI results show no
significant differences between the female Maasai and the Turkana (p = .394), and the
Datoga (p = 1.000). The Datoga show no significant differences between Turkana
with ap = .521. This is different than found with the male BMI comparison between
pastoralist populations in table 6.12. This relationship will be examined in chapter 7.
Based on the results of our analysis, we fail to reject both hypothesis 3
predictions of no significant differences between three samples of East African rural
pastoralist male and female youth populations ages 7 through 16.
72


CHAPTER 7
DISCUSSION
This study examines rural Maasai pastoralist growth patterns in comparison to
urban Mwanzan dwelling youth, and between non-settled and settled rural pastoralist
youth inhabiting the Endulen Village area. Research results regarding growth
trajectories for both rural pastoralist versus urban dwelling, and non-settled versus
settled pastoralist youth are mixed across age groups. It appears that the growth
patterns discovered in this study are consistent with the original premise that growth
of rural pastoralist youth is influenced by their unique lifestyle behaviors and
environment. Review of the uniformity and fit of the rural Maasai data in comparison
with two other East African nomadic pastoralist groups showed that mean growth
measured by anthropometries of the Maasai were not significantly different from two
other pastoralist groups.
Growth and Environmental Differences between Rural
Pastoralists vs. Urban Dwellers and Non-settled
vs. Settled Rural Pastoralists
Differences in youth growth patterns have been noted between those living in
rural environments in contrast to those living in urban environments. Numerous
studies have documented increased height, weight, and Body Mass Index (BM1) in
urban populations when compared to their rural counterparts (Aspray et al., 2000;
Carlin et al., 2001). Scientists who study individuals residing in rural and urban
73


environments argue that the different growth pattern curves are a result of differing
lifestyles and environments. For urban dwellers, increased carbohydrates along with
decreased protein in the diet, and increased quantity and frequency of food
consumption along with decreased physical activity have been identified as important
factors that have contributed to height and weight increases. In addition, the
existence of a positive relationship between obesity and non-communicable diseases
(NCD) has been demonstrated in previous studies world-wide as result of rural to
urban migration (Aspray et al., 2000; Carlin et al., 2001; Knowler, Pettitt, Saad, & al.,
1991; Shell-Duncan & Obiero, 2000; Sobngwi et al., 2004).
In general, results of this study were consistent with previous findings. Male
and female rural Maasai non-settle pastoralist youth were both found generally to
have less height stature, weight and BMI when compared to their settled Endulen
Village pastoralist counterparts. Differences found to exist in height, weight and BMI
between rural versus urban males (Figures 6.1, 6.2 and 6.3) and females (Figures 6.4,
6.5, and 6.6), and between non-settled versus settled males (Figures 6.7, 6.8 and 6.9)
and females (Figures 6.10, 6.11 and 6.12) conformed to previous research predictions
and results (Aspray et al., 2000; Carlin et al., 2001; M. A. Little & Gray, 1990; L.R.
Pawloski, 2002; Sellen, 1999). The unique feature of this study is that the significant
differences were found at specific ages and in a specific pattern in both rural
populations.
74


Similarities between growth trajectories of the rural or non-settled pastoralists,
when compared to either the urban dwelling or settled population indicate a slower
growth specific to the rural lifestyle and environment, more pronounced in standing
height than weight or BMI. Less plastic than weight, height increases over time, and
is a better indicator of longer term total nutrition and health history than weight.
Human height has been determined to become greater along with more rapid growth
in proportion to a countrys wealth, generally representative of long-term
nourishment, healthcare, labor, and housing (Eveleth & Tanner, 1990). The
environmental factors combine in various proportions in response to a populations
lifestyle and environment, the majority of them generally are categorized under level
of nutrition and physical activity, prevalence of childhood infection, and availability
of health care (Eveleth & Tanner, 1990).
Differences in weight-for-age and especially BMI-for-age also are observed in
males and females in both rural versus urban and non-settled versus settled population
calculations. Because weight increases and decreases over time, it is a better
indicator of more recent or short-term nutrition, wellness and general status of health,
while BMI is often used as a proxy for current nutritional status (Bogin, 1995,, 1999).
Greater differences are present between BMI trajectories than differences found in
height and weight for both males and females. Taken together, these data imply that
75


the current negative lifestyle and environment of the non-settled Maasai is
measurably more stressful than that of the settled populations.
Between the ages of 7 to 9 years, height, weight, and BMI growth curve
trajectories are virtually indistinguishable in both sexes between the rural versus
urban dwelling or settled male and female youth. This lack of growth curve
differentiation during these early ages reflects similarity in long-term levels of
nutrition (protein energy) and physical activity. This is especially true for male and
female height. A period of growth faltering in height then occurs for rural males
between ages 10 to 12 and for rural females between ages 9 to 11. In like manner,
growth curve differences in weight and BMI between rural and urban males reach
statistical significance beginning at age 11 (BMI) and continue through age 16 (Table
6.2), and for females, beginning at age 11 and continuing through age 14 (Table 6.4).
Preadolescent Dip or Nadir
As mentioned above, the period of growth faltering observed in both sexes of
the rural Maasai pastoralist, is unusual in magnitude when contrasted to the normal
occurrence of a slight preadolescent dip or nadir in the height, weight, and BMI
growth curves of various African populations in Eveleth and Tanner (1990). A slight
preadolescent dip or nadir is a normal part of the growth process and is evident in
both male (see Figures 6.1, 6.2 and 6.3) and female (see Figures 6.4, 6.5 and 6.6)
urban dwelling youth. The cause of the leveling off or slight decrease in weight and
76


BMI of the growth curve is not currently known, but has been characterized as a
normal phenomenon. What is known is that this slowing of the growth velocity
marks the end of the juvenile stage (Bogin, 1999; Rogol, Clark, & Roemmich, 2000).
A possible explanation for the more pronounced nadir of the rural Maasai
youth population is that this rural population may be reflecting the magnitude of
effect an environmental insult has on a marginalized versus a well-off population.
Within the better-off populations of industrialized countries where resources are more
abundant overall, differences in growth between well-off and poor are relatively small,
while in developing countries with scarce resources, the gap in growth between well-
off and poor is greater. Eveleth and Tanner (1990) note that a considerable
proportion of the mean differences in body size among populations as well as
differences between individuals within populations are due to the differences in
environmental conditions.
Liminality Lifestyle Transition
from Youth to Adult Status
In general, following the occurrence of the nadir, the individual enters the
adolescent stage of accelerated growth (Bogin, 1999; Rogol, Clark, & Roemmich,
2000). This reversal in the rate of growth from deceleration to acceleration signals
the transition from juvenile to the adolescent stage of development, which normally
begins around the age of 13. In females, the juvenile period including the growth
77


nadir ends about the age of 10, three years before males. This difference in timing
between males and females is reflective of the earlier onset of adolescence in girls
(Bogin, 1999). The period of adolescence in humans is also the stage of life where
social and sexual maturation take place (Bogin, 1999). This transitional period of
stress and change could be contributing to the growth curve differences observed
between rural and urban male (see Figures 6.1, 6.2 and 6.3) and female (see Figures
6.4, 6.5 and 6.6) youth, and in response to cultural and behavioral environmental
differences. The unusual magnitude of growth faltering occurring between the ages
of 11 to 15 in rural but not in urban populations raises questions regarding the
influence of bio-cultural factors and complexities on human growth and development
in Maasai pastoralist youth. As previously mentioned, research by Shell-Duncan
(2000) and Sellen (1999) identifies traditional differences in nutritional resources and
health care access as strong contributors to different child growth trajectories. As
anthropologists, it is critical to look beyond these obvious broad categories and
explore the contextual influences within the rural Maasai pastoralist environment for
possible contributions to this pattern of growth faltering. Of particular interest is the
consideration that the unusual growth faltering of the Maasai youth may be
influenced by bio-cultural traditions and lifestyle practices, in particular, cultural rites
of passage that can result in physical and psychological stress on youth through their
adolescent years.
78


Within the rural nomadic pastoralist cultural settings of the Maasai, rites of
passage provide a continuity that link youth with adulthood and the individual to the
larger community (Turner, 1967,, 1969; van Gennep, 1960). During this period of
time of liminality, the child either boy or girl, is viewed within particular social
situations that reflect the childs qualities or temperament and provide a means to
analyze whether or not a child exhibits maturity when compared to other children
(Pratt, 2003). This liminality phase incorporates the combination of age, ability and
maturity which, taken together, serve to propel a child toward the end of childhood,
toward circumcision and eventual self-ownership as being an adult (Pratt, 2003).
These practices instill responsibility and value systems in youth, while providing the
transition from youth to adulthood. Arnold van Gennep (1960) proposes that The
life of an individual in any society is a series of passages from one age to another and
from one occupation to another (pg 2). This state of being between phases is defined
as the transitional stage of liminality (Turner, 1967,, 1969; van Gennep, 1960).
Quoting van Gennep (1960) further on liminality, Whoever passes from one to the
other finds himself physically and mag/o-religiously in a special situation for a certain
length of time: he wavers between two worlds (pg 18); ...be interpreted as a direct
Rite of Passage by means of which a person leaves one world behind and enters a
new one (pg 19).
79


Within rural Maasai pastoral societies, both boys and girls undergo a long
stressful process of liminality as they navigate this transition between two worlds to
a new way of adult life. The ages at which this period of lifestyle transition occurs
for boys and for girls matches the ages at which the growth faltering occurred in this
study.
Among the Maasai, puberty occurs about the age of 12 for boys, with
circumcision of the boys occurring as soon as they are sufficiently strong, generally
between the ages of 12 and 16 (Pratt, 2003). Sometimes the ceremony is performed
sooner if the parents are wealthy, but if they are poor the circumcision is put off until
they are able to pay for the ceremony. The circumcision of boys takes place every
four or five years and those who are circumcised at that time form an age cohort
bearing a special name chosen by the chief (Pratt, 2003).
Circumcision rites of girls are similar to boys, generally occurring between the
ages of 11 and 15 (Kouba & Muasher, 1985), but differ in the following respects.
Several girls are excised at the same time, and after scar tissue forms on the wound,
the marriage takes place as soon as the fiance is able to pay what he owes on the
dowry (Pratt, 2003). This marks the transition of the girl into the next age group, that
of a married woman. Even though the purpose of the girls operation is marriage,
marriage is viewed within the community as a social institution, not as a sexual union
(van Gennep, 1960). Circumcision ceremonies are independent of puberty, they are
80


of a sexual nature, since they incorporate both boys and girls into the adult society of
the sexes (van Gennep, 1960). Female circumcision, like male circumcision is an
initiation rite of passage signaling the passage of a child from youth to adulthood thus,
becoming a full member of the tribe (Kouba & Muasher, 1985).
Urban dwellers in Tanzania have in some part moved away from at least one
aspect of the rites of passage such as the ritualistic practice of male and female
circumcision. In lieu of rural pastoralist traditional rites of passage customs, todays
urban parents and families seek to give their children the basic nurturing, guidance,
resources and information needed to make a successful transition from childhood to
adulthood (Warfield-Coppock, 1992).
Based on distinctive growth curves found in this study, the possibility exists
that stressors experienced during rites of passage in rural environments may
contribute to the growth faltering. Phenotypic differences found between rural and
urban school age youth are possibly influenced by both bio-cultural traditions and
lifestyle practices, in particular cultural rites of passage resulting in physical and
psychological stress on youth through their adolescent years.
Evidence exists that in some youth, negative emotional or psychological stress
causes relative failure to grow, or in some cases may even stop growth in height often
by affecting the secretion of growth hormone (Bogin, 1999). Starvation is speculated
to reduce insulin-like growth factor (IGF) secretion and activity as a means of saving
81


body stores of protein for the maintenance and repair of tissues at the expense of
cellular growth (Bogin, 1999). When the stressor is removed, secretion of growth
hormone returns to normal and in clinical cases a catch-up occurs, which is
indistinguishable from the catch-up following administration of human growth
hormone to a child permanently deficient in growth hormone for physiological
reasons (Eveleth & Tanner, 1990).
Future research needs to look beyond the obvious broad categories of nutrition,
physical activity and healthcare access, and explore the origin of the effect of
contextual cultural and behavioral environmental influences may contribute toward
phenotypic variation.
Growth Trends in Rural Non-settled
vs. Settled Pastoralists
Growth patterns found are reflective of the unique Maasai pastoralist lifestyle,
behavior and environment. Data results are mixed for rural male and female Maasai
pastoralists, with a general trend indicating that rural non-settled pastoralists have
reduced height, weight and BMI when compared to their rural settled Endulen Village
counterparts. A notable difference between rural non-settled and settled youth in
BMI was found to exist only at age 11 for both sexes (see Figures 6.7, 6.8, 6.9, 6.10,
6.11 and 6.12). This lack of difference is contrary to previous research and results
where differences in height and weight were found (Brainard, 1990; Corbett, Gray,
82


Campbell, & Leslie, 2003; M. A. Little & Gray, 1990). Even though physical
differences between rural non-settled and settled pastoralist school youth lacks
statistical significance, figures 6.7, 6.8, 6.9, 6.10, 6.11 and 6.12 graphically show that
rural settled youth living in the village setting are indeed taller in height, weight and
BMI than youth experiencing a rural non-settled pastoralist lifestyle.
The general growth trend for both sexes shows growth differences of rural
settled youth progressively increasing in size over the rural non-settled youth from
age 7 until at age 11 (significant differences in all three indices only appears in rural
females). Further discussion of this proposed trend or lack of trend is not possible at
this time due to the lack of measurement data for children/youth age 12 though 16 in
the rural settled pastoralist group, since youth in the older age groups were not in
school the day measurements were taken. However, recalling data results from the
rural pastoralist/urban dweller comparison, significant growth differences begin at
age 11 in females and indications point to the impression that growth differences may
begin in males around age 13. This delayed occurrence of significant growth
differences in rural males is predictive because females normally start pubescence
approximately two years earlier than males (Bogin, 1999). Further research including
growth parameters of older youth to age 16 will be needed in order to discover if the
same pattern of growth faltering and significance exists between non-settled and
settled rural youth, as was found between rural pastoralist and urban dwelling youth.
83


An explanation of the apparent trend of larger overall growth found in rural
settled pastoralists follows similar findings in the rural versus urban population
comparisons. Similar lifestyle and environmental differences between rural and
urban youth are found within the rural settled and non-settled environments,
particularly, the protein/carbohydrate variation in diet and nutrition composition and
quantity. As previous studies have found comparing the rural pastoralist to the urban
dwelling population (Aspray et al., 2000; Carlin et al., 2001), nutrition for rural non-
settled youth consists of higher levels of protein from animal products, mostly milk,
while the rural settled youths diet is higher in carbohydrate composition, consisting
of grain products such as maize and millet (Corbett, Gray, Campbell, & Leslie, 2003).
In addition, the rural non-settled pastoralists diet is less dependable than that of the
rural settled youth related to seasonality and rainfall, which determine the availability
and quantity of nutritional substances for consumption, resulting in the rural non-
settled nomadic pastoralist youth incurring more frequent caloric deficits at older ages.
Older rural settled youth diet consists of the more reliable grain based cereal, which
allows the rural settled youth to become better nourished over the older rural non-
settled youth; thus contributing to rural settled youth being larger in height, weight
and BMI. These influences are drawn from research review and on-site observations.
No specific data were gathered on caloric availability.
84


School Lunches
There is a question of whether or not provisioned school lunches could
possibly be contributing to the trend of rural settled school youth of Endulen Village
being larger than their non-settled counterparts. Previous studies of school lunch
programs have produced mixed results. Some studies found no change in
anthropometric status after providing snack biscuits at school (van Stuijvenberg et al.,
1999), while other studies determined that students attending schools providing daily
lunches are larger when compared to youth attending schools without daily lunch
provision (Brainard, 1990; Powell, Walker, Chang, & Grantham-McGregor, 1998).
As discussed in chapter 3, Study Population, rural settled students living in Endulen
Village attending Endulen primary school are allowed to go home for lunch, while all
boarding students, as well as those students who travel to school from the countryside,
are provided a daily lunch generally consisting of maize meal gruel. It is possible that
the level of nutrition from the maize meal gruel lunch received by the rural Endulen
primary school youth is so minimal that anthropometric status was not impacted to a
significant degree.
Physical Activity
Another possible contributor to growth trend differences in physical size is the
difference in the level of physical activity experienced between the rural settled and
non-settled youth. The impressions regarding physical activity come from
85


observation and discussion with other local inhabitants. No systematic data were
available for this discussion. Endulen primary school is located within Endulen
Village; as such the rural settled school youth only have to walk a short distance in
comparison to the rural non-settled youth who walk a considerable distance each day
in order to attend school. Included in the settled youth category are the 3% of the
student body who live onsite in the dormitory facilities (see chapter 3). Since these
students only go home during school closure, their caloric expenditure is similar to
their settled counterparts. Rural non-settled school youth, in addition to walking long
distances to school each day, perform chores when they return home. Nomadic boys
walk considerable distances herding animals, and nomadic girls work with their
mothers collecting firewood, fetching water, and tending animals (M. A. Little &
Gray, 1990). There appears to be little doubt that non-settled youth expend
considerably more energy than settled youth, contributing to their smaller body size.
Another aspect of differing physical activity levels may include a social
behavior characteristic, the sexual division of labor. The ability of boys to become
independent cattle herders during the school attendance years places them at higher
risk of under nutrition than girls who are generally involved in domestic work and
food preparation. Over time, young males become increasingly involved in herding
activities that involve both high level of physical activity and allocation of time to
work rather than to food consumption (Sellen, 1999; D.W. Sellen, 2000). The social
86


roles of male and female youths differ in rural African populations and this may be
reflected in their different growth patterns.
Genetic Variation
Debate continues between researchers on the relative contributions of genes
and environment to human development (Bogin, 1999; Eveleth & Tanner, 1990). In
reality, human development is a product of the interactions between genes and
environments, thus it is erroneous to consider whether one or the other is more
important. There are acknowledged genetic differences between the rural and urban
populations that cannot be ignored when considering what factors may be
contributing to the phenotypic differences found. However, both rural non-settled
and rural settled populations inhabiting the Endulen Village area are genetically
similar, thus lowering the genetic influence while placing higher contributory levels
on lifestyle, behavior and the environment. Because the human phenotype is the
outcome of these interactions, it seems reasonable to conclude that differences in
lifestyle behaviors and environments between rural non-settled and rural settled
pastoralists, (as between rural and urban youth), specifically diet, nutrition and
physical activity levels are contributors to the trends observed in phenotypic variation.
Comparisons among East African Pastoralists
Growth pattern differences found in the rural Maasai pastoralist populations
when compared to urban (Mwanza) or more settled (Endulen Village) populations are
87